Hospital Medicine

From Western Michigan University, Homer Stryker MD School of Medicine, Kalamazoo, MI (Dr. Vaillant and Dr. Kavanaugh), Ferris State University, Grand Rapids, MI (Dr. Mersfelder), and Bronson Methodist Hospital, Kalamazoo, MI (Dr. Maynard).

Abstract

• Background: Procalcitonin has emerged as an important marker of sepsis and lung infections of bacterial origin. The role of procalcitonin in guiding antibiotic stewardship in lower respiratory tract infections and sepsis has been extensively studied, and use of this biomarker has been shown to decrease antibiotic usage in clinical trials. We sought to evaluate the impact of a pharmacist-driven initiative regarding discontinuation of antibiotics utilizing procalcitonin levels at a community teaching hospital.
• Methods: We retrospectively gathered baseline data on adult patients admitted to a community teaching hospital who were 18 years of age and older, under the care of an inpatient service, and had a single procalcitonin level < 0.25 mcg/L obtained during admission. We then prospectively identified an intervention group of similar patients using a web-based, real-time clinical surveillance system. When a low procalcitonin level was identified in the intervention group, the participating clinical pharmacists screened for antibiotic use and the indication(s), determined whether the antibiotic could be discontinued based on the low procalcitonin level and the absence of another indication for antibiotics, and, when appropriate, contacted the patient’s health care provider via telephone to discuss possible antibiotic discontinuation. The total antibiotic treatment duration was compared between the baseline and intervention groups.
• Results: A total of 172 patients were included in this study (86 in each group). The duration of antibiotic use was not significantly different between the baseline (3.14 ± 4.04 days) and the intervention (3.34 ± 2.8 days) groups (P = 0.1083). Other patient demographics did not influence antibiotic duration.
• Conclusion: Our study did not demonstrate a difference in total antibiotic treatment duration with the utilization of procalcitonin and an oral communication intervention made by a clinical pharmacist at a community-based teaching hospital. Outside of clinical trials, and in the absence of an algorithmic approach, procalcitonin has not consistently been shown to aid in the diagnosis and treatment of infectious diseases. It is important to have a comprehensive antimicrobial stewardship program to reduce antibiotic use and effectively use laboratory values.

Keywords: antibiotic use; bacterial infection; biomarkers; procalcitonin.

Procalcitonin is the precursor of the hormone calcitonin, which is normally produced in the parafollicular cells of the thyroid gland under physiological conditions.¹ However, procalcitonin is also released in response to a proinflammatory stimulus, especially that of bacterial origin.¹ The source of the procalcitonin surge seen during proinflammatory states is not the parafollicular cells of the thyroid, but rather the neuroendocrine cells of the lung and intestine.¹ Stimulants of procalcitonin in these scenarios include bacterial endotoxin, tumor necrosis factor, and interleukin-6.^1,2 Due to these observations, procalcitonin has emerged as an important marker of sepsis and lung infections of bacterial origin.³

The role of procalcitonin in guiding antibiotic stewardship in lower respiratory tract infections and sepsis has been extensively studied.^4,5 Various randomized controlled trials have shown that antibiotic stewardship guided by procalcitonin levels resulted in lower rates of antibiotic initiation and shorter duration of antibiotic use.^4-6 Similar results were obtained in prospective studies evaluating its role in patients with chronic obstructive pulmonary disease and sepsis.^7,8 Based on these data, protocol-driven procalcitonin-guided antibiotic stewardship appears beneficial.

Many of these studies employed rigorous protocols. Studies of procalcitonin use in a so-called real-world setting, in which the provider can order and use procalcitonin levels without the use of protocols, are limited. The objective of our study was to evaluate the impact of a pharmacist-driven initiative on discontinuing antibiotics, if indicated, utilizing single procalcitonin measurement results of < 0.25 mcg/L at a community teaching hospital.

Methods

Our study utilized a 2-phase approach. The first phase was a retrospective chart review to establish baseline data regarding adult inpatients with a low procalcitonin level; these patients were randomly selected over a 1-year period (2017). Patients were included if they were 18 years of age or older, under the care of an inpatient service, and had a single procalcitonin level < 0.25 mcg/L obtained during their admission. Patients admitted to the intensive care unit were excluded. In the second phase, we prospectively identified similar patients admitted between January and March 2018 using a web-based, real-time clinical surveillance system. When patients with low procalcitonin levels were identified, 2 participating clinical pharmacists screened for antibiotic use and indication. If it was determined that the antibiotic could be discontinued as a result of the low procalcitonin level and no additional indication for antibiotics was present, the pharmacist contacted the patient’s health care provider via telephone to discuss possible antibiotic discontinuation. Data collected before and after the intervention included total antibiotic treatment duration, white blood cell count, maximum temperature, age, and procalcitonin level.

A sample size of 86 was calculated to provide an alpha of 0.05 and a power of 0.8. A nonparametric Wilcoxon 2-sample test was used to test for a difference in duration of antibiotic treatment between the baseline and intervention groups. A nonparametric test was used due to right-skewed data. All patients were included in the group analysis, regardless of antibiotic use, as the procalcitonin level may have been used in the decision to initiate antibiotics, and this is more representative of a real-world application of the test. This allowed for detection of a significant decrease of 2 days in antibiotic duration post intervention, with a 10% margin to compensate for potential missing data. Data from 86 patients obtained prior to the pharmacist intervention acted as a control comparison group. Statistical analysis was performed using SAS 9.4.

Results

A total of 172 patients were included in this study: 86 patients prior to the intervention, and 86 after implementation. Baseline demographics, laboratory values, vitals, and principal diagnoses for both groups are shown in Table 1 and Table 2. The most common indications for procalcitonin measurement were pneumonia (45.9%), chronic obstructive pulmonary disease (15.7%), and sepsis (14.5%). The remaining diagnoses were encephalopathy, fever and leukocytosis, skin and soft tissue infection, urinary tract infection or pyelonephritis, bone and joint infection, meningitis, intra-abdominal infection, and asthma exacerbation.

Antibiotic therapy was initiated in 68% of the patients overall, 59% in the baseline group and 76% in the intervention group. The duration of antibiotic use was not significantly different between the baseline (3.14 ± 4.04 days) and intervention (3.34 ± 2.8 days) groups (P = 0.1083). Furthermore, antibiotic treatment duration did not vary significantly with patient age, white blood cell count, maximum temperature, or procalcitonin level in either group. Although there was no difference in total antibiotic treatment duration, a post-hoc analysis revealed a 0.6-day decrease in the interval between the date of procalcitonin measurement and the stop date of antibiotics in the intervention group. The average time from admission to obtaining a procalcitonin level was 3 days in the baseline group and 2 days in the intervention group.

Discussion

Our study did not demonstrate a difference in total antibiotic treatment duration with procalcitonin measurement and an oral communication intervention made by a clinical pharmacist at a community teaching hospital with a well-established antimicrobial stewardship program. This may be due to several factors. First, the providers did not receive ongoing education regarding the appropriate use or interpretation of procalcitonin. The procalcitonin result in the electronic health record references the risk for progression to severe sepsis and/or septic shock, but does not indicate how to use procalcitonin as an aid in antibiotic decision-making. However, a recent study in patients with lower respiratory tract infections treated by providers who had been educated on the use of procalcitonin failed to find a reduction in total antibiotic use.⁹ Second, our study included hospital-wide use of procalcitonin, and was not limited to infections for which procalcitonin use has the strongest evidence (eg, upper respiratory tract infections, pneumonia, sepsis). Thus, providers may have been less likely to use protocolized guidelines. Last, we did not limit the data on antibiotic duration to patients with a procalcitonin level obtained within a defined time frame from antibiotic initiation or time of admission, and some patients had procalcitonin levels measured several days into their hospital stay. While this is likely to have skewed the data in favor of longer antibiotic treatment courses, it also represents a more realistic way in which this laboratory test is being used. Our post-hoc finding of earlier discontinuation of antibiotics after procalcitonin measurement suggests that our intervention may have influenced the decision to discontinue antibiotics. Such an effect may be augmented if procalcitonin is measured earlier in a hospital admission.

Previous studies have also failed to show that the use of procalcitonin decreased duration of antibiotics.^9,10 In the aforementioned study regarding real-world outcomes in patients with lower respiratory tract infections, antibiotic duration was not reduced, despite provider education.⁹ A large observational study that evaluated real-world outcomes in intensive care unit patients did not find decreased antibiotic use or improved outcomes with procalcitonin use.¹⁰ With these large studies evaluating the 2 most common infectious diseases for which procalcitonin has previously been found to have clinical benefit, it is important for institutions to re-evaluate how procalcitonin is being utilized by providers. Furthermore, institutions should explore ways to optimize procalcitonin use and decrease unnecessary health care costs. Notably, the current community-acquired pneumonia guidelines recommend against routine use of procalcitonin.¹¹

Conclusion

Outside of clinical trials, and in the absence of an algorithmic approach, procalcitonin has not consistently been shown to aid in the diagnosis or treatment of infectious diseases. It is important to have a comprehensive antimicrobial stewardship program that includes an algorithmic protocol to promote appropriate laboratory testing and reduce total antibiotic use. In addition to improved communication with providers, other interventions need to be investigated to effectively use this biomarker or limit its use.

Acknowledgment: The authors thank the Western Michigan University Department of Epidemiology and Biostatistics for their assistance in preparing this article.

Corresponding author: James Vaillant, MD, Western Michigan University, Homer Stryker MD School of Medicine, 1000 Oakland Drive, Kalamazoo, MI, 49008; james.vaillant@med.wmich.edu.

Financial disclosures: None.

From Western Michigan University, Homer Stryker MD School of Medicine, Kalamazoo, MI (Dr. Vaillant and Dr. Kavanaugh), Ferris State University, Grand Rapids, MI (Dr. Mersfelder), and Bronson Methodist Hospital, Kalamazoo, MI (Dr. Maynard).

Abstract

• Background: Procalcitonin has emerged as an important marker of sepsis and lung infections of bacterial origin. The role of procalcitonin in guiding antibiotic stewardship in lower respiratory tract infections and sepsis has been extensively studied, and use of this biomarker has been shown to decrease antibiotic usage in clinical trials. We sought to evaluate the impact of a pharmacist-driven initiative regarding discontinuation of antibiotics utilizing procalcitonin levels at a community teaching hospital.
• Methods: We retrospectively gathered baseline data on adult patients admitted to a community teaching hospital who were 18 years of age and older, under the care of an inpatient service, and had a single procalcitonin level < 0.25 mcg/L obtained during admission. We then prospectively identified an intervention group of similar patients using a web-based, real-time clinical surveillance system. When a low procalcitonin level was identified in the intervention group, the participating clinical pharmacists screened for antibiotic use and the indication(s), determined whether the antibiotic could be discontinued based on the low procalcitonin level and the absence of another indication for antibiotics, and, when appropriate, contacted the patient’s health care provider via telephone to discuss possible antibiotic discontinuation. The total antibiotic treatment duration was compared between the baseline and intervention groups.
• Results: A total of 172 patients were included in this study (86 in each group). The duration of antibiotic use was not significantly different between the baseline (3.14 ± 4.04 days) and the intervention (3.34 ± 2.8 days) groups (P = 0.1083). Other patient demographics did not influence antibiotic duration.
• Conclusion: Our study did not demonstrate a difference in total antibiotic treatment duration with the utilization of procalcitonin and an oral communication intervention made by a clinical pharmacist at a community-based teaching hospital. Outside of clinical trials, and in the absence of an algorithmic approach, procalcitonin has not consistently been shown to aid in the diagnosis and treatment of infectious diseases. It is important to have a comprehensive antimicrobial stewardship program to reduce antibiotic use and effectively use laboratory values.

Keywords: antibiotic use; bacterial infection; biomarkers; procalcitonin.

Procalcitonin is the precursor of the hormone calcitonin, which is normally produced in the parafollicular cells of the thyroid gland under physiological conditions.¹ However, procalcitonin is also released in response to a proinflammatory stimulus, especially that of bacterial origin.¹ The source of the procalcitonin surge seen during proinflammatory states is not the parafollicular cells of the thyroid, but rather the neuroendocrine cells of the lung and intestine.¹ Stimulants of procalcitonin in these scenarios include bacterial endotoxin, tumor necrosis factor, and interleukin-6.^1,2 Due to these observations, procalcitonin has emerged as an important marker of sepsis and lung infections of bacterial origin.³

The role of procalcitonin in guiding antibiotic stewardship in lower respiratory tract infections and sepsis has been extensively studied.^4,5 Various randomized controlled trials have shown that antibiotic stewardship guided by procalcitonin levels resulted in lower rates of antibiotic initiation and shorter duration of antibiotic use.^4-6 Similar results were obtained in prospective studies evaluating its role in patients with chronic obstructive pulmonary disease and sepsis.^7,8 Based on these data, protocol-driven procalcitonin-guided antibiotic stewardship appears beneficial.

Many of these studies employed rigorous protocols. Studies of procalcitonin use in a so-called real-world setting, in which the provider can order and use procalcitonin levels without the use of protocols, are limited. The objective of our study was to evaluate the impact of a pharmacist-driven initiative on discontinuing antibiotics, if indicated, utilizing single procalcitonin measurement results of < 0.25 mcg/L at a community teaching hospital.

Methods

Our study utilized a 2-phase approach. The first phase was a retrospective chart review to establish baseline data regarding adult inpatients with a low procalcitonin level; these patients were randomly selected over a 1-year period (2017). Patients were included if they were 18 years of age or older, under the care of an inpatient service, and had a single procalcitonin level < 0.25 mcg/L obtained during their admission. Patients admitted to the intensive care unit were excluded. In the second phase, we prospectively identified similar patients admitted between January and March 2018 using a web-based, real-time clinical surveillance system. When patients with low procalcitonin levels were identified, 2 participating clinical pharmacists screened for antibiotic use and indication. If it was determined that the antibiotic could be discontinued as a result of the low procalcitonin level and no additional indication for antibiotics was present, the pharmacist contacted the patient’s health care provider via telephone to discuss possible antibiotic discontinuation. Data collected before and after the intervention included total antibiotic treatment duration, white blood cell count, maximum temperature, age, and procalcitonin level.

A sample size of 86 was calculated to provide an alpha of 0.05 and a power of 0.8. A nonparametric Wilcoxon 2-sample test was used to test for a difference in duration of antibiotic treatment between the baseline and intervention groups. A nonparametric test was used due to right-skewed data. All patients were included in the group analysis, regardless of antibiotic use, as the procalcitonin level may have been used in the decision to initiate antibiotics, and this is more representative of a real-world application of the test. This allowed for detection of a significant decrease of 2 days in antibiotic duration post intervention, with a 10% margin to compensate for potential missing data. Data from 86 patients obtained prior to the pharmacist intervention acted as a control comparison group. Statistical analysis was performed using SAS 9.4.

Results

A total of 172 patients were included in this study: 86 patients prior to the intervention, and 86 after implementation. Baseline demographics, laboratory values, vitals, and principal diagnoses for both groups are shown in Table 1 and Table 2. The most common indications for procalcitonin measurement were pneumonia (45.9%), chronic obstructive pulmonary disease (15.7%), and sepsis (14.5%). The remaining diagnoses were encephalopathy, fever and leukocytosis, skin and soft tissue infection, urinary tract infection or pyelonephritis, bone and joint infection, meningitis, intra-abdominal infection, and asthma exacerbation.

Antibiotic therapy was initiated in 68% of the patients overall, 59% in the baseline group and 76% in the intervention group. The duration of antibiotic use was not significantly different between the baseline (3.14 ± 4.04 days) and intervention (3.34 ± 2.8 days) groups (P = 0.1083). Furthermore, antibiotic treatment duration did not vary significantly with patient age, white blood cell count, maximum temperature, or procalcitonin level in either group. Although there was no difference in total antibiotic treatment duration, a post-hoc analysis revealed a 0.6-day decrease in the interval between the date of procalcitonin measurement and the stop date of antibiotics in the intervention group. The average time from admission to obtaining a procalcitonin level was 3 days in the baseline group and 2 days in the intervention group.

Discussion

Our study did not demonstrate a difference in total antibiotic treatment duration with procalcitonin measurement and an oral communication intervention made by a clinical pharmacist at a community teaching hospital with a well-established antimicrobial stewardship program. This may be due to several factors. First, the providers did not receive ongoing education regarding the appropriate use or interpretation of procalcitonin. The procalcitonin result in the electronic health record references the risk for progression to severe sepsis and/or septic shock, but does not indicate how to use procalcitonin as an aid in antibiotic decision-making. However, a recent study in patients with lower respiratory tract infections treated by providers who had been educated on the use of procalcitonin failed to find a reduction in total antibiotic use.⁹ Second, our study included hospital-wide use of procalcitonin, and was not limited to infections for which procalcitonin use has the strongest evidence (eg, upper respiratory tract infections, pneumonia, sepsis). Thus, providers may have been less likely to use protocolized guidelines. Last, we did not limit the data on antibiotic duration to patients with a procalcitonin level obtained within a defined time frame from antibiotic initiation or time of admission, and some patients had procalcitonin levels measured several days into their hospital stay. While this is likely to have skewed the data in favor of longer antibiotic treatment courses, it also represents a more realistic way in which this laboratory test is being used. Our post-hoc finding of earlier discontinuation of antibiotics after procalcitonin measurement suggests that our intervention may have influenced the decision to discontinue antibiotics. Such an effect may be augmented if procalcitonin is measured earlier in a hospital admission.

Previous studies have also failed to show that the use of procalcitonin decreased duration of antibiotics.^9,10 In the aforementioned study regarding real-world outcomes in patients with lower respiratory tract infections, antibiotic duration was not reduced, despite provider education.⁹ A large observational study that evaluated real-world outcomes in intensive care unit patients did not find decreased antibiotic use or improved outcomes with procalcitonin use.¹⁰ With these large studies evaluating the 2 most common infectious diseases for which procalcitonin has previously been found to have clinical benefit, it is important for institutions to re-evaluate how procalcitonin is being utilized by providers. Furthermore, institutions should explore ways to optimize procalcitonin use and decrease unnecessary health care costs. Notably, the current community-acquired pneumonia guidelines recommend against routine use of procalcitonin.¹¹

Conclusion

Outside of clinical trials, and in the absence of an algorithmic approach, procalcitonin has not consistently been shown to aid in the diagnosis or treatment of infectious diseases. It is important to have a comprehensive antimicrobial stewardship program that includes an algorithmic protocol to promote appropriate laboratory testing and reduce total antibiotic use. In addition to improved communication with providers, other interventions need to be investigated to effectively use this biomarker or limit its use.

Acknowledgment: The authors thank the Western Michigan University Department of Epidemiology and Biostatistics for their assistance in preparing this article.

Corresponding author: James Vaillant, MD, Western Michigan University, Homer Stryker MD School of Medicine, 1000 Oakland Drive, Kalamazoo, MI, 49008; james.vaillant@med.wmich.edu.

Financial disclosures: None.

From Western Michigan University, Homer Stryker MD School of Medicine, Kalamazoo, MI (Dr. Vaillant and Dr. Kavanaugh), Ferris State University, Grand Rapids, MI (Dr. Mersfelder), and Bronson Methodist Hospital, Kalamazoo, MI (Dr. Maynard).

Abstract

• Background: Procalcitonin has emerged as an important marker of sepsis and lung infections of bacterial origin. The role of procalcitonin in guiding antibiotic stewardship in lower respiratory tract infections and sepsis has been extensively studied, and use of this biomarker has been shown to decrease antibiotic usage in clinical trials. We sought to evaluate the impact of a pharmacist-driven initiative regarding discontinuation of antibiotics utilizing procalcitonin levels at a community teaching hospital.
• Methods: We retrospectively gathered baseline data on adult patients admitted to a community teaching hospital who were 18 years of age and older, under the care of an inpatient service, and had a single procalcitonin level < 0.25 mcg/L obtained during admission. We then prospectively identified an intervention group of similar patients using a web-based, real-time clinical surveillance system. When a low procalcitonin level was identified in the intervention group, the participating clinical pharmacists screened for antibiotic use and the indication(s), determined whether the antibiotic could be discontinued based on the low procalcitonin level and the absence of another indication for antibiotics, and, when appropriate, contacted the patient’s health care provider via telephone to discuss possible antibiotic discontinuation. The total antibiotic treatment duration was compared between the baseline and intervention groups.
• Results: A total of 172 patients were included in this study (86 in each group). The duration of antibiotic use was not significantly different between the baseline (3.14 ± 4.04 days) and the intervention (3.34 ± 2.8 days) groups (P = 0.1083). Other patient demographics did not influence antibiotic duration.
• Conclusion: Our study did not demonstrate a difference in total antibiotic treatment duration with the utilization of procalcitonin and an oral communication intervention made by a clinical pharmacist at a community-based teaching hospital. Outside of clinical trials, and in the absence of an algorithmic approach, procalcitonin has not consistently been shown to aid in the diagnosis and treatment of infectious diseases. It is important to have a comprehensive antimicrobial stewardship program to reduce antibiotic use and effectively use laboratory values.

Keywords: antibiotic use; bacterial infection; biomarkers; procalcitonin.

Procalcitonin is the precursor of the hormone calcitonin, which is normally produced in the parafollicular cells of the thyroid gland under physiological conditions.¹ However, procalcitonin is also released in response to a proinflammatory stimulus, especially that of bacterial origin.¹ The source of the procalcitonin surge seen during proinflammatory states is not the parafollicular cells of the thyroid, but rather the neuroendocrine cells of the lung and intestine.¹ Stimulants of procalcitonin in these scenarios include bacterial endotoxin, tumor necrosis factor, and interleukin-6.^1,2 Due to these observations, procalcitonin has emerged as an important marker of sepsis and lung infections of bacterial origin.³

The role of procalcitonin in guiding antibiotic stewardship in lower respiratory tract infections and sepsis has been extensively studied.^4,5 Various randomized controlled trials have shown that antibiotic stewardship guided by procalcitonin levels resulted in lower rates of antibiotic initiation and shorter duration of antibiotic use.^4-6 Similar results were obtained in prospective studies evaluating its role in patients with chronic obstructive pulmonary disease and sepsis.^7,8 Based on these data, protocol-driven procalcitonin-guided antibiotic stewardship appears beneficial.

Many of these studies employed rigorous protocols. Studies of procalcitonin use in a so-called real-world setting, in which the provider can order and use procalcitonin levels without the use of protocols, are limited. The objective of our study was to evaluate the impact of a pharmacist-driven initiative on discontinuing antibiotics, if indicated, utilizing single procalcitonin measurement results of < 0.25 mcg/L at a community teaching hospital.

Methods

Our study utilized a 2-phase approach. The first phase was a retrospective chart review to establish baseline data regarding adult inpatients with a low procalcitonin level; these patients were randomly selected over a 1-year period (2017). Patients were included if they were 18 years of age or older, under the care of an inpatient service, and had a single procalcitonin level < 0.25 mcg/L obtained during their admission. Patients admitted to the intensive care unit were excluded. In the second phase, we prospectively identified similar patients admitted between January and March 2018 using a web-based, real-time clinical surveillance system. When patients with low procalcitonin levels were identified, 2 participating clinical pharmacists screened for antibiotic use and indication. If it was determined that the antibiotic could be discontinued as a result of the low procalcitonin level and no additional indication for antibiotics was present, the pharmacist contacted the patient’s health care provider via telephone to discuss possible antibiotic discontinuation. Data collected before and after the intervention included total antibiotic treatment duration, white blood cell count, maximum temperature, age, and procalcitonin level.

A sample size of 86 was calculated to provide an alpha of 0.05 and a power of 0.8. A nonparametric Wilcoxon 2-sample test was used to test for a difference in duration of antibiotic treatment between the baseline and intervention groups. A nonparametric test was used due to right-skewed data. All patients were included in the group analysis, regardless of antibiotic use, as the procalcitonin level may have been used in the decision to initiate antibiotics, and this is more representative of a real-world application of the test. This allowed for detection of a significant decrease of 2 days in antibiotic duration post intervention, with a 10% margin to compensate for potential missing data. Data from 86 patients obtained prior to the pharmacist intervention acted as a control comparison group. Statistical analysis was performed using SAS 9.4.

Results

A total of 172 patients were included in this study: 86 patients prior to the intervention, and 86 after implementation. Baseline demographics, laboratory values, vitals, and principal diagnoses for both groups are shown in Table 1 and Table 2. The most common indications for procalcitonin measurement were pneumonia (45.9%), chronic obstructive pulmonary disease (15.7%), and sepsis (14.5%). The remaining diagnoses were encephalopathy, fever and leukocytosis, skin and soft tissue infection, urinary tract infection or pyelonephritis, bone and joint infection, meningitis, intra-abdominal infection, and asthma exacerbation.

Antibiotic therapy was initiated in 68% of the patients overall, 59% in the baseline group and 76% in the intervention group. The duration of antibiotic use was not significantly different between the baseline (3.14 ± 4.04 days) and intervention (3.34 ± 2.8 days) groups (P = 0.1083). Furthermore, antibiotic treatment duration did not vary significantly with patient age, white blood cell count, maximum temperature, or procalcitonin level in either group. Although there was no difference in total antibiotic treatment duration, a post-hoc analysis revealed a 0.6-day decrease in the interval between the date of procalcitonin measurement and the stop date of antibiotics in the intervention group. The average time from admission to obtaining a procalcitonin level was 3 days in the baseline group and 2 days in the intervention group.

Discussion

Our study did not demonstrate a difference in total antibiotic treatment duration with procalcitonin measurement and an oral communication intervention made by a clinical pharmacist at a community teaching hospital with a well-established antimicrobial stewardship program. This may be due to several factors. First, the providers did not receive ongoing education regarding the appropriate use or interpretation of procalcitonin. The procalcitonin result in the electronic health record references the risk for progression to severe sepsis and/or septic shock, but does not indicate how to use procalcitonin as an aid in antibiotic decision-making. However, a recent study in patients with lower respiratory tract infections treated by providers who had been educated on the use of procalcitonin failed to find a reduction in total antibiotic use.⁹ Second, our study included hospital-wide use of procalcitonin, and was not limited to infections for which procalcitonin use has the strongest evidence (eg, upper respiratory tract infections, pneumonia, sepsis). Thus, providers may have been less likely to use protocolized guidelines. Last, we did not limit the data on antibiotic duration to patients with a procalcitonin level obtained within a defined time frame from antibiotic initiation or time of admission, and some patients had procalcitonin levels measured several days into their hospital stay. While this is likely to have skewed the data in favor of longer antibiotic treatment courses, it also represents a more realistic way in which this laboratory test is being used. Our post-hoc finding of earlier discontinuation of antibiotics after procalcitonin measurement suggests that our intervention may have influenced the decision to discontinue antibiotics. Such an effect may be augmented if procalcitonin is measured earlier in a hospital admission.

Previous studies have also failed to show that the use of procalcitonin decreased duration of antibiotics.^9,10 In the aforementioned study regarding real-world outcomes in patients with lower respiratory tract infections, antibiotic duration was not reduced, despite provider education.⁹ A large observational study that evaluated real-world outcomes in intensive care unit patients did not find decreased antibiotic use or improved outcomes with procalcitonin use.¹⁰ With these large studies evaluating the 2 most common infectious diseases for which procalcitonin has previously been found to have clinical benefit, it is important for institutions to re-evaluate how procalcitonin is being utilized by providers. Furthermore, institutions should explore ways to optimize procalcitonin use and decrease unnecessary health care costs. Notably, the current community-acquired pneumonia guidelines recommend against routine use of procalcitonin.¹¹

Conclusion

Outside of clinical trials, and in the absence of an algorithmic approach, procalcitonin has not consistently been shown to aid in the diagnosis or treatment of infectious diseases. It is important to have a comprehensive antimicrobial stewardship program that includes an algorithmic protocol to promote appropriate laboratory testing and reduce total antibiotic use. In addition to improved communication with providers, other interventions need to be investigated to effectively use this biomarker or limit its use.

Acknowledgment: The authors thank the Western Michigan University Department of Epidemiology and Biostatistics for their assistance in preparing this article.

Corresponding author: James Vaillant, MD, Western Michigan University, Homer Stryker MD School of Medicine, 1000 Oakland Drive, Kalamazoo, MI, 49008; james.vaillant@med.wmich.edu.

Financial disclosures: None.

The American Heart Association/American Stroke Association has developed a “conceptual framework” to assist emergency medical service (EMS) providers and in-hospital triage teams handle suspected cases of acute stroke during the ongoing COVID-19 crisis and future pandemics. A key goal is to ensure timely transfer of patients while minimizing the risk of infectious exposure for EMS personnel, coworkers, and other patients, the writing group says.

“Acute ischemic stroke is still a highly devastating disease and the Time Is Brain paradigm remains true during the COVID-19 pandemic as well,” said writing group chair Mayank Goyal, MD, of the University of Calgary (Alta.)

“We have highly effective and proven treatments available. As such, treatment delays due to additional screening requirements and personal protection equipment (PPE) should be kept at a minimum,” Dr. Goyal said.

“Practicing COVID-19 stroke work flows, through simulation training, can help to reduce treatment delays, minimize the risk of infectious exposure for patients and staff, and help alleviate stress,” he added.
 

A new layer of complexity

The guidance statement, Prehospital Triage of Acute Stroke Patients During the COVID-19 Pandemic, was published online May 13 in the journal Stroke.

“The need to limit infectious spread during the COVID-19 pandemic has added a new layer of complexity to prehospital stroke triage and transfer,” the writing group noted. “Timely and enhanced” communication between EMS, hospitals, and local coordinating authorities are critical, especially ambulance-and facility-based telestroke networks, they wrote.

The main factors to guide the triage decision are the likelihood of a large vessel occlusion; the magnitude of additional delays because of interhospital transfer and work flow efficiency at the primary stroke center or acute stroke ready hospital; the need for advanced critical care resources; and the available bed, staff, and PPE resources at the hospitals.

The group said it “seems reasonable” to lower the threshold to bypass hospitals that can’t provide acute stroke treatment in favor of transporting to a hospital that is “stroke ready,” particularly in patients likely to require advanced care. They cautioned, however, that taking all acute stroke patients to a comprehensive stroke center could overwhelm these centers and lead to clustering of COVID-19 patients.

They said it is equally important to ensure “necessary transfers” of stroke patients who would benefit from endovascular therapy or neurocritical care and avoid unnecessary patient transfers. “Doing so will likely require local hospital boards and health care authorities to collaborate and establish local guidelines and protocols,” the writing group said.

“During the COVID-19 pandemic, it is more important than ever to ensure that stroke patients are taken to the right hospital that can meet their urgent needs at the outset,” Dr. Goyal commented in an AHA news release.

The writing group emphasized that the principles put forth in the document are intended as suggestions rather than strict rules and will be adapted and updated to meet the evolving needs during the COVID-19 crisis and future pandemics.

“The process of improving stroke work flow and getting the correct patient to the correct hospital fast is dependent on training, protocols, simulation, technology, and – probably most importantly – teamwork. These principles are extremely important during the current pandemic but will be useful in improving stroke care afterwards as well,” Dr. Goyal said.

This research had no commercial funding. Members of the writing committee are on several AHA/ASA Council Science Subcommittees, including the Emergency Neurovascular Care, the Telestroke, and the Neurovascular Intervention committees. Goyal is a consultant for Medtronic, Stryker, Microvention, GE Healthcare, and Mentice. A complete list of author disclosures is available with the original article.

This article first appeared on Medscape.com.

The American Heart Association/American Stroke Association has developed a “conceptual framework” to assist emergency medical service (EMS) providers and in-hospital triage teams handle suspected cases of acute stroke during the ongoing COVID-19 crisis and future pandemics. A key goal is to ensure timely transfer of patients while minimizing the risk of infectious exposure for EMS personnel, coworkers, and other patients, the writing group says.

“Acute ischemic stroke is still a highly devastating disease and the Time Is Brain paradigm remains true during the COVID-19 pandemic as well,” said writing group chair Mayank Goyal, MD, of the University of Calgary (Alta.)

“We have highly effective and proven treatments available. As such, treatment delays due to additional screening requirements and personal protection equipment (PPE) should be kept at a minimum,” Dr. Goyal said.

“Practicing COVID-19 stroke work flows, through simulation training, can help to reduce treatment delays, minimize the risk of infectious exposure for patients and staff, and help alleviate stress,” he added.
 

A new layer of complexity

The guidance statement, Prehospital Triage of Acute Stroke Patients During the COVID-19 Pandemic, was published online May 13 in the journal Stroke.

“The need to limit infectious spread during the COVID-19 pandemic has added a new layer of complexity to prehospital stroke triage and transfer,” the writing group noted. “Timely and enhanced” communication between EMS, hospitals, and local coordinating authorities are critical, especially ambulance-and facility-based telestroke networks, they wrote.

The main factors to guide the triage decision are the likelihood of a large vessel occlusion; the magnitude of additional delays because of interhospital transfer and work flow efficiency at the primary stroke center or acute stroke ready hospital; the need for advanced critical care resources; and the available bed, staff, and PPE resources at the hospitals.

The group said it “seems reasonable” to lower the threshold to bypass hospitals that can’t provide acute stroke treatment in favor of transporting to a hospital that is “stroke ready,” particularly in patients likely to require advanced care. They cautioned, however, that taking all acute stroke patients to a comprehensive stroke center could overwhelm these centers and lead to clustering of COVID-19 patients.

They said it is equally important to ensure “necessary transfers” of stroke patients who would benefit from endovascular therapy or neurocritical care and avoid unnecessary patient transfers. “Doing so will likely require local hospital boards and health care authorities to collaborate and establish local guidelines and protocols,” the writing group said.

“During the COVID-19 pandemic, it is more important than ever to ensure that stroke patients are taken to the right hospital that can meet their urgent needs at the outset,” Dr. Goyal commented in an AHA news release.

The writing group emphasized that the principles put forth in the document are intended as suggestions rather than strict rules and will be adapted and updated to meet the evolving needs during the COVID-19 crisis and future pandemics.

“The process of improving stroke work flow and getting the correct patient to the correct hospital fast is dependent on training, protocols, simulation, technology, and – probably most importantly – teamwork. These principles are extremely important during the current pandemic but will be useful in improving stroke care afterwards as well,” Dr. Goyal said.

This research had no commercial funding. Members of the writing committee are on several AHA/ASA Council Science Subcommittees, including the Emergency Neurovascular Care, the Telestroke, and the Neurovascular Intervention committees. Goyal is a consultant for Medtronic, Stryker, Microvention, GE Healthcare, and Mentice. A complete list of author disclosures is available with the original article.

This article first appeared on Medscape.com.

The American Heart Association/American Stroke Association has developed a “conceptual framework” to assist emergency medical service (EMS) providers and in-hospital triage teams handle suspected cases of acute stroke during the ongoing COVID-19 crisis and future pandemics. A key goal is to ensure timely transfer of patients while minimizing the risk of infectious exposure for EMS personnel, coworkers, and other patients, the writing group says.

“Acute ischemic stroke is still a highly devastating disease and the Time Is Brain paradigm remains true during the COVID-19 pandemic as well,” said writing group chair Mayank Goyal, MD, of the University of Calgary (Alta.)

“We have highly effective and proven treatments available. As such, treatment delays due to additional screening requirements and personal protection equipment (PPE) should be kept at a minimum,” Dr. Goyal said.

“Practicing COVID-19 stroke work flows, through simulation training, can help to reduce treatment delays, minimize the risk of infectious exposure for patients and staff, and help alleviate stress,” he added.
 

A new layer of complexity

The guidance statement, Prehospital Triage of Acute Stroke Patients During the COVID-19 Pandemic, was published online May 13 in the journal Stroke.

“The need to limit infectious spread during the COVID-19 pandemic has added a new layer of complexity to prehospital stroke triage and transfer,” the writing group noted. “Timely and enhanced” communication between EMS, hospitals, and local coordinating authorities are critical, especially ambulance-and facility-based telestroke networks, they wrote.

The main factors to guide the triage decision are the likelihood of a large vessel occlusion; the magnitude of additional delays because of interhospital transfer and work flow efficiency at the primary stroke center or acute stroke ready hospital; the need for advanced critical care resources; and the available bed, staff, and PPE resources at the hospitals.

The group said it “seems reasonable” to lower the threshold to bypass hospitals that can’t provide acute stroke treatment in favor of transporting to a hospital that is “stroke ready,” particularly in patients likely to require advanced care. They cautioned, however, that taking all acute stroke patients to a comprehensive stroke center could overwhelm these centers and lead to clustering of COVID-19 patients.

They said it is equally important to ensure “necessary transfers” of stroke patients who would benefit from endovascular therapy or neurocritical care and avoid unnecessary patient transfers. “Doing so will likely require local hospital boards and health care authorities to collaborate and establish local guidelines and protocols,” the writing group said.

“During the COVID-19 pandemic, it is more important than ever to ensure that stroke patients are taken to the right hospital that can meet their urgent needs at the outset,” Dr. Goyal commented in an AHA news release.

The writing group emphasized that the principles put forth in the document are intended as suggestions rather than strict rules and will be adapted and updated to meet the evolving needs during the COVID-19 crisis and future pandemics.

“The process of improving stroke work flow and getting the correct patient to the correct hospital fast is dependent on training, protocols, simulation, technology, and – probably most importantly – teamwork. These principles are extremely important during the current pandemic but will be useful in improving stroke care afterwards as well,” Dr. Goyal said.

This research had no commercial funding. Members of the writing committee are on several AHA/ASA Council Science Subcommittees, including the Emergency Neurovascular Care, the Telestroke, and the Neurovascular Intervention committees. Goyal is a consultant for Medtronic, Stryker, Microvention, GE Healthcare, and Mentice. A complete list of author disclosures is available with the original article.

This article first appeared on Medscape.com.

From the University of Florida College of Medicine, Division of Infectious Diseases and Global Medicine, Gainesville, FL.

Abstract

• Objective: To review current reports on atypical manifestations of coronavirus disease 2019 (COVID-19).
• Methods: Review of the literature.
• Results: Evidence regarding atypical features of COVID-19 is accumulating. SARS-CoV-2 can infect human cells that express the angiotensin-converting enzyme 2 receptor, which would allow for a broad spectrum of illnesses affecting the renal, cardiac, and gastrointestinal organ systems. Neurologic, cutaneous, and musculoskeletal manifestations have also been reported. The potential for SARS-CoV-2 to induce a hypercoagulable state provides another avenue for the virus to indirectly damage various organ systems, as evidenced by reports of cerebrovascular disease, myocardial injury, and a chilblain-like rash in patients with COVID-19.
• Conclusion: Because the signs and symptoms of COVID-19 may occur with varying frequency across populations, it is important to keep differentials broad when assessing patients with a clinical illness that may indeed be COVID-19.

Keywords: coronavirus; severe acute respiratory syndrome coronavirus-2; SARS-CoV-2; pandemic.

Coronavirus disease 2019 (COVID-19), the syndrome caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), was first reported in Wuhan, China, in early December 2019.¹ Since then, the virus has spread quickly around the world, with the World Health Organization (WHO) declaring the coronavirus outbreak a global pandemic on March 11, 2020. As of May 21, 2020, more than 5,000,000 cases of COVID-19 have been confirmed, and more than 328,000 deaths related to COVID-19 have been reported globally.² These numbers are expected to increase, due to the reproduction number (R0) of SARS-CoV-2. R0 represents the number of new infections generated by an infectious person in a totally naïve population.³ The WHO estimates that the R0 of SARS-CoV-2 is 1.95, with other estimates ranging from 1.4 to 6.49.³ To control the pathogen, the R0 needs to be brought under a value of 1.

A fundamental tool in lowering the R0 is prompt testing and isolation of those who display signs and symptoms of infection. SARS-CoV-2 is still a novel pathogen about which we know relatively little. The common symptoms of COVID-19 are now well known—including fever, fatigue, anorexia, cough, and shortness of breath—but atypical manifestations of this viral continue to be reported and described. To help clinicians across specialties and settings identify patients with possible infection, we have summarized findings from current reports on COVID-19 manifestations involving the renal, cardiac, gastrointestinal (GI), and other organ systems.

Renal

During the 2003 SARS-CoV-1 outbreak, acute kidney injury (AKI) was an uncommon complication of the infection, but early reports suggest that AKI may occur more commonly with COVID-19.⁴ In a study of 193 patients with laboratory-confirmed COVID-19 treated in 3 Chinese hospitals, 59% presented with proteinuria, 44% with hematuria, 14% with increased blood urea nitrogen, and 10% with increased levels of serum creatinine.⁴ These markers, indicative of AKI, may be associated with increased mortality. Among this cohort, those with AKI had a mortality risk 5.3 times higher than those who did not have AKI.⁴ The pathophysiology of renal disease in COVID-19 may be related to dehydration or inflammatory mediators, causing decreased renal perfusion and cytokine storm, but evidence also suggests that SARS-CoV-2 is able to directly infect kidney cells.⁵ The virus infects cells by using angiotensin-converting enzyme 2 (ACE2) on the cell membrane as a cell entry receptor; ACE2 is expressed on the kidney, heart, and GI cells, and this may allow SARS-CoV-2 to directly infect and damage these organs. Other potential mechanisms of renal injury include overproduction of proinflammatory cytokines and administration of nephrotoxic drugs. No matter the mechanism, however, increased serum creatinine and blood urea nitrogen correlate with an increased likelihood of requiring intensive care unit (ICU) admission.⁶ Therefore, clinicians should carefully monitor renal function in patients with COVID-19.

Cardiac

In a report of 138 Chinese patients hospitalized for COVID-19, 36 required ICU admission: 44.4% of these had arrhythmias and 22.2% had developed acute cardiac injury.⁶ In addition, the cardiac cell injury biomarker troponin I was more likely to be elevated in ICU patients.⁶ A study of 21 patients admitted to the ICU in Washington State found elevated levels of brain natriuretic peptide.⁷ These biomarkers reflect the presence of myocardial stress, but do not necessarily indicate direct myocardial infection. Case reports of fulminant myocarditis in those with COVID-19 have begun to surface, however.^8,9 An examination of 68 deaths in persons with COVID-19 concluded that 7% were caused by myocarditis with circulatory failure.¹⁰

The pathophysiology of myocardial injury in COVID-19 is likely multifactorial. This includes increased inflammatory mediators, hypoxemia, and metabolic changes that can directly damage myocardial tissue. These factors can also exacerbate comorbid conditions, such as coronary artery disease, leading to ischemia and dysfunction of preexisting electrical conduction abnormalities. However, pathologic evidence of myocarditis and the presence of the ACE2 receptor, which may be a mediator of cardiac function, on cardiac muscle cells suggest that SARS-CoV-2 is capable of directly infecting and damaging myocardial cells. Other proposed mechanisms include infection-mediated downregulation of ACE2, causing cardiac dysfunction, or thrombus formation.¹¹ Although respiratory failure is the most common source of advanced illness in COVID-19 patients, myocarditis and arrhythmias can be life-threatening manifestations of the disease.

Gastrointestinal

As noted, ACE2 is expressed in the GI tract. In 73 patients hospitalized for COVID-19, 53.4% tested positive for SARS-CoV-2 RNA in stool, and 23.4% continued to have RNA-positive stool samples even after their respiratory samples tested negative.¹² These findings suggest the potential for SARS-CoV-2 to spread through fecal-oral transmission in those who are asymptomatic, pre-symptomatic, or symptomatic. This mode of transmission has yet to be determined conclusively, and more research is needed. However, GI symptoms have been reported in persons with COVID-19. Among 138 hospitalized patients, 10.1% had complaints of diarrhea and nausea and 3.6% reported vomiting.⁶ Those who reported nausea and diarrhea noted that they developed these symptoms 1 to 2 days before they developed fever.⁶ Also, among a cohort of 1099 Chinese patients with COVID-19, 3.8% complained of diarrhea.¹³ Although diarrhea does not occur in a majority of patients, GI complaints, such as nausea, vomiting, or diarrhea, should raise clinical suspicion for COVID-19, and in known areas of active transmission, testing of patients with GI symptoms is likely warranted.

Ocular

Ocular manifestations of COVID-19 are now being described, and should be taken into consideration when examining a patient. In a study of 38 patients with COVID-19 from Hubei province, China, 31.6% had ocular findings consistent with conjunctivitis, including conjunctival hyperemia, chemosis, epiphora, and increased ocular secretions.¹⁴ SARS-CoV-2 was detected in conjunctival and nasopharyngeal samples in 2 patients from this cohort. Conjunctival congestion was reported in a cohort of 1099 patients with COVID-19 treated at multiple centers throughout China, but at a much lower incidence, approximately 0.8%.¹³ Because SARS-CoV-2 can cause conjunctival disease and has been detected in samples from the external surface of the eye, it appears the virus is transmissible from tears or contact with the eye itself.

Neurologic

Common reported neurologic symptoms include dizziness, headache, impaired consciousness, ataxia, and cerebrovascular events. In a cohort of 214 patients from Wuhan, China, 36.4% had some form of neurological insult.¹⁵ These symptoms were more common in those with severe illness (P = 0.02).¹⁵ Two interesting neurologic symptoms that have been described are anosmia (loss of smell) and ageusia (loss of taste), which are being found primarily in tandem. It is still unclear how many people with COVID-19 are experiencing these symptoms, but a report from Italy estimates 19.4% of 320 patients examined had chemosensory dysfunction.¹⁶ The aforementioned report from Wuhan, China, found that 5.1% had anosmia and 5.6% had ageusia.¹⁵ The presence of anosmia/ageusia in some patients suggests that SARS-CoV-2 may enter the central nervous system (CNS) through a retrograde neuronal route.¹⁵ In addition, a case report from Japan described a 24-year-old man who presented with meningitis/encephalitis and had SARS-CoV-2 RNA present in his cerebrospinal fluid, showing that SARS-CoV-2 can penetrate into the CNS.¹⁷

SARS-CoV-2 may also have an association with Guillain–Barré syndrome, as this condition was reported in 5 patients from 3 hospitals in Northern Italy.¹⁸ The symptoms of Guillain–Barré syndrome presented 5 to 10 days after the typical COVID-19 symptoms, and evolved over 36 hours to 4 days afterwards. Four of the 5 patients experienced flaccid tetraparesis or tetraplegia, and 3 required mechanical ventilation.¹⁸

Another possible cause of neurologic injury in COVID-19 is damage to endothelial cells in cerebral blood vessels, causing thrombus formation and possibly increasing the risk of acute ischemic stroke.^15,19 Supporting this mechanism of injury, significantly lower platelet counts were noted in patients with CNS symptoms (P = 0.005).¹⁵ Other hematological impacts of COVID-19 have been reported, particularly hypercoagulability, as evidenced by elevated D-dimer levels.^13,20 This hypercoagulable state is linked to overproduction of proinflammatory cytokines (cytokine storm), leading to dysregulation of coagulation pathways and reduced concentrations of anticoagulants, such as protein C, antithrombin III, and tissue factor pathway inhibitor.²¹

Cutaneous

Cutaneous findings emerging in persons with COVID-19 demonstrate features of small-vessel and capillary occlusion, including erythematous skin eruptions and petechial rash. One report from Italy noted that 20.4% of patients with COVID-19 (n = 88) had a cutaneous finding, with a cutaneous manifestation developing in 8 at the onset of illness and in 10 following hospital admission.²² Fourteen patients had an erythematous rash, primarily on the trunk, with 3 patients having a diffuse urticarial appearing rash, and 1 patient developing vesicles.²² The severity of illness did not appear to correlate with the cutaneous manifestation, and the lesions healed within a few days.

One case report described a patient from Bangkok who was thought to be suffering from dengue fever, but was found to have SARS-CoV-2 infection. He initially presented with skin rash and petechiae, and later developed respiratory disease.²³

Other dermatologic findings of COVID-19 resemble chilblains disease, colloquially referred to as “COVID toes.” Two women, 27 and 35 years old, presented to a dermatology clinic in Qatar with a chief complaint of skin rash, described as red-purple papules on the dorsal aspects of the fingers bilaterally.²² Both patients had an unremarkable medical and drug history, but recent travel to the United Kingdom dictated SARS-CoV-2 screening, which was positive.²⁴ An Italian case report describes a 23-year-old man who tested positive for SARS-CoV-2 and had violaceous plaques on an erythematous background on his feet, without any lesions on his hands.²⁵ Since chilblains is less common in the warmer months and these events correspond with the COVID-19 pandemic, SARS-CoV-2 infection is the suspected etiology. The pathophysiology of these lesions is unclear, and more research is needed. As more data become available, we may see cutaneous manifestations in patients with COVID-19 similar to those commonly reported with other viral infectious processes.

Musculoskeletal

Of 138 patients hospitalized in Wuhan, China, for COVID-19, 34.8% presented with myalgia; the presence of myalgia does not appear to be correlated with an increased likelihood of ICU admission.⁶ Myalgia or arthralgia was also reported in 14.9% among the cohort of 1099 COVID-19 patients in China.¹³ These musculoskeletal symptoms are described among large muscle groups found in the extremities, trunk, and back, and should raise suspicion in patients who present with other signs and symptoms concerning for COVID-19.

Evidence regarding atypical features of COVID-19 is accumulating. SARS-CoV-2 can infect a human cells that express the ACE2 receptor, which would allow for a broad spectrum of illnesses. The potential for SARS-CoV-2 to induce a hypercoagulable state allows it to indirectly damage various organ systems,²⁰ leading to cerebrovascular disease, myocardial injury, and a chilblain-like rash. Clinicians must be aware of these unique features, as early recognition of persons who present with COVID-19 will allow for prompt testing, institution of infection control and isolation practices, and treatment, as needed, among those infected. Also, this is a pandemic involving a novel virus affecting different populations throughout the world, and these signs and symptoms may occur with varying frequency across populations. Therefore, it is important to keep differentials broad when assessing patients with a clinical illness that may indeed be COVID-19.

Corresponding author: Norman L. Beatty, MD, norman.beatty@medicine.ufl.edu.

Financial disclosures: None.

From the University of Florida College of Medicine, Division of Infectious Diseases and Global Medicine, Gainesville, FL.

Abstract

• Objective: To review current reports on atypical manifestations of coronavirus disease 2019 (COVID-19).
• Methods: Review of the literature.
• Results: Evidence regarding atypical features of COVID-19 is accumulating. SARS-CoV-2 can infect human cells that express the angiotensin-converting enzyme 2 receptor, which would allow for a broad spectrum of illnesses affecting the renal, cardiac, and gastrointestinal organ systems. Neurologic, cutaneous, and musculoskeletal manifestations have also been reported. The potential for SARS-CoV-2 to induce a hypercoagulable state provides another avenue for the virus to indirectly damage various organ systems, as evidenced by reports of cerebrovascular disease, myocardial injury, and a chilblain-like rash in patients with COVID-19.
• Conclusion: Because the signs and symptoms of COVID-19 may occur with varying frequency across populations, it is important to keep differentials broad when assessing patients with a clinical illness that may indeed be COVID-19.

Keywords: coronavirus; severe acute respiratory syndrome coronavirus-2; SARS-CoV-2; pandemic.

Coronavirus disease 2019 (COVID-19), the syndrome caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), was first reported in Wuhan, China, in early December 2019.¹ Since then, the virus has spread quickly around the world, with the World Health Organization (WHO) declaring the coronavirus outbreak a global pandemic on March 11, 2020. As of May 21, 2020, more than 5,000,000 cases of COVID-19 have been confirmed, and more than 328,000 deaths related to COVID-19 have been reported globally.² These numbers are expected to increase, due to the reproduction number (R0) of SARS-CoV-2. R0 represents the number of new infections generated by an infectious person in a totally naïve population.³ The WHO estimates that the R0 of SARS-CoV-2 is 1.95, with other estimates ranging from 1.4 to 6.49.³ To control the pathogen, the R0 needs to be brought under a value of 1.

A fundamental tool in lowering the R0 is prompt testing and isolation of those who display signs and symptoms of infection. SARS-CoV-2 is still a novel pathogen about which we know relatively little. The common symptoms of COVID-19 are now well known—including fever, fatigue, anorexia, cough, and shortness of breath—but atypical manifestations of this viral continue to be reported and described. To help clinicians across specialties and settings identify patients with possible infection, we have summarized findings from current reports on COVID-19 manifestations involving the renal, cardiac, gastrointestinal (GI), and other organ systems.

Renal

During the 2003 SARS-CoV-1 outbreak, acute kidney injury (AKI) was an uncommon complication of the infection, but early reports suggest that AKI may occur more commonly with COVID-19.⁴ In a study of 193 patients with laboratory-confirmed COVID-19 treated in 3 Chinese hospitals, 59% presented with proteinuria, 44% with hematuria, 14% with increased blood urea nitrogen, and 10% with increased levels of serum creatinine.⁴ These markers, indicative of AKI, may be associated with increased mortality. Among this cohort, those with AKI had a mortality risk 5.3 times higher than those who did not have AKI.⁴ The pathophysiology of renal disease in COVID-19 may be related to dehydration or inflammatory mediators, causing decreased renal perfusion and cytokine storm, but evidence also suggests that SARS-CoV-2 is able to directly infect kidney cells.⁵ The virus infects cells by using angiotensin-converting enzyme 2 (ACE2) on the cell membrane as a cell entry receptor; ACE2 is expressed on the kidney, heart, and GI cells, and this may allow SARS-CoV-2 to directly infect and damage these organs. Other potential mechanisms of renal injury include overproduction of proinflammatory cytokines and administration of nephrotoxic drugs. No matter the mechanism, however, increased serum creatinine and blood urea nitrogen correlate with an increased likelihood of requiring intensive care unit (ICU) admission.⁶ Therefore, clinicians should carefully monitor renal function in patients with COVID-19.

Cardiac

In a report of 138 Chinese patients hospitalized for COVID-19, 36 required ICU admission: 44.4% of these had arrhythmias and 22.2% had developed acute cardiac injury.⁶ In addition, the cardiac cell injury biomarker troponin I was more likely to be elevated in ICU patients.⁶ A study of 21 patients admitted to the ICU in Washington State found elevated levels of brain natriuretic peptide.⁷ These biomarkers reflect the presence of myocardial stress, but do not necessarily indicate direct myocardial infection. Case reports of fulminant myocarditis in those with COVID-19 have begun to surface, however.^8,9 An examination of 68 deaths in persons with COVID-19 concluded that 7% were caused by myocarditis with circulatory failure.¹⁰

The pathophysiology of myocardial injury in COVID-19 is likely multifactorial. This includes increased inflammatory mediators, hypoxemia, and metabolic changes that can directly damage myocardial tissue. These factors can also exacerbate comorbid conditions, such as coronary artery disease, leading to ischemia and dysfunction of preexisting electrical conduction abnormalities. However, pathologic evidence of myocarditis and the presence of the ACE2 receptor, which may be a mediator of cardiac function, on cardiac muscle cells suggest that SARS-CoV-2 is capable of directly infecting and damaging myocardial cells. Other proposed mechanisms include infection-mediated downregulation of ACE2, causing cardiac dysfunction, or thrombus formation.¹¹ Although respiratory failure is the most common source of advanced illness in COVID-19 patients, myocarditis and arrhythmias can be life-threatening manifestations of the disease.

Gastrointestinal

As noted, ACE2 is expressed in the GI tract. In 73 patients hospitalized for COVID-19, 53.4% tested positive for SARS-CoV-2 RNA in stool, and 23.4% continued to have RNA-positive stool samples even after their respiratory samples tested negative.¹² These findings suggest the potential for SARS-CoV-2 to spread through fecal-oral transmission in those who are asymptomatic, pre-symptomatic, or symptomatic. This mode of transmission has yet to be determined conclusively, and more research is needed. However, GI symptoms have been reported in persons with COVID-19. Among 138 hospitalized patients, 10.1% had complaints of diarrhea and nausea and 3.6% reported vomiting.⁶ Those who reported nausea and diarrhea noted that they developed these symptoms 1 to 2 days before they developed fever.⁶ Also, among a cohort of 1099 Chinese patients with COVID-19, 3.8% complained of diarrhea.¹³ Although diarrhea does not occur in a majority of patients, GI complaints, such as nausea, vomiting, or diarrhea, should raise clinical suspicion for COVID-19, and in known areas of active transmission, testing of patients with GI symptoms is likely warranted.

Ocular

Ocular manifestations of COVID-19 are now being described, and should be taken into consideration when examining a patient. In a study of 38 patients with COVID-19 from Hubei province, China, 31.6% had ocular findings consistent with conjunctivitis, including conjunctival hyperemia, chemosis, epiphora, and increased ocular secretions.¹⁴ SARS-CoV-2 was detected in conjunctival and nasopharyngeal samples in 2 patients from this cohort. Conjunctival congestion was reported in a cohort of 1099 patients with COVID-19 treated at multiple centers throughout China, but at a much lower incidence, approximately 0.8%.¹³ Because SARS-CoV-2 can cause conjunctival disease and has been detected in samples from the external surface of the eye, it appears the virus is transmissible from tears or contact with the eye itself.

Neurologic

Common reported neurologic symptoms include dizziness, headache, impaired consciousness, ataxia, and cerebrovascular events. In a cohort of 214 patients from Wuhan, China, 36.4% had some form of neurological insult.¹⁵ These symptoms were more common in those with severe illness (P = 0.02).¹⁵ Two interesting neurologic symptoms that have been described are anosmia (loss of smell) and ageusia (loss of taste), which are being found primarily in tandem. It is still unclear how many people with COVID-19 are experiencing these symptoms, but a report from Italy estimates 19.4% of 320 patients examined had chemosensory dysfunction.¹⁶ The aforementioned report from Wuhan, China, found that 5.1% had anosmia and 5.6% had ageusia.¹⁵ The presence of anosmia/ageusia in some patients suggests that SARS-CoV-2 may enter the central nervous system (CNS) through a retrograde neuronal route.¹⁵ In addition, a case report from Japan described a 24-year-old man who presented with meningitis/encephalitis and had SARS-CoV-2 RNA present in his cerebrospinal fluid, showing that SARS-CoV-2 can penetrate into the CNS.¹⁷

SARS-CoV-2 may also have an association with Guillain–Barré syndrome, as this condition was reported in 5 patients from 3 hospitals in Northern Italy.¹⁸ The symptoms of Guillain–Barré syndrome presented 5 to 10 days after the typical COVID-19 symptoms, and evolved over 36 hours to 4 days afterwards. Four of the 5 patients experienced flaccid tetraparesis or tetraplegia, and 3 required mechanical ventilation.¹⁸

Another possible cause of neurologic injury in COVID-19 is damage to endothelial cells in cerebral blood vessels, causing thrombus formation and possibly increasing the risk of acute ischemic stroke.^15,19 Supporting this mechanism of injury, significantly lower platelet counts were noted in patients with CNS symptoms (P = 0.005).¹⁵ Other hematological impacts of COVID-19 have been reported, particularly hypercoagulability, as evidenced by elevated D-dimer levels.^13,20 This hypercoagulable state is linked to overproduction of proinflammatory cytokines (cytokine storm), leading to dysregulation of coagulation pathways and reduced concentrations of anticoagulants, such as protein C, antithrombin III, and tissue factor pathway inhibitor.²¹

Cutaneous

Cutaneous findings emerging in persons with COVID-19 demonstrate features of small-vessel and capillary occlusion, including erythematous skin eruptions and petechial rash. One report from Italy noted that 20.4% of patients with COVID-19 (n = 88) had a cutaneous finding, with a cutaneous manifestation developing in 8 at the onset of illness and in 10 following hospital admission.²² Fourteen patients had an erythematous rash, primarily on the trunk, with 3 patients having a diffuse urticarial appearing rash, and 1 patient developing vesicles.²² The severity of illness did not appear to correlate with the cutaneous manifestation, and the lesions healed within a few days.

One case report described a patient from Bangkok who was thought to be suffering from dengue fever, but was found to have SARS-CoV-2 infection. He initially presented with skin rash and petechiae, and later developed respiratory disease.²³

Other dermatologic findings of COVID-19 resemble chilblains disease, colloquially referred to as “COVID toes.” Two women, 27 and 35 years old, presented to a dermatology clinic in Qatar with a chief complaint of skin rash, described as red-purple papules on the dorsal aspects of the fingers bilaterally.²² Both patients had an unremarkable medical and drug history, but recent travel to the United Kingdom dictated SARS-CoV-2 screening, which was positive.²⁴ An Italian case report describes a 23-year-old man who tested positive for SARS-CoV-2 and had violaceous plaques on an erythematous background on his feet, without any lesions on his hands.²⁵ Since chilblains is less common in the warmer months and these events correspond with the COVID-19 pandemic, SARS-CoV-2 infection is the suspected etiology. The pathophysiology of these lesions is unclear, and more research is needed. As more data become available, we may see cutaneous manifestations in patients with COVID-19 similar to those commonly reported with other viral infectious processes.

Musculoskeletal

Of 138 patients hospitalized in Wuhan, China, for COVID-19, 34.8% presented with myalgia; the presence of myalgia does not appear to be correlated with an increased likelihood of ICU admission.⁶ Myalgia or arthralgia was also reported in 14.9% among the cohort of 1099 COVID-19 patients in China.¹³ These musculoskeletal symptoms are described among large muscle groups found in the extremities, trunk, and back, and should raise suspicion in patients who present with other signs and symptoms concerning for COVID-19.

Evidence regarding atypical features of COVID-19 is accumulating. SARS-CoV-2 can infect a human cells that express the ACE2 receptor, which would allow for a broad spectrum of illnesses. The potential for SARS-CoV-2 to induce a hypercoagulable state allows it to indirectly damage various organ systems,²⁰ leading to cerebrovascular disease, myocardial injury, and a chilblain-like rash. Clinicians must be aware of these unique features, as early recognition of persons who present with COVID-19 will allow for prompt testing, institution of infection control and isolation practices, and treatment, as needed, among those infected. Also, this is a pandemic involving a novel virus affecting different populations throughout the world, and these signs and symptoms may occur with varying frequency across populations. Therefore, it is important to keep differentials broad when assessing patients with a clinical illness that may indeed be COVID-19.

Corresponding author: Norman L. Beatty, MD, norman.beatty@medicine.ufl.edu.

Financial disclosures: None.

From the University of Florida College of Medicine, Division of Infectious Diseases and Global Medicine, Gainesville, FL.

Abstract

• Objective: To review current reports on atypical manifestations of coronavirus disease 2019 (COVID-19).
• Methods: Review of the literature.
• Results: Evidence regarding atypical features of COVID-19 is accumulating. SARS-CoV-2 can infect human cells that express the angiotensin-converting enzyme 2 receptor, which would allow for a broad spectrum of illnesses affecting the renal, cardiac, and gastrointestinal organ systems. Neurologic, cutaneous, and musculoskeletal manifestations have also been reported. The potential for SARS-CoV-2 to induce a hypercoagulable state provides another avenue for the virus to indirectly damage various organ systems, as evidenced by reports of cerebrovascular disease, myocardial injury, and a chilblain-like rash in patients with COVID-19.
• Conclusion: Because the signs and symptoms of COVID-19 may occur with varying frequency across populations, it is important to keep differentials broad when assessing patients with a clinical illness that may indeed be COVID-19.

Keywords: coronavirus; severe acute respiratory syndrome coronavirus-2; SARS-CoV-2; pandemic.

Coronavirus disease 2019 (COVID-19), the syndrome caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), was first reported in Wuhan, China, in early December 2019.¹ Since then, the virus has spread quickly around the world, with the World Health Organization (WHO) declaring the coronavirus outbreak a global pandemic on March 11, 2020. As of May 21, 2020, more than 5,000,000 cases of COVID-19 have been confirmed, and more than 328,000 deaths related to COVID-19 have been reported globally.² These numbers are expected to increase, due to the reproduction number (R0) of SARS-CoV-2. R0 represents the number of new infections generated by an infectious person in a totally naïve population.³ The WHO estimates that the R0 of SARS-CoV-2 is 1.95, with other estimates ranging from 1.4 to 6.49.³ To control the pathogen, the R0 needs to be brought under a value of 1.

A fundamental tool in lowering the R0 is prompt testing and isolation of those who display signs and symptoms of infection. SARS-CoV-2 is still a novel pathogen about which we know relatively little. The common symptoms of COVID-19 are now well known—including fever, fatigue, anorexia, cough, and shortness of breath—but atypical manifestations of this viral continue to be reported and described. To help clinicians across specialties and settings identify patients with possible infection, we have summarized findings from current reports on COVID-19 manifestations involving the renal, cardiac, gastrointestinal (GI), and other organ systems.

Renal

During the 2003 SARS-CoV-1 outbreak, acute kidney injury (AKI) was an uncommon complication of the infection, but early reports suggest that AKI may occur more commonly with COVID-19.⁴ In a study of 193 patients with laboratory-confirmed COVID-19 treated in 3 Chinese hospitals, 59% presented with proteinuria, 44% with hematuria, 14% with increased blood urea nitrogen, and 10% with increased levels of serum creatinine.⁴ These markers, indicative of AKI, may be associated with increased mortality. Among this cohort, those with AKI had a mortality risk 5.3 times higher than those who did not have AKI.⁴ The pathophysiology of renal disease in COVID-19 may be related to dehydration or inflammatory mediators, causing decreased renal perfusion and cytokine storm, but evidence also suggests that SARS-CoV-2 is able to directly infect kidney cells.⁵ The virus infects cells by using angiotensin-converting enzyme 2 (ACE2) on the cell membrane as a cell entry receptor; ACE2 is expressed on the kidney, heart, and GI cells, and this may allow SARS-CoV-2 to directly infect and damage these organs. Other potential mechanisms of renal injury include overproduction of proinflammatory cytokines and administration of nephrotoxic drugs. No matter the mechanism, however, increased serum creatinine and blood urea nitrogen correlate with an increased likelihood of requiring intensive care unit (ICU) admission.⁶ Therefore, clinicians should carefully monitor renal function in patients with COVID-19.

Cardiac

In a report of 138 Chinese patients hospitalized for COVID-19, 36 required ICU admission: 44.4% of these had arrhythmias and 22.2% had developed acute cardiac injury.⁶ In addition, the cardiac cell injury biomarker troponin I was more likely to be elevated in ICU patients.⁶ A study of 21 patients admitted to the ICU in Washington State found elevated levels of brain natriuretic peptide.⁷ These biomarkers reflect the presence of myocardial stress, but do not necessarily indicate direct myocardial infection. Case reports of fulminant myocarditis in those with COVID-19 have begun to surface, however.^8,9 An examination of 68 deaths in persons with COVID-19 concluded that 7% were caused by myocarditis with circulatory failure.¹⁰

The pathophysiology of myocardial injury in COVID-19 is likely multifactorial. This includes increased inflammatory mediators, hypoxemia, and metabolic changes that can directly damage myocardial tissue. These factors can also exacerbate comorbid conditions, such as coronary artery disease, leading to ischemia and dysfunction of preexisting electrical conduction abnormalities. However, pathologic evidence of myocarditis and the presence of the ACE2 receptor, which may be a mediator of cardiac function, on cardiac muscle cells suggest that SARS-CoV-2 is capable of directly infecting and damaging myocardial cells. Other proposed mechanisms include infection-mediated downregulation of ACE2, causing cardiac dysfunction, or thrombus formation.¹¹ Although respiratory failure is the most common source of advanced illness in COVID-19 patients, myocarditis and arrhythmias can be life-threatening manifestations of the disease.

Gastrointestinal

As noted, ACE2 is expressed in the GI tract. In 73 patients hospitalized for COVID-19, 53.4% tested positive for SARS-CoV-2 RNA in stool, and 23.4% continued to have RNA-positive stool samples even after their respiratory samples tested negative.¹² These findings suggest the potential for SARS-CoV-2 to spread through fecal-oral transmission in those who are asymptomatic, pre-symptomatic, or symptomatic. This mode of transmission has yet to be determined conclusively, and more research is needed. However, GI symptoms have been reported in persons with COVID-19. Among 138 hospitalized patients, 10.1% had complaints of diarrhea and nausea and 3.6% reported vomiting.⁶ Those who reported nausea and diarrhea noted that they developed these symptoms 1 to 2 days before they developed fever.⁶ Also, among a cohort of 1099 Chinese patients with COVID-19, 3.8% complained of diarrhea.¹³ Although diarrhea does not occur in a majority of patients, GI complaints, such as nausea, vomiting, or diarrhea, should raise clinical suspicion for COVID-19, and in known areas of active transmission, testing of patients with GI symptoms is likely warranted.

Ocular

Ocular manifestations of COVID-19 are now being described, and should be taken into consideration when examining a patient. In a study of 38 patients with COVID-19 from Hubei province, China, 31.6% had ocular findings consistent with conjunctivitis, including conjunctival hyperemia, chemosis, epiphora, and increased ocular secretions.¹⁴ SARS-CoV-2 was detected in conjunctival and nasopharyngeal samples in 2 patients from this cohort. Conjunctival congestion was reported in a cohort of 1099 patients with COVID-19 treated at multiple centers throughout China, but at a much lower incidence, approximately 0.8%.¹³ Because SARS-CoV-2 can cause conjunctival disease and has been detected in samples from the external surface of the eye, it appears the virus is transmissible from tears or contact with the eye itself.

Neurologic

Common reported neurologic symptoms include dizziness, headache, impaired consciousness, ataxia, and cerebrovascular events. In a cohort of 214 patients from Wuhan, China, 36.4% had some form of neurological insult.¹⁵ These symptoms were more common in those with severe illness (P = 0.02).¹⁵ Two interesting neurologic symptoms that have been described are anosmia (loss of smell) and ageusia (loss of taste), which are being found primarily in tandem. It is still unclear how many people with COVID-19 are experiencing these symptoms, but a report from Italy estimates 19.4% of 320 patients examined had chemosensory dysfunction.¹⁶ The aforementioned report from Wuhan, China, found that 5.1% had anosmia and 5.6% had ageusia.¹⁵ The presence of anosmia/ageusia in some patients suggests that SARS-CoV-2 may enter the central nervous system (CNS) through a retrograde neuronal route.¹⁵ In addition, a case report from Japan described a 24-year-old man who presented with meningitis/encephalitis and had SARS-CoV-2 RNA present in his cerebrospinal fluid, showing that SARS-CoV-2 can penetrate into the CNS.¹⁷

SARS-CoV-2 may also have an association with Guillain–Barré syndrome, as this condition was reported in 5 patients from 3 hospitals in Northern Italy.¹⁸ The symptoms of Guillain–Barré syndrome presented 5 to 10 days after the typical COVID-19 symptoms, and evolved over 36 hours to 4 days afterwards. Four of the 5 patients experienced flaccid tetraparesis or tetraplegia, and 3 required mechanical ventilation.¹⁸

Another possible cause of neurologic injury in COVID-19 is damage to endothelial cells in cerebral blood vessels, causing thrombus formation and possibly increasing the risk of acute ischemic stroke.^15,19 Supporting this mechanism of injury, significantly lower platelet counts were noted in patients with CNS symptoms (P = 0.005).¹⁵ Other hematological impacts of COVID-19 have been reported, particularly hypercoagulability, as evidenced by elevated D-dimer levels.^13,20 This hypercoagulable state is linked to overproduction of proinflammatory cytokines (cytokine storm), leading to dysregulation of coagulation pathways and reduced concentrations of anticoagulants, such as protein C, antithrombin III, and tissue factor pathway inhibitor.²¹

Cutaneous

Cutaneous findings emerging in persons with COVID-19 demonstrate features of small-vessel and capillary occlusion, including erythematous skin eruptions and petechial rash. One report from Italy noted that 20.4% of patients with COVID-19 (n = 88) had a cutaneous finding, with a cutaneous manifestation developing in 8 at the onset of illness and in 10 following hospital admission.²² Fourteen patients had an erythematous rash, primarily on the trunk, with 3 patients having a diffuse urticarial appearing rash, and 1 patient developing vesicles.²² The severity of illness did not appear to correlate with the cutaneous manifestation, and the lesions healed within a few days.

One case report described a patient from Bangkok who was thought to be suffering from dengue fever, but was found to have SARS-CoV-2 infection. He initially presented with skin rash and petechiae, and later developed respiratory disease.²³

Other dermatologic findings of COVID-19 resemble chilblains disease, colloquially referred to as “COVID toes.” Two women, 27 and 35 years old, presented to a dermatology clinic in Qatar with a chief complaint of skin rash, described as red-purple papules on the dorsal aspects of the fingers bilaterally.²² Both patients had an unremarkable medical and drug history, but recent travel to the United Kingdom dictated SARS-CoV-2 screening, which was positive.²⁴ An Italian case report describes a 23-year-old man who tested positive for SARS-CoV-2 and had violaceous plaques on an erythematous background on his feet, without any lesions on his hands.²⁵ Since chilblains is less common in the warmer months and these events correspond with the COVID-19 pandemic, SARS-CoV-2 infection is the suspected etiology. The pathophysiology of these lesions is unclear, and more research is needed. As more data become available, we may see cutaneous manifestations in patients with COVID-19 similar to those commonly reported with other viral infectious processes.

Musculoskeletal

Of 138 patients hospitalized in Wuhan, China, for COVID-19, 34.8% presented with myalgia; the presence of myalgia does not appear to be correlated with an increased likelihood of ICU admission.⁶ Myalgia or arthralgia was also reported in 14.9% among the cohort of 1099 COVID-19 patients in China.¹³ These musculoskeletal symptoms are described among large muscle groups found in the extremities, trunk, and back, and should raise suspicion in patients who present with other signs and symptoms concerning for COVID-19.

Evidence regarding atypical features of COVID-19 is accumulating. SARS-CoV-2 can infect a human cells that express the ACE2 receptor, which would allow for a broad spectrum of illnesses. The potential for SARS-CoV-2 to induce a hypercoagulable state allows it to indirectly damage various organ systems,²⁰ leading to cerebrovascular disease, myocardial injury, and a chilblain-like rash. Clinicians must be aware of these unique features, as early recognition of persons who present with COVID-19 will allow for prompt testing, institution of infection control and isolation practices, and treatment, as needed, among those infected. Also, this is a pandemic involving a novel virus affecting different populations throughout the world, and these signs and symptoms may occur with varying frequency across populations. Therefore, it is important to keep differentials broad when assessing patients with a clinical illness that may indeed be COVID-19.

Corresponding author: Norman L. Beatty, MD, norman.beatty@medicine.ufl.edu.

Financial disclosures: None.

Study Overview

Objective. To assess the efficacy, safety, and clinical benefit of remdesivir in hospitalized adults with confirmed pneumonia due to severe SARS-CoV-2 infection.

Design. Randomized, investigator-initiated, placebo-controlled, double-blind, multicenter trial.

Setting and participants. The trial took place between February 6, 2020 and March 12, 2020, at 10 hospitals in Wuhan, China. Study participants included adult patients (aged ≥ 18 years) admitted to hospital who tested positive for SARS-CoV-2 by reverse transcription polymerase chain reaction assay and had the following clinical characteristics: radiographic evidence of pneumonia; hypoxia with oxygen saturation ≤ 94% on room air or a ratio of arterial oxygen partial pressure to fractional inspired oxygen ≤ 300 mm Hg; and symptom onset to enrollment ≤ 12 days. Some of the exclusion criteria for participation in the study were pregnancy or breast feeding, liver cirrhosis, abnormal liver enzymes ≥ 5 times the upper limit of normal, severe renal impairment or receipt of renal replacement therapy, plan for transfer to a non-study hospital, and enrollment in a trial for COVID-19 within the previous month.

Intervention. Participants were randomized in a 2:1 ratio to the remdesivir group or the placebo group and were administered either intravenous infusions of remdesivir (200 mg on day 1 followed by 100 mg daily on days 2-10) or the same volume of placebo for 10 days. Clinical and safety data assessed included laboratory testing, electrocardiogram, and medication adverse effects. Testing of oropharyngeal and nasopharyngeal swab samples, anal swab samples, sputum, and stool was performed for viral RNA detection and quantification on days 1, 3, 5, 7, 10, 14, 21, and 28.

Main outcome measures. The primary endpoint of this study was time to clinical improvement within 28 days after randomization. Clinical improvement was defined as a 2-point reduction in participants’ admission status on a 6-point ordinal scale (1 = discharged or clinical recovery, 6 = death) or live discharge from hospital, whichever came first. Secondary outcomes included all-cause mortality at day 28 and duration of hospital admission, oxygen support, and invasive mechanical ventilation. Virological measures and safety outcomes ascertained included treatment-emergent adverse events, serious adverse events, and premature discontinuation of remdesivir.

The sample size estimate for the original study design was a total of 453 patients (302 in the remdesivir group and 151 in the placebo group). This sample size would provide 80% power, assuming a hazard ratio (HR) of 1.4 comparing remdesivir to placebo, and corresponding to a change in time to clinical improvement of 6 days. The analysis of primary outcome was performed on an intention-to-treat basis. Time to clinical improvement within 28 days was assessed with Kaplan-Meier plots.

Main results. A total of 255 patients were screened, of whom 237 were enrolled and randomized to remdesivir (158) or placebo (79) group. Of the participants in the remdesivir group, 155 started study treatment and 150 completed treatment per protocol. For the participants in the placebo group, 78 started study treatment and 76 completed treatment per-protocol. Study enrollment was terminated after March 12, 2020, before attaining the prespecified sample size, because no additional patients met study eligibility criteria due to various public health measures implemented in Wuhan. The median age of participants was 65 years (IQR, 56-71), the majority were men (56% in remdesivir group vs 65% in placebo group), and the most common comorbidities included hypertension, diabetes, and coronary artery disease. Median time from symptom onset to study enrollment was 10 days (IQR, 9-12). The time to clinical improvement between treatments (21 days for remdesivir group vs 23 days for placebo group) was not significantly different (HR, 1.23; 95% confidence interval [CI], 0.87-1.75). In addition, in participants who received treatment within 10 days of symptom onset, those who were administered remdesivir had a nonsignificant (HR, 1.52; 95% CI, 0.95-2.43) but faster time (18 days) to clinical improvement, compared to those administered placebo (23 days). Moreover, treatment with remdesivir versus placebo did not lead to differences in secondary outcomes (eg, 28-day mortality and duration of hospital stay, oxygen support, and invasive mechanical ventilation), changes in viral load over time, or adverse events between the groups.

Conclusion. This study found that, compared with placebo, intravenous remdesivir did not significantly improve the time to clinical improvement, mortality, or time to clearance of SARS-CoV-2 in hospitalized adults with severe COVID-19. A numeric reduction in time to clinical improvement with early remdesivir treatment (ie, within 10 days of symptom onset) that approached statistical significance was observed in this underpowered study.

Commentary

Within a few short months since its emergence. SARS-CoV-2 infection has caused a global pandemic, posing a dire threat to public health due to its adverse effects on morbidity (eg, respiratory failure, thromboembolic diseases, multiorgan failure) and mortality. To date, no pharmacologic treatment has been shown to effectively improve clinical outcomes in patients with COVID-19. Multiple ongoing clinical trials are being conducted globally to determine potential therapeutic treatments for severe COVID-19. The first clinical trials of hydroxychloroquine and lopinavir-ritonavir, agents traditionally used for other indications, such as malaria and HIV, did not show a clear benefit in COVID-19.^1,2 Remdesivir, a nucleoside analogue prodrug, is a broad-spectrum antiviral agent that was previously used for treatment of Ebola and has been shown to have inhibitory effects on pathogenic coronaviruses. The study reported by Wang and colleagues was the first randomized controlled trial (RCT) aimed at evaluating whether remdesivir improves outcomes in patients with severe COVID-19. Thus, the worsening COVID-19 pandemic, coupled with the absence of a curative treatment, underscore the urgency of this trial.

The study was grounded on observational data from several recent case reports and case series centering on the potential efficacy of remdesivir in treating COVID-19.³ The study itself was designed well (ie, randomized, placebo-controlled, double-blind, multicenter) and carefully implemented (ie, high protocol adherence to treatments, no loss to follow-up). The principal limitation of this study was its inability to reach the estimated statistical power of study. Due to successful epidemic control in Wuhan, which led to marked reductions in hospital admission of patients with COVID-19, and implementation of stringent termination criteria per the study protocol, only 237 participants were enrolled, instead of the 453, as specified by the sample estimate. This corresponded to a reduction of statistical power from 80% to 58%. Due to this limitation, the study was underpowered, rendering its findings inconclusive.

Despite this limitation, the study found that those treated with remdesivir within 10 days of symptom onset had a numerically faster time (although not statistically significant) to clinical improvement. This leads to an interesting question: whether remdesivir administration early in COVID-19 course could improve clinical outcomes, a question that warrants further investigation by an adequately powered trial. Also, data from this study provided evidence that intravenous remdesivir administration is likely safe in adults during the treatment period, although the long-term drug effects, as well as the safety profile in pediatric patients, remain unknown at this time.

While the study reported by Wang and colleagues was underpowered and is thus inconclusive, several other ongoing RCTs are evaluating the potential clinical benefit of remdesivir treatment in patients hospitalized with COVID-19. On the date of online publication of this report in The Lancet, the National Institutes of Health (NIH) published a news release summarizing preliminary findings from the Adaptive COVID-19 Treatment Trial (ACTT), which showed positive effects of remdesivir on clinical recovery from advanced COVID-19.⁴ The ACTT, the first RCT launched in the United States to evaluate experimental treatment for COVID-19, included 1063 hospitalized participants with advanced COVID-19 and lung involvement. Participants who were administered remdesivir had a 31% faster time to recovery compared to those in the placebo group (median time to recovery, 11 days vs 15 days, respectively; P < 0.001), and had near statistically significant improved survival (mortality rate, 8.0% vs 11.6%, respectively; P = 0.059). In response to these findings, the US Food and Drug Administration (FDA) issued an emergency use authorization for remdesivir on May 1, 2020, for the treatment of suspected or laboratory-confirmed COVID-19 in adults and children hospitalized with severe disease.⁵ While the findings noted from the NIH news release are very encouraging and provide the first evidence of a potentially beneficial antiviral treatment for severe COVID-19 in humans, the scientific community awaits the peer-reviewed publication of the ACTT to better assess the safety and effectiveness of remdesivir therapy and determine the trial’s implications in the management of COVID-19.

The discovery of an effective pharmacologic intervention for COVID-19 is of utmost urgency. While the present study was unable to answer the question of whether remdesivir is effective in improving clinical outcomes in patients with severe COVID-19, other ongoing or completed (ie, ACTT) studies will likely address this knowledge gap in the coming months. The FDA’s emergency use authorization for remdesivir provides a glimpse into this possibility.

–Katerina Oikonomou, MD, Brookdale Department of Geriatrics & Palliative Medicine, Icahn School of Medicine at Mount Sinai, New York, NY

–Fred Ko, MD

Study Overview

Objective. To assess the efficacy, safety, and clinical benefit of remdesivir in hospitalized adults with confirmed pneumonia due to severe SARS-CoV-2 infection.

Design. Randomized, investigator-initiated, placebo-controlled, double-blind, multicenter trial.

Setting and participants. The trial took place between February 6, 2020 and March 12, 2020, at 10 hospitals in Wuhan, China. Study participants included adult patients (aged ≥ 18 years) admitted to hospital who tested positive for SARS-CoV-2 by reverse transcription polymerase chain reaction assay and had the following clinical characteristics: radiographic evidence of pneumonia; hypoxia with oxygen saturation ≤ 94% on room air or a ratio of arterial oxygen partial pressure to fractional inspired oxygen ≤ 300 mm Hg; and symptom onset to enrollment ≤ 12 days. Some of the exclusion criteria for participation in the study were pregnancy or breast feeding, liver cirrhosis, abnormal liver enzymes ≥ 5 times the upper limit of normal, severe renal impairment or receipt of renal replacement therapy, plan for transfer to a non-study hospital, and enrollment in a trial for COVID-19 within the previous month.

Intervention. Participants were randomized in a 2:1 ratio to the remdesivir group or the placebo group and were administered either intravenous infusions of remdesivir (200 mg on day 1 followed by 100 mg daily on days 2-10) or the same volume of placebo for 10 days. Clinical and safety data assessed included laboratory testing, electrocardiogram, and medication adverse effects. Testing of oropharyngeal and nasopharyngeal swab samples, anal swab samples, sputum, and stool was performed for viral RNA detection and quantification on days 1, 3, 5, 7, 10, 14, 21, and 28.

Main outcome measures. The primary endpoint of this study was time to clinical improvement within 28 days after randomization. Clinical improvement was defined as a 2-point reduction in participants’ admission status on a 6-point ordinal scale (1 = discharged or clinical recovery, 6 = death) or live discharge from hospital, whichever came first. Secondary outcomes included all-cause mortality at day 28 and duration of hospital admission, oxygen support, and invasive mechanical ventilation. Virological measures and safety outcomes ascertained included treatment-emergent adverse events, serious adverse events, and premature discontinuation of remdesivir.

The sample size estimate for the original study design was a total of 453 patients (302 in the remdesivir group and 151 in the placebo group). This sample size would provide 80% power, assuming a hazard ratio (HR) of 1.4 comparing remdesivir to placebo, and corresponding to a change in time to clinical improvement of 6 days. The analysis of primary outcome was performed on an intention-to-treat basis. Time to clinical improvement within 28 days was assessed with Kaplan-Meier plots.

Main results. A total of 255 patients were screened, of whom 237 were enrolled and randomized to remdesivir (158) or placebo (79) group. Of the participants in the remdesivir group, 155 started study treatment and 150 completed treatment per protocol. For the participants in the placebo group, 78 started study treatment and 76 completed treatment per-protocol. Study enrollment was terminated after March 12, 2020, before attaining the prespecified sample size, because no additional patients met study eligibility criteria due to various public health measures implemented in Wuhan. The median age of participants was 65 years (IQR, 56-71), the majority were men (56% in remdesivir group vs 65% in placebo group), and the most common comorbidities included hypertension, diabetes, and coronary artery disease. Median time from symptom onset to study enrollment was 10 days (IQR, 9-12). The time to clinical improvement between treatments (21 days for remdesivir group vs 23 days for placebo group) was not significantly different (HR, 1.23; 95% confidence interval [CI], 0.87-1.75). In addition, in participants who received treatment within 10 days of symptom onset, those who were administered remdesivir had a nonsignificant (HR, 1.52; 95% CI, 0.95-2.43) but faster time (18 days) to clinical improvement, compared to those administered placebo (23 days). Moreover, treatment with remdesivir versus placebo did not lead to differences in secondary outcomes (eg, 28-day mortality and duration of hospital stay, oxygen support, and invasive mechanical ventilation), changes in viral load over time, or adverse events between the groups.

Conclusion. This study found that, compared with placebo, intravenous remdesivir did not significantly improve the time to clinical improvement, mortality, or time to clearance of SARS-CoV-2 in hospitalized adults with severe COVID-19. A numeric reduction in time to clinical improvement with early remdesivir treatment (ie, within 10 days of symptom onset) that approached statistical significance was observed in this underpowered study.

Commentary

Within a few short months since its emergence. SARS-CoV-2 infection has caused a global pandemic, posing a dire threat to public health due to its adverse effects on morbidity (eg, respiratory failure, thromboembolic diseases, multiorgan failure) and mortality. To date, no pharmacologic treatment has been shown to effectively improve clinical outcomes in patients with COVID-19. Multiple ongoing clinical trials are being conducted globally to determine potential therapeutic treatments for severe COVID-19. The first clinical trials of hydroxychloroquine and lopinavir-ritonavir, agents traditionally used for other indications, such as malaria and HIV, did not show a clear benefit in COVID-19.^1,2 Remdesivir, a nucleoside analogue prodrug, is a broad-spectrum antiviral agent that was previously used for treatment of Ebola and has been shown to have inhibitory effects on pathogenic coronaviruses. The study reported by Wang and colleagues was the first randomized controlled trial (RCT) aimed at evaluating whether remdesivir improves outcomes in patients with severe COVID-19. Thus, the worsening COVID-19 pandemic, coupled with the absence of a curative treatment, underscore the urgency of this trial.

The study was grounded on observational data from several recent case reports and case series centering on the potential efficacy of remdesivir in treating COVID-19.³ The study itself was designed well (ie, randomized, placebo-controlled, double-blind, multicenter) and carefully implemented (ie, high protocol adherence to treatments, no loss to follow-up). The principal limitation of this study was its inability to reach the estimated statistical power of study. Due to successful epidemic control in Wuhan, which led to marked reductions in hospital admission of patients with COVID-19, and implementation of stringent termination criteria per the study protocol, only 237 participants were enrolled, instead of the 453, as specified by the sample estimate. This corresponded to a reduction of statistical power from 80% to 58%. Due to this limitation, the study was underpowered, rendering its findings inconclusive.

Despite this limitation, the study found that those treated with remdesivir within 10 days of symptom onset had a numerically faster time (although not statistically significant) to clinical improvement. This leads to an interesting question: whether remdesivir administration early in COVID-19 course could improve clinical outcomes, a question that warrants further investigation by an adequately powered trial. Also, data from this study provided evidence that intravenous remdesivir administration is likely safe in adults during the treatment period, although the long-term drug effects, as well as the safety profile in pediatric patients, remain unknown at this time.

While the study reported by Wang and colleagues was underpowered and is thus inconclusive, several other ongoing RCTs are evaluating the potential clinical benefit of remdesivir treatment in patients hospitalized with COVID-19. On the date of online publication of this report in The Lancet, the National Institutes of Health (NIH) published a news release summarizing preliminary findings from the Adaptive COVID-19 Treatment Trial (ACTT), which showed positive effects of remdesivir on clinical recovery from advanced COVID-19.⁴ The ACTT, the first RCT launched in the United States to evaluate experimental treatment for COVID-19, included 1063 hospitalized participants with advanced COVID-19 and lung involvement. Participants who were administered remdesivir had a 31% faster time to recovery compared to those in the placebo group (median time to recovery, 11 days vs 15 days, respectively; P < 0.001), and had near statistically significant improved survival (mortality rate, 8.0% vs 11.6%, respectively; P = 0.059). In response to these findings, the US Food and Drug Administration (FDA) issued an emergency use authorization for remdesivir on May 1, 2020, for the treatment of suspected or laboratory-confirmed COVID-19 in adults and children hospitalized with severe disease.⁵ While the findings noted from the NIH news release are very encouraging and provide the first evidence of a potentially beneficial antiviral treatment for severe COVID-19 in humans, the scientific community awaits the peer-reviewed publication of the ACTT to better assess the safety and effectiveness of remdesivir therapy and determine the trial’s implications in the management of COVID-19.

The discovery of an effective pharmacologic intervention for COVID-19 is of utmost urgency. While the present study was unable to answer the question of whether remdesivir is effective in improving clinical outcomes in patients with severe COVID-19, other ongoing or completed (ie, ACTT) studies will likely address this knowledge gap in the coming months. The FDA’s emergency use authorization for remdesivir provides a glimpse into this possibility.

–Katerina Oikonomou, MD, Brookdale Department of Geriatrics & Palliative Medicine, Icahn School of Medicine at Mount Sinai, New York, NY

–Fred Ko, MD

Study Overview

Objective. To assess the efficacy, safety, and clinical benefit of remdesivir in hospitalized adults with confirmed pneumonia due to severe SARS-CoV-2 infection.

Design. Randomized, investigator-initiated, placebo-controlled, double-blind, multicenter trial.

Setting and participants. The trial took place between February 6, 2020 and March 12, 2020, at 10 hospitals in Wuhan, China. Study participants included adult patients (aged ≥ 18 years) admitted to hospital who tested positive for SARS-CoV-2 by reverse transcription polymerase chain reaction assay and had the following clinical characteristics: radiographic evidence of pneumonia; hypoxia with oxygen saturation ≤ 94% on room air or a ratio of arterial oxygen partial pressure to fractional inspired oxygen ≤ 300 mm Hg; and symptom onset to enrollment ≤ 12 days. Some of the exclusion criteria for participation in the study were pregnancy or breast feeding, liver cirrhosis, abnormal liver enzymes ≥ 5 times the upper limit of normal, severe renal impairment or receipt of renal replacement therapy, plan for transfer to a non-study hospital, and enrollment in a trial for COVID-19 within the previous month.

Intervention. Participants were randomized in a 2:1 ratio to the remdesivir group or the placebo group and were administered either intravenous infusions of remdesivir (200 mg on day 1 followed by 100 mg daily on days 2-10) or the same volume of placebo for 10 days. Clinical and safety data assessed included laboratory testing, electrocardiogram, and medication adverse effects. Testing of oropharyngeal and nasopharyngeal swab samples, anal swab samples, sputum, and stool was performed for viral RNA detection and quantification on days 1, 3, 5, 7, 10, 14, 21, and 28.

Main outcome measures. The primary endpoint of this study was time to clinical improvement within 28 days after randomization. Clinical improvement was defined as a 2-point reduction in participants’ admission status on a 6-point ordinal scale (1 = discharged or clinical recovery, 6 = death) or live discharge from hospital, whichever came first. Secondary outcomes included all-cause mortality at day 28 and duration of hospital admission, oxygen support, and invasive mechanical ventilation. Virological measures and safety outcomes ascertained included treatment-emergent adverse events, serious adverse events, and premature discontinuation of remdesivir.

The sample size estimate for the original study design was a total of 453 patients (302 in the remdesivir group and 151 in the placebo group). This sample size would provide 80% power, assuming a hazard ratio (HR) of 1.4 comparing remdesivir to placebo, and corresponding to a change in time to clinical improvement of 6 days. The analysis of primary outcome was performed on an intention-to-treat basis. Time to clinical improvement within 28 days was assessed with Kaplan-Meier plots.

Main results. A total of 255 patients were screened, of whom 237 were enrolled and randomized to remdesivir (158) or placebo (79) group. Of the participants in the remdesivir group, 155 started study treatment and 150 completed treatment per protocol. For the participants in the placebo group, 78 started study treatment and 76 completed treatment per-protocol. Study enrollment was terminated after March 12, 2020, before attaining the prespecified sample size, because no additional patients met study eligibility criteria due to various public health measures implemented in Wuhan. The median age of participants was 65 years (IQR, 56-71), the majority were men (56% in remdesivir group vs 65% in placebo group), and the most common comorbidities included hypertension, diabetes, and coronary artery disease. Median time from symptom onset to study enrollment was 10 days (IQR, 9-12). The time to clinical improvement between treatments (21 days for remdesivir group vs 23 days for placebo group) was not significantly different (HR, 1.23; 95% confidence interval [CI], 0.87-1.75). In addition, in participants who received treatment within 10 days of symptom onset, those who were administered remdesivir had a nonsignificant (HR, 1.52; 95% CI, 0.95-2.43) but faster time (18 days) to clinical improvement, compared to those administered placebo (23 days). Moreover, treatment with remdesivir versus placebo did not lead to differences in secondary outcomes (eg, 28-day mortality and duration of hospital stay, oxygen support, and invasive mechanical ventilation), changes in viral load over time, or adverse events between the groups.

Conclusion. This study found that, compared with placebo, intravenous remdesivir did not significantly improve the time to clinical improvement, mortality, or time to clearance of SARS-CoV-2 in hospitalized adults with severe COVID-19. A numeric reduction in time to clinical improvement with early remdesivir treatment (ie, within 10 days of symptom onset) that approached statistical significance was observed in this underpowered study.

Commentary

Within a few short months since its emergence. SARS-CoV-2 infection has caused a global pandemic, posing a dire threat to public health due to its adverse effects on morbidity (eg, respiratory failure, thromboembolic diseases, multiorgan failure) and mortality. To date, no pharmacologic treatment has been shown to effectively improve clinical outcomes in patients with COVID-19. Multiple ongoing clinical trials are being conducted globally to determine potential therapeutic treatments for severe COVID-19. The first clinical trials of hydroxychloroquine and lopinavir-ritonavir, agents traditionally used for other indications, such as malaria and HIV, did not show a clear benefit in COVID-19.^1,2 Remdesivir, a nucleoside analogue prodrug, is a broad-spectrum antiviral agent that was previously used for treatment of Ebola and has been shown to have inhibitory effects on pathogenic coronaviruses. The study reported by Wang and colleagues was the first randomized controlled trial (RCT) aimed at evaluating whether remdesivir improves outcomes in patients with severe COVID-19. Thus, the worsening COVID-19 pandemic, coupled with the absence of a curative treatment, underscore the urgency of this trial.

The study was grounded on observational data from several recent case reports and case series centering on the potential efficacy of remdesivir in treating COVID-19.³ The study itself was designed well (ie, randomized, placebo-controlled, double-blind, multicenter) and carefully implemented (ie, high protocol adherence to treatments, no loss to follow-up). The principal limitation of this study was its inability to reach the estimated statistical power of study. Due to successful epidemic control in Wuhan, which led to marked reductions in hospital admission of patients with COVID-19, and implementation of stringent termination criteria per the study protocol, only 237 participants were enrolled, instead of the 453, as specified by the sample estimate. This corresponded to a reduction of statistical power from 80% to 58%. Due to this limitation, the study was underpowered, rendering its findings inconclusive.

Despite this limitation, the study found that those treated with remdesivir within 10 days of symptom onset had a numerically faster time (although not statistically significant) to clinical improvement. This leads to an interesting question: whether remdesivir administration early in COVID-19 course could improve clinical outcomes, a question that warrants further investigation by an adequately powered trial. Also, data from this study provided evidence that intravenous remdesivir administration is likely safe in adults during the treatment period, although the long-term drug effects, as well as the safety profile in pediatric patients, remain unknown at this time.

While the study reported by Wang and colleagues was underpowered and is thus inconclusive, several other ongoing RCTs are evaluating the potential clinical benefit of remdesivir treatment in patients hospitalized with COVID-19. On the date of online publication of this report in The Lancet, the National Institutes of Health (NIH) published a news release summarizing preliminary findings from the Adaptive COVID-19 Treatment Trial (ACTT), which showed positive effects of remdesivir on clinical recovery from advanced COVID-19.⁴ The ACTT, the first RCT launched in the United States to evaluate experimental treatment for COVID-19, included 1063 hospitalized participants with advanced COVID-19 and lung involvement. Participants who were administered remdesivir had a 31% faster time to recovery compared to those in the placebo group (median time to recovery, 11 days vs 15 days, respectively; P < 0.001), and had near statistically significant improved survival (mortality rate, 8.0% vs 11.6%, respectively; P = 0.059). In response to these findings, the US Food and Drug Administration (FDA) issued an emergency use authorization for remdesivir on May 1, 2020, for the treatment of suspected or laboratory-confirmed COVID-19 in adults and children hospitalized with severe disease.⁵ While the findings noted from the NIH news release are very encouraging and provide the first evidence of a potentially beneficial antiviral treatment for severe COVID-19 in humans, the scientific community awaits the peer-reviewed publication of the ACTT to better assess the safety and effectiveness of remdesivir therapy and determine the trial’s implications in the management of COVID-19.

The discovery of an effective pharmacologic intervention for COVID-19 is of utmost urgency. While the present study was unable to answer the question of whether remdesivir is effective in improving clinical outcomes in patients with severe COVID-19, other ongoing or completed (ie, ACTT) studies will likely address this knowledge gap in the coming months. The FDA’s emergency use authorization for remdesivir provides a glimpse into this possibility.

–Katerina Oikonomou, MD, Brookdale Department of Geriatrics & Palliative Medicine, Icahn School of Medicine at Mount Sinai, New York, NY

–Fred Ko, MD

Children with neurologic impairment (NI; eg, hypoxic-­ischemic encephalopathy, muscular dystrophy) are characterized by functional and/or intellectual impairments resulting from a variety of neurologic diseases.¹ These children commonly have respiratory comorbidities, including central hypoventilation, impaired cough, and oromotor dysfunction, that may lead to chronic respiratory insufficiency and a need for chronic respiratory support at baseline.^2,3 Baseline respiratory support modalities can include supplemental oxygen, noninvasive positive pressure ventilation, or invasive mechanical ventilation.

Acute respiratory infections (ARI; eg, pneumonia, bronchiolitis) are the most common cause of hospitalization, intensive care unit (ICU) admission, and death for children with NI.^1,3 Discharge criteria for otherwise healthy children admitted to the hospital with ARI often include return to respiratory baseline.⁴ Children with complex chronic conditions have longer hospitalizations when hospitalized with respiratory infections,^5-7 because, in part, comorbidities and complications prolong the time to return to baseline. This prolonged return to respiratory baseline in combination with family knowledge, comfort, and skill in managing respiratory support and other complexities at home may alter discharge practices in the population of children with NI. In our clinical experience, discharge before return to baseline respiratory support occurs more frequently in children with NI than in otherwise healthy children when hospitalized with ARI. However, the consequences of discharging children with NI prior to return to respiratory baseline are unknown.

In this study, we sought to determine if discharge prior to return to baseline respiratory support is associated with reutilization among children with NI hospitalized with ARI. We hypothesized that patients discharged prior to return to respiratory baseline would have higher rates of 30-day hospital reutilization.

Study Design and Data Source

This single-center, retrospective cohort study of children hospitalized at Cincinnati Children’s Hospital Medical Center (CCHMC) used data from the Pediatric Health Information System (PHIS) and the electronic medical record (EMR). PHIS, an administrative database of 45 not-for-profit, tertiary care, US pediatric hospitals managed by Children’s Hospital Association (Lenexa, Kansas), was used to identify eligible children, examine demographic and clinical variables, and define outcomes. PHIS contains data regarding patient demographics, inpatient resource utilization, and diagnoses. Encrypted medical record numbers in PHIS allowed for local identification of patients’ medical records to complete EMR review to confirm eligibility and obtain detailed patient-level clinical information (eg, respiratory support needs) not available in PHIS.

Pilot medical record reviews allowed for standardized study definitions and procedures. All study staff underwent training with the primary investigator, including detailed review of 10 initial abstractions. Two investigators (K.M. and S.C.) performed repeat abstractions from 40 randomly selected records to enable assessment of interrater reliability. Average reliability, indicated by the κ statistic, indicated substantial to near-perfect reliability⁸ (κ = 0.97, 95% CI 0.90-1.00) for the primary exposure. EMR data were managed using Research Electronic Data Capture (REDCap, Nashville, Tennessee)⁹ and subsequently merged with PHIS data.

Study Population

Hospitalizations of children with NI aged 1 to 18 years at CCHMC between January 2010 and September 2015 were eligible for inclusion if they had a principal discharge diagnosis indicative of ARI and required increased respiratory support from baseline during hospitalization. NI was defined as a high-intensity, chronic neurological diagnosis with substantial functional impairments according to previously defined diagnosis codes.^1,10 ARI was identified using codes in the Clinical Classification Software (Agency for Healthcare Research and Quality, Rockville, MD) respiratory group indicative of ARI (eg, pneumonia, bronchiolitis, influenza; Appendix Table).

Children transferred to CCHMC were excluded because records from their initial illness presentation and management were not available. Because of expected differences in management and outcomes, children with a known diagnosis of tuberculosis or human immunodeficiency virus were excluded. Because exposure criteria were dependent on hospital discharge status, hospitalizations for children who died during admission (4 of 632 hospitalizations, 0.63%) were excluded from the final cohort (Appendix Figure).

Study Definitions

Baseline respiratory support (ie, “respiratory baseline”) was defined as the child’s highest level of respiratory support needed prior to admission when well (ie, no support, supplemental oxygen, continuous positive airway pressure [CPAP] or bilevel positive airway pressure [BiPAP], or ventilator support), and further characterized by night or day/night requirement. Respiratory baseline was identified using EMR documentation of home respiratory support use at the time of index admission. Return to respiratory baseline was defined as the date on which the child achieved documented home respiratory support settings, regardless of clinical symptoms.

Children may have required increased respiratory support from baseline at any time during hospitalization. Maximum respiratory support required was categorized as one of the following: (1) initiation of supplemental oxygen or increase in oxygen flow or duration; (2) initiation of CPAP or BiPAP; (3) increase in pressure settings or duration of pressure support for those with baseline CPAP, BiPAP, or ventilator use; and (4) initiation of full mechanical ventilation. Respiratory support categories were mutually exclusive: children requiring multiple types of increased respiratory support were classified for analysis by the most invasive form of respiratory support used (eg, a child requiring increase in both oxygen flow and pressure settings was categorized as an increase in pressure settings). Children who received heated high-flow nasal cannula therapy as maximum support were categorized as initiation or increase in oxygen support.

Time to return to respiratory baseline was defined as the difference in days between date of return to respiratory baseline and date of admission. Time to return to respiratory baseline was determined only for children who were discharged at respiratory baseline.

Primary Exposure and Outcome Measures

The primary exposure was hospital discharge before return to respiratory baseline (ie, discharge on higher respiratory support than at baseline settings). At our institution, standardized discharge criteria for children with NI do not exist. The primary outcome was all-cause, 30-day hospital reutilization, including hospital readmissions and emergency department (ED) revisits. Secondary outcomes included 30-day reutilization for ARI and hospital length of stay (LOS) in days.

Patient Demographics and Clinical Characteristics

Demographic and patient characteristics that might influence hospital discharge before return to respiratory baseline or readmission were obtained from PHIS (eg, demographic information, age, insurance type, measures of clinical complexity, illness severity) and by EMR review (eg, baseline respiratory support needs, maximum respiratory support during hospitalization). Measures of clinical complexity included comorbid complex chronic conditions (CCCs)^11-14 and technology dependence^14-16 using previously defined diagnostic codes. Measures of illness severity included sepsis¹⁷ and ICU-level care. At our institution, children with baseline ventilator use do not require admission to the ICU unless they are clinically unstable.

Statistical Analysis

Continuous variables were described using medians and interquartile ranges (IQR). Categorical variables were described using counts and percentages. Patient characteristics and outcomes were stratified by primary exposure and compared using chi-square test or Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables.

To examine the independent association between discharge before return to respiratory baseline and hospital reutilization, a generalized estimating equation was used that included potential confounders while accounting for within-patient clustering. Patient demographics included age, race, ethnicity, and insurance type; measures of clinical complexity included number of CCCs, technology dependence, and baseline respiratory support; and measures of acute illness severity included ARI diagnosis, degree of increase in respiratory support during hospitalization, and ICU-level care. LOS was also included in the model as a covariate because of its expected association with both exposure and outcome.

Secondary analyses were conducted using the outcome of 30-day reutilization for ARI. Subgroup analysis excluding hospitalizations of children lost to follow-up (ie, no encounters in the 6 months after hospital discharge) was also conducted. All analyses were performed with SAS v9.3 (SAS Institute, Cary, North Carolina). P values < .05 were considered statistically significant. This study was approved by the Institutional Review Board.

Study Cohort

A total of 632 hospitalizations experienced by 366 children with NI who were admitted with ARI were included (Appendix Figure). Most children (66.4%) in the cohort experienced only one hospitalization, 17.5% had two hospitalizations, 7.9% had three hospitalizations, and 8.2% had four or more hospitalizations. The median age at hospitalization was 5.0 years (IQR 2.8-10.5) and most hospitalizations were for children who were male (56.6%), white (78.3%), non-Hispanic (96.0%), and publicly insured (51.7%; Table 1). More than one-quarter (28.6%) of hospitalizations were for children with four or more CCCs, and in 73.4% of hospitalizations, children were technology dependent (Table 1). Baseline respiratory support was common (46.8%), including home mechanical ventilation in 11.1% of hospitalizations (Table 1). Bacterial pneumonia, including aspiration pneumonia, was the most common discharge diagnosis (50.5%, Table 1).

Demographic and Clinical Characteristics

Children were discharged before return to respiratory baseline in 30.4% of hospitalizations (Appendix Figure). Children discharged before return to respiratory baseline were older (median age 5.7 years, IQR 3.1-11.0, vs 4.9 years, IQR 2.6-9.7; P = .04) and more likely to be privately insured (54.2% vs 43.4%; P = .04), compared with children discharged at respiratory baseline (Table 1). Children discharged before return to respiratory baseline were also more likely to have a respiratory CCC (59.9% vs 30.9%; P < .001), have a respiratory technology dependence diagnosis code (44.8% vs 24.1%; P < .001), and have baseline respiratory support needs on EMR review (67.7% vs 37.7%; P < .001), compared with children discharged at baseline (Table 1).

Children discharged before return to respiratory baseline required significantly greater escalation in respiratory support during hospitalization, compared with children discharged at respiratory baseline, including higher rates of initiation of CPAP or BiPAP, increased pressure settings from baseline (for home CPAP, BiPAP, or ventilator users), and initiation of full mechanical ventilation (Table 1). Hospitalizations in which children were discharged before return to respiratory baseline were also more likely to include ICU care than were those for children discharged at baseline (52.1% vs 35.2%; P < .001; Table 1).

Clinical Outcomes and Utilization

Reutilization within 30 days occurred after 32.1% of hospitalizations, with 26.1% requiring hospital readmission and 6.0% requiring ED revisit (Table 2). There was no statistical association in either unadjusted (Table 2) or adjusted (Table 3) analysis between children discharged before return to respiratory baseline and 30-day all-cause hospital reutilizations, hospital readmissions, or ED revisits.

In analysis of secondary outcomes, 30-day reutilization because of ARI occurred after 21.5% of hospitalizations, with 19.0% requiring hospital readmission and 2.5% requiring ED revisit. Median hospital LOS for the cohort was 4 days (IQR 2-8; Table 2). Hospitalizations in which children were discharged before return to respiratory baseline were longer than in those discharged at baseline (median 6 days, IQR 3-11, vs 4 days, IQR 2-7; P < .001; Table 2).

For hospitalizations of children discharged at respiratory baseline, the median time to return to respiratory baseline was 3 days (IQR 1-5, complete range 0-80). In these encounters, discharge occurred soon after return to respiratory baseline (median 1 day, IQR 0-1.5, complete range 0-54).

In subgroup analysis excluding the 18 hospitalizations in which children were lost to follow-up (2.8% of the total cohort), discharge before return to respiratory baseline was not associated with 30-day all-cause hospital reutilization (Table 4).

In this retrospective cohort study, children with NI hospitalized with ARI were frequently discharged using increased respiratory support from baseline. However, those discharged before return to respiratory baseline, despite their greater clinical complexity and acute illness severity, did not have increased hospital reutilization, compared with children discharged at respiratory baseline. Our findings suggest that discharge before return to baseline respiratory support after ARI may be clinically appropriate in some children with NI.

With the growing emphasis on decreasing hospital costs, concern exists that patients are being discharged from hospitals “quicker and sicker,”^18,19 with shortening lengths of stay and higher patient instability at discharge. Clinical instability at discharge has been associated with adverse postdischarge outcomes in adults with pneumonia^20-23; however, studies evaluating discharge readiness have not examined the population of children with NI. Our findings of no difference in reutilization for children with NI discharged before return to respiratory baseline, which would be expected to approximate one or more clinical instabilities, contrast these concerns.

Clinicians caring for children with NI hospitalized with ARI may find it difficult to determine a child’s discharge readiness, in part because many children with NI have longer times to return to respiratory baseline and some never return to their pre-­illness baseline.²⁴ In otherwise healthy children hospitalized with respiratory infections such as pneumonia, discharge criteria typically include complete wean from respiratory support prior to discharge.^4,25 In our study’s more complex children, whose parents already manage respiratory support at home, we hypothesize that discharging providers may be comfortable with discharge when the child has certain types of increased respiratory support compatible with home equipment, a parent skilled with monitoring the child’s respiratory status, and the support of an experienced outpatient provider and home nursing providers. At our institution, outpatient respiratory support weans are primarily performed by pediatric pulmonologists and, for isolated weaning of supplemental oxygen or time using support, by parents and outpatient pediatricians.

Another important factor in determining a child’s discharge readiness is the perspective of the child’s parent. Berry et al found that children whose parents believe they are not healthy enough for discharge are more likely to experience unplanned hospital readmissions,²⁴ signaling the role of child- and family-­specific factors in safe discharge decisions. Therefore, parents of children with NI should be proactively involved throughout the multidisciplinary discharge process,^26,27 including the decision to discharge before return to respiratory baseline. Parents have identified ongoing provider support, opportunities to practice home care skills, and written instructions with contingency plans as important components of discharge readiness.²⁸ Further work to create partnerships with these highly skilled caregivers in discharge decision making and transition planning are needed to promote safe discharge practices in this complex population.

In our study, children discharged before return to respiratory baseline were more likely to be older and privately insured compared with children discharged at respiratory baseline. Prior studies have found that social factors including low socioeconomic status influence ED provider admissions decisions for children with pneumonia.^29,30 However, the role of socioeconomic factors in provider discharge decisions for children with NI has not been assessed. These traits may also be proxy markers of other sociodemographic factors, such as parent education level, financial hardship influencing ability to participate in a child’s care at the bedside, access to comprehensive outpatient primary care, and availability of private home nursing. We hypothesize that these related characteristics directly and indirectly influence provider discharge decisions.

Discharging providers are likely more comfortable with discharge prior to return to respiratory baseline when the family has private duty nursing in the home. Home nurses can assist families in providing increased respiratory support and recognizing respiratory problems that may arise following discharge. However, home nursing shortages are common nationwide.^31,32 Low-income children, children with respiratory technology use, and children without Medicaid have been found to have larger gaps in home nursing availability.^31,32 Further studies are needed to understand the role of home nursing availability in provider discharge decisions in this population.

This study has several limitations. The retrospective design of this study creates the potential for residual confounding; there may be other clinical or demographic factors influencing hospital discharge decisions that we are unable to capture using EMR review, including parental knowledge and comfort managing illness, quality of discharge instructions, frequency of follow-up visits, and presence of skilled home nursing services. Categorization of children based on respiratory support status at discharge lends potential for misclassification of exposure; however, our substantial interrater reliability suggests that misclassification bias is small. This study’s primary finding indicated no difference between exposure groups; although we may be unable to detect small differences, we had sufficient power with our sample size to detect meaningful differences in reutilization outcomes.

This study was not designed to capture outpatient time to return to respiratory baseline; prospective studies are needed to identify rates of return to respiratory baseline following ARI in children with NI. We did not measure the level of respiratory support used by children at the time of discharge and, therefore, are unable to estimate the amount of respiratory support weaning needed following discharge or the compatibility of support with home equipment using our data. In addition, this study focused on respiratory support modalities and, thus, did not measure inpatient utilization of mucociliary clearance technologies that might be hypothesized to decrease the time to return to baseline respiratory support. Next steps in evaluating treatment of ARI include investigating the effect of mucociliary clearance on both exposure and outcome in this population.

There may be other clinically meaningful outcomes for this population apart from reutilization that we have not assessed in this study, including increased respiratory support required following discharge, primary care reutilization, healthcare costs, or parent satisfaction with timing of and outcomes after discharge. Finally, although our hospital has reutilization rates for children with NI that are similar to other institutions in the United States,³³ our results may not be generalizable to children with NI hospitalized at other institutions because local discharge processes and systems of care may be different. Prospective, multicenter investigation is needed to evaluate the clinical consequences of discharge before return to respiratory baseline more broadly.

At our institution, approximately one-quarter of children with NI hospitalized with ARI were discharged before return to respiratory baseline, but these children were not at increased risk of reutilization, compared with children discharged at respiratory baseline. Our findings suggest that return to baseline respiratory support might not be a necessary component of hospital discharge criteria. In otherwise clinically stable children with NI, discharge before return to respiratory baseline may be reasonable if their parents are comfortable managing respiratory support at home.

The authors thank Jonathan Rodean of the Children’s Hospital Association for his assistance with abstraction of PHIS data.

Children with neurologic impairment (NI; eg, hypoxic-­ischemic encephalopathy, muscular dystrophy) are characterized by functional and/or intellectual impairments resulting from a variety of neurologic diseases.¹ These children commonly have respiratory comorbidities, including central hypoventilation, impaired cough, and oromotor dysfunction, that may lead to chronic respiratory insufficiency and a need for chronic respiratory support at baseline.^2,3 Baseline respiratory support modalities can include supplemental oxygen, noninvasive positive pressure ventilation, or invasive mechanical ventilation.

Acute respiratory infections (ARI; eg, pneumonia, bronchiolitis) are the most common cause of hospitalization, intensive care unit (ICU) admission, and death for children with NI.^1,3 Discharge criteria for otherwise healthy children admitted to the hospital with ARI often include return to respiratory baseline.⁴ Children with complex chronic conditions have longer hospitalizations when hospitalized with respiratory infections,^5-7 because, in part, comorbidities and complications prolong the time to return to baseline. This prolonged return to respiratory baseline in combination with family knowledge, comfort, and skill in managing respiratory support and other complexities at home may alter discharge practices in the population of children with NI. In our clinical experience, discharge before return to baseline respiratory support occurs more frequently in children with NI than in otherwise healthy children when hospitalized with ARI. However, the consequences of discharging children with NI prior to return to respiratory baseline are unknown.

In this study, we sought to determine if discharge prior to return to baseline respiratory support is associated with reutilization among children with NI hospitalized with ARI. We hypothesized that patients discharged prior to return to respiratory baseline would have higher rates of 30-day hospital reutilization.

Study Design and Data Source

This single-center, retrospective cohort study of children hospitalized at Cincinnati Children’s Hospital Medical Center (CCHMC) used data from the Pediatric Health Information System (PHIS) and the electronic medical record (EMR). PHIS, an administrative database of 45 not-for-profit, tertiary care, US pediatric hospitals managed by Children’s Hospital Association (Lenexa, Kansas), was used to identify eligible children, examine demographic and clinical variables, and define outcomes. PHIS contains data regarding patient demographics, inpatient resource utilization, and diagnoses. Encrypted medical record numbers in PHIS allowed for local identification of patients’ medical records to complete EMR review to confirm eligibility and obtain detailed patient-level clinical information (eg, respiratory support needs) not available in PHIS.

Pilot medical record reviews allowed for standardized study definitions and procedures. All study staff underwent training with the primary investigator, including detailed review of 10 initial abstractions. Two investigators (K.M. and S.C.) performed repeat abstractions from 40 randomly selected records to enable assessment of interrater reliability. Average reliability, indicated by the κ statistic, indicated substantial to near-perfect reliability⁸ (κ = 0.97, 95% CI 0.90-1.00) for the primary exposure. EMR data were managed using Research Electronic Data Capture (REDCap, Nashville, Tennessee)⁹ and subsequently merged with PHIS data.

Study Population

Hospitalizations of children with NI aged 1 to 18 years at CCHMC between January 2010 and September 2015 were eligible for inclusion if they had a principal discharge diagnosis indicative of ARI and required increased respiratory support from baseline during hospitalization. NI was defined as a high-intensity, chronic neurological diagnosis with substantial functional impairments according to previously defined diagnosis codes.^1,10 ARI was identified using codes in the Clinical Classification Software (Agency for Healthcare Research and Quality, Rockville, MD) respiratory group indicative of ARI (eg, pneumonia, bronchiolitis, influenza; Appendix Table).

Children transferred to CCHMC were excluded because records from their initial illness presentation and management were not available. Because of expected differences in management and outcomes, children with a known diagnosis of tuberculosis or human immunodeficiency virus were excluded. Because exposure criteria were dependent on hospital discharge status, hospitalizations for children who died during admission (4 of 632 hospitalizations, 0.63%) were excluded from the final cohort (Appendix Figure).

Study Definitions

Baseline respiratory support (ie, “respiratory baseline”) was defined as the child’s highest level of respiratory support needed prior to admission when well (ie, no support, supplemental oxygen, continuous positive airway pressure [CPAP] or bilevel positive airway pressure [BiPAP], or ventilator support), and further characterized by night or day/night requirement. Respiratory baseline was identified using EMR documentation of home respiratory support use at the time of index admission. Return to respiratory baseline was defined as the date on which the child achieved documented home respiratory support settings, regardless of clinical symptoms.

Children may have required increased respiratory support from baseline at any time during hospitalization. Maximum respiratory support required was categorized as one of the following: (1) initiation of supplemental oxygen or increase in oxygen flow or duration; (2) initiation of CPAP or BiPAP; (3) increase in pressure settings or duration of pressure support for those with baseline CPAP, BiPAP, or ventilator use; and (4) initiation of full mechanical ventilation. Respiratory support categories were mutually exclusive: children requiring multiple types of increased respiratory support were classified for analysis by the most invasive form of respiratory support used (eg, a child requiring increase in both oxygen flow and pressure settings was categorized as an increase in pressure settings). Children who received heated high-flow nasal cannula therapy as maximum support were categorized as initiation or increase in oxygen support.

Time to return to respiratory baseline was defined as the difference in days between date of return to respiratory baseline and date of admission. Time to return to respiratory baseline was determined only for children who were discharged at respiratory baseline.

Primary Exposure and Outcome Measures

The primary exposure was hospital discharge before return to respiratory baseline (ie, discharge on higher respiratory support than at baseline settings). At our institution, standardized discharge criteria for children with NI do not exist. The primary outcome was all-cause, 30-day hospital reutilization, including hospital readmissions and emergency department (ED) revisits. Secondary outcomes included 30-day reutilization for ARI and hospital length of stay (LOS) in days.

Patient Demographics and Clinical Characteristics

Demographic and patient characteristics that might influence hospital discharge before return to respiratory baseline or readmission were obtained from PHIS (eg, demographic information, age, insurance type, measures of clinical complexity, illness severity) and by EMR review (eg, baseline respiratory support needs, maximum respiratory support during hospitalization). Measures of clinical complexity included comorbid complex chronic conditions (CCCs)^11-14 and technology dependence^14-16 using previously defined diagnostic codes. Measures of illness severity included sepsis¹⁷ and ICU-level care. At our institution, children with baseline ventilator use do not require admission to the ICU unless they are clinically unstable.

Statistical Analysis

Continuous variables were described using medians and interquartile ranges (IQR). Categorical variables were described using counts and percentages. Patient characteristics and outcomes were stratified by primary exposure and compared using chi-square test or Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables.

To examine the independent association between discharge before return to respiratory baseline and hospital reutilization, a generalized estimating equation was used that included potential confounders while accounting for within-patient clustering. Patient demographics included age, race, ethnicity, and insurance type; measures of clinical complexity included number of CCCs, technology dependence, and baseline respiratory support; and measures of acute illness severity included ARI diagnosis, degree of increase in respiratory support during hospitalization, and ICU-level care. LOS was also included in the model as a covariate because of its expected association with both exposure and outcome.

Secondary analyses were conducted using the outcome of 30-day reutilization for ARI. Subgroup analysis excluding hospitalizations of children lost to follow-up (ie, no encounters in the 6 months after hospital discharge) was also conducted. All analyses were performed with SAS v9.3 (SAS Institute, Cary, North Carolina). P values < .05 were considered statistically significant. This study was approved by the Institutional Review Board.

Study Cohort

A total of 632 hospitalizations experienced by 366 children with NI who were admitted with ARI were included (Appendix Figure). Most children (66.4%) in the cohort experienced only one hospitalization, 17.5% had two hospitalizations, 7.9% had three hospitalizations, and 8.2% had four or more hospitalizations. The median age at hospitalization was 5.0 years (IQR 2.8-10.5) and most hospitalizations were for children who were male (56.6%), white (78.3%), non-Hispanic (96.0%), and publicly insured (51.7%; Table 1). More than one-quarter (28.6%) of hospitalizations were for children with four or more CCCs, and in 73.4% of hospitalizations, children were technology dependent (Table 1). Baseline respiratory support was common (46.8%), including home mechanical ventilation in 11.1% of hospitalizations (Table 1). Bacterial pneumonia, including aspiration pneumonia, was the most common discharge diagnosis (50.5%, Table 1).

Demographic and Clinical Characteristics

Children were discharged before return to respiratory baseline in 30.4% of hospitalizations (Appendix Figure). Children discharged before return to respiratory baseline were older (median age 5.7 years, IQR 3.1-11.0, vs 4.9 years, IQR 2.6-9.7; P = .04) and more likely to be privately insured (54.2% vs 43.4%; P = .04), compared with children discharged at respiratory baseline (Table 1). Children discharged before return to respiratory baseline were also more likely to have a respiratory CCC (59.9% vs 30.9%; P < .001), have a respiratory technology dependence diagnosis code (44.8% vs 24.1%; P < .001), and have baseline respiratory support needs on EMR review (67.7% vs 37.7%; P < .001), compared with children discharged at baseline (Table 1).

Children discharged before return to respiratory baseline required significantly greater escalation in respiratory support during hospitalization, compared with children discharged at respiratory baseline, including higher rates of initiation of CPAP or BiPAP, increased pressure settings from baseline (for home CPAP, BiPAP, or ventilator users), and initiation of full mechanical ventilation (Table 1). Hospitalizations in which children were discharged before return to respiratory baseline were also more likely to include ICU care than were those for children discharged at baseline (52.1% vs 35.2%; P < .001; Table 1).

Clinical Outcomes and Utilization

Reutilization within 30 days occurred after 32.1% of hospitalizations, with 26.1% requiring hospital readmission and 6.0% requiring ED revisit (Table 2). There was no statistical association in either unadjusted (Table 2) or adjusted (Table 3) analysis between children discharged before return to respiratory baseline and 30-day all-cause hospital reutilizations, hospital readmissions, or ED revisits.

In analysis of secondary outcomes, 30-day reutilization because of ARI occurred after 21.5% of hospitalizations, with 19.0% requiring hospital readmission and 2.5% requiring ED revisit. Median hospital LOS for the cohort was 4 days (IQR 2-8; Table 2). Hospitalizations in which children were discharged before return to respiratory baseline were longer than in those discharged at baseline (median 6 days, IQR 3-11, vs 4 days, IQR 2-7; P < .001; Table 2).

For hospitalizations of children discharged at respiratory baseline, the median time to return to respiratory baseline was 3 days (IQR 1-5, complete range 0-80). In these encounters, discharge occurred soon after return to respiratory baseline (median 1 day, IQR 0-1.5, complete range 0-54).

In subgroup analysis excluding the 18 hospitalizations in which children were lost to follow-up (2.8% of the total cohort), discharge before return to respiratory baseline was not associated with 30-day all-cause hospital reutilization (Table 4).

In this retrospective cohort study, children with NI hospitalized with ARI were frequently discharged using increased respiratory support from baseline. However, those discharged before return to respiratory baseline, despite their greater clinical complexity and acute illness severity, did not have increased hospital reutilization, compared with children discharged at respiratory baseline. Our findings suggest that discharge before return to baseline respiratory support after ARI may be clinically appropriate in some children with NI.

With the growing emphasis on decreasing hospital costs, concern exists that patients are being discharged from hospitals “quicker and sicker,”^18,19 with shortening lengths of stay and higher patient instability at discharge. Clinical instability at discharge has been associated with adverse postdischarge outcomes in adults with pneumonia^20-23; however, studies evaluating discharge readiness have not examined the population of children with NI. Our findings of no difference in reutilization for children with NI discharged before return to respiratory baseline, which would be expected to approximate one or more clinical instabilities, contrast these concerns.

Clinicians caring for children with NI hospitalized with ARI may find it difficult to determine a child’s discharge readiness, in part because many children with NI have longer times to return to respiratory baseline and some never return to their pre-­illness baseline.²⁴ In otherwise healthy children hospitalized with respiratory infections such as pneumonia, discharge criteria typically include complete wean from respiratory support prior to discharge.^4,25 In our study’s more complex children, whose parents already manage respiratory support at home, we hypothesize that discharging providers may be comfortable with discharge when the child has certain types of increased respiratory support compatible with home equipment, a parent skilled with monitoring the child’s respiratory status, and the support of an experienced outpatient provider and home nursing providers. At our institution, outpatient respiratory support weans are primarily performed by pediatric pulmonologists and, for isolated weaning of supplemental oxygen or time using support, by parents and outpatient pediatricians.

Another important factor in determining a child’s discharge readiness is the perspective of the child’s parent. Berry et al found that children whose parents believe they are not healthy enough for discharge are more likely to experience unplanned hospital readmissions,²⁴ signaling the role of child- and family-­specific factors in safe discharge decisions. Therefore, parents of children with NI should be proactively involved throughout the multidisciplinary discharge process,^26,27 including the decision to discharge before return to respiratory baseline. Parents have identified ongoing provider support, opportunities to practice home care skills, and written instructions with contingency plans as important components of discharge readiness.²⁸ Further work to create partnerships with these highly skilled caregivers in discharge decision making and transition planning are needed to promote safe discharge practices in this complex population.

In our study, children discharged before return to respiratory baseline were more likely to be older and privately insured compared with children discharged at respiratory baseline. Prior studies have found that social factors including low socioeconomic status influence ED provider admissions decisions for children with pneumonia.^29,30 However, the role of socioeconomic factors in provider discharge decisions for children with NI has not been assessed. These traits may also be proxy markers of other sociodemographic factors, such as parent education level, financial hardship influencing ability to participate in a child’s care at the bedside, access to comprehensive outpatient primary care, and availability of private home nursing. We hypothesize that these related characteristics directly and indirectly influence provider discharge decisions.

Discharging providers are likely more comfortable with discharge prior to return to respiratory baseline when the family has private duty nursing in the home. Home nurses can assist families in providing increased respiratory support and recognizing respiratory problems that may arise following discharge. However, home nursing shortages are common nationwide.^31,32 Low-income children, children with respiratory technology use, and children without Medicaid have been found to have larger gaps in home nursing availability.^31,32 Further studies are needed to understand the role of home nursing availability in provider discharge decisions in this population.

This study has several limitations. The retrospective design of this study creates the potential for residual confounding; there may be other clinical or demographic factors influencing hospital discharge decisions that we are unable to capture using EMR review, including parental knowledge and comfort managing illness, quality of discharge instructions, frequency of follow-up visits, and presence of skilled home nursing services. Categorization of children based on respiratory support status at discharge lends potential for misclassification of exposure; however, our substantial interrater reliability suggests that misclassification bias is small. This study’s primary finding indicated no difference between exposure groups; although we may be unable to detect small differences, we had sufficient power with our sample size to detect meaningful differences in reutilization outcomes.

This study was not designed to capture outpatient time to return to respiratory baseline; prospective studies are needed to identify rates of return to respiratory baseline following ARI in children with NI. We did not measure the level of respiratory support used by children at the time of discharge and, therefore, are unable to estimate the amount of respiratory support weaning needed following discharge or the compatibility of support with home equipment using our data. In addition, this study focused on respiratory support modalities and, thus, did not measure inpatient utilization of mucociliary clearance technologies that might be hypothesized to decrease the time to return to baseline respiratory support. Next steps in evaluating treatment of ARI include investigating the effect of mucociliary clearance on both exposure and outcome in this population.

There may be other clinically meaningful outcomes for this population apart from reutilization that we have not assessed in this study, including increased respiratory support required following discharge, primary care reutilization, healthcare costs, or parent satisfaction with timing of and outcomes after discharge. Finally, although our hospital has reutilization rates for children with NI that are similar to other institutions in the United States,³³ our results may not be generalizable to children with NI hospitalized at other institutions because local discharge processes and systems of care may be different. Prospective, multicenter investigation is needed to evaluate the clinical consequences of discharge before return to respiratory baseline more broadly.

At our institution, approximately one-quarter of children with NI hospitalized with ARI were discharged before return to respiratory baseline, but these children were not at increased risk of reutilization, compared with children discharged at respiratory baseline. Our findings suggest that return to baseline respiratory support might not be a necessary component of hospital discharge criteria. In otherwise clinically stable children with NI, discharge before return to respiratory baseline may be reasonable if their parents are comfortable managing respiratory support at home.

The authors thank Jonathan Rodean of the Children’s Hospital Association for his assistance with abstraction of PHIS data.

Children with neurologic impairment (NI; eg, hypoxic-­ischemic encephalopathy, muscular dystrophy) are characterized by functional and/or intellectual impairments resulting from a variety of neurologic diseases.¹ These children commonly have respiratory comorbidities, including central hypoventilation, impaired cough, and oromotor dysfunction, that may lead to chronic respiratory insufficiency and a need for chronic respiratory support at baseline.^2,3 Baseline respiratory support modalities can include supplemental oxygen, noninvasive positive pressure ventilation, or invasive mechanical ventilation.

Acute respiratory infections (ARI; eg, pneumonia, bronchiolitis) are the most common cause of hospitalization, intensive care unit (ICU) admission, and death for children with NI.^1,3 Discharge criteria for otherwise healthy children admitted to the hospital with ARI often include return to respiratory baseline.⁴ Children with complex chronic conditions have longer hospitalizations when hospitalized with respiratory infections,^5-7 because, in part, comorbidities and complications prolong the time to return to baseline. This prolonged return to respiratory baseline in combination with family knowledge, comfort, and skill in managing respiratory support and other complexities at home may alter discharge practices in the population of children with NI. In our clinical experience, discharge before return to baseline respiratory support occurs more frequently in children with NI than in otherwise healthy children when hospitalized with ARI. However, the consequences of discharging children with NI prior to return to respiratory baseline are unknown.

In this study, we sought to determine if discharge prior to return to baseline respiratory support is associated with reutilization among children with NI hospitalized with ARI. We hypothesized that patients discharged prior to return to respiratory baseline would have higher rates of 30-day hospital reutilization.

Study Design and Data Source

This single-center, retrospective cohort study of children hospitalized at Cincinnati Children’s Hospital Medical Center (CCHMC) used data from the Pediatric Health Information System (PHIS) and the electronic medical record (EMR). PHIS, an administrative database of 45 not-for-profit, tertiary care, US pediatric hospitals managed by Children’s Hospital Association (Lenexa, Kansas), was used to identify eligible children, examine demographic and clinical variables, and define outcomes. PHIS contains data regarding patient demographics, inpatient resource utilization, and diagnoses. Encrypted medical record numbers in PHIS allowed for local identification of patients’ medical records to complete EMR review to confirm eligibility and obtain detailed patient-level clinical information (eg, respiratory support needs) not available in PHIS.

Pilot medical record reviews allowed for standardized study definitions and procedures. All study staff underwent training with the primary investigator, including detailed review of 10 initial abstractions. Two investigators (K.M. and S.C.) performed repeat abstractions from 40 randomly selected records to enable assessment of interrater reliability. Average reliability, indicated by the κ statistic, indicated substantial to near-perfect reliability⁸ (κ = 0.97, 95% CI 0.90-1.00) for the primary exposure. EMR data were managed using Research Electronic Data Capture (REDCap, Nashville, Tennessee)⁹ and subsequently merged with PHIS data.

Study Population

Hospitalizations of children with NI aged 1 to 18 years at CCHMC between January 2010 and September 2015 were eligible for inclusion if they had a principal discharge diagnosis indicative of ARI and required increased respiratory support from baseline during hospitalization. NI was defined as a high-intensity, chronic neurological diagnosis with substantial functional impairments according to previously defined diagnosis codes.^1,10 ARI was identified using codes in the Clinical Classification Software (Agency for Healthcare Research and Quality, Rockville, MD) respiratory group indicative of ARI (eg, pneumonia, bronchiolitis, influenza; Appendix Table).

Children transferred to CCHMC were excluded because records from their initial illness presentation and management were not available. Because of expected differences in management and outcomes, children with a known diagnosis of tuberculosis or human immunodeficiency virus were excluded. Because exposure criteria were dependent on hospital discharge status, hospitalizations for children who died during admission (4 of 632 hospitalizations, 0.63%) were excluded from the final cohort (Appendix Figure).

Study Definitions

Baseline respiratory support (ie, “respiratory baseline”) was defined as the child’s highest level of respiratory support needed prior to admission when well (ie, no support, supplemental oxygen, continuous positive airway pressure [CPAP] or bilevel positive airway pressure [BiPAP], or ventilator support), and further characterized by night or day/night requirement. Respiratory baseline was identified using EMR documentation of home respiratory support use at the time of index admission. Return to respiratory baseline was defined as the date on which the child achieved documented home respiratory support settings, regardless of clinical symptoms.

Children may have required increased respiratory support from baseline at any time during hospitalization. Maximum respiratory support required was categorized as one of the following: (1) initiation of supplemental oxygen or increase in oxygen flow or duration; (2) initiation of CPAP or BiPAP; (3) increase in pressure settings or duration of pressure support for those with baseline CPAP, BiPAP, or ventilator use; and (4) initiation of full mechanical ventilation. Respiratory support categories were mutually exclusive: children requiring multiple types of increased respiratory support were classified for analysis by the most invasive form of respiratory support used (eg, a child requiring increase in both oxygen flow and pressure settings was categorized as an increase in pressure settings). Children who received heated high-flow nasal cannula therapy as maximum support were categorized as initiation or increase in oxygen support.

Time to return to respiratory baseline was defined as the difference in days between date of return to respiratory baseline and date of admission. Time to return to respiratory baseline was determined only for children who were discharged at respiratory baseline.

Primary Exposure and Outcome Measures

The primary exposure was hospital discharge before return to respiratory baseline (ie, discharge on higher respiratory support than at baseline settings). At our institution, standardized discharge criteria for children with NI do not exist. The primary outcome was all-cause, 30-day hospital reutilization, including hospital readmissions and emergency department (ED) revisits. Secondary outcomes included 30-day reutilization for ARI and hospital length of stay (LOS) in days.

Patient Demographics and Clinical Characteristics

Demographic and patient characteristics that might influence hospital discharge before return to respiratory baseline or readmission were obtained from PHIS (eg, demographic information, age, insurance type, measures of clinical complexity, illness severity) and by EMR review (eg, baseline respiratory support needs, maximum respiratory support during hospitalization). Measures of clinical complexity included comorbid complex chronic conditions (CCCs)^11-14 and technology dependence^14-16 using previously defined diagnostic codes. Measures of illness severity included sepsis¹⁷ and ICU-level care. At our institution, children with baseline ventilator use do not require admission to the ICU unless they are clinically unstable.

Statistical Analysis

Continuous variables were described using medians and interquartile ranges (IQR). Categorical variables were described using counts and percentages. Patient characteristics and outcomes were stratified by primary exposure and compared using chi-square test or Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables.

To examine the independent association between discharge before return to respiratory baseline and hospital reutilization, a generalized estimating equation was used that included potential confounders while accounting for within-patient clustering. Patient demographics included age, race, ethnicity, and insurance type; measures of clinical complexity included number of CCCs, technology dependence, and baseline respiratory support; and measures of acute illness severity included ARI diagnosis, degree of increase in respiratory support during hospitalization, and ICU-level care. LOS was also included in the model as a covariate because of its expected association with both exposure and outcome.

Secondary analyses were conducted using the outcome of 30-day reutilization for ARI. Subgroup analysis excluding hospitalizations of children lost to follow-up (ie, no encounters in the 6 months after hospital discharge) was also conducted. All analyses were performed with SAS v9.3 (SAS Institute, Cary, North Carolina). P values < .05 were considered statistically significant. This study was approved by the Institutional Review Board.

Study Cohort

A total of 632 hospitalizations experienced by 366 children with NI who were admitted with ARI were included (Appendix Figure). Most children (66.4%) in the cohort experienced only one hospitalization, 17.5% had two hospitalizations, 7.9% had three hospitalizations, and 8.2% had four or more hospitalizations. The median age at hospitalization was 5.0 years (IQR 2.8-10.5) and most hospitalizations were for children who were male (56.6%), white (78.3%), non-Hispanic (96.0%), and publicly insured (51.7%; Table 1). More than one-quarter (28.6%) of hospitalizations were for children with four or more CCCs, and in 73.4% of hospitalizations, children were technology dependent (Table 1). Baseline respiratory support was common (46.8%), including home mechanical ventilation in 11.1% of hospitalizations (Table 1). Bacterial pneumonia, including aspiration pneumonia, was the most common discharge diagnosis (50.5%, Table 1).

Demographic and Clinical Characteristics

Children were discharged before return to respiratory baseline in 30.4% of hospitalizations (Appendix Figure). Children discharged before return to respiratory baseline were older (median age 5.7 years, IQR 3.1-11.0, vs 4.9 years, IQR 2.6-9.7; P = .04) and more likely to be privately insured (54.2% vs 43.4%; P = .04), compared with children discharged at respiratory baseline (Table 1). Children discharged before return to respiratory baseline were also more likely to have a respiratory CCC (59.9% vs 30.9%; P < .001), have a respiratory technology dependence diagnosis code (44.8% vs 24.1%; P < .001), and have baseline respiratory support needs on EMR review (67.7% vs 37.7%; P < .001), compared with children discharged at baseline (Table 1).

Children discharged before return to respiratory baseline required significantly greater escalation in respiratory support during hospitalization, compared with children discharged at respiratory baseline, including higher rates of initiation of CPAP or BiPAP, increased pressure settings from baseline (for home CPAP, BiPAP, or ventilator users), and initiation of full mechanical ventilation (Table 1). Hospitalizations in which children were discharged before return to respiratory baseline were also more likely to include ICU care than were those for children discharged at baseline (52.1% vs 35.2%; P < .001; Table 1).

Clinical Outcomes and Utilization

Reutilization within 30 days occurred after 32.1% of hospitalizations, with 26.1% requiring hospital readmission and 6.0% requiring ED revisit (Table 2). There was no statistical association in either unadjusted (Table 2) or adjusted (Table 3) analysis between children discharged before return to respiratory baseline and 30-day all-cause hospital reutilizations, hospital readmissions, or ED revisits.

In analysis of secondary outcomes, 30-day reutilization because of ARI occurred after 21.5% of hospitalizations, with 19.0% requiring hospital readmission and 2.5% requiring ED revisit. Median hospital LOS for the cohort was 4 days (IQR 2-8; Table 2). Hospitalizations in which children were discharged before return to respiratory baseline were longer than in those discharged at baseline (median 6 days, IQR 3-11, vs 4 days, IQR 2-7; P < .001; Table 2).

For hospitalizations of children discharged at respiratory baseline, the median time to return to respiratory baseline was 3 days (IQR 1-5, complete range 0-80). In these encounters, discharge occurred soon after return to respiratory baseline (median 1 day, IQR 0-1.5, complete range 0-54).

In subgroup analysis excluding the 18 hospitalizations in which children were lost to follow-up (2.8% of the total cohort), discharge before return to respiratory baseline was not associated with 30-day all-cause hospital reutilization (Table 4).

In this retrospective cohort study, children with NI hospitalized with ARI were frequently discharged using increased respiratory support from baseline. However, those discharged before return to respiratory baseline, despite their greater clinical complexity and acute illness severity, did not have increased hospital reutilization, compared with children discharged at respiratory baseline. Our findings suggest that discharge before return to baseline respiratory support after ARI may be clinically appropriate in some children with NI.

With the growing emphasis on decreasing hospital costs, concern exists that patients are being discharged from hospitals “quicker and sicker,”^18,19 with shortening lengths of stay and higher patient instability at discharge. Clinical instability at discharge has been associated with adverse postdischarge outcomes in adults with pneumonia^20-23; however, studies evaluating discharge readiness have not examined the population of children with NI. Our findings of no difference in reutilization for children with NI discharged before return to respiratory baseline, which would be expected to approximate one or more clinical instabilities, contrast these concerns.

Clinicians caring for children with NI hospitalized with ARI may find it difficult to determine a child’s discharge readiness, in part because many children with NI have longer times to return to respiratory baseline and some never return to their pre-­illness baseline.²⁴ In otherwise healthy children hospitalized with respiratory infections such as pneumonia, discharge criteria typically include complete wean from respiratory support prior to discharge.^4,25 In our study’s more complex children, whose parents already manage respiratory support at home, we hypothesize that discharging providers may be comfortable with discharge when the child has certain types of increased respiratory support compatible with home equipment, a parent skilled with monitoring the child’s respiratory status, and the support of an experienced outpatient provider and home nursing providers. At our institution, outpatient respiratory support weans are primarily performed by pediatric pulmonologists and, for isolated weaning of supplemental oxygen or time using support, by parents and outpatient pediatricians.

Another important factor in determining a child’s discharge readiness is the perspective of the child’s parent. Berry et al found that children whose parents believe they are not healthy enough for discharge are more likely to experience unplanned hospital readmissions,²⁴ signaling the role of child- and family-­specific factors in safe discharge decisions. Therefore, parents of children with NI should be proactively involved throughout the multidisciplinary discharge process,^26,27 including the decision to discharge before return to respiratory baseline. Parents have identified ongoing provider support, opportunities to practice home care skills, and written instructions with contingency plans as important components of discharge readiness.²⁸ Further work to create partnerships with these highly skilled caregivers in discharge decision making and transition planning are needed to promote safe discharge practices in this complex population.

In our study, children discharged before return to respiratory baseline were more likely to be older and privately insured compared with children discharged at respiratory baseline. Prior studies have found that social factors including low socioeconomic status influence ED provider admissions decisions for children with pneumonia.^29,30 However, the role of socioeconomic factors in provider discharge decisions for children with NI has not been assessed. These traits may also be proxy markers of other sociodemographic factors, such as parent education level, financial hardship influencing ability to participate in a child’s care at the bedside, access to comprehensive outpatient primary care, and availability of private home nursing. We hypothesize that these related characteristics directly and indirectly influence provider discharge decisions.

Discharging providers are likely more comfortable with discharge prior to return to respiratory baseline when the family has private duty nursing in the home. Home nurses can assist families in providing increased respiratory support and recognizing respiratory problems that may arise following discharge. However, home nursing shortages are common nationwide.^31,32 Low-income children, children with respiratory technology use, and children without Medicaid have been found to have larger gaps in home nursing availability.^31,32 Further studies are needed to understand the role of home nursing availability in provider discharge decisions in this population.

This study has several limitations. The retrospective design of this study creates the potential for residual confounding; there may be other clinical or demographic factors influencing hospital discharge decisions that we are unable to capture using EMR review, including parental knowledge and comfort managing illness, quality of discharge instructions, frequency of follow-up visits, and presence of skilled home nursing services. Categorization of children based on respiratory support status at discharge lends potential for misclassification of exposure; however, our substantial interrater reliability suggests that misclassification bias is small. This study’s primary finding indicated no difference between exposure groups; although we may be unable to detect small differences, we had sufficient power with our sample size to detect meaningful differences in reutilization outcomes.

This study was not designed to capture outpatient time to return to respiratory baseline; prospective studies are needed to identify rates of return to respiratory baseline following ARI in children with NI. We did not measure the level of respiratory support used by children at the time of discharge and, therefore, are unable to estimate the amount of respiratory support weaning needed following discharge or the compatibility of support with home equipment using our data. In addition, this study focused on respiratory support modalities and, thus, did not measure inpatient utilization of mucociliary clearance technologies that might be hypothesized to decrease the time to return to baseline respiratory support. Next steps in evaluating treatment of ARI include investigating the effect of mucociliary clearance on both exposure and outcome in this population.

There may be other clinically meaningful outcomes for this population apart from reutilization that we have not assessed in this study, including increased respiratory support required following discharge, primary care reutilization, healthcare costs, or parent satisfaction with timing of and outcomes after discharge. Finally, although our hospital has reutilization rates for children with NI that are similar to other institutions in the United States,³³ our results may not be generalizable to children with NI hospitalized at other institutions because local discharge processes and systems of care may be different. Prospective, multicenter investigation is needed to evaluate the clinical consequences of discharge before return to respiratory baseline more broadly.

At our institution, approximately one-quarter of children with NI hospitalized with ARI were discharged before return to respiratory baseline, but these children were not at increased risk of reutilization, compared with children discharged at respiratory baseline. Our findings suggest that return to baseline respiratory support might not be a necessary component of hospital discharge criteria. In otherwise clinically stable children with NI, discharge before return to respiratory baseline may be reasonable if their parents are comfortable managing respiratory support at home.

The authors thank Jonathan Rodean of the Children’s Hospital Association for his assistance with abstraction of PHIS data.

Increasing numbers of children and adolescents are presenting to children’s hospitals with acute mental health crises requiring emergent or inpatient treatment.^1-5 As a result, children’s hospitals are experiencing additional financial challenges because specialty mental health services are often reimbursed at lower rates than other medical services.^6-9 Poor reimbursement has also been cited as a deterrent to the provision of mental health specialty care, including emergency mental health crisis services.¹⁰ The cumulative financial impact of recent trends in the provision of mental health crisis services at children’s hospitals, however, is unknown. We conducted this study to assess children’s hospitals’ costs, reimbursement, and net profits or losses when delivering inpatient mental health care.

We conducted a retrospective cohort study of the Children’s Hospital Association’s Pediatric Health Information System (PHIS) and Revenue Management Program (RMP) databases. PHIS is an administrative and billing database that collects International Classification of Disease, 10th Revision (ICD-10) diagnoses, procedure codes, and hospital charges from encounters at 52 US children’s hospitals. Costs are estimated from charges using hospital-, year-, and department-specific cost-to-charge ratios. The RMP database is an add-on module to the PHIS database that captures reimbursement data submitted quarterly from 17 participating hospitals based on actual reimbursement amounts collected for each encounter.

Among the 17 participating hospitals, we included all medical (ie, not surgical or intensive care) encounters during calendar year 2017 for children older than 6 years. We stratified encounters into three diagnosis types: primary mental health diagnosis,⁵ suicide attempt,¹¹ or other medical hospitalizations. We separated suicide attempts since these encounters often require care for both mental health concerns and medical complications. Eating disorders were excluded because these programs at children’s hospitals primarily focus on medical complications, require complex multispecialty support, have significantly longer hospitalizations and made up a small volume of overall mental health hospitalizations.

We stratified all analyses by inpatient or observation encounter and determined the proportion of encounters and hospital days attributed to primary mental health, suicide attempt, and other medical conditions at each hospital. One of the 17 children’s hospitals does not use observation status billing, so the observation encounters dataset includes 16 hospitals.

We summarized patients’ demographic and clinical characteristics using frequencies and percentages, comparing across diagnosis groups using chi-square tests. We calculated mean cost per day as (total cost) ÷ (total length of stay [LOS]), reimbursement per day as (total reimbursement) ÷ (total LOS) for each hospital and patient group, and margin per day as (reimbursement per day) – (cost per day). We then determined the total margin difference of caring for mental health vs caring for other medical encounters as ([margin per day for mental health] – [margin per day other medical]) × (number of mental health days). Similarly, we calculated the total margin loss for suicide attempts vs other medical encounters. After calculating profits and losses at individual hospitals, we summed total annual profits and losses to calculate cumulative annual differences. We summarized these profits and losses across all hospitals with medians and interquartile ranges (IQR).

This study of deidentified administrative data was approved by the Internal Review Board at Vanderbilt University as non-human subjects research. All statistical analyses were performed using SAS v.9.4 (SAS Institute, Cary, North Carolina), and P values < .05 were considered statistically significant.

Study Population

Across the 17 included children’s hospitals, there were 8,521 (7.6%) mental health encounters, 3,247 (2.9%) suicide attempt encounters, and 99,937 (89.5%) other medical encounters. LOS was significantly longer for mental health hospitalizations than for suicide attempts and for other medical hospitalizations.

Hospital Characteristics

All 17 free-standing children’s hospitals in the study had an inpatient behavioral health/psychiatric consultation service, and 7 of the 17 had an inpatient behavioral health/psychiatric unit. The total number of discharges for mental health, suicide attempt, and other medical conditions per year varied (range, 2,868-13,214) across the hospitals.

Hospital Daily Profits and Losses for Mental Health, Suicide Attempt, and Other Medical Admissions

For inpatient status mental health hospitalizations, the median margin was $376/day (IQR, $23-$618). For inpatient status suicide attempt hospitalizations, the median margin was $685/day (IQR, $3-$1,117), and for other medical hospitalizations the median margin was $603/day (IQR, $240-$991). With regard to observation status admissions, mental health hospitalizations had a median margin of –$453/day (IQR, –$806 to $362), suicide attempts of –$103/day (IQR, –$639 to $264), and other medical conditions of $353/day (IQR, –$616 to $658; Figure).

Hospital Annual Profits and Losses for Mental Health and Suicide Attempt Admissions, Compared With Other Medical Admissions

The Table shows daily and annual profits and losses for inpatient and observation status. The total annual loss across all hospitals for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, including both inpatient and observation status, was –$26,658,255 when taking both profits and losses into account. For the seven hospitals with net profits for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, the median net profit for combined inpatient and observation status encounters was $119,361 (IQR, $82,818-$195,543), and the total net profit was $5,872,665. For the 10 hospitals with net losses for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, the median net loss for combined inpatient and observation status was –$2,169,357 (IQR, –$4,034,085 to –$511,755), and the total net loss was –$27,419,379.

Hospitalizations for mental health disorders and suicide attempts accounted for 10.5% of hospitalizations at 17 US children’s hospitals in 2017. Overall, mental health and suicide attempt hospitalizations had lower financial margins than did other medical hospitalizations, and they accounted for a total margin loss of more than $26 million across 17 hospitals. Seven hospitals generated a profit for mental health and suicide attempt admissions; 10 hospitals reported losses. Only three hospitals generated a higher net profit for mental health admissions than for other medical admissions. More hospitals had net profits for inpatient status mental health and suicide attempt admissions than for observation status mental health and suicide attempt admissions.

For a minority of children’s hospitals, mental health hospitalizations had higher profit margins than for other medical hospitalizations. This raises questions about patient outcomes and the type of care models employed. One potential explanation is that these hospitals have negotiated favorable agreements with payers. Another possibility could be variations in case-mix and payer mix. Certain mental health services, such as crisis response teams, social workers, and child life specialists, may also be funded from nonpayer sources, so estimates may not fully reflect the cost of providing mental health services. A worst-case view is that hospitals with higher profit margins are providing less or poorer care because of lower reimbursement.

Mental health and suicide attempt hospitalizations were associated with smaller margins but counterintuitively generally wider IQRs for cost. This might be related to variation in care models, but our study was not positioned to examine reasons for this variation. The relationship between reimbursement or margins and patient outcomes, as well as specific mechanisms which may drive costs and outcomes, are areas for future research.

Health insurance plays a crucial role in mental health care. In our study, hospitals were more likely to report positive margins from inpatient status mental health hospitalizations rather than from observation status ones. This is unsurprising because payments for observation status are generally lower than for inpatient status.¹² Less is known about what influences billing and payment for inpatient versus observation at individual hospitals, particularly for mental health hospitalizations. In many cases, billing status is not strictly under the hospital’s control and may be determined by payers during or after the hospitalization. Significant variability in the percentage of patients billed as observation status and the impact of lower, often negative, margins for observation mental health encounters, will have a disproportionate effect on some hospitals. Future work could investigate how these differences may influence overall costs and delivery of care.

This study has several limitations that deserve attention. Costs reported are based on cost to charge ratios, which may generate imperfect estimates. Data was limited to 17 freestanding children’s hospitals, and our findings may not generalize to other hospitals. We also compared mental health and suicide attempt hospitalizations with “other medical” hospitalizations. This broad group contains certain medical conditions that may have higher or lower profit margins than average, and estimates of the margins could be over- or underestimated. We assumed that mental health and suicide attempt admissions were displacing admissions with non–mental health medical conditions (ie, not an empty bed). If those beds would otherwise be unoccupied, raw margins are better estimates of the financial impact than margin differences between mental health/suicide attempt and other medical hospitalizations.

Children’s hospitals are more likely to have significantly lower financial margins for mental health and suicide attempt hospitalizations than for other medical hospitalizations. Future work to investigate how quality of care is associated with reimbursement can help ensure that funding for children’s acute mental health care services is commensurate with resources required to provide high quality services.

The authors had no financial relationships relevant to this article to disclose.

Research reported in this publication was supported by the National Institute of Mental Health of the National Institutes of Health under Award Number K23MH115162 (Doupnik).

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Increasing numbers of children and adolescents are presenting to children’s hospitals with acute mental health crises requiring emergent or inpatient treatment.^1-5 As a result, children’s hospitals are experiencing additional financial challenges because specialty mental health services are often reimbursed at lower rates than other medical services.^6-9 Poor reimbursement has also been cited as a deterrent to the provision of mental health specialty care, including emergency mental health crisis services.¹⁰ The cumulative financial impact of recent trends in the provision of mental health crisis services at children’s hospitals, however, is unknown. We conducted this study to assess children’s hospitals’ costs, reimbursement, and net profits or losses when delivering inpatient mental health care.

We conducted a retrospective cohort study of the Children’s Hospital Association’s Pediatric Health Information System (PHIS) and Revenue Management Program (RMP) databases. PHIS is an administrative and billing database that collects International Classification of Disease, 10th Revision (ICD-10) diagnoses, procedure codes, and hospital charges from encounters at 52 US children’s hospitals. Costs are estimated from charges using hospital-, year-, and department-specific cost-to-charge ratios. The RMP database is an add-on module to the PHIS database that captures reimbursement data submitted quarterly from 17 participating hospitals based on actual reimbursement amounts collected for each encounter.

Among the 17 participating hospitals, we included all medical (ie, not surgical or intensive care) encounters during calendar year 2017 for children older than 6 years. We stratified encounters into three diagnosis types: primary mental health diagnosis,⁵ suicide attempt,¹¹ or other medical hospitalizations. We separated suicide attempts since these encounters often require care for both mental health concerns and medical complications. Eating disorders were excluded because these programs at children’s hospitals primarily focus on medical complications, require complex multispecialty support, have significantly longer hospitalizations and made up a small volume of overall mental health hospitalizations.

We stratified all analyses by inpatient or observation encounter and determined the proportion of encounters and hospital days attributed to primary mental health, suicide attempt, and other medical conditions at each hospital. One of the 17 children’s hospitals does not use observation status billing, so the observation encounters dataset includes 16 hospitals.

We summarized patients’ demographic and clinical characteristics using frequencies and percentages, comparing across diagnosis groups using chi-square tests. We calculated mean cost per day as (total cost) ÷ (total length of stay [LOS]), reimbursement per day as (total reimbursement) ÷ (total LOS) for each hospital and patient group, and margin per day as (reimbursement per day) – (cost per day). We then determined the total margin difference of caring for mental health vs caring for other medical encounters as ([margin per day for mental health] – [margin per day other medical]) × (number of mental health days). Similarly, we calculated the total margin loss for suicide attempts vs other medical encounters. After calculating profits and losses at individual hospitals, we summed total annual profits and losses to calculate cumulative annual differences. We summarized these profits and losses across all hospitals with medians and interquartile ranges (IQR).

This study of deidentified administrative data was approved by the Internal Review Board at Vanderbilt University as non-human subjects research. All statistical analyses were performed using SAS v.9.4 (SAS Institute, Cary, North Carolina), and P values < .05 were considered statistically significant.

Study Population

Across the 17 included children’s hospitals, there were 8,521 (7.6%) mental health encounters, 3,247 (2.9%) suicide attempt encounters, and 99,937 (89.5%) other medical encounters. LOS was significantly longer for mental health hospitalizations than for suicide attempts and for other medical hospitalizations.

Hospital Characteristics

All 17 free-standing children’s hospitals in the study had an inpatient behavioral health/psychiatric consultation service, and 7 of the 17 had an inpatient behavioral health/psychiatric unit. The total number of discharges for mental health, suicide attempt, and other medical conditions per year varied (range, 2,868-13,214) across the hospitals.

Hospital Daily Profits and Losses for Mental Health, Suicide Attempt, and Other Medical Admissions

For inpatient status mental health hospitalizations, the median margin was $376/day (IQR, $23-$618). For inpatient status suicide attempt hospitalizations, the median margin was $685/day (IQR, $3-$1,117), and for other medical hospitalizations the median margin was $603/day (IQR, $240-$991). With regard to observation status admissions, mental health hospitalizations had a median margin of –$453/day (IQR, –$806 to $362), suicide attempts of –$103/day (IQR, –$639 to $264), and other medical conditions of $353/day (IQR, –$616 to $658; Figure).

Hospital Annual Profits and Losses for Mental Health and Suicide Attempt Admissions, Compared With Other Medical Admissions

The Table shows daily and annual profits and losses for inpatient and observation status. The total annual loss across all hospitals for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, including both inpatient and observation status, was –$26,658,255 when taking both profits and losses into account. For the seven hospitals with net profits for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, the median net profit for combined inpatient and observation status encounters was $119,361 (IQR, $82,818-$195,543), and the total net profit was $5,872,665. For the 10 hospitals with net losses for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, the median net loss for combined inpatient and observation status was –$2,169,357 (IQR, –$4,034,085 to –$511,755), and the total net loss was –$27,419,379.

Hospitalizations for mental health disorders and suicide attempts accounted for 10.5% of hospitalizations at 17 US children’s hospitals in 2017. Overall, mental health and suicide attempt hospitalizations had lower financial margins than did other medical hospitalizations, and they accounted for a total margin loss of more than $26 million across 17 hospitals. Seven hospitals generated a profit for mental health and suicide attempt admissions; 10 hospitals reported losses. Only three hospitals generated a higher net profit for mental health admissions than for other medical admissions. More hospitals had net profits for inpatient status mental health and suicide attempt admissions than for observation status mental health and suicide attempt admissions.

For a minority of children’s hospitals, mental health hospitalizations had higher profit margins than for other medical hospitalizations. This raises questions about patient outcomes and the type of care models employed. One potential explanation is that these hospitals have negotiated favorable agreements with payers. Another possibility could be variations in case-mix and payer mix. Certain mental health services, such as crisis response teams, social workers, and child life specialists, may also be funded from nonpayer sources, so estimates may not fully reflect the cost of providing mental health services. A worst-case view is that hospitals with higher profit margins are providing less or poorer care because of lower reimbursement.

Mental health and suicide attempt hospitalizations were associated with smaller margins but counterintuitively generally wider IQRs for cost. This might be related to variation in care models, but our study was not positioned to examine reasons for this variation. The relationship between reimbursement or margins and patient outcomes, as well as specific mechanisms which may drive costs and outcomes, are areas for future research.

Health insurance plays a crucial role in mental health care. In our study, hospitals were more likely to report positive margins from inpatient status mental health hospitalizations rather than from observation status ones. This is unsurprising because payments for observation status are generally lower than for inpatient status.¹² Less is known about what influences billing and payment for inpatient versus observation at individual hospitals, particularly for mental health hospitalizations. In many cases, billing status is not strictly under the hospital’s control and may be determined by payers during or after the hospitalization. Significant variability in the percentage of patients billed as observation status and the impact of lower, often negative, margins for observation mental health encounters, will have a disproportionate effect on some hospitals. Future work could investigate how these differences may influence overall costs and delivery of care.

This study has several limitations that deserve attention. Costs reported are based on cost to charge ratios, which may generate imperfect estimates. Data was limited to 17 freestanding children’s hospitals, and our findings may not generalize to other hospitals. We also compared mental health and suicide attempt hospitalizations with “other medical” hospitalizations. This broad group contains certain medical conditions that may have higher or lower profit margins than average, and estimates of the margins could be over- or underestimated. We assumed that mental health and suicide attempt admissions were displacing admissions with non–mental health medical conditions (ie, not an empty bed). If those beds would otherwise be unoccupied, raw margins are better estimates of the financial impact than margin differences between mental health/suicide attempt and other medical hospitalizations.

Children’s hospitals are more likely to have significantly lower financial margins for mental health and suicide attempt hospitalizations than for other medical hospitalizations. Future work to investigate how quality of care is associated with reimbursement can help ensure that funding for children’s acute mental health care services is commensurate with resources required to provide high quality services.

The authors had no financial relationships relevant to this article to disclose.

Research reported in this publication was supported by the National Institute of Mental Health of the National Institutes of Health under Award Number K23MH115162 (Doupnik).

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Increasing numbers of children and adolescents are presenting to children’s hospitals with acute mental health crises requiring emergent or inpatient treatment.^1-5 As a result, children’s hospitals are experiencing additional financial challenges because specialty mental health services are often reimbursed at lower rates than other medical services.^6-9 Poor reimbursement has also been cited as a deterrent to the provision of mental health specialty care, including emergency mental health crisis services.¹⁰ The cumulative financial impact of recent trends in the provision of mental health crisis services at children’s hospitals, however, is unknown. We conducted this study to assess children’s hospitals’ costs, reimbursement, and net profits or losses when delivering inpatient mental health care.

We conducted a retrospective cohort study of the Children’s Hospital Association’s Pediatric Health Information System (PHIS) and Revenue Management Program (RMP) databases. PHIS is an administrative and billing database that collects International Classification of Disease, 10th Revision (ICD-10) diagnoses, procedure codes, and hospital charges from encounters at 52 US children’s hospitals. Costs are estimated from charges using hospital-, year-, and department-specific cost-to-charge ratios. The RMP database is an add-on module to the PHIS database that captures reimbursement data submitted quarterly from 17 participating hospitals based on actual reimbursement amounts collected for each encounter.

Among the 17 participating hospitals, we included all medical (ie, not surgical or intensive care) encounters during calendar year 2017 for children older than 6 years. We stratified encounters into three diagnosis types: primary mental health diagnosis,⁵ suicide attempt,¹¹ or other medical hospitalizations. We separated suicide attempts since these encounters often require care for both mental health concerns and medical complications. Eating disorders were excluded because these programs at children’s hospitals primarily focus on medical complications, require complex multispecialty support, have significantly longer hospitalizations and made up a small volume of overall mental health hospitalizations.

We stratified all analyses by inpatient or observation encounter and determined the proportion of encounters and hospital days attributed to primary mental health, suicide attempt, and other medical conditions at each hospital. One of the 17 children’s hospitals does not use observation status billing, so the observation encounters dataset includes 16 hospitals.

We summarized patients’ demographic and clinical characteristics using frequencies and percentages, comparing across diagnosis groups using chi-square tests. We calculated mean cost per day as (total cost) ÷ (total length of stay [LOS]), reimbursement per day as (total reimbursement) ÷ (total LOS) for each hospital and patient group, and margin per day as (reimbursement per day) – (cost per day). We then determined the total margin difference of caring for mental health vs caring for other medical encounters as ([margin per day for mental health] – [margin per day other medical]) × (number of mental health days). Similarly, we calculated the total margin loss for suicide attempts vs other medical encounters. After calculating profits and losses at individual hospitals, we summed total annual profits and losses to calculate cumulative annual differences. We summarized these profits and losses across all hospitals with medians and interquartile ranges (IQR).

This study of deidentified administrative data was approved by the Internal Review Board at Vanderbilt University as non-human subjects research. All statistical analyses were performed using SAS v.9.4 (SAS Institute, Cary, North Carolina), and P values < .05 were considered statistically significant.

Study Population

Across the 17 included children’s hospitals, there were 8,521 (7.6%) mental health encounters, 3,247 (2.9%) suicide attempt encounters, and 99,937 (89.5%) other medical encounters. LOS was significantly longer for mental health hospitalizations than for suicide attempts and for other medical hospitalizations.

Hospital Characteristics

All 17 free-standing children’s hospitals in the study had an inpatient behavioral health/psychiatric consultation service, and 7 of the 17 had an inpatient behavioral health/psychiatric unit. The total number of discharges for mental health, suicide attempt, and other medical conditions per year varied (range, 2,868-13,214) across the hospitals.

Hospital Daily Profits and Losses for Mental Health, Suicide Attempt, and Other Medical Admissions

For inpatient status mental health hospitalizations, the median margin was $376/day (IQR, $23-$618). For inpatient status suicide attempt hospitalizations, the median margin was $685/day (IQR, $3-$1,117), and for other medical hospitalizations the median margin was $603/day (IQR, $240-$991). With regard to observation status admissions, mental health hospitalizations had a median margin of –$453/day (IQR, –$806 to $362), suicide attempts of –$103/day (IQR, –$639 to $264), and other medical conditions of $353/day (IQR, –$616 to $658; Figure).

Hospital Annual Profits and Losses for Mental Health and Suicide Attempt Admissions, Compared With Other Medical Admissions

The Table shows daily and annual profits and losses for inpatient and observation status. The total annual loss across all hospitals for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, including both inpatient and observation status, was –$26,658,255 when taking both profits and losses into account. For the seven hospitals with net profits for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, the median net profit for combined inpatient and observation status encounters was $119,361 (IQR, $82,818-$195,543), and the total net profit was $5,872,665. For the 10 hospitals with net losses for mental health and suicide attempt hospitalizations, compared with other medical hospitalizations, the median net loss for combined inpatient and observation status was –$2,169,357 (IQR, –$4,034,085 to –$511,755), and the total net loss was –$27,419,379.

Hospitalizations for mental health disorders and suicide attempts accounted for 10.5% of hospitalizations at 17 US children’s hospitals in 2017. Overall, mental health and suicide attempt hospitalizations had lower financial margins than did other medical hospitalizations, and they accounted for a total margin loss of more than $26 million across 17 hospitals. Seven hospitals generated a profit for mental health and suicide attempt admissions; 10 hospitals reported losses. Only three hospitals generated a higher net profit for mental health admissions than for other medical admissions. More hospitals had net profits for inpatient status mental health and suicide attempt admissions than for observation status mental health and suicide attempt admissions.

For a minority of children’s hospitals, mental health hospitalizations had higher profit margins than for other medical hospitalizations. This raises questions about patient outcomes and the type of care models employed. One potential explanation is that these hospitals have negotiated favorable agreements with payers. Another possibility could be variations in case-mix and payer mix. Certain mental health services, such as crisis response teams, social workers, and child life specialists, may also be funded from nonpayer sources, so estimates may not fully reflect the cost of providing mental health services. A worst-case view is that hospitals with higher profit margins are providing less or poorer care because of lower reimbursement.

Mental health and suicide attempt hospitalizations were associated with smaller margins but counterintuitively generally wider IQRs for cost. This might be related to variation in care models, but our study was not positioned to examine reasons for this variation. The relationship between reimbursement or margins and patient outcomes, as well as specific mechanisms which may drive costs and outcomes, are areas for future research.

Health insurance plays a crucial role in mental health care. In our study, hospitals were more likely to report positive margins from inpatient status mental health hospitalizations rather than from observation status ones. This is unsurprising because payments for observation status are generally lower than for inpatient status.¹² Less is known about what influences billing and payment for inpatient versus observation at individual hospitals, particularly for mental health hospitalizations. In many cases, billing status is not strictly under the hospital’s control and may be determined by payers during or after the hospitalization. Significant variability in the percentage of patients billed as observation status and the impact of lower, often negative, margins for observation mental health encounters, will have a disproportionate effect on some hospitals. Future work could investigate how these differences may influence overall costs and delivery of care.

This study has several limitations that deserve attention. Costs reported are based on cost to charge ratios, which may generate imperfect estimates. Data was limited to 17 freestanding children’s hospitals, and our findings may not generalize to other hospitals. We also compared mental health and suicide attempt hospitalizations with “other medical” hospitalizations. This broad group contains certain medical conditions that may have higher or lower profit margins than average, and estimates of the margins could be over- or underestimated. We assumed that mental health and suicide attempt admissions were displacing admissions with non–mental health medical conditions (ie, not an empty bed). If those beds would otherwise be unoccupied, raw margins are better estimates of the financial impact than margin differences between mental health/suicide attempt and other medical hospitalizations.

Children’s hospitals are more likely to have significantly lower financial margins for mental health and suicide attempt hospitalizations than for other medical hospitalizations. Future work to investigate how quality of care is associated with reimbursement can help ensure that funding for children’s acute mental health care services is commensurate with resources required to provide high quality services.

The authors had no financial relationships relevant to this article to disclose.

Research reported in this publication was supported by the National Institute of Mental Health of the National Institutes of Health under Award Number K23MH115162 (Doupnik).

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Adverse Childhood Experiences, or ACEs, include exposure to abuse, neglect, or household dysfunction (eg, having a parent who is mentally ill) as a child.¹ Exposure to ACEs affects health into adulthood, with a dose-response relationship between ACEs and a range of comorbidities.¹ Adults with 6 or more ACEs have a 20-year shorter life expectancy than do those with no ACEs.¹ Still, ACEs are static; once experienced, that experience cannot be undone. However, resilience, or positive adaptation in the context of adversity, can be protective, buffering the negative effects of ACEs.^2,3 Protective factors that promote resilience include social capital, such as positive relationships with caregivers and peers.³

With their clear link to health outcomes across the life-course, there is a movement for pediatricians to screen children for ACEs⁴ and to develop strategies that promote resilience in children, parents, and families. However, screening a child for adversity has challenges because younger children may not have experienced an adverse exposure, or they may be unable to voice their experiences. Studies have demonstrated that parental adversity, or ACEs, may be a marker for childhood adversity.^5,6 Biological models also support this potential intergenerational effect of ACEs. Chronic exposure to stress, including ACEs, results in elevated cortisol via a dysregulated hypothalamic-pituitary-adrenal axis, which results in chronic inflammation.⁷ This “toxic stress” is prolonged, severe in intensity, and can lead to epigenetic changes that may be passed on to the next generation.^8,9

Hospitalization of an ill child, and the transition to home after that hospitalization, is a stressful event for children and families.¹⁰ This stress may be relevant to parents that have a history of a high rate of ACEs or a current low degree of resilience. Our previous work demonstrated that, in the inpatient setting, parents with high ACEs (≥4) or low resilience have increased coping difficulty 14 days after their child’s hospital discharge.¹¹ Our objective here was to evaluate whether a parent’s ACEs and/or resilience would also be associated with that child’s likelihood of reutilization. We hypothesized that more parental ACEs and/or lower parental resilience would be associated with revisits the emergency room, urgent care, or hospital readmissions.

Participants and Study Design

We conducted a prospective cohort study of parents of hospitalized children recruited from the “Hospital-to-Home Outcomes” Studies (H2O I and H2O II).^12,13 H2O I and II were prospective, single-center, randomized controlled trials designed to determine the effectiveness of either a nurse-led transitional home visit (H2O I) or telephone call (H2O II) on 30-day unplanned healthcare reutilization. The trials and this study were approved by the Cincinnati Children’s Institutional Review Board. All parents provided written informed consent.

Details of H2O I and II recruitment and design have been described previously.^12,13 Briefly, children were eligible for inclusion in either study if they were admitted to our institution’s general Hospital Medicine or the Hospital Medicine Complex Care Services; for H2O I, children hospitalized on the Neurology and Neurosurgery services were also eligible.^12,13 Patients were excluded if they were discharged to a residential facility, if they lived outside the home healthcare nurse service area, if they were eligible for skilled home healthcare services (eg, intravenous antibiotics), or if the participating caregiver was non-English speaking.^12,13 In H2O I, families were randomized either to receive a single nurse home visit within 96 hours of discharge or standard of care. In H2O II, families enrolled were randomized to receive a telephone call by a nurse within 96 hours of discharge or standard of care. As we have previously published, randomization in both trials successfully balanced the intervention and control arms with respect to key demographic characteristics.^12,13 For the analyses presented here, we focused on a subset of caregivers 18 years and older whose children were enrolled in either H2O I or II between August 2015 and October 2016. In both H2O trials, face-to-face and paper-based questionnaires were completed by parents during the index hospitalization.

Outcome and Predictors

Our primary outcome was unanticipated healthcare reutilization defined as return to the emergency room, urgent care, or unplanned readmission within 30 days of hospital discharge, consistent with the H2O trials. This was measured using the primary institution’s administrative data supplemented by a utilization database shared across regional hospitals.¹⁴ Readmissions were identified as “unplanned” using a previously validated algorithm,¹⁵ and treated as a dichotomous yes/no variable.

Our primary predictors were parental ACEs and resilience (see Appendix Tables). The ACE questionnaire addresses abuse, neglect, and household dysfunction in the first 18 years of life.¹ It is composed of 10 questions, each with a yes/no response.¹ We defined parents as low (ACE 0), moderate (ACE 1-3), or high (ACE ≥4) risk a priori because previous literature has described poor outcomes in adults with 4 or more ACEs.¹⁶

Given the sensitive nature of the questions, respondents independently completed the ACE questionnaire on paper instead of via the face-to-face survey. Respondents returned the completed questionnaire to the research assistant in a sealed envelope. All families received educational information on relevant hospital and community-based resources (eg, social work).

Parental resilience was measured using the Brief Resilience Scale (BRS). The BRS is 6 items, each on a 5-point Likert scale. Responses were averaged, providing a total score of 1-5; higher scores are representative of higher resilience.¹⁷ We treated the BRS score as a continuous variable. BRS has been used in clinical settings; it has demonstrated positive correlation with social support and negative correlation with fatigue.¹⁷ Parents answered BRS questions during the index pediatric hospitalization in a face-to-face interview.

Parent and Child Characteristics

Parent and child sociodemographic variables were also obtained during the face-to-face interview. Parental variables included age, gender, educational attainment, household income, employment status, and financial and social strain.¹¹ Educational attainment was analyzed in 2 categories—high school or less vs more than high school—because most discharge instructions are written at a high school reading level.¹⁸ Parents reported their annual household income in the following categories: <$15,000; $15,000-$29,999; $30,000-$44,999; $45,000-$59,999; $60,000-$89,999; $90,000-$119,999; ≥$120,000. Employment was dichotomized as not employed/student vs any employment. Financial and social strain were assessed using a series of 9 previously described questions.¹⁹ These questions assessed, via self-report, a family’s ability to make ends meet, ability to pay rent/mortgage or utilities, need to move in with others because of financial reasons, and ability to borrow money if needed, as well as home ownership and parental marital status.^15,19 Strain questions were all dichotomous (yes/no, single/not single). A composite variable was then constructed that categorized those reporting no strain items, 1 to 2 items, 3 to 4 items, and 5 or more items.²⁰

Child variables included race, ethnicity, age, primary care access,²¹ payer, and H2O treatment arm. Race categories were white/Caucasian, black/African American, American Indian or Alaskan Native, Asian or Pacific Islander, and other; ethnicity categories were Hispanic/Latino, non-Hispanic/Latino, and unknown. Given relatively low numbers of children reported to be Hispanic/Latino, we combined race and ethnicity into a single variable, categorized as non-Hispanic/white, non-Hispanic/black, and multiracial/Hispanic/other. Primary care access was assessed using the access subscale to the Parent’s Perception of Primary Care questionnaire. This includes assessment of a family’s ability to travel to their doctor, to see their doctor for routine or sick care, and to get help or advice on evenings or weekends. Scores were categorized as always adequate, almost always adequate, or sometimes/never adequate.²¹ Payer was dichotomized to private or public/self-pay.

Statistical Analyses

We examined the distribution of outcomes, predictors, and covariates. We compared sociodemographic characteristics of those respondents and nonrespondents to the ACE screen using the chi-square test for categorical variables or the t test for continuous variables. We used logistic regression to assess for associations between the independent variables of interest and reutilization, adjusting for potential confounders. To build our adjusted, multivariable model, we decided a priori to include child race/ethnicity, primary care access, financial and social strain, and trial treatment arm. We treated the H2O I control group as the referent group. Other covariates considered for inclusion were caregiver education, household income, employment, and payer. These were included in multivariable models if bivariate associations were significant at the P < .1 level. We assessed an ACE-by-resilience interaction term because we hypothesized that those with more ACEs and lower resilience may have more reutilization outcomes than parents with fewer ACEs and higher resilience. We also evaluated interaction terms between trial arm assignment and predictors to assess effects that may be introduced by the randomization. Predictors in the final logistic regression model were significant at the P < .05 level. Logistic regression assumption of little or no multicollinearity among the independent variables was verified in the final models. All analyses were performed with Stata v16 (Stata Corp, College Station, Texas).

There were a total of 1,787 parent-child dyads enrolled in the H2O I and II during the study period; 1,320 parents (74%) completed the ACE questionnaire and were included the analysis. Included parents were primarily female and employed, as well as educated beyond high school (Table 1). Overall, 64% reported one or more ACEs (range 0 to 9); 45% reported 1to 3, and 19% reported 4 or more ACEs. The most commonly reported ACEs were divorce (n = 573, 43%), exposure to alcoholism (n = 306, 23%), and exposure to mental illness (n = 281, 21%; Figure 1). Parents had a mean BRS score of 3.97 (range 1.17-5.00), with the distribution shown in Figure 2.

Of the 1,320 included patients, the average length of stay was 2.5 days, and 82% of hospitalizations were caused by acute medical issues (eg, bronchiolitis). A total of 211 children experienced a reutilization event within 30 days of discharge. In bivariate analysis, children with parents with 4 or more ACEs had a 2.02-times (95% CI 1.35-3.02) higher odds of experiencing a reutilization event than did those with parents reporting no ACEs. Parents with higher resilience scores had children with a lower odds of reutilization (odds ratio [OR] 0.77 95% CI 0.63-0.95).

In addition to our a priori variables, parental education, employment, and insurance met our significance threshold for inclusion in the multivariable model. The ACE-by-resilience interaction term was not significant and not included in the model. Similarly, there was no significant interaction between ACE and resilience and H2O treatment arm; the interaction terms were not included in the final adjusted model, but treatment arm assignment was kept as a covariate. A total of 1,292 children, out of the 1,320 respondents, remained in the final multivariable model; the excluded 28 had incomplete covariate data but were not otherwise different. In this final adjusted model, children with parents reporting 4 or more ACEs had a 1.69-times (95% CI 1.11-2.60) greater odds of reutilization than did those with parents reporting no ACEs (Table 2). Resilience failed to reach statistical significance in the adjusted model (OR 0.86, 95% CI 0.70-1.07).

We found that high-risk parents (4 or more ACEs) had children with an increased odds of healthcare reutilization, suggesting intergenerational effects of ACEs. We did not find a similar effect relating to parental resilience. We also did not find an interaction between parental ACEs and resilience, suggesting that a parent’s reported degree of resilience does not modify the effect of ACEs on reutilization risk.

Parental adversity may be a risk factor for a child’s unanticipated reutilization. We previously demonstrated that parents with 4 or more ACEs have more coping difficulty than a parent with no ACEs after a child’s hospitalization.¹¹ It is possible that parents with high adversity may have poorer coping mechanisms when dealing with a stressful situation, such as a child’s hospitalization. This may have resulted in inequitable outcomes (eg, increased reutilization) for their children. Other studies have confirmed such an intergenerational effect of adversity, linking a parent’s ACEs with poor developmental, behavioral, and health outcomes in their children.^6,22,23 O’Malley et al showed an association of parental ACEs to current adversities,²⁴ such as insurance or housing concerns, that affect the entirety of the household, including children. In short, it appears that parental ACEs may be a compelling predictor of current childhood adversity.

Resilience buffers the negative effects of ACEs; however, we did not find significant associations between resilience and reutilization or an interaction between ACEs and resilience. The factors that may contribute to reutilization are complex. In our previous work, parental resilience was associated with coping difficulty after discharge; but again, did not interact with parental ACEs.¹¹ Here, we suggest that while resilience may buffer the negative effects of ACEs, that buffering may not affect the likelihood of reutilization. It is also possible that the BRS tool is of less relevance on how one handles the stress of a child’s hospitalization. While the BRS is one measure of resilience, there are many other relevant constructs to resilience, such as connection to social supports, that also may also contribute to risk of reutilization.²⁵

Reducing the stress of a hospitalization itself and promoting a safe transition from hospital to home is critical to improving child health outcomes. Our data here, and in our previous work, demonstrate that a history of adversity and one’s current coping ability may drive a parent’s response to a child’s hospitalization and affect their capacity to care for that child after hospital discharge.¹¹ Additional in-hospital supports like child life, behavioral health, or pastoral care could reduce the stress of the hospitalization while also building positive coping mechanisms.^26-29 A meta-analysis demonstrated that such coping interventions can help alleviate the stress of a hospitalization.³⁰ Hill et al demonstrated successful stress reduction in parents of hospitalized children using a “Coping Kit for Parents.”³¹ Further studies are warranted to understand which interventions are most effective for children and families and whether they could be more effectively deployed if the inpatient team knew more about parental ACEs.

Screening for parental ACEs could help to identify patients at highest risk for a poor transition to home. Therefore, screening for parental adversity in clinical settings, including inpatient settings, may be relevant and valuable.³² Additionally, by recognizing the high prevalence of ACEs in an inpatient setting, hospitals and healthcare organizations could be motivated to develop and enact trauma-informed approaches. A trauma-informed care approach recognizes the intersection of trauma with health and social problems. With this recognition, care teams can more sensitively address the trauma as they provide relevant services.³³ Trauma-informed care is a secondary public health prevention approach that would help team members identify the prevalence and effects of trauma via screening, recognize the signs of a maladaptive response to stress, and respond by integrating awareness of trauma into practice management.^28,34 Both the National Academy of Medicine and the Agency for Healthcare Research and Quality have called for such a trauma-informed approach in primary care.³⁵ In response, many healthcare organizations have developed trauma-informed practices to better address the needs of the populations they serve. For example, provider training on this approach has led to improved rapport in patient-provider relationships.³⁶

Although ACE awareness is a component of trauma-informed care, there are still limitations of the original ACE questionnaire developed by Felitti et al. The existing tool is not inclusive of all adversities a parent or child may face. Moreover, its focus is on past exposures and experiences and not current health-related social needs (eg, food insecurity) which have known linkages with a range of health outcomes and health disparities.³⁷ Additionally, the original ACE questionnaire was created as a population level tool and not as a screening tool. If used as a screening tool, providers may view the questions as too sensitive to ask, and parents may have difficulty responding to and understanding the relevance to their child’s care. Therefore, we suggest that more evidence is required to understand how to best adapt ACE questions into a screening processes that may be implemented in a medical setting.

More evidence is also needed to determine when and where such screening may be most useful. A primary care provider would be best equipped to screen caregivers for ACEs given their established relationship with parents and patients. Given the potential relevance of such information for inpatient care provision, information could then flow from primary care to the inpatient team. However, because not all patients have established primary care providers and only 4% of pediatricians screen for ACEs,³⁸ it is important for inpatient medical teams to understand their role in identifying and addressing ACEs during hospital stays. Development of a screening tool, with input from all stakeholders—including parents—that is valid and feasible for use in a pediatric inpatient setting would be an important step forward. This tool should be paired with training in how to discuss these topics in a trauma-informed, nonjudgmental, empathic manner. We see this as a way in which providers can more effectively elicit an accurate response while simultaneously educating parents on the relevance of such sensitive topics during an acute hospital stay. We also recommend that screening should always be paired with response capabilities that connect those who screen positive with resources that could help them to navigate the stress experienced during and after a child’s hospitalization. Furthermore, communication with primary care providers about parents that screen positive should be integrated into the transition process.

This work has several limitations. First, our study was a part of randomized controlled trials conducted in one academic setting, which thereby limits generalizability. For example, we limited our cohort to those who were English-speaking patients only. This may bias our results because respondents with limited English proficiency may have different risk profiles than their English-speaking peers. In addition, the administration of the both the ACE and resilience questionnaires occurred during an acutely stressful period, which may influence how a parent responds to these questions. Also, both of the surveys are self-reported by parents, which may be susceptible to memory and response biases. Relatedly, we had a high number of nonrespondents, particularly to the ACE questionnaire. Our results are therefore only relevant to those who chose to respond and cannot be applied to nonrespondents. Further work assessing why one does or does not respond to such sensitive questions is an important area for future inquiry. Lastly, our cohort had limited medical complexity; future studies may consider links between parental ACEs (and resilience) and morbidity experienced by children with medical complexity.

Parents history of adversity is linked to their children’s unanticipated healthcare reutilization after a hospital discharge. Screening for parental stressors during a hospitalization may be an important first step to connecting parents and children to evidence-based interventions capable of mitigating the stress of hospitalization and promoting better, more seamless transitions from hospital to home.

Group Members: The following H2O members are nonauthor contributors: JoAnne Bachus, BSN, RN; Monica Borell, BSN, RN; Lenisa V Chang, MA, PhD; Patricia Crawford, RN; Sarah Ferris, BA; Jennifer Gold, BSN, RN; Judy A Heilman, BSN, RN; Jane C Khoury, PhD; Pierce Kuhnell, MS; Karen Lawley, BSN, RN; Margo Moore, MS, BSN, RN; Lynne O’Donnell, BSN, RN; Sarah Riddle, MD; Susan N Sherman, DPA; Angela M Statile, MD, MEd; Karen P Sullivan, BSN, RN; Heather Tubbs-Cooley, PhD, RN; Susan Wade-Murphy, MSN, RN; and Christine M White, MD, MAT.

The authors also thank David Keller, MD, for his guidance on the study.

The authors have no financial relationships or conflicts of interest relevant to this article to disclose.

Supported by funds from the Academic Pediatric Young Investigator Award (Dr A Shah) and the Patient-Centered Outcomes Research Institute Award (IHS-1306-0081, to Dr K Auger, Dr S Shah, Dr H Sucharew, Dr J Simmons), the National Institutes of Health (1K23AI112916, to Dr AF Beck), and the Agency for Healthcare Research and Quality (1K12HS026393-01, to Dr A Shah, K08-HS024735- 01A1, to Dr K Auger). Dr J Haney received Summer Undergraduate Research Fellowship funding through the Summer Undergraduate Research Fellowship at Cincinnati Children’s Hospital Medical Center.

All statements in this report, including findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors, or the Methodology Committee.

Adverse Childhood Experiences, or ACEs, include exposure to abuse, neglect, or household dysfunction (eg, having a parent who is mentally ill) as a child.¹ Exposure to ACEs affects health into adulthood, with a dose-response relationship between ACEs and a range of comorbidities.¹ Adults with 6 or more ACEs have a 20-year shorter life expectancy than do those with no ACEs.¹ Still, ACEs are static; once experienced, that experience cannot be undone. However, resilience, or positive adaptation in the context of adversity, can be protective, buffering the negative effects of ACEs.^2,3 Protective factors that promote resilience include social capital, such as positive relationships with caregivers and peers.³

With their clear link to health outcomes across the life-course, there is a movement for pediatricians to screen children for ACEs⁴ and to develop strategies that promote resilience in children, parents, and families. However, screening a child for adversity has challenges because younger children may not have experienced an adverse exposure, or they may be unable to voice their experiences. Studies have demonstrated that parental adversity, or ACEs, may be a marker for childhood adversity.^5,6 Biological models also support this potential intergenerational effect of ACEs. Chronic exposure to stress, including ACEs, results in elevated cortisol via a dysregulated hypothalamic-pituitary-adrenal axis, which results in chronic inflammation.⁷ This “toxic stress” is prolonged, severe in intensity, and can lead to epigenetic changes that may be passed on to the next generation.^8,9

Hospitalization of an ill child, and the transition to home after that hospitalization, is a stressful event for children and families.¹⁰ This stress may be relevant to parents that have a history of a high rate of ACEs or a current low degree of resilience. Our previous work demonstrated that, in the inpatient setting, parents with high ACEs (≥4) or low resilience have increased coping difficulty 14 days after their child’s hospital discharge.¹¹ Our objective here was to evaluate whether a parent’s ACEs and/or resilience would also be associated with that child’s likelihood of reutilization. We hypothesized that more parental ACEs and/or lower parental resilience would be associated with revisits the emergency room, urgent care, or hospital readmissions.

Participants and Study Design

We conducted a prospective cohort study of parents of hospitalized children recruited from the “Hospital-to-Home Outcomes” Studies (H2O I and H2O II).^12,13 H2O I and II were prospective, single-center, randomized controlled trials designed to determine the effectiveness of either a nurse-led transitional home visit (H2O I) or telephone call (H2O II) on 30-day unplanned healthcare reutilization. The trials and this study were approved by the Cincinnati Children’s Institutional Review Board. All parents provided written informed consent.

Details of H2O I and II recruitment and design have been described previously.^12,13 Briefly, children were eligible for inclusion in either study if they were admitted to our institution’s general Hospital Medicine or the Hospital Medicine Complex Care Services; for H2O I, children hospitalized on the Neurology and Neurosurgery services were also eligible.^12,13 Patients were excluded if they were discharged to a residential facility, if they lived outside the home healthcare nurse service area, if they were eligible for skilled home healthcare services (eg, intravenous antibiotics), or if the participating caregiver was non-English speaking.^12,13 In H2O I, families were randomized either to receive a single nurse home visit within 96 hours of discharge or standard of care. In H2O II, families enrolled were randomized to receive a telephone call by a nurse within 96 hours of discharge or standard of care. As we have previously published, randomization in both trials successfully balanced the intervention and control arms with respect to key demographic characteristics.^12,13 For the analyses presented here, we focused on a subset of caregivers 18 years and older whose children were enrolled in either H2O I or II between August 2015 and October 2016. In both H2O trials, face-to-face and paper-based questionnaires were completed by parents during the index hospitalization.

Outcome and Predictors

Our primary outcome was unanticipated healthcare reutilization defined as return to the emergency room, urgent care, or unplanned readmission within 30 days of hospital discharge, consistent with the H2O trials. This was measured using the primary institution’s administrative data supplemented by a utilization database shared across regional hospitals.¹⁴ Readmissions were identified as “unplanned” using a previously validated algorithm,¹⁵ and treated as a dichotomous yes/no variable.

Our primary predictors were parental ACEs and resilience (see Appendix Tables). The ACE questionnaire addresses abuse, neglect, and household dysfunction in the first 18 years of life.¹ It is composed of 10 questions, each with a yes/no response.¹ We defined parents as low (ACE 0), moderate (ACE 1-3), or high (ACE ≥4) risk a priori because previous literature has described poor outcomes in adults with 4 or more ACEs.¹⁶

Given the sensitive nature of the questions, respondents independently completed the ACE questionnaire on paper instead of via the face-to-face survey. Respondents returned the completed questionnaire to the research assistant in a sealed envelope. All families received educational information on relevant hospital and community-based resources (eg, social work).

Parental resilience was measured using the Brief Resilience Scale (BRS). The BRS is 6 items, each on a 5-point Likert scale. Responses were averaged, providing a total score of 1-5; higher scores are representative of higher resilience.¹⁷ We treated the BRS score as a continuous variable. BRS has been used in clinical settings; it has demonstrated positive correlation with social support and negative correlation with fatigue.¹⁷ Parents answered BRS questions during the index pediatric hospitalization in a face-to-face interview.

Parent and Child Characteristics

Parent and child sociodemographic variables were also obtained during the face-to-face interview. Parental variables included age, gender, educational attainment, household income, employment status, and financial and social strain.¹¹ Educational attainment was analyzed in 2 categories—high school or less vs more than high school—because most discharge instructions are written at a high school reading level.¹⁸ Parents reported their annual household income in the following categories: <$15,000; $15,000-$29,999; $30,000-$44,999; $45,000-$59,999; $60,000-$89,999; $90,000-$119,999; ≥$120,000. Employment was dichotomized as not employed/student vs any employment. Financial and social strain were assessed using a series of 9 previously described questions.¹⁹ These questions assessed, via self-report, a family’s ability to make ends meet, ability to pay rent/mortgage or utilities, need to move in with others because of financial reasons, and ability to borrow money if needed, as well as home ownership and parental marital status.^15,19 Strain questions were all dichotomous (yes/no, single/not single). A composite variable was then constructed that categorized those reporting no strain items, 1 to 2 items, 3 to 4 items, and 5 or more items.²⁰

Child variables included race, ethnicity, age, primary care access,²¹ payer, and H2O treatment arm. Race categories were white/Caucasian, black/African American, American Indian or Alaskan Native, Asian or Pacific Islander, and other; ethnicity categories were Hispanic/Latino, non-Hispanic/Latino, and unknown. Given relatively low numbers of children reported to be Hispanic/Latino, we combined race and ethnicity into a single variable, categorized as non-Hispanic/white, non-Hispanic/black, and multiracial/Hispanic/other. Primary care access was assessed using the access subscale to the Parent’s Perception of Primary Care questionnaire. This includes assessment of a family’s ability to travel to their doctor, to see their doctor for routine or sick care, and to get help or advice on evenings or weekends. Scores were categorized as always adequate, almost always adequate, or sometimes/never adequate.²¹ Payer was dichotomized to private or public/self-pay.

Statistical Analyses

We examined the distribution of outcomes, predictors, and covariates. We compared sociodemographic characteristics of those respondents and nonrespondents to the ACE screen using the chi-square test for categorical variables or the t test for continuous variables. We used logistic regression to assess for associations between the independent variables of interest and reutilization, adjusting for potential confounders. To build our adjusted, multivariable model, we decided a priori to include child race/ethnicity, primary care access, financial and social strain, and trial treatment arm. We treated the H2O I control group as the referent group. Other covariates considered for inclusion were caregiver education, household income, employment, and payer. These were included in multivariable models if bivariate associations were significant at the P < .1 level. We assessed an ACE-by-resilience interaction term because we hypothesized that those with more ACEs and lower resilience may have more reutilization outcomes than parents with fewer ACEs and higher resilience. We also evaluated interaction terms between trial arm assignment and predictors to assess effects that may be introduced by the randomization. Predictors in the final logistic regression model were significant at the P < .05 level. Logistic regression assumption of little or no multicollinearity among the independent variables was verified in the final models. All analyses were performed with Stata v16 (Stata Corp, College Station, Texas).

There were a total of 1,787 parent-child dyads enrolled in the H2O I and II during the study period; 1,320 parents (74%) completed the ACE questionnaire and were included the analysis. Included parents were primarily female and employed, as well as educated beyond high school (Table 1). Overall, 64% reported one or more ACEs (range 0 to 9); 45% reported 1to 3, and 19% reported 4 or more ACEs. The most commonly reported ACEs were divorce (n = 573, 43%), exposure to alcoholism (n = 306, 23%), and exposure to mental illness (n = 281, 21%; Figure 1). Parents had a mean BRS score of 3.97 (range 1.17-5.00), with the distribution shown in Figure 2.

Of the 1,320 included patients, the average length of stay was 2.5 days, and 82% of hospitalizations were caused by acute medical issues (eg, bronchiolitis). A total of 211 children experienced a reutilization event within 30 days of discharge. In bivariate analysis, children with parents with 4 or more ACEs had a 2.02-times (95% CI 1.35-3.02) higher odds of experiencing a reutilization event than did those with parents reporting no ACEs. Parents with higher resilience scores had children with a lower odds of reutilization (odds ratio [OR] 0.77 95% CI 0.63-0.95).

In addition to our a priori variables, parental education, employment, and insurance met our significance threshold for inclusion in the multivariable model. The ACE-by-resilience interaction term was not significant and not included in the model. Similarly, there was no significant interaction between ACE and resilience and H2O treatment arm; the interaction terms were not included in the final adjusted model, but treatment arm assignment was kept as a covariate. A total of 1,292 children, out of the 1,320 respondents, remained in the final multivariable model; the excluded 28 had incomplete covariate data but were not otherwise different. In this final adjusted model, children with parents reporting 4 or more ACEs had a 1.69-times (95% CI 1.11-2.60) greater odds of reutilization than did those with parents reporting no ACEs (Table 2). Resilience failed to reach statistical significance in the adjusted model (OR 0.86, 95% CI 0.70-1.07).

We found that high-risk parents (4 or more ACEs) had children with an increased odds of healthcare reutilization, suggesting intergenerational effects of ACEs. We did not find a similar effect relating to parental resilience. We also did not find an interaction between parental ACEs and resilience, suggesting that a parent’s reported degree of resilience does not modify the effect of ACEs on reutilization risk.

Parental adversity may be a risk factor for a child’s unanticipated reutilization. We previously demonstrated that parents with 4 or more ACEs have more coping difficulty than a parent with no ACEs after a child’s hospitalization.¹¹ It is possible that parents with high adversity may have poorer coping mechanisms when dealing with a stressful situation, such as a child’s hospitalization. This may have resulted in inequitable outcomes (eg, increased reutilization) for their children. Other studies have confirmed such an intergenerational effect of adversity, linking a parent’s ACEs with poor developmental, behavioral, and health outcomes in their children.^6,22,23 O’Malley et al showed an association of parental ACEs to current adversities,²⁴ such as insurance or housing concerns, that affect the entirety of the household, including children. In short, it appears that parental ACEs may be a compelling predictor of current childhood adversity.

Resilience buffers the negative effects of ACEs; however, we did not find significant associations between resilience and reutilization or an interaction between ACEs and resilience. The factors that may contribute to reutilization are complex. In our previous work, parental resilience was associated with coping difficulty after discharge; but again, did not interact with parental ACEs.¹¹ Here, we suggest that while resilience may buffer the negative effects of ACEs, that buffering may not affect the likelihood of reutilization. It is also possible that the BRS tool is of less relevance on how one handles the stress of a child’s hospitalization. While the BRS is one measure of resilience, there are many other relevant constructs to resilience, such as connection to social supports, that also may also contribute to risk of reutilization.²⁵

Reducing the stress of a hospitalization itself and promoting a safe transition from hospital to home is critical to improving child health outcomes. Our data here, and in our previous work, demonstrate that a history of adversity and one’s current coping ability may drive a parent’s response to a child’s hospitalization and affect their capacity to care for that child after hospital discharge.¹¹ Additional in-hospital supports like child life, behavioral health, or pastoral care could reduce the stress of the hospitalization while also building positive coping mechanisms.^26-29 A meta-analysis demonstrated that such coping interventions can help alleviate the stress of a hospitalization.³⁰ Hill et al demonstrated successful stress reduction in parents of hospitalized children using a “Coping Kit for Parents.”³¹ Further studies are warranted to understand which interventions are most effective for children and families and whether they could be more effectively deployed if the inpatient team knew more about parental ACEs.

Screening for parental ACEs could help to identify patients at highest risk for a poor transition to home. Therefore, screening for parental adversity in clinical settings, including inpatient settings, may be relevant and valuable.³² Additionally, by recognizing the high prevalence of ACEs in an inpatient setting, hospitals and healthcare organizations could be motivated to develop and enact trauma-informed approaches. A trauma-informed care approach recognizes the intersection of trauma with health and social problems. With this recognition, care teams can more sensitively address the trauma as they provide relevant services.³³ Trauma-informed care is a secondary public health prevention approach that would help team members identify the prevalence and effects of trauma via screening, recognize the signs of a maladaptive response to stress, and respond by integrating awareness of trauma into practice management.^28,34 Both the National Academy of Medicine and the Agency for Healthcare Research and Quality have called for such a trauma-informed approach in primary care.³⁵ In response, many healthcare organizations have developed trauma-informed practices to better address the needs of the populations they serve. For example, provider training on this approach has led to improved rapport in patient-provider relationships.³⁶

Although ACE awareness is a component of trauma-informed care, there are still limitations of the original ACE questionnaire developed by Felitti et al. The existing tool is not inclusive of all adversities a parent or child may face. Moreover, its focus is on past exposures and experiences and not current health-related social needs (eg, food insecurity) which have known linkages with a range of health outcomes and health disparities.³⁷ Additionally, the original ACE questionnaire was created as a population level tool and not as a screening tool. If used as a screening tool, providers may view the questions as too sensitive to ask, and parents may have difficulty responding to and understanding the relevance to their child’s care. Therefore, we suggest that more evidence is required to understand how to best adapt ACE questions into a screening processes that may be implemented in a medical setting.

More evidence is also needed to determine when and where such screening may be most useful. A primary care provider would be best equipped to screen caregivers for ACEs given their established relationship with parents and patients. Given the potential relevance of such information for inpatient care provision, information could then flow from primary care to the inpatient team. However, because not all patients have established primary care providers and only 4% of pediatricians screen for ACEs,³⁸ it is important for inpatient medical teams to understand their role in identifying and addressing ACEs during hospital stays. Development of a screening tool, with input from all stakeholders—including parents—that is valid and feasible for use in a pediatric inpatient setting would be an important step forward. This tool should be paired with training in how to discuss these topics in a trauma-informed, nonjudgmental, empathic manner. We see this as a way in which providers can more effectively elicit an accurate response while simultaneously educating parents on the relevance of such sensitive topics during an acute hospital stay. We also recommend that screening should always be paired with response capabilities that connect those who screen positive with resources that could help them to navigate the stress experienced during and after a child’s hospitalization. Furthermore, communication with primary care providers about parents that screen positive should be integrated into the transition process.

This work has several limitations. First, our study was a part of randomized controlled trials conducted in one academic setting, which thereby limits generalizability. For example, we limited our cohort to those who were English-speaking patients only. This may bias our results because respondents with limited English proficiency may have different risk profiles than their English-speaking peers. In addition, the administration of the both the ACE and resilience questionnaires occurred during an acutely stressful period, which may influence how a parent responds to these questions. Also, both of the surveys are self-reported by parents, which may be susceptible to memory and response biases. Relatedly, we had a high number of nonrespondents, particularly to the ACE questionnaire. Our results are therefore only relevant to those who chose to respond and cannot be applied to nonrespondents. Further work assessing why one does or does not respond to such sensitive questions is an important area for future inquiry. Lastly, our cohort had limited medical complexity; future studies may consider links between parental ACEs (and resilience) and morbidity experienced by children with medical complexity.

Parents history of adversity is linked to their children’s unanticipated healthcare reutilization after a hospital discharge. Screening for parental stressors during a hospitalization may be an important first step to connecting parents and children to evidence-based interventions capable of mitigating the stress of hospitalization and promoting better, more seamless transitions from hospital to home.

Group Members: The following H2O members are nonauthor contributors: JoAnne Bachus, BSN, RN; Monica Borell, BSN, RN; Lenisa V Chang, MA, PhD; Patricia Crawford, RN; Sarah Ferris, BA; Jennifer Gold, BSN, RN; Judy A Heilman, BSN, RN; Jane C Khoury, PhD; Pierce Kuhnell, MS; Karen Lawley, BSN, RN; Margo Moore, MS, BSN, RN; Lynne O’Donnell, BSN, RN; Sarah Riddle, MD; Susan N Sherman, DPA; Angela M Statile, MD, MEd; Karen P Sullivan, BSN, RN; Heather Tubbs-Cooley, PhD, RN; Susan Wade-Murphy, MSN, RN; and Christine M White, MD, MAT.

The authors also thank David Keller, MD, for his guidance on the study.

The authors have no financial relationships or conflicts of interest relevant to this article to disclose.

Supported by funds from the Academic Pediatric Young Investigator Award (Dr A Shah) and the Patient-Centered Outcomes Research Institute Award (IHS-1306-0081, to Dr K Auger, Dr S Shah, Dr H Sucharew, Dr J Simmons), the National Institutes of Health (1K23AI112916, to Dr AF Beck), and the Agency for Healthcare Research and Quality (1K12HS026393-01, to Dr A Shah, K08-HS024735- 01A1, to Dr K Auger). Dr J Haney received Summer Undergraduate Research Fellowship funding through the Summer Undergraduate Research Fellowship at Cincinnati Children’s Hospital Medical Center.

All statements in this report, including findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors, or the Methodology Committee.

Adverse Childhood Experiences, or ACEs, include exposure to abuse, neglect, or household dysfunction (eg, having a parent who is mentally ill) as a child.¹ Exposure to ACEs affects health into adulthood, with a dose-response relationship between ACEs and a range of comorbidities.¹ Adults with 6 or more ACEs have a 20-year shorter life expectancy than do those with no ACEs.¹ Still, ACEs are static; once experienced, that experience cannot be undone. However, resilience, or positive adaptation in the context of adversity, can be protective, buffering the negative effects of ACEs.^2,3 Protective factors that promote resilience include social capital, such as positive relationships with caregivers and peers.³

With their clear link to health outcomes across the life-course, there is a movement for pediatricians to screen children for ACEs⁴ and to develop strategies that promote resilience in children, parents, and families. However, screening a child for adversity has challenges because younger children may not have experienced an adverse exposure, or they may be unable to voice their experiences. Studies have demonstrated that parental adversity, or ACEs, may be a marker for childhood adversity.^5,6 Biological models also support this potential intergenerational effect of ACEs. Chronic exposure to stress, including ACEs, results in elevated cortisol via a dysregulated hypothalamic-pituitary-adrenal axis, which results in chronic inflammation.⁷ This “toxic stress” is prolonged, severe in intensity, and can lead to epigenetic changes that may be passed on to the next generation.^8,9

Hospitalization of an ill child, and the transition to home after that hospitalization, is a stressful event for children and families.¹⁰ This stress may be relevant to parents that have a history of a high rate of ACEs or a current low degree of resilience. Our previous work demonstrated that, in the inpatient setting, parents with high ACEs (≥4) or low resilience have increased coping difficulty 14 days after their child’s hospital discharge.¹¹ Our objective here was to evaluate whether a parent’s ACEs and/or resilience would also be associated with that child’s likelihood of reutilization. We hypothesized that more parental ACEs and/or lower parental resilience would be associated with revisits the emergency room, urgent care, or hospital readmissions.

Participants and Study Design

We conducted a prospective cohort study of parents of hospitalized children recruited from the “Hospital-to-Home Outcomes” Studies (H2O I and H2O II).^12,13 H2O I and II were prospective, single-center, randomized controlled trials designed to determine the effectiveness of either a nurse-led transitional home visit (H2O I) or telephone call (H2O II) on 30-day unplanned healthcare reutilization. The trials and this study were approved by the Cincinnati Children’s Institutional Review Board. All parents provided written informed consent.

Details of H2O I and II recruitment and design have been described previously.^12,13 Briefly, children were eligible for inclusion in either study if they were admitted to our institution’s general Hospital Medicine or the Hospital Medicine Complex Care Services; for H2O I, children hospitalized on the Neurology and Neurosurgery services were also eligible.^12,13 Patients were excluded if they were discharged to a residential facility, if they lived outside the home healthcare nurse service area, if they were eligible for skilled home healthcare services (eg, intravenous antibiotics), or if the participating caregiver was non-English speaking.^12,13 In H2O I, families were randomized either to receive a single nurse home visit within 96 hours of discharge or standard of care. In H2O II, families enrolled were randomized to receive a telephone call by a nurse within 96 hours of discharge or standard of care. As we have previously published, randomization in both trials successfully balanced the intervention and control arms with respect to key demographic characteristics.^12,13 For the analyses presented here, we focused on a subset of caregivers 18 years and older whose children were enrolled in either H2O I or II between August 2015 and October 2016. In both H2O trials, face-to-face and paper-based questionnaires were completed by parents during the index hospitalization.

Outcome and Predictors

Our primary outcome was unanticipated healthcare reutilization defined as return to the emergency room, urgent care, or unplanned readmission within 30 days of hospital discharge, consistent with the H2O trials. This was measured using the primary institution’s administrative data supplemented by a utilization database shared across regional hospitals.¹⁴ Readmissions were identified as “unplanned” using a previously validated algorithm,¹⁵ and treated as a dichotomous yes/no variable.

Our primary predictors were parental ACEs and resilience (see Appendix Tables). The ACE questionnaire addresses abuse, neglect, and household dysfunction in the first 18 years of life.¹ It is composed of 10 questions, each with a yes/no response.¹ We defined parents as low (ACE 0), moderate (ACE 1-3), or high (ACE ≥4) risk a priori because previous literature has described poor outcomes in adults with 4 or more ACEs.¹⁶

Given the sensitive nature of the questions, respondents independently completed the ACE questionnaire on paper instead of via the face-to-face survey. Respondents returned the completed questionnaire to the research assistant in a sealed envelope. All families received educational information on relevant hospital and community-based resources (eg, social work).

Parental resilience was measured using the Brief Resilience Scale (BRS). The BRS is 6 items, each on a 5-point Likert scale. Responses were averaged, providing a total score of 1-5; higher scores are representative of higher resilience.¹⁷ We treated the BRS score as a continuous variable. BRS has been used in clinical settings; it has demonstrated positive correlation with social support and negative correlation with fatigue.¹⁷ Parents answered BRS questions during the index pediatric hospitalization in a face-to-face interview.

Parent and Child Characteristics

Parent and child sociodemographic variables were also obtained during the face-to-face interview. Parental variables included age, gender, educational attainment, household income, employment status, and financial and social strain.¹¹ Educational attainment was analyzed in 2 categories—high school or less vs more than high school—because most discharge instructions are written at a high school reading level.¹⁸ Parents reported their annual household income in the following categories: <$15,000; $15,000-$29,999; $30,000-$44,999; $45,000-$59,999; $60,000-$89,999; $90,000-$119,999; ≥$120,000. Employment was dichotomized as not employed/student vs any employment. Financial and social strain were assessed using a series of 9 previously described questions.¹⁹ These questions assessed, via self-report, a family’s ability to make ends meet, ability to pay rent/mortgage or utilities, need to move in with others because of financial reasons, and ability to borrow money if needed, as well as home ownership and parental marital status.^15,19 Strain questions were all dichotomous (yes/no, single/not single). A composite variable was then constructed that categorized those reporting no strain items, 1 to 2 items, 3 to 4 items, and 5 or more items.²⁰

Child variables included race, ethnicity, age, primary care access,²¹ payer, and H2O treatment arm. Race categories were white/Caucasian, black/African American, American Indian or Alaskan Native, Asian or Pacific Islander, and other; ethnicity categories were Hispanic/Latino, non-Hispanic/Latino, and unknown. Given relatively low numbers of children reported to be Hispanic/Latino, we combined race and ethnicity into a single variable, categorized as non-Hispanic/white, non-Hispanic/black, and multiracial/Hispanic/other. Primary care access was assessed using the access subscale to the Parent’s Perception of Primary Care questionnaire. This includes assessment of a family’s ability to travel to their doctor, to see their doctor for routine or sick care, and to get help or advice on evenings or weekends. Scores were categorized as always adequate, almost always adequate, or sometimes/never adequate.²¹ Payer was dichotomized to private or public/self-pay.

Statistical Analyses

We examined the distribution of outcomes, predictors, and covariates. We compared sociodemographic characteristics of those respondents and nonrespondents to the ACE screen using the chi-square test for categorical variables or the t test for continuous variables. We used logistic regression to assess for associations between the independent variables of interest and reutilization, adjusting for potential confounders. To build our adjusted, multivariable model, we decided a priori to include child race/ethnicity, primary care access, financial and social strain, and trial treatment arm. We treated the H2O I control group as the referent group. Other covariates considered for inclusion were caregiver education, household income, employment, and payer. These were included in multivariable models if bivariate associations were significant at the P < .1 level. We assessed an ACE-by-resilience interaction term because we hypothesized that those with more ACEs and lower resilience may have more reutilization outcomes than parents with fewer ACEs and higher resilience. We also evaluated interaction terms between trial arm assignment and predictors to assess effects that may be introduced by the randomization. Predictors in the final logistic regression model were significant at the P < .05 level. Logistic regression assumption of little or no multicollinearity among the independent variables was verified in the final models. All analyses were performed with Stata v16 (Stata Corp, College Station, Texas).

There were a total of 1,787 parent-child dyads enrolled in the H2O I and II during the study period; 1,320 parents (74%) completed the ACE questionnaire and were included the analysis. Included parents were primarily female and employed, as well as educated beyond high school (Table 1). Overall, 64% reported one or more ACEs (range 0 to 9); 45% reported 1to 3, and 19% reported 4 or more ACEs. The most commonly reported ACEs were divorce (n = 573, 43%), exposure to alcoholism (n = 306, 23%), and exposure to mental illness (n = 281, 21%; Figure 1). Parents had a mean BRS score of 3.97 (range 1.17-5.00), with the distribution shown in Figure 2.

Of the 1,320 included patients, the average length of stay was 2.5 days, and 82% of hospitalizations were caused by acute medical issues (eg, bronchiolitis). A total of 211 children experienced a reutilization event within 30 days of discharge. In bivariate analysis, children with parents with 4 or more ACEs had a 2.02-times (95% CI 1.35-3.02) higher odds of experiencing a reutilization event than did those with parents reporting no ACEs. Parents with higher resilience scores had children with a lower odds of reutilization (odds ratio [OR] 0.77 95% CI 0.63-0.95).

In addition to our a priori variables, parental education, employment, and insurance met our significance threshold for inclusion in the multivariable model. The ACE-by-resilience interaction term was not significant and not included in the model. Similarly, there was no significant interaction between ACE and resilience and H2O treatment arm; the interaction terms were not included in the final adjusted model, but treatment arm assignment was kept as a covariate. A total of 1,292 children, out of the 1,320 respondents, remained in the final multivariable model; the excluded 28 had incomplete covariate data but were not otherwise different. In this final adjusted model, children with parents reporting 4 or more ACEs had a 1.69-times (95% CI 1.11-2.60) greater odds of reutilization than did those with parents reporting no ACEs (Table 2). Resilience failed to reach statistical significance in the adjusted model (OR 0.86, 95% CI 0.70-1.07).

We found that high-risk parents (4 or more ACEs) had children with an increased odds of healthcare reutilization, suggesting intergenerational effects of ACEs. We did not find a similar effect relating to parental resilience. We also did not find an interaction between parental ACEs and resilience, suggesting that a parent’s reported degree of resilience does not modify the effect of ACEs on reutilization risk.

Parental adversity may be a risk factor for a child’s unanticipated reutilization. We previously demonstrated that parents with 4 or more ACEs have more coping difficulty than a parent with no ACEs after a child’s hospitalization.¹¹ It is possible that parents with high adversity may have poorer coping mechanisms when dealing with a stressful situation, such as a child’s hospitalization. This may have resulted in inequitable outcomes (eg, increased reutilization) for their children. Other studies have confirmed such an intergenerational effect of adversity, linking a parent’s ACEs with poor developmental, behavioral, and health outcomes in their children.^6,22,23 O’Malley et al showed an association of parental ACEs to current adversities,²⁴ such as insurance or housing concerns, that affect the entirety of the household, including children. In short, it appears that parental ACEs may be a compelling predictor of current childhood adversity.

Resilience buffers the negative effects of ACEs; however, we did not find significant associations between resilience and reutilization or an interaction between ACEs and resilience. The factors that may contribute to reutilization are complex. In our previous work, parental resilience was associated with coping difficulty after discharge; but again, did not interact with parental ACEs.¹¹ Here, we suggest that while resilience may buffer the negative effects of ACEs, that buffering may not affect the likelihood of reutilization. It is also possible that the BRS tool is of less relevance on how one handles the stress of a child’s hospitalization. While the BRS is one measure of resilience, there are many other relevant constructs to resilience, such as connection to social supports, that also may also contribute to risk of reutilization.²⁵

Reducing the stress of a hospitalization itself and promoting a safe transition from hospital to home is critical to improving child health outcomes. Our data here, and in our previous work, demonstrate that a history of adversity and one’s current coping ability may drive a parent’s response to a child’s hospitalization and affect their capacity to care for that child after hospital discharge.¹¹ Additional in-hospital supports like child life, behavioral health, or pastoral care could reduce the stress of the hospitalization while also building positive coping mechanisms.^26-29 A meta-analysis demonstrated that such coping interventions can help alleviate the stress of a hospitalization.³⁰ Hill et al demonstrated successful stress reduction in parents of hospitalized children using a “Coping Kit for Parents.”³¹ Further studies are warranted to understand which interventions are most effective for children and families and whether they could be more effectively deployed if the inpatient team knew more about parental ACEs.

Screening for parental ACEs could help to identify patients at highest risk for a poor transition to home. Therefore, screening for parental adversity in clinical settings, including inpatient settings, may be relevant and valuable.³² Additionally, by recognizing the high prevalence of ACEs in an inpatient setting, hospitals and healthcare organizations could be motivated to develop and enact trauma-informed approaches. A trauma-informed care approach recognizes the intersection of trauma with health and social problems. With this recognition, care teams can more sensitively address the trauma as they provide relevant services.³³ Trauma-informed care is a secondary public health prevention approach that would help team members identify the prevalence and effects of trauma via screening, recognize the signs of a maladaptive response to stress, and respond by integrating awareness of trauma into practice management.^28,34 Both the National Academy of Medicine and the Agency for Healthcare Research and Quality have called for such a trauma-informed approach in primary care.³⁵ In response, many healthcare organizations have developed trauma-informed practices to better address the needs of the populations they serve. For example, provider training on this approach has led to improved rapport in patient-provider relationships.³⁶

Although ACE awareness is a component of trauma-informed care, there are still limitations of the original ACE questionnaire developed by Felitti et al. The existing tool is not inclusive of all adversities a parent or child may face. Moreover, its focus is on past exposures and experiences and not current health-related social needs (eg, food insecurity) which have known linkages with a range of health outcomes and health disparities.³⁷ Additionally, the original ACE questionnaire was created as a population level tool and not as a screening tool. If used as a screening tool, providers may view the questions as too sensitive to ask, and parents may have difficulty responding to and understanding the relevance to their child’s care. Therefore, we suggest that more evidence is required to understand how to best adapt ACE questions into a screening processes that may be implemented in a medical setting.

More evidence is also needed to determine when and where such screening may be most useful. A primary care provider would be best equipped to screen caregivers for ACEs given their established relationship with parents and patients. Given the potential relevance of such information for inpatient care provision, information could then flow from primary care to the inpatient team. However, because not all patients have established primary care providers and only 4% of pediatricians screen for ACEs,³⁸ it is important for inpatient medical teams to understand their role in identifying and addressing ACEs during hospital stays. Development of a screening tool, with input from all stakeholders—including parents—that is valid and feasible for use in a pediatric inpatient setting would be an important step forward. This tool should be paired with training in how to discuss these topics in a trauma-informed, nonjudgmental, empathic manner. We see this as a way in which providers can more effectively elicit an accurate response while simultaneously educating parents on the relevance of such sensitive topics during an acute hospital stay. We also recommend that screening should always be paired with response capabilities that connect those who screen positive with resources that could help them to navigate the stress experienced during and after a child’s hospitalization. Furthermore, communication with primary care providers about parents that screen positive should be integrated into the transition process.

This work has several limitations. First, our study was a part of randomized controlled trials conducted in one academic setting, which thereby limits generalizability. For example, we limited our cohort to those who were English-speaking patients only. This may bias our results because respondents with limited English proficiency may have different risk profiles than their English-speaking peers. In addition, the administration of the both the ACE and resilience questionnaires occurred during an acutely stressful period, which may influence how a parent responds to these questions. Also, both of the surveys are self-reported by parents, which may be susceptible to memory and response biases. Relatedly, we had a high number of nonrespondents, particularly to the ACE questionnaire. Our results are therefore only relevant to those who chose to respond and cannot be applied to nonrespondents. Further work assessing why one does or does not respond to such sensitive questions is an important area for future inquiry. Lastly, our cohort had limited medical complexity; future studies may consider links between parental ACEs (and resilience) and morbidity experienced by children with medical complexity.

Parents history of adversity is linked to their children’s unanticipated healthcare reutilization after a hospital discharge. Screening for parental stressors during a hospitalization may be an important first step to connecting parents and children to evidence-based interventions capable of mitigating the stress of hospitalization and promoting better, more seamless transitions from hospital to home.

Group Members: The following H2O members are nonauthor contributors: JoAnne Bachus, BSN, RN; Monica Borell, BSN, RN; Lenisa V Chang, MA, PhD; Patricia Crawford, RN; Sarah Ferris, BA; Jennifer Gold, BSN, RN; Judy A Heilman, BSN, RN; Jane C Khoury, PhD; Pierce Kuhnell, MS; Karen Lawley, BSN, RN; Margo Moore, MS, BSN, RN; Lynne O’Donnell, BSN, RN; Sarah Riddle, MD; Susan N Sherman, DPA; Angela M Statile, MD, MEd; Karen P Sullivan, BSN, RN; Heather Tubbs-Cooley, PhD, RN; Susan Wade-Murphy, MSN, RN; and Christine M White, MD, MAT.

The authors also thank David Keller, MD, for his guidance on the study.

The authors have no financial relationships or conflicts of interest relevant to this article to disclose.

Supported by funds from the Academic Pediatric Young Investigator Award (Dr A Shah) and the Patient-Centered Outcomes Research Institute Award (IHS-1306-0081, to Dr K Auger, Dr S Shah, Dr H Sucharew, Dr J Simmons), the National Institutes of Health (1K23AI112916, to Dr AF Beck), and the Agency for Healthcare Research and Quality (1K12HS026393-01, to Dr A Shah, K08-HS024735- 01A1, to Dr K Auger). Dr J Haney received Summer Undergraduate Research Fellowship funding through the Summer Undergraduate Research Fellowship at Cincinnati Children’s Hospital Medical Center.

All statements in this report, including findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors, or the Methodology Committee.

A 73-year-old man presented to primary care for an annual examination. Four days prior, he noted right-sided sharp jaw pain such that he could not open his mouth nor chew solid food; it radiated from the right mandible to the ipsilateral temple. He also noted bilateral aching hip pain for several years that increased in severity in the prior 2 months. He reported an intentional weight loss of 9 kg over the past year, achieved through dietary modification. He denied fever, chills, and visual disturbance.

Acute onset of unilateral jaw pain that is worsened by chewing is a feature consistent with a temporomandibular disorder (TMD). TMD consists of musculoskeletal and neuromuscular conditions that affect the temporomandibular joints (TMJs), masticatory muscles, and associated tissues. Common symptoms of TMD include facial or ear pain, temporal headache, and TMJ dysfunction or discomfort. In addition to TMD, craniofacial pain has many possible etiologies such as dental pathology, neuralgias, sinus and otologic disorders, headache and migraine disorders, infections, rheumatologic conditions, and neoplasms.

Systemic etiologies for this patient’s symptoms are a consideration given his age and concomitant worsening of chronic hip pain. Rheumatologic conditions such as giant cell arteritis (GCA) and polymyalgia rheumatica (PMR) are more common in adults older than 50 years of age and cause headache, jaw claudication, and pelvic girdle pain. Rarely, hematologic malignancies (eg, lymphoma), solid tumor metastases (eg, breast cancer, melanoma), and primary tumors of the head and neck (eg, nasopharyngeal carcinoma) can involve the mandible, TMJ, or parotid gland and result in symptoms of TMD.

Medical history was notable for hypertension and type 2 diabetes mellitus complicated by peripheral neuropathy. He smoked one pack of cigarettes daily for 40 years but quit 15 years prior. He drank 4 ounces of vodka each night.

On examination, temperature was 36.5°C, heart rate 92 beats per minute, blood pressure 127/60 mmHg, respiratory rate 12 breaths per minute, oxygen saturation 98% on ambient air, and weight 118 kg. Extraocular movements were intact, pupils were equal and reactive to light and accommodation, and there were no visual field deficits. Nondilated funduscopic examination revealed normal blood vessels, optic disc, and optic cup-to-disc ratio. Dentition was good with pink gingiva. Bilateral temples were nontender. There was normal range of motion and strength in the shoulders, hips, and lower extremities with no tenderness over the trochanters. Patellar and ankle reflexes were present and symmetric bilaterally. He had no rashes or ecchymoses.

The history of smoking, especially with concomitant alcohol intake, is a risk factor for head and neck cancer, and these malignancies can lead to facial pain. While the normal oral cavity exam argues against localized oral and dental causes of the patient’s symptoms, direct fiberoptic endoscopy should be considered. The neck should be examined for lymphadenopathy. Normal vital signs point away from severe infection. The lack of findings in the head and musculoskeletal regions does not exclude systemic etiologies such as rheumatologic conditions or neoplasm. Complete blood cell count and markers of inflammation including erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels should be obtained. Hip and pelvic radiographs should be obtained to evaluate for hip osteoarthritis, fractures, or osseous lesions.

The appointment occurred during evening hours and the patient declined further evaluation until the following morning, at which time laboratory studies revealed normal serum levels of electrolytes, blood urea nitrogen, and creatinine. White blood cell (WBC) count was 6,800/mm³ with an immature granulocyte ratio of 1.8% (normal, 0.0-0.5%), hemoglobin 13.2 g/dL, and platelet count 163,000/mm³. ESR was 118 mm/hr (normal, 0-15 mm/hr) and CRP was 1.5 mg/dL (normal, 0-0.75 mg/dL). Radiographs of the hips and pelvis showed osteoarthritis of the bilateral hip joints and degenerative disc disease of the lower lumbar spine.

Granulocytosis may occur in response to infection, rheumatologic conditions, and hematologic malignancies such as chronic myelogenous leukemia. While infectious etiologies (eg, abscess, osteomyelitis) are the most common cause of an extremely elevated ESR level, this patient does not have other signs or symptoms of infection such as fever or leukocytosis. Therefore, other common causes for an extremely elevated ESR level should be considered, including malignancy (eg, multiple myeloma, lymphoma, metastatic solid tumor) and autoimmune conditions (eg, rheumatoid arthritis, vasculitis). While multiple myeloma is the most common malignant etiology for extremely elevated ESR, the patient lacks signs of this condition such as anemia, elevated creatinine, or osteolytic lesions on radiographic imaging. Osteoarthritis identified on the radiographs may contribute to the patient’s hip pain but would not explain the patient’s jaw pain, weight loss, granulocytosis, and elevated ESR. These findings, taken together with the patient’s age, are most suggestive of GCA with possible coexisting PMR. Temporal artery biopsy should be obtained as it is the gold standard test for diagnosing GCA.

The patient was contacted by telephone that same day with laboratory test results. During the call, he endorsed increased jaw and temple pain. He was advised to proceed to the emergency department (ED) for timely evaluation and treatment.

Because GCA was being considered, ophthalmology performed an ocular examination in the ED, which demonstrated no signs of optic nerve or retinal ischemia. Computed tomography (CT) scan of the head and neck with intravenous contrast revealed no abscess or soft tissue abnormalities. Right temporal artery biopsy was performed.

The normal ocular examination does not exclude GCA, and temporal artery biopsy is appropriate. The mainstay of treatment for GCA is high-dose systemic glucocorticoids, which should not be withheld while awaiting biopsy results since ophthalmic artery inflammation may occur and threaten vision.

While GCA remains the leading diagnosis, malignant etiologies warrant further consideration because they are a common cause of extreme ESR elevation, particularly among older patients. The patient’s cancer screening history should be reviewed. The normal CT scan of the head and neck reduces the likelihood of localized solid tumor etiologies; however, additional CT imaging of the chest, abdomen, and pelvis is warranted to evaluate for metastatic solid tumors or lymphoma.

A 10-day course of prednisone 60 mg daily was prescribed for empiric treatment of GCA. The patient was discharged home with follow-up scheduled in rheumatology and primary care clinics. Pain in the jaw and temple resolved within several days.

Two weeks later, he presented to the rheumatology clinic. He noted 1 week of lower right back pain described as dull, aching, radiating to the lateral right hip, and occurring when transitioning from sitting to standing. He had no leg numbness, weakness, or change in bowel habits. Bladder habits were also unchanged, although he reported chronic urinary frequency and occasional incontinence. He reported further weight loss, this time an unintentional loss of 9 kg. He noted frequent sweating but no fever.

He reported a normal colonoscopy within the prior 5 years. Because these records were not available for review, a fecal immunochemical test was obtained and negative for hemoglobin. He had previously declined prostate cancer screening.

The resolution of jaw and temple pain with prednisone supports the presumed diagnosis of GCA. Up to half of patients with GCA may also have PMR, which can cause aching and stiffness in the arms, hips, and lumbar region, and pain may be abrupt in onset. However, PMR-related pain would be expected to improve rather than develop or worsen in the setting of high-dose glucocorticoid use. Therefore, other causes of acute-onset back pain must be considered.

While localized musculoskeletal etiologies such as lumbar muscle strain, radiculopathy, and vertebral compression fracture are possible, co-occurrence of unintentional weight loss and diaphoresis with elevated inflammatory markers suggests a systemic etiology. A neoplastic process with bony metastasis is possible. The reportedly normal colonoscopy and the negative fecal immunochemical test make colorectal cancer less likely. Inflammatory conditions such as ankylosing spondylitis and rheumatoid arthritis are also possible. Ankylosing spondylitis usually presents at a much younger age, however, and axial skeletal involvement in rheumatoid arthritis often involves the cervical spine and is usually seen after longstanding disease. Additionally, the hallmark of inflammatory back pain is morning stiffness which the patient does not endorse. Nonetheless, additional laboratory testing should include antinuclear antibody, rheumatoid factor, and anti-cyclic citrullinated peptide (anti-CCP) antibody. Vertebral osteomyelitis remains on the differential diagnosis, and repeat WBC count and inflammatory markers should be assessed. Lumbosacral radiographs should be obtained to rule out fracture.

Physical examination in the rheumatology clinic revealed a temperature of 37.0°C, heart rate 100 beats per minute, blood pressure 146/72 mmHg, respiratory rate 12 breaths per minute, and oxygen saturation 98% on ambient air. Weight was 109 kg. He was pale and diaphoretic. There was diffuse tenderness to palpation of the right-sided lumbar paraspinal muscles. Straight leg raise was negative bilaterally. Patellar reflexes and gait were normal.

Blood chemistries, renal function, and aminotransferase levels were normal. WBC count was 7,100/mm³, hemoglobin 8.0 g/dL, mean corpuscular volume 88.9 fL, platelet count 128,000/mm³, ESR 66 mm/hr, CRP 0.57 mg/dL, alkaline phosphatase 438 IU/L (normal, 30-130 IU/L), and thyroid-stimulating hormone 0.925 mU/L (normal, 0.34-5.60 mU/L). Testing for antinuclear antibodies, rheumatoid factor, and anti-CCP antibody was unremarkable. Prostate-specific antigen (PSA) level was 2.2 ng/mL (normal, 0-4 ng/mL). Urinalysis was unremarkable. Antibodies to hepatitis C and Treponema pallidum were negative. Interferon gamma release assay was negative.

Findings of new onset anemia and thrombocytopenia, in combination with elevated ESR and alkaline phosphatase level, are concerning for disseminated intravascular coagulation (DIC) and microangiopathic hemolytic anemia (MAHA), bone marrow infiltration of a metastatic neoplasm, or ineffective hematopoiesis caused by myelodysplastic syndromes or myelofibrosis.

Laboratory evaluation should include iron studies, lactate dehydrogenase (LDH), haptoglobin, fibrinogen, D-dimer, reticulocyte count, and peripheral blood smear to assess for hemolysis and erythrocyte morphology. Advanced imaging with lumbosacral magnetic resonance imaging (MRI) should be obtained to evaluate for focal etiologies of back pain such as disc herniation, abscess, marrow infiltration, and infarction.

Additional laboratory studies revealed a gamma-glutamyl transferase level of 49 IU/L (normal, 8-56 IU/L), LDH 288 IU/L (normal, 98-192 IU/L), haptoglobin 495 mg/dL (normal, 32-240 mg/dL), fibrinogen >700 mg/dL (normal, 225-550 mg/dL), D-dimer 693 ng/mL (normal, 200-250 ng/mL), serum iron 57 mcg/dL (normal, 33-150 mcg/dL), total iron binding capacity 286 mcg/dL (normal, 250-450 mcg/dL), ferritin 1,012 ng/mL (normal, 17.9-464 ng/mL), and reticulocyte count 2.9% (normal, 0.5-2.5%). Coagulation studies and serum protein electrophoresis were normal. Erythropoietin level was 109 mIU/mL (normal, 4.0-20.0 mIU/mL). Peripheral blood smear demonstrated moderate anemia with 8% nucleated erythrocytes per white blood cell (normal, 0%) and no circulating blasts.

MRI of the thoracolumbar spine and pelvis revealed diffusely abnormal bone marrow signal with multiple superimposed focal and poorly defined enhancing lesions along the lumbar spine marrow, sacrum, and bilateral iliac bones (Figure 1). Positron emission tomography/computed tomography (PET/CT) scan showed no scintigraphic evidence of metabolically active neoplastic, paraneoplastic, or inflammatory disorder.

The elevated haptoglobin, normal coagulation studies, and absence of fragmented erythrocytes on peripheral smear exclude an intravascular hemolytic process. The patient’s lower than expected reticulocyte count for the degree of anemia, elevated erythropoietin, and nucleated erythrocytes constitute a pattern that can be seen with bone marrow infiltration. There are no circulating blasts, making leukemia less likely. A solid organ tumor with bone metastases may cause enhancing lesions on MRI since this form of imaging is more sensitive than radiography for detecting skeletal malignancies. The negative PET/CT, however, does not reveal a primary tumor. Myelofibrosis is an infiltrative myeloproliferative disorder associated with nonspecific laboratory abnormalities, bone pain, weight loss, and night sweats that could cause diffuse MRI bone marrow signal alterations with normal PET/CT findings. However, myelofibrosis would not typically cause a significantly elevated ESR, and thus would be an unlikely cause for this patient’s presentation.

Given the constellation of symptoms, hematologic abnormalities, and bone marrow infiltration on imaging, hematology should be consulted to perform a bone marrow biopsy to assist with definitive diagnosis.

Bone marrow biopsy demonstrated metastatic adenocarcinoma consistent with prostatic origin (Figure 2). Bone scan demonstrated widespread osteoblastic metastases, which included the skull and temporal regions. These lesions were thought to be the cause of the patient’s original presenting symptom of jaw pain.

The patient was started on androgen deprivation therapy, initially with degarelix and subsequently leuprolide shots and abiraterone with prednisone. PSA was 0.08 ng/mL after 3 months of androgen deprivation therapy. His back and hip pain slowly improved.

Prostate cancer is the most common cancer in men with one out of every nine men diagnosed in his lifetime.¹ While most men initially present with localized, curable disease,¹ 4% present with metastatic disease, an incidence that has been increasing since 2004.² Despite available treatments, metastatic prostate cancer has a poor prognosis, with an average overall survival of approximately 5 years.³

Prostate cancer can be challenging to diagnose. Men with prostate cancer are commonly asymptomatic. Rarely, patients may present with hematuria, bony pain caused by metastasis, or obstructive urinary symptoms like hesitancy or incomplete bladder emptying. Our patient presented with jaw pain, which was ultimately attributed to osteoblastic lesions of the skull. Additionally, his history of urinary frequency and incontinence may have been clues to his underlying diagnosis of prostate cancer.

Prostate cancer screening remains highly nuanced and relies on shared decision-making between patients and healthcare providers. Clinical practice guidelines for early detection of prostate cancer recommend individualized PSA-based serologic screening.^4,5 Specifically, the United States Preventive Services Task Force recommends screening men aged 55 to 69 years who desire screening and understand the potential harms associated with a positive test result. These harms may include psychological distress and complications from prostate biopsy (eg, pain or infection) or prostate cancer treatment (eg, erectile, urinary, and/or bowel dysfunction).^4-6 The decision to screen can be guided by individuals’ risk factors including African American race, family history, and older age.

While our patient elected not to undergo routine prostate cancer screening, a PSA level was obtained during his diagnostic evaluation and highlights the limitations of PSA-based screening. A PSA level ≤4.0 ng/mL has 21% sensitivity and 91% specificity for detecting prostate cancer.⁷ PSA levels above 4.0 ng/mL warrant repeat testing and, if persistently elevated, referral to urology for possible prostate biopsy. PSA levels often correlate with burden of disease, and patients with PSA levels >20 ng/mL are referred for CT imaging to evaluate for metastatic disease.⁸ PSA’s poor sensitivity was underscored in a study by Thompson et al who evaluated the incidence of prostate cancer in men participating in the Prostate Cancer Prevention Trial with PSA levels of <4 ng/mL.⁹ In this study, 15% of men diagnosed with prostate cancer never had a PSA level >4 ng/mL.⁹ While most of the cancers in this study were low grade and may have been clinically insignificant, 15% demonstrated histologic signs of at least intermediate-risk disease. Our patient’s PSA level of 2.2 ng/mL was below the threshold that triggers additional evaluation even though he had widely metastatic prostate cancer.

Our patient’s severe jaw and temple pain, weight loss, and progressive hip pain were concerning for GCA. This vasculitis of large- and medium-sized arteries predominantly affects older adults with greatest incidence among those 70 years of age and older.¹⁰ Symptoms occur because of cranial artery inflammation and may include headache, visual disturbance, erythema or tenderness of the temporal artery, and jaw claudication. Extracranial inflammation may affect the thoracic aorta and its branches and rarely the abdominal aorta and lower limb arteries. Pelvic girdle pain more typically results from associated PMR. Patients may also note systemic symptoms such as fever, weight loss, and fatigue.

Prompt diagnostic testing is important when considering GCA. Most patients with GCA have ESR levels greater than 40 mm/hr.¹¹ ESR is a laboratory test that measures the vertical distance erythrocytes travel in a column of blood over 1 hour; in the setting of inflammation, cells form clumps and travel more quickly than individual cells, resulting in a higher value. While moderate elevations in ESR may occur without an identifiable cause, extreme ESR levels—those above 100 mm/hr, as observed in our patient—are highly suggestive of certain serious conditions, including infection, malignancy, and autoimmune disease such as GCA.^12,13 Temporal artery biopsy is the gold standard test to diagnose GCA. However, because of noncontiguous inflammation of the temporal artery, biopsies may be falsely negative. Thus, sampling of the contralateral temporal artery may be warranted if suspicion remains high.

As was the case for our patient, PET/CT is not reliable for diagnosing prostate cancer. In contrast to other malignancies (eg, lymphoma, lung cancer), prostate cancer typically does not display increased glucose metabolism. Moreover, the close proximity of the bladder and prostate can interfere with imaging interpretation because the fluorodeoxyglucose (FDG) tracer is excreted in the urine.¹⁴ The reported sensitivity of PET/CT for the diagnosis of prostate cancer ranges from 17%-65%.^15,16 In a small study of men with metastatic prostate cancer, only 18% of bony metastases were FDG avid, and there was no correlation between FDG avidity and PSA level.¹⁵ Notably, although PET/CT includes CT imaging, this CT is used to map anatomic landmarks and is not separately interpreted by the radiologist. Thus, even if evidence of prostate cancer was apparent on traditional CT, it may be overlooked on PET/CT.

Several important points regarding diagnostic testing are raised by this case. First, PSA-based screening for prostate cancer may be falsely negative, even in the setting of widely metastatic disease. Second, extreme ESR elevation is a marker for serious underlying disease and warrants a thorough diagnostic evaluation. Finally, PET/CT has limited diagnostic utility in evaluating metastatic prostate cancer because of the normal rates of glucose metabolism. Our patient initially presented with jaw pain, yet his progressive physical symptoms and laboratory abnormalities prompted an evaluation which ultimately revealed the jaw-dropping diagnosis of PSA-negative, metastatic prostate cancer.

• ESR levels greater than 100 mm/hr are highly suggestive of certain serious conditions including infection, autoimmune disease, and malignancy.
• PSA-based screening for prostate cancer can result in false negative test results. In one study, 15% of men diagnosed with prostate cancer never had a PSA level greater than 4 ng/mL (ie, the level at which repeat laboratory testing and/or referral to urology for possible prostate biopsy is advisable).
• PET/CT has limited diagnostic utility in evaluating metastatic prostate cancer, because prostate cancer cells typically demonstrate normal glucose metabolism.

Drs Griauzde, Northway, Yentz, and Houchens have nothing to disclose. Dr Saint reports personal fees from ISMIE Mutual Insurance Company during the conduct of the study, as well as personal fees from Jvion and Doximity outside the submitted work.

A 73-year-old man presented to primary care for an annual examination. Four days prior, he noted right-sided sharp jaw pain such that he could not open his mouth nor chew solid food; it radiated from the right mandible to the ipsilateral temple. He also noted bilateral aching hip pain for several years that increased in severity in the prior 2 months. He reported an intentional weight loss of 9 kg over the past year, achieved through dietary modification. He denied fever, chills, and visual disturbance.

Acute onset of unilateral jaw pain that is worsened by chewing is a feature consistent with a temporomandibular disorder (TMD). TMD consists of musculoskeletal and neuromuscular conditions that affect the temporomandibular joints (TMJs), masticatory muscles, and associated tissues. Common symptoms of TMD include facial or ear pain, temporal headache, and TMJ dysfunction or discomfort. In addition to TMD, craniofacial pain has many possible etiologies such as dental pathology, neuralgias, sinus and otologic disorders, headache and migraine disorders, infections, rheumatologic conditions, and neoplasms.

Systemic etiologies for this patient’s symptoms are a consideration given his age and concomitant worsening of chronic hip pain. Rheumatologic conditions such as giant cell arteritis (GCA) and polymyalgia rheumatica (PMR) are more common in adults older than 50 years of age and cause headache, jaw claudication, and pelvic girdle pain. Rarely, hematologic malignancies (eg, lymphoma), solid tumor metastases (eg, breast cancer, melanoma), and primary tumors of the head and neck (eg, nasopharyngeal carcinoma) can involve the mandible, TMJ, or parotid gland and result in symptoms of TMD.

Medical history was notable for hypertension and type 2 diabetes mellitus complicated by peripheral neuropathy. He smoked one pack of cigarettes daily for 40 years but quit 15 years prior. He drank 4 ounces of vodka each night.

On examination, temperature was 36.5°C, heart rate 92 beats per minute, blood pressure 127/60 mmHg, respiratory rate 12 breaths per minute, oxygen saturation 98% on ambient air, and weight 118 kg. Extraocular movements were intact, pupils were equal and reactive to light and accommodation, and there were no visual field deficits. Nondilated funduscopic examination revealed normal blood vessels, optic disc, and optic cup-to-disc ratio. Dentition was good with pink gingiva. Bilateral temples were nontender. There was normal range of motion and strength in the shoulders, hips, and lower extremities with no tenderness over the trochanters. Patellar and ankle reflexes were present and symmetric bilaterally. He had no rashes or ecchymoses.

The history of smoking, especially with concomitant alcohol intake, is a risk factor for head and neck cancer, and these malignancies can lead to facial pain. While the normal oral cavity exam argues against localized oral and dental causes of the patient’s symptoms, direct fiberoptic endoscopy should be considered. The neck should be examined for lymphadenopathy. Normal vital signs point away from severe infection. The lack of findings in the head and musculoskeletal regions does not exclude systemic etiologies such as rheumatologic conditions or neoplasm. Complete blood cell count and markers of inflammation including erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels should be obtained. Hip and pelvic radiographs should be obtained to evaluate for hip osteoarthritis, fractures, or osseous lesions.

The appointment occurred during evening hours and the patient declined further evaluation until the following morning, at which time laboratory studies revealed normal serum levels of electrolytes, blood urea nitrogen, and creatinine. White blood cell (WBC) count was 6,800/mm³ with an immature granulocyte ratio of 1.8% (normal, 0.0-0.5%), hemoglobin 13.2 g/dL, and platelet count 163,000/mm³. ESR was 118 mm/hr (normal, 0-15 mm/hr) and CRP was 1.5 mg/dL (normal, 0-0.75 mg/dL). Radiographs of the hips and pelvis showed osteoarthritis of the bilateral hip joints and degenerative disc disease of the lower lumbar spine.

Granulocytosis may occur in response to infection, rheumatologic conditions, and hematologic malignancies such as chronic myelogenous leukemia. While infectious etiologies (eg, abscess, osteomyelitis) are the most common cause of an extremely elevated ESR level, this patient does not have other signs or symptoms of infection such as fever or leukocytosis. Therefore, other common causes for an extremely elevated ESR level should be considered, including malignancy (eg, multiple myeloma, lymphoma, metastatic solid tumor) and autoimmune conditions (eg, rheumatoid arthritis, vasculitis). While multiple myeloma is the most common malignant etiology for extremely elevated ESR, the patient lacks signs of this condition such as anemia, elevated creatinine, or osteolytic lesions on radiographic imaging. Osteoarthritis identified on the radiographs may contribute to the patient’s hip pain but would not explain the patient’s jaw pain, weight loss, granulocytosis, and elevated ESR. These findings, taken together with the patient’s age, are most suggestive of GCA with possible coexisting PMR. Temporal artery biopsy should be obtained as it is the gold standard test for diagnosing GCA.

The patient was contacted by telephone that same day with laboratory test results. During the call, he endorsed increased jaw and temple pain. He was advised to proceed to the emergency department (ED) for timely evaluation and treatment.

Because GCA was being considered, ophthalmology performed an ocular examination in the ED, which demonstrated no signs of optic nerve or retinal ischemia. Computed tomography (CT) scan of the head and neck with intravenous contrast revealed no abscess or soft tissue abnormalities. Right temporal artery biopsy was performed.

The normal ocular examination does not exclude GCA, and temporal artery biopsy is appropriate. The mainstay of treatment for GCA is high-dose systemic glucocorticoids, which should not be withheld while awaiting biopsy results since ophthalmic artery inflammation may occur and threaten vision.

While GCA remains the leading diagnosis, malignant etiologies warrant further consideration because they are a common cause of extreme ESR elevation, particularly among older patients. The patient’s cancer screening history should be reviewed. The normal CT scan of the head and neck reduces the likelihood of localized solid tumor etiologies; however, additional CT imaging of the chest, abdomen, and pelvis is warranted to evaluate for metastatic solid tumors or lymphoma.

A 10-day course of prednisone 60 mg daily was prescribed for empiric treatment of GCA. The patient was discharged home with follow-up scheduled in rheumatology and primary care clinics. Pain in the jaw and temple resolved within several days.

Two weeks later, he presented to the rheumatology clinic. He noted 1 week of lower right back pain described as dull, aching, radiating to the lateral right hip, and occurring when transitioning from sitting to standing. He had no leg numbness, weakness, or change in bowel habits. Bladder habits were also unchanged, although he reported chronic urinary frequency and occasional incontinence. He reported further weight loss, this time an unintentional loss of 9 kg. He noted frequent sweating but no fever.

He reported a normal colonoscopy within the prior 5 years. Because these records were not available for review, a fecal immunochemical test was obtained and negative for hemoglobin. He had previously declined prostate cancer screening.

The resolution of jaw and temple pain with prednisone supports the presumed diagnosis of GCA. Up to half of patients with GCA may also have PMR, which can cause aching and stiffness in the arms, hips, and lumbar region, and pain may be abrupt in onset. However, PMR-related pain would be expected to improve rather than develop or worsen in the setting of high-dose glucocorticoid use. Therefore, other causes of acute-onset back pain must be considered.

While localized musculoskeletal etiologies such as lumbar muscle strain, radiculopathy, and vertebral compression fracture are possible, co-occurrence of unintentional weight loss and diaphoresis with elevated inflammatory markers suggests a systemic etiology. A neoplastic process with bony metastasis is possible. The reportedly normal colonoscopy and the negative fecal immunochemical test make colorectal cancer less likely. Inflammatory conditions such as ankylosing spondylitis and rheumatoid arthritis are also possible. Ankylosing spondylitis usually presents at a much younger age, however, and axial skeletal involvement in rheumatoid arthritis often involves the cervical spine and is usually seen after longstanding disease. Additionally, the hallmark of inflammatory back pain is morning stiffness which the patient does not endorse. Nonetheless, additional laboratory testing should include antinuclear antibody, rheumatoid factor, and anti-cyclic citrullinated peptide (anti-CCP) antibody. Vertebral osteomyelitis remains on the differential diagnosis, and repeat WBC count and inflammatory markers should be assessed. Lumbosacral radiographs should be obtained to rule out fracture.

Physical examination in the rheumatology clinic revealed a temperature of 37.0°C, heart rate 100 beats per minute, blood pressure 146/72 mmHg, respiratory rate 12 breaths per minute, and oxygen saturation 98% on ambient air. Weight was 109 kg. He was pale and diaphoretic. There was diffuse tenderness to palpation of the right-sided lumbar paraspinal muscles. Straight leg raise was negative bilaterally. Patellar reflexes and gait were normal.

Blood chemistries, renal function, and aminotransferase levels were normal. WBC count was 7,100/mm³, hemoglobin 8.0 g/dL, mean corpuscular volume 88.9 fL, platelet count 128,000/mm³, ESR 66 mm/hr, CRP 0.57 mg/dL, alkaline phosphatase 438 IU/L (normal, 30-130 IU/L), and thyroid-stimulating hormone 0.925 mU/L (normal, 0.34-5.60 mU/L). Testing for antinuclear antibodies, rheumatoid factor, and anti-CCP antibody was unremarkable. Prostate-specific antigen (PSA) level was 2.2 ng/mL (normal, 0-4 ng/mL). Urinalysis was unremarkable. Antibodies to hepatitis C and Treponema pallidum were negative. Interferon gamma release assay was negative.

Findings of new onset anemia and thrombocytopenia, in combination with elevated ESR and alkaline phosphatase level, are concerning for disseminated intravascular coagulation (DIC) and microangiopathic hemolytic anemia (MAHA), bone marrow infiltration of a metastatic neoplasm, or ineffective hematopoiesis caused by myelodysplastic syndromes or myelofibrosis.

Laboratory evaluation should include iron studies, lactate dehydrogenase (LDH), haptoglobin, fibrinogen, D-dimer, reticulocyte count, and peripheral blood smear to assess for hemolysis and erythrocyte morphology. Advanced imaging with lumbosacral magnetic resonance imaging (MRI) should be obtained to evaluate for focal etiologies of back pain such as disc herniation, abscess, marrow infiltration, and infarction.

Additional laboratory studies revealed a gamma-glutamyl transferase level of 49 IU/L (normal, 8-56 IU/L), LDH 288 IU/L (normal, 98-192 IU/L), haptoglobin 495 mg/dL (normal, 32-240 mg/dL), fibrinogen >700 mg/dL (normal, 225-550 mg/dL), D-dimer 693 ng/mL (normal, 200-250 ng/mL), serum iron 57 mcg/dL (normal, 33-150 mcg/dL), total iron binding capacity 286 mcg/dL (normal, 250-450 mcg/dL), ferritin 1,012 ng/mL (normal, 17.9-464 ng/mL), and reticulocyte count 2.9% (normal, 0.5-2.5%). Coagulation studies and serum protein electrophoresis were normal. Erythropoietin level was 109 mIU/mL (normal, 4.0-20.0 mIU/mL). Peripheral blood smear demonstrated moderate anemia with 8% nucleated erythrocytes per white blood cell (normal, 0%) and no circulating blasts.

MRI of the thoracolumbar spine and pelvis revealed diffusely abnormal bone marrow signal with multiple superimposed focal and poorly defined enhancing lesions along the lumbar spine marrow, sacrum, and bilateral iliac bones (Figure 1). Positron emission tomography/computed tomography (PET/CT) scan showed no scintigraphic evidence of metabolically active neoplastic, paraneoplastic, or inflammatory disorder.

The elevated haptoglobin, normal coagulation studies, and absence of fragmented erythrocytes on peripheral smear exclude an intravascular hemolytic process. The patient’s lower than expected reticulocyte count for the degree of anemia, elevated erythropoietin, and nucleated erythrocytes constitute a pattern that can be seen with bone marrow infiltration. There are no circulating blasts, making leukemia less likely. A solid organ tumor with bone metastases may cause enhancing lesions on MRI since this form of imaging is more sensitive than radiography for detecting skeletal malignancies. The negative PET/CT, however, does not reveal a primary tumor. Myelofibrosis is an infiltrative myeloproliferative disorder associated with nonspecific laboratory abnormalities, bone pain, weight loss, and night sweats that could cause diffuse MRI bone marrow signal alterations with normal PET/CT findings. However, myelofibrosis would not typically cause a significantly elevated ESR, and thus would be an unlikely cause for this patient’s presentation.

Given the constellation of symptoms, hematologic abnormalities, and bone marrow infiltration on imaging, hematology should be consulted to perform a bone marrow biopsy to assist with definitive diagnosis.

Bone marrow biopsy demonstrated metastatic adenocarcinoma consistent with prostatic origin (Figure 2). Bone scan demonstrated widespread osteoblastic metastases, which included the skull and temporal regions. These lesions were thought to be the cause of the patient’s original presenting symptom of jaw pain.

The patient was started on androgen deprivation therapy, initially with degarelix and subsequently leuprolide shots and abiraterone with prednisone. PSA was 0.08 ng/mL after 3 months of androgen deprivation therapy. His back and hip pain slowly improved.

Prostate cancer is the most common cancer in men with one out of every nine men diagnosed in his lifetime.¹ While most men initially present with localized, curable disease,¹ 4% present with metastatic disease, an incidence that has been increasing since 2004.² Despite available treatments, metastatic prostate cancer has a poor prognosis, with an average overall survival of approximately 5 years.³

Prostate cancer can be challenging to diagnose. Men with prostate cancer are commonly asymptomatic. Rarely, patients may present with hematuria, bony pain caused by metastasis, or obstructive urinary symptoms like hesitancy or incomplete bladder emptying. Our patient presented with jaw pain, which was ultimately attributed to osteoblastic lesions of the skull. Additionally, his history of urinary frequency and incontinence may have been clues to his underlying diagnosis of prostate cancer.

Prostate cancer screening remains highly nuanced and relies on shared decision-making between patients and healthcare providers. Clinical practice guidelines for early detection of prostate cancer recommend individualized PSA-based serologic screening.^4,5 Specifically, the United States Preventive Services Task Force recommends screening men aged 55 to 69 years who desire screening and understand the potential harms associated with a positive test result. These harms may include psychological distress and complications from prostate biopsy (eg, pain or infection) or prostate cancer treatment (eg, erectile, urinary, and/or bowel dysfunction).^4-6 The decision to screen can be guided by individuals’ risk factors including African American race, family history, and older age.

While our patient elected not to undergo routine prostate cancer screening, a PSA level was obtained during his diagnostic evaluation and highlights the limitations of PSA-based screening. A PSA level ≤4.0 ng/mL has 21% sensitivity and 91% specificity for detecting prostate cancer.⁷ PSA levels above 4.0 ng/mL warrant repeat testing and, if persistently elevated, referral to urology for possible prostate biopsy. PSA levels often correlate with burden of disease, and patients with PSA levels >20 ng/mL are referred for CT imaging to evaluate for metastatic disease.⁸ PSA’s poor sensitivity was underscored in a study by Thompson et al who evaluated the incidence of prostate cancer in men participating in the Prostate Cancer Prevention Trial with PSA levels of <4 ng/mL.⁹ In this study, 15% of men diagnosed with prostate cancer never had a PSA level >4 ng/mL.⁹ While most of the cancers in this study were low grade and may have been clinically insignificant, 15% demonstrated histologic signs of at least intermediate-risk disease. Our patient’s PSA level of 2.2 ng/mL was below the threshold that triggers additional evaluation even though he had widely metastatic prostate cancer.

Our patient’s severe jaw and temple pain, weight loss, and progressive hip pain were concerning for GCA. This vasculitis of large- and medium-sized arteries predominantly affects older adults with greatest incidence among those 70 years of age and older.¹⁰ Symptoms occur because of cranial artery inflammation and may include headache, visual disturbance, erythema or tenderness of the temporal artery, and jaw claudication. Extracranial inflammation may affect the thoracic aorta and its branches and rarely the abdominal aorta and lower limb arteries. Pelvic girdle pain more typically results from associated PMR. Patients may also note systemic symptoms such as fever, weight loss, and fatigue.

Prompt diagnostic testing is important when considering GCA. Most patients with GCA have ESR levels greater than 40 mm/hr.¹¹ ESR is a laboratory test that measures the vertical distance erythrocytes travel in a column of blood over 1 hour; in the setting of inflammation, cells form clumps and travel more quickly than individual cells, resulting in a higher value. While moderate elevations in ESR may occur without an identifiable cause, extreme ESR levels—those above 100 mm/hr, as observed in our patient—are highly suggestive of certain serious conditions, including infection, malignancy, and autoimmune disease such as GCA.^12,13 Temporal artery biopsy is the gold standard test to diagnose GCA. However, because of noncontiguous inflammation of the temporal artery, biopsies may be falsely negative. Thus, sampling of the contralateral temporal artery may be warranted if suspicion remains high.

As was the case for our patient, PET/CT is not reliable for diagnosing prostate cancer. In contrast to other malignancies (eg, lymphoma, lung cancer), prostate cancer typically does not display increased glucose metabolism. Moreover, the close proximity of the bladder and prostate can interfere with imaging interpretation because the fluorodeoxyglucose (FDG) tracer is excreted in the urine.¹⁴ The reported sensitivity of PET/CT for the diagnosis of prostate cancer ranges from 17%-65%.^15,16 In a small study of men with metastatic prostate cancer, only 18% of bony metastases were FDG avid, and there was no correlation between FDG avidity and PSA level.¹⁵ Notably, although PET/CT includes CT imaging, this CT is used to map anatomic landmarks and is not separately interpreted by the radiologist. Thus, even if evidence of prostate cancer was apparent on traditional CT, it may be overlooked on PET/CT.

Several important points regarding diagnostic testing are raised by this case. First, PSA-based screening for prostate cancer may be falsely negative, even in the setting of widely metastatic disease. Second, extreme ESR elevation is a marker for serious underlying disease and warrants a thorough diagnostic evaluation. Finally, PET/CT has limited diagnostic utility in evaluating metastatic prostate cancer because of the normal rates of glucose metabolism. Our patient initially presented with jaw pain, yet his progressive physical symptoms and laboratory abnormalities prompted an evaluation which ultimately revealed the jaw-dropping diagnosis of PSA-negative, metastatic prostate cancer.

• ESR levels greater than 100 mm/hr are highly suggestive of certain serious conditions including infection, autoimmune disease, and malignancy.
• PSA-based screening for prostate cancer can result in false negative test results. In one study, 15% of men diagnosed with prostate cancer never had a PSA level greater than 4 ng/mL (ie, the level at which repeat laboratory testing and/or referral to urology for possible prostate biopsy is advisable).
• PET/CT has limited diagnostic utility in evaluating metastatic prostate cancer, because prostate cancer cells typically demonstrate normal glucose metabolism.

Drs Griauzde, Northway, Yentz, and Houchens have nothing to disclose. Dr Saint reports personal fees from ISMIE Mutual Insurance Company during the conduct of the study, as well as personal fees from Jvion and Doximity outside the submitted work.

A 73-year-old man presented to primary care for an annual examination. Four days prior, he noted right-sided sharp jaw pain such that he could not open his mouth nor chew solid food; it radiated from the right mandible to the ipsilateral temple. He also noted bilateral aching hip pain for several years that increased in severity in the prior 2 months. He reported an intentional weight loss of 9 kg over the past year, achieved through dietary modification. He denied fever, chills, and visual disturbance.

Acute onset of unilateral jaw pain that is worsened by chewing is a feature consistent with a temporomandibular disorder (TMD). TMD consists of musculoskeletal and neuromuscular conditions that affect the temporomandibular joints (TMJs), masticatory muscles, and associated tissues. Common symptoms of TMD include facial or ear pain, temporal headache, and TMJ dysfunction or discomfort. In addition to TMD, craniofacial pain has many possible etiologies such as dental pathology, neuralgias, sinus and otologic disorders, headache and migraine disorders, infections, rheumatologic conditions, and neoplasms.

Systemic etiologies for this patient’s symptoms are a consideration given his age and concomitant worsening of chronic hip pain. Rheumatologic conditions such as giant cell arteritis (GCA) and polymyalgia rheumatica (PMR) are more common in adults older than 50 years of age and cause headache, jaw claudication, and pelvic girdle pain. Rarely, hematologic malignancies (eg, lymphoma), solid tumor metastases (eg, breast cancer, melanoma), and primary tumors of the head and neck (eg, nasopharyngeal carcinoma) can involve the mandible, TMJ, or parotid gland and result in symptoms of TMD.

Medical history was notable for hypertension and type 2 diabetes mellitus complicated by peripheral neuropathy. He smoked one pack of cigarettes daily for 40 years but quit 15 years prior. He drank 4 ounces of vodka each night.

On examination, temperature was 36.5°C, heart rate 92 beats per minute, blood pressure 127/60 mmHg, respiratory rate 12 breaths per minute, oxygen saturation 98% on ambient air, and weight 118 kg. Extraocular movements were intact, pupils were equal and reactive to light and accommodation, and there were no visual field deficits. Nondilated funduscopic examination revealed normal blood vessels, optic disc, and optic cup-to-disc ratio. Dentition was good with pink gingiva. Bilateral temples were nontender. There was normal range of motion and strength in the shoulders, hips, and lower extremities with no tenderness over the trochanters. Patellar and ankle reflexes were present and symmetric bilaterally. He had no rashes or ecchymoses.

The history of smoking, especially with concomitant alcohol intake, is a risk factor for head and neck cancer, and these malignancies can lead to facial pain. While the normal oral cavity exam argues against localized oral and dental causes of the patient’s symptoms, direct fiberoptic endoscopy should be considered. The neck should be examined for lymphadenopathy. Normal vital signs point away from severe infection. The lack of findings in the head and musculoskeletal regions does not exclude systemic etiologies such as rheumatologic conditions or neoplasm. Complete blood cell count and markers of inflammation including erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels should be obtained. Hip and pelvic radiographs should be obtained to evaluate for hip osteoarthritis, fractures, or osseous lesions.

The appointment occurred during evening hours and the patient declined further evaluation until the following morning, at which time laboratory studies revealed normal serum levels of electrolytes, blood urea nitrogen, and creatinine. White blood cell (WBC) count was 6,800/mm³ with an immature granulocyte ratio of 1.8% (normal, 0.0-0.5%), hemoglobin 13.2 g/dL, and platelet count 163,000/mm³. ESR was 118 mm/hr (normal, 0-15 mm/hr) and CRP was 1.5 mg/dL (normal, 0-0.75 mg/dL). Radiographs of the hips and pelvis showed osteoarthritis of the bilateral hip joints and degenerative disc disease of the lower lumbar spine.

Granulocytosis may occur in response to infection, rheumatologic conditions, and hematologic malignancies such as chronic myelogenous leukemia. While infectious etiologies (eg, abscess, osteomyelitis) are the most common cause of an extremely elevated ESR level, this patient does not have other signs or symptoms of infection such as fever or leukocytosis. Therefore, other common causes for an extremely elevated ESR level should be considered, including malignancy (eg, multiple myeloma, lymphoma, metastatic solid tumor) and autoimmune conditions (eg, rheumatoid arthritis, vasculitis). While multiple myeloma is the most common malignant etiology for extremely elevated ESR, the patient lacks signs of this condition such as anemia, elevated creatinine, or osteolytic lesions on radiographic imaging. Osteoarthritis identified on the radiographs may contribute to the patient’s hip pain but would not explain the patient’s jaw pain, weight loss, granulocytosis, and elevated ESR. These findings, taken together with the patient’s age, are most suggestive of GCA with possible coexisting PMR. Temporal artery biopsy should be obtained as it is the gold standard test for diagnosing GCA.

The patient was contacted by telephone that same day with laboratory test results. During the call, he endorsed increased jaw and temple pain. He was advised to proceed to the emergency department (ED) for timely evaluation and treatment.

Because GCA was being considered, ophthalmology performed an ocular examination in the ED, which demonstrated no signs of optic nerve or retinal ischemia. Computed tomography (CT) scan of the head and neck with intravenous contrast revealed no abscess or soft tissue abnormalities. Right temporal artery biopsy was performed.

The normal ocular examination does not exclude GCA, and temporal artery biopsy is appropriate. The mainstay of treatment for GCA is high-dose systemic glucocorticoids, which should not be withheld while awaiting biopsy results since ophthalmic artery inflammation may occur and threaten vision.

While GCA remains the leading diagnosis, malignant etiologies warrant further consideration because they are a common cause of extreme ESR elevation, particularly among older patients. The patient’s cancer screening history should be reviewed. The normal CT scan of the head and neck reduces the likelihood of localized solid tumor etiologies; however, additional CT imaging of the chest, abdomen, and pelvis is warranted to evaluate for metastatic solid tumors or lymphoma.

A 10-day course of prednisone 60 mg daily was prescribed for empiric treatment of GCA. The patient was discharged home with follow-up scheduled in rheumatology and primary care clinics. Pain in the jaw and temple resolved within several days.

Two weeks later, he presented to the rheumatology clinic. He noted 1 week of lower right back pain described as dull, aching, radiating to the lateral right hip, and occurring when transitioning from sitting to standing. He had no leg numbness, weakness, or change in bowel habits. Bladder habits were also unchanged, although he reported chronic urinary frequency and occasional incontinence. He reported further weight loss, this time an unintentional loss of 9 kg. He noted frequent sweating but no fever.

He reported a normal colonoscopy within the prior 5 years. Because these records were not available for review, a fecal immunochemical test was obtained and negative for hemoglobin. He had previously declined prostate cancer screening.

The resolution of jaw and temple pain with prednisone supports the presumed diagnosis of GCA. Up to half of patients with GCA may also have PMR, which can cause aching and stiffness in the arms, hips, and lumbar region, and pain may be abrupt in onset. However, PMR-related pain would be expected to improve rather than develop or worsen in the setting of high-dose glucocorticoid use. Therefore, other causes of acute-onset back pain must be considered.

While localized musculoskeletal etiologies such as lumbar muscle strain, radiculopathy, and vertebral compression fracture are possible, co-occurrence of unintentional weight loss and diaphoresis with elevated inflammatory markers suggests a systemic etiology. A neoplastic process with bony metastasis is possible. The reportedly normal colonoscopy and the negative fecal immunochemical test make colorectal cancer less likely. Inflammatory conditions such as ankylosing spondylitis and rheumatoid arthritis are also possible. Ankylosing spondylitis usually presents at a much younger age, however, and axial skeletal involvement in rheumatoid arthritis often involves the cervical spine and is usually seen after longstanding disease. Additionally, the hallmark of inflammatory back pain is morning stiffness which the patient does not endorse. Nonetheless, additional laboratory testing should include antinuclear antibody, rheumatoid factor, and anti-cyclic citrullinated peptide (anti-CCP) antibody. Vertebral osteomyelitis remains on the differential diagnosis, and repeat WBC count and inflammatory markers should be assessed. Lumbosacral radiographs should be obtained to rule out fracture.

Physical examination in the rheumatology clinic revealed a temperature of 37.0°C, heart rate 100 beats per minute, blood pressure 146/72 mmHg, respiratory rate 12 breaths per minute, and oxygen saturation 98% on ambient air. Weight was 109 kg. He was pale and diaphoretic. There was diffuse tenderness to palpation of the right-sided lumbar paraspinal muscles. Straight leg raise was negative bilaterally. Patellar reflexes and gait were normal.

Blood chemistries, renal function, and aminotransferase levels were normal. WBC count was 7,100/mm³, hemoglobin 8.0 g/dL, mean corpuscular volume 88.9 fL, platelet count 128,000/mm³, ESR 66 mm/hr, CRP 0.57 mg/dL, alkaline phosphatase 438 IU/L (normal, 30-130 IU/L), and thyroid-stimulating hormone 0.925 mU/L (normal, 0.34-5.60 mU/L). Testing for antinuclear antibodies, rheumatoid factor, and anti-CCP antibody was unremarkable. Prostate-specific antigen (PSA) level was 2.2 ng/mL (normal, 0-4 ng/mL). Urinalysis was unremarkable. Antibodies to hepatitis C and Treponema pallidum were negative. Interferon gamma release assay was negative.

Findings of new onset anemia and thrombocytopenia, in combination with elevated ESR and alkaline phosphatase level, are concerning for disseminated intravascular coagulation (DIC) and microangiopathic hemolytic anemia (MAHA), bone marrow infiltration of a metastatic neoplasm, or ineffective hematopoiesis caused by myelodysplastic syndromes or myelofibrosis.

Laboratory evaluation should include iron studies, lactate dehydrogenase (LDH), haptoglobin, fibrinogen, D-dimer, reticulocyte count, and peripheral blood smear to assess for hemolysis and erythrocyte morphology. Advanced imaging with lumbosacral magnetic resonance imaging (MRI) should be obtained to evaluate for focal etiologies of back pain such as disc herniation, abscess, marrow infiltration, and infarction.

Additional laboratory studies revealed a gamma-glutamyl transferase level of 49 IU/L (normal, 8-56 IU/L), LDH 288 IU/L (normal, 98-192 IU/L), haptoglobin 495 mg/dL (normal, 32-240 mg/dL), fibrinogen >700 mg/dL (normal, 225-550 mg/dL), D-dimer 693 ng/mL (normal, 200-250 ng/mL), serum iron 57 mcg/dL (normal, 33-150 mcg/dL), total iron binding capacity 286 mcg/dL (normal, 250-450 mcg/dL), ferritin 1,012 ng/mL (normal, 17.9-464 ng/mL), and reticulocyte count 2.9% (normal, 0.5-2.5%). Coagulation studies and serum protein electrophoresis were normal. Erythropoietin level was 109 mIU/mL (normal, 4.0-20.0 mIU/mL). Peripheral blood smear demonstrated moderate anemia with 8% nucleated erythrocytes per white blood cell (normal, 0%) and no circulating blasts.

MRI of the thoracolumbar spine and pelvis revealed diffusely abnormal bone marrow signal with multiple superimposed focal and poorly defined enhancing lesions along the lumbar spine marrow, sacrum, and bilateral iliac bones (Figure 1). Positron emission tomography/computed tomography (PET/CT) scan showed no scintigraphic evidence of metabolically active neoplastic, paraneoplastic, or inflammatory disorder.

The elevated haptoglobin, normal coagulation studies, and absence of fragmented erythrocytes on peripheral smear exclude an intravascular hemolytic process. The patient’s lower than expected reticulocyte count for the degree of anemia, elevated erythropoietin, and nucleated erythrocytes constitute a pattern that can be seen with bone marrow infiltration. There are no circulating blasts, making leukemia less likely. A solid organ tumor with bone metastases may cause enhancing lesions on MRI since this form of imaging is more sensitive than radiography for detecting skeletal malignancies. The negative PET/CT, however, does not reveal a primary tumor. Myelofibrosis is an infiltrative myeloproliferative disorder associated with nonspecific laboratory abnormalities, bone pain, weight loss, and night sweats that could cause diffuse MRI bone marrow signal alterations with normal PET/CT findings. However, myelofibrosis would not typically cause a significantly elevated ESR, and thus would be an unlikely cause for this patient’s presentation.

Given the constellation of symptoms, hematologic abnormalities, and bone marrow infiltration on imaging, hematology should be consulted to perform a bone marrow biopsy to assist with definitive diagnosis.

Bone marrow biopsy demonstrated metastatic adenocarcinoma consistent with prostatic origin (Figure 2). Bone scan demonstrated widespread osteoblastic metastases, which included the skull and temporal regions. These lesions were thought to be the cause of the patient’s original presenting symptom of jaw pain.

The patient was started on androgen deprivation therapy, initially with degarelix and subsequently leuprolide shots and abiraterone with prednisone. PSA was 0.08 ng/mL after 3 months of androgen deprivation therapy. His back and hip pain slowly improved.

Prostate cancer is the most common cancer in men with one out of every nine men diagnosed in his lifetime.¹ While most men initially present with localized, curable disease,¹ 4% present with metastatic disease, an incidence that has been increasing since 2004.² Despite available treatments, metastatic prostate cancer has a poor prognosis, with an average overall survival of approximately 5 years.³

Prostate cancer can be challenging to diagnose. Men with prostate cancer are commonly asymptomatic. Rarely, patients may present with hematuria, bony pain caused by metastasis, or obstructive urinary symptoms like hesitancy or incomplete bladder emptying. Our patient presented with jaw pain, which was ultimately attributed to osteoblastic lesions of the skull. Additionally, his history of urinary frequency and incontinence may have been clues to his underlying diagnosis of prostate cancer.

Prostate cancer screening remains highly nuanced and relies on shared decision-making between patients and healthcare providers. Clinical practice guidelines for early detection of prostate cancer recommend individualized PSA-based serologic screening.^4,5 Specifically, the United States Preventive Services Task Force recommends screening men aged 55 to 69 years who desire screening and understand the potential harms associated with a positive test result. These harms may include psychological distress and complications from prostate biopsy (eg, pain or infection) or prostate cancer treatment (eg, erectile, urinary, and/or bowel dysfunction).^4-6 The decision to screen can be guided by individuals’ risk factors including African American race, family history, and older age.

While our patient elected not to undergo routine prostate cancer screening, a PSA level was obtained during his diagnostic evaluation and highlights the limitations of PSA-based screening. A PSA level ≤4.0 ng/mL has 21% sensitivity and 91% specificity for detecting prostate cancer.⁷ PSA levels above 4.0 ng/mL warrant repeat testing and, if persistently elevated, referral to urology for possible prostate biopsy. PSA levels often correlate with burden of disease, and patients with PSA levels >20 ng/mL are referred for CT imaging to evaluate for metastatic disease.⁸ PSA’s poor sensitivity was underscored in a study by Thompson et al who evaluated the incidence of prostate cancer in men participating in the Prostate Cancer Prevention Trial with PSA levels of <4 ng/mL.⁹ In this study, 15% of men diagnosed with prostate cancer never had a PSA level >4 ng/mL.⁹ While most of the cancers in this study were low grade and may have been clinically insignificant, 15% demonstrated histologic signs of at least intermediate-risk disease. Our patient’s PSA level of 2.2 ng/mL was below the threshold that triggers additional evaluation even though he had widely metastatic prostate cancer.

Our patient’s severe jaw and temple pain, weight loss, and progressive hip pain were concerning for GCA. This vasculitis of large- and medium-sized arteries predominantly affects older adults with greatest incidence among those 70 years of age and older.¹⁰ Symptoms occur because of cranial artery inflammation and may include headache, visual disturbance, erythema or tenderness of the temporal artery, and jaw claudication. Extracranial inflammation may affect the thoracic aorta and its branches and rarely the abdominal aorta and lower limb arteries. Pelvic girdle pain more typically results from associated PMR. Patients may also note systemic symptoms such as fever, weight loss, and fatigue.

Prompt diagnostic testing is important when considering GCA. Most patients with GCA have ESR levels greater than 40 mm/hr.¹¹ ESR is a laboratory test that measures the vertical distance erythrocytes travel in a column of blood over 1 hour; in the setting of inflammation, cells form clumps and travel more quickly than individual cells, resulting in a higher value. While moderate elevations in ESR may occur without an identifiable cause, extreme ESR levels—those above 100 mm/hr, as observed in our patient—are highly suggestive of certain serious conditions, including infection, malignancy, and autoimmune disease such as GCA.^12,13 Temporal artery biopsy is the gold standard test to diagnose GCA. However, because of noncontiguous inflammation of the temporal artery, biopsies may be falsely negative. Thus, sampling of the contralateral temporal artery may be warranted if suspicion remains high.

As was the case for our patient, PET/CT is not reliable for diagnosing prostate cancer. In contrast to other malignancies (eg, lymphoma, lung cancer), prostate cancer typically does not display increased glucose metabolism. Moreover, the close proximity of the bladder and prostate can interfere with imaging interpretation because the fluorodeoxyglucose (FDG) tracer is excreted in the urine.¹⁴ The reported sensitivity of PET/CT for the diagnosis of prostate cancer ranges from 17%-65%.^15,16 In a small study of men with metastatic prostate cancer, only 18% of bony metastases were FDG avid, and there was no correlation between FDG avidity and PSA level.¹⁵ Notably, although PET/CT includes CT imaging, this CT is used to map anatomic landmarks and is not separately interpreted by the radiologist. Thus, even if evidence of prostate cancer was apparent on traditional CT, it may be overlooked on PET/CT.

Several important points regarding diagnostic testing are raised by this case. First, PSA-based screening for prostate cancer may be falsely negative, even in the setting of widely metastatic disease. Second, extreme ESR elevation is a marker for serious underlying disease and warrants a thorough diagnostic evaluation. Finally, PET/CT has limited diagnostic utility in evaluating metastatic prostate cancer because of the normal rates of glucose metabolism. Our patient initially presented with jaw pain, yet his progressive physical symptoms and laboratory abnormalities prompted an evaluation which ultimately revealed the jaw-dropping diagnosis of PSA-negative, metastatic prostate cancer.

• ESR levels greater than 100 mm/hr are highly suggestive of certain serious conditions including infection, autoimmune disease, and malignancy.
• PSA-based screening for prostate cancer can result in false negative test results. In one study, 15% of men diagnosed with prostate cancer never had a PSA level greater than 4 ng/mL (ie, the level at which repeat laboratory testing and/or referral to urology for possible prostate biopsy is advisable).
• PET/CT has limited diagnostic utility in evaluating metastatic prostate cancer, because prostate cancer cells typically demonstrate normal glucose metabolism.

Drs Griauzde, Northway, Yentz, and Houchens have nothing to disclose. Dr Saint reports personal fees from ISMIE Mutual Insurance Company during the conduct of the study, as well as personal fees from Jvion and Doximity outside the submitted work.

Xerostomia, the subjective sensation of dry mouth, is a common problem developed by geriatric patients. In practice, xerostomia can impair swallowing, speech, and oral hygiene, and if left unchecked, symptoms such as dysphagia and dysarthria can diminish patients’ quality of life (QOL). Salivary gland hypofunction (SGH) is the objective measure of decreased saliva production, determined by sialometry. Although xerostomia and SGH can coexist, the 2 conditions are not necessarily related.^1-4 For this discussion, the term xerostomia will denote dry mouth with or without a concomitant diagnosis of SGH.

Xerostomia is seen in a wide variety of patients with varied comorbidities. It is commonly associated with Sjögren syndrome and head and neck irradiation. The diagnosis and treatment of xerostomia often involves rheumatologists, dentists, otolaryngologists, and oncologists. Additionally, most of the scientific literature about this topic exists in dental journals, such as the Journal of the American Dental Association and the British Dental Journal. Rarer still are studies in the veteran population.⁵

Faced with increasing time pressure to treat the many chronic diseases affecting aging veterans, health care providers (HCPs) tend to deprioritize diagnosing dry mouth. To that point, saliva is often not considered in the same category as other bodily fluids. According to Mandel, “It lacks the drama of blood, the sincerity of sweat…[and] the emotional appeal of tears.”⁶ In reality, saliva plays a critical role in the oral-digestive tract and in swallowing. It contains the first digestive enzymes in the gastrointestinal tract and is key for maintaining homeostasis in the oral cavity.⁷ Decreased saliva production results in difficulties with speech and mastication as well as problems of dysphagia, esophageal dysfunction, dysgeusia, nutritional compromises, new and recurrent dental caries, candidiasis, glossitis, impaired use of dentures, halitosis, and susceptibility to mucosal injury.^7,8 Problems with the production of saliva may lead to loss of QOL, such as enjoying a meal or conversing with others.⁴

Although xerostomia is often associated with advanced age, it is more often explained by the diseases that afflict geriatric patients and the arsenal of medications used to treat them.^2,9-16 Polypharmacy, the simultaneous use of multiple drugs by a single patient for ≥ 1 conditions, is an independent risk factor for xerostomia regardless of the types of medication taken.¹⁶ From 2005 to 2011, older adults in the US significantly increased their prescription medication use and dietary supplements. More than one-third of older adults used ≥ 5 prescription medications concurrently, and two-thirds of older adults used combinations of prescribed medications, over-the-counter medications, and dietary supplements.¹⁷ Several drug classes have the capacity to induce xerostomia, such as antihypertensives, antiulcer agents, anticholinergics, and antidepressants.^2,5,12 Prevalence of dry mouth also can range from 10% to 46%, and women typically are more medicated and symptomatic.^{2,3,9,13,14,16} Xerostomia can also lead to depression and even reduce patients’ will to live.¹⁸ Despite xerostomia’s prevalence and impact on QOL, few patients report it as their chief symptom, and few physicians attempt to treat it.¹⁹

In order to target polypharmacy as a cause of dry mouth, the objectives for this study were to evaluate (1) the prevalence of xerostomia; (2) the relationship between xerostomia and other oral conditions; and (3) the impact of polypharmacy on dry mouth in a veteran population.

Methods

This is a retrospective cross-sectional study of all outpatient visits in fiscal year (FY) 2015 (October 1, 2014 to September 30, 2015) at the VA Palo Alto Health Care System (VAPAHCS), a tertiary care US Department of Veterans Affairs (VA) hospital. This study was approved by the Stanford University Institutional Review Board. All patients diagnosed with xerostomia in the 1-year study period were identified using ICD-9 diagnosis codes for dry mouth or disturbance of salivary gland secretion (527.7, 527.8, R68.2) and Systemized Nomenclature of Medicine Clinical Terms (SNOMED CT) codes covering dry mouth, xerostomia, aptyalism, absent salivary secretion, and disturbance of salivary secretion (87715008, 78948009). Data analysts in the VA Office of Business Analytics assisted in gathering data from the Veterans Information Systems and Technology Architecture (VistA) electronic health record.

The statistical analysis of that data was performed using Microsoft Excel. Age and gender distributions were determined for the patients. The relationship between xerostomia and the number and types of medications taken by patients also was examined. A previous Swedish study examining the link between dry mouth and quantities of medications used a scale ranging from 0 to ≥ 7 medications.¹⁶ The scale for this study was made wider to include the following groups: 0-2, 3-5, 6-8, 9-11, and ≥ 12 medications. Items that do not have xerogenic risks, such as medical supplies (eg, gloves, syringes, etc) and topical medications, were excluded from the analysis. Finally, the number of subjects with comorbid problems with speech, dentition, or swallowing (SDS) was recorded. Non-VA medications were included to capture any self- or externally prescribed xerogenic medications.

Results

Of the patients seen at VAPAHCS during FY 2015, 138 had a diagnostic code for xerostomia, including 129 men (93.5%) and 9 women (6.5%). The average (SD) age of this xerostomia cohort was 69.3 (12.6) years, and the 3 most common age groups were 60 to 69 years (37.7%), 70 to 79 years (28.3%), and 80 to 89 years (13.0%) (Table 1). Of those 138 patients with a xerostomia diagnosis, a majority (84; 60.9%) had at least 1 documented SDS problem (Table 2). Conversely, during FY 2015, although 4,971 patients seen at VAPAHCS had documented SDS problems, only 77 (1.5%) had a recorded diagnosis of xerostomia.

Of the 138 patients with xerostomia, 55 (39.9%) were taking ≥ 12 medications, more than twice as many patients as in any of the other groups studied (0-2, 3-5, 6-8, and 9-11 medications taken) (Table 3). On average, each patient with xerostomia filled prescriptions for 10.4 (SD, 7.2) different drugs. In this cohort of 138 patients diagnosed with xerostomia, antihypertensive medications or analgesics were taken by > 50% of patients, while statins, psychiatric medications, antibiotics, proton pump inhibitors, or drugs known to have anticholinergic activity were taken by > 40%. Antihistamines, anticonvulsants, diuretics, or inhaled respiratory agents were used by > 20% of the patients in this cohort (Table 4).

Data on each individual medication were split into 2 categories: the percentage of patients that filled ≥ 1 prescription for that drug, and the total number of prescriptions filled and/or refilled for that drug (ie, including all fills and refills made by individual patients). The 5 most widely used medications in this cohort were omeprazole (39.1%), docusate sodium (29.7%), gabapentin (29.7%), aspirin (27.5%), and hydrocodone/acetaminophen (26.1%) (Table 5). The 5 prescriptions that were cumulatively most filled and/or refilled were omeprazole (128), sildenafil citrate (108), gabapentin (101), hydrocodone/acetaminophen (100), and oxycodone (92) (Table 6). Though sildenafil citrate and oxycodone were among the most-filled prescriptions, these were not included in Table 5 as neither was taken by > 15% of the patients studied. These prescriptions were filled multiple times by a small subset of patients.

Regarding treatment for dry mouth, artificial saliva spray was one of the most widely used (23.2%) and the seventh most-filled prescription within this cohort (86). The only other medication taken by > 15% of patients in a formulation other than a tablet or capsule was chlorhexidine, a germicidal mouthwash used to improve oral care.

Discussion

Saliva is an essential component for the maintenance of normal oral health.^20,21 Decreased saliva production causes problems, including difficulties with speech, mastication, dysphagia, changes in taste, dental caries, impaired use of prostheses, recurrent infections, halitosis, deterioration of soft tissues, and compromised QOL.^22,23 Among patients with a diagnosed SDS abnormality who were seen at this facility during FY 2015, the prevalence of xerostomia was only 1.5%. However, the true prevalence and incidence of xerostomia among veterans is not known. Given the role of xerostomia as a common risk factor for SDS problems and the polypharmacy exhibited by those presented here with SDS problems, it is probable that xerostomia was underreported in this veteran cohort.

Additionally, although salivary acinar cells are known to atrophy with age, as is consistent with this xerostomia cohort’s average age (SD) of 69.3 (12.6) years, the development of dry mouth is a multifactorial process. The current scientific literature asserts that most salivary loss is due to local and systemic diseases, immunologic disorders, external radiation, and as was highlighted by this study, multiple prescription and nonprescription medications.^24-26

It has also been demonstrated previously that dry mouth complaints and low salivary flow rates are directly proportional to the number of medications taken by patients.^2,27-30 Polypharmacy is therefore an area of great interest, and ≥ 40 categories of xerogenic medications have been identified by investigators such as Sreebny and Schwartz.³¹ Among those, some of the most xerogenic medication classes include antihypertensives, antiulcer agents, anticholinergics, and antidepressants, are all very commonly consumed in this cohort of patients with dry mouth (58.7%, 42.0%, 47.1%, and 38.4%, respectively). The medication regimens within this cohort of veterans with xerostomia were prime examples of polypharmacy as each patient took an average (SD) of 10.4 (7.2) medications, 39.9% took ≥ 12, and 72.5% of patients with xerostomia were taking ≥ 6 prescription drugs during a 1-year period.

Given the dangers of polypharmacy, a more conservative approach to prescribing medications could feasibly help with preventing xerostomia and SGH. In practice, while clinicians try to avoid prescribing anticholinergics, antimuscarinics, and antihistaminergic drugs for geriatric patients, they are tasked with the complex management of medication adverse effects (AEs) when dealing with multiple health conditions. The clinicians’ primary responsibilities are to follow the standard of care and not to introduce unnecessary harm when managing patients, but they also must push for, stay abreast of, and conduct more basic research and clinical trials to inform, adjust, and improve our current standard.

Research into polypharmacy and its role in inducing dry mouth is ongoing. Twenty years ago, Thomson and colleagues identified reduced salivary flow in patients who used antianginals, thyroxine, diuretics, antidepressants, and medications for asthma, while only 5 years earlier Loesche and colleagues reported the role of antiulcer medications, such as proton pump inhibitors, in the development of xerostomia.^2,32 Within the past 5 years, Viljakainen and colleagues and Ohara and colleagues have echoed some of those findings by identifying associations between xerostomia and agents that impact digestive organs.^33,34 A strong association recently was identified between the use of antipsychotic drugs and xerostomia.³⁵ Additionally, when attributing xerostomia to polypharmacy, the interaction between medications is often overlooked in favor of considering the raw number of prescriptions taken. Whereas 1 medication alone may not have drying properties, combinations of medications might be more likely to induce xerostomia. Thomson and colleagues suggested such a situation regarding the interaction between thyroxine and diuretics.³⁶ Future studies should focus on identifying viable substitutes for existing medications that reduce risk for xerostomia without compromising the management of other serious conditions.

Treatment

Another pressing question for clinicians concerns artificial saliva. Although 23.2% of patients with dry mouth in this xerostomia cohort used artificial saliva, the efficacy of this treatment is still unproven. Saliva substitutes are often used by patients who cannot produce sufficient amounts of natural saliva. In practice, artificial saliva produces, at best, modest temporary improvement in dry mouth symptoms in up to 40% of patients. At worst, as put forth by the Cochrane Review, artificial saliva may be no better than placebo in treating dry mouth.^37,38 The volumes needed for symptom relief are large, ranging between 40 mL and 150 mL per day depending on the substitute’s composition. Saliva substitutes also must be reapplied throughout each day. This is particularly bothersome when patients must wake up repeatedly to reapply the treatment at night.³⁷ In short, these substances do not seem wholly effective in managing dry mouth, and other options must be made available to patients with refractory xerostomia when artificial saliva and lifestyle modifications fail.

For now, few alternatives exist. Chewing gums and lozenges help to stimulate salivary flow, as do muscarinic agonists like pilocarpine. Unfortunately, muscarinic agonists are seldom used due to cholinergic AEs. Humidifiers are effective in increasing nighttime moisture but are contraindicated in patients with dust mite allergies.³⁹ Reservoir-based devices with automated pumps funnel water and/or salivary substitutes from a fanny pack into patients’ mouths for lubrication.³⁷ Other more esoteric pharmacologic treatments include D-limonene, yohimbine, and amifostine, which purportedly protect salivary progenitor cells, increase peripheral cholinergic activity, and protect salivary glands from free-radical damage during radiation treatment, respectively. Although these agents have shown some promise, D-limonene is difficult to administer given the high dosage required for treatment, yohimbine hasn’t been seriously investigated for improving salivary secretion since 1997, and amifostine isn’t used widely due to its AE profile despite its US Food and Drug Administration approval for prevention of xerostomia.³⁹

Substance Abuse

The impact of smoking on xerostomia remains controversial. Some studies report an association between active smoking and xerostomia; others suggest that the local irritant effect of tobacco smoke may increase salivary gland output.^40,41 The same may be true for chronic alcohol use as there are no epidemiologic studies showing a causal relationship between alcohol use and xerostomia. Studies with rats that are chronically exposed to ethanol have found increased salivary flow rates.⁴² In the xerostomia cohort presented here, 30 patients (21.7%) had a documented history of substance misuse. That percentage is likely underestimated, as substance misuse is often underreported, and frequent use may not always constitute misuse. Therefore, nicotine exposure, alcohol exposure, illicit drug use, and vaping all should be considered during the workup of a patient with xerostomia.

Limitations

It is common for medications to remain in a patient’s health record long after that patient stops taking them. Developing methods to track when patients discontinue their prescriptions will be essential for clearing up uncertainty in our data and in other similar studies. This study also did not account for patients’ medication adherence and the duration of exposure to medications and illicit substances. Furthermore, the results of this veteran study are not easily generalizable as this cohort is disproportionately male, of advanced age, and especially prone to exhibiting both substance use and psychiatric diagnoses relative to the general population. As described by Viljakainen and colleagues, risk factors for xerostomia include advanced age, female gender, low body mass index (BMI), malnutrition, and depressive symptoms, but because the demographic scope of this veteran population was narrow, it was not possible to discern the impact of, for example, gender.³³ Data on variables like BMI, malnutrition, and depressive symptoms were not available. For this study, xerostomia could only be considered as an all-or-nothing phenomenon because the dataset did not describe different levels of dry mouth severity (eg, mild, moderate, severe).

Additionally, past polypharmacy studies have acknowledged an inability to tell whether xerostomia is mainly due to medications or to underlying medical conditions. For example, for emphysema, ß-adrenergic stimulation from bronchodilators could cause dry mouth by thickening saliva and decreasing salivary volume, but the pathophysiology and/or cardinal symptoms of emphysema, including chronic obstructive pulmonary disease-associated tachypnea, might contribute independently to dryness.

Though we can make inferences based on the medications taken by this cohort (eg, those taking antihypertensives have high blood pressure), this dataset did not explicitly detail comorbid conditions and ICD codes for chronic diseases that commonly arise with xerostomia. Those conditions, however, are of great clinical importance. Diabetes mellitus, HIV/AIDS, and, classically, Sjögren syndrome, all are known to cause dry mouth.⁴³ Identifying new conditions that co-occur with xerostomia would allow clinicians to describe the root causes of and risk factors for dry mouth and SDS conditions in greater detail. Patients with dry mouth without SDS problems in this dataset are of particular interest as closer examination of their medications and comorbid conditions could help us understand why some individuals and not others develop SDS problems. The subjects of how comorbidities contribute to dry mouth and how their influences can be judged independently from the effects of medications are of great interest to us and will be investigated rigorously in our future studies.

Conclusions

In this cohort, few patients with SDS problems had documentation of a concomitant xerostomia diagnosis. This could represent a true infrequency of dry mouth or more likely, an underrecognition by clinicians. Heightened physician awareness regarding the signs and symptoms of and risk factors for xerostomia is needed to improve providers’ ability to diagnose this condition. 

In particular, polypharmacy should be a major consideration when evaluating patients for xerostomia. This continues to be an important area of research, and some of the latest data on polypharmacy among older patients were compiled in a recent meta-analysis by Tan and colleagues. The authors of that systematic review reiterated the significant association between salivary gland hypofunction and the number of medications taken by patients. They also advocated for the creation of a risk score for medication-induced dry mouth to aid in medication management.⁴⁴ Per their recommendations, it is now as crucial as ever to consider the numbers and types of medications taken by patients, to discontinue unnecessary prescriptions when possible, and to continue developing new strategies for preventing and treating xerostomia.

Xerostomia, the subjective sensation of dry mouth, is a common problem developed by geriatric patients. In practice, xerostomia can impair swallowing, speech, and oral hygiene, and if left unchecked, symptoms such as dysphagia and dysarthria can diminish patients’ quality of life (QOL). Salivary gland hypofunction (SGH) is the objective measure of decreased saliva production, determined by sialometry. Although xerostomia and SGH can coexist, the 2 conditions are not necessarily related.^1-4 For this discussion, the term xerostomia will denote dry mouth with or without a concomitant diagnosis of SGH.

Xerostomia is seen in a wide variety of patients with varied comorbidities. It is commonly associated with Sjögren syndrome and head and neck irradiation. The diagnosis and treatment of xerostomia often involves rheumatologists, dentists, otolaryngologists, and oncologists. Additionally, most of the scientific literature about this topic exists in dental journals, such as the Journal of the American Dental Association and the British Dental Journal. Rarer still are studies in the veteran population.⁵

Faced with increasing time pressure to treat the many chronic diseases affecting aging veterans, health care providers (HCPs) tend to deprioritize diagnosing dry mouth. To that point, saliva is often not considered in the same category as other bodily fluids. According to Mandel, “It lacks the drama of blood, the sincerity of sweat…[and] the emotional appeal of tears.”⁶ In reality, saliva plays a critical role in the oral-digestive tract and in swallowing. It contains the first digestive enzymes in the gastrointestinal tract and is key for maintaining homeostasis in the oral cavity.⁷ Decreased saliva production results in difficulties with speech and mastication as well as problems of dysphagia, esophageal dysfunction, dysgeusia, nutritional compromises, new and recurrent dental caries, candidiasis, glossitis, impaired use of dentures, halitosis, and susceptibility to mucosal injury.^7,8 Problems with the production of saliva may lead to loss of QOL, such as enjoying a meal or conversing with others.⁴

Although xerostomia is often associated with advanced age, it is more often explained by the diseases that afflict geriatric patients and the arsenal of medications used to treat them.^2,9-16 Polypharmacy, the simultaneous use of multiple drugs by a single patient for ≥ 1 conditions, is an independent risk factor for xerostomia regardless of the types of medication taken.¹⁶ From 2005 to 2011, older adults in the US significantly increased their prescription medication use and dietary supplements. More than one-third of older adults used ≥ 5 prescription medications concurrently, and two-thirds of older adults used combinations of prescribed medications, over-the-counter medications, and dietary supplements.¹⁷ Several drug classes have the capacity to induce xerostomia, such as antihypertensives, antiulcer agents, anticholinergics, and antidepressants.^2,5,12 Prevalence of dry mouth also can range from 10% to 46%, and women typically are more medicated and symptomatic.^{2,3,9,13,14,16} Xerostomia can also lead to depression and even reduce patients’ will to live.¹⁸ Despite xerostomia’s prevalence and impact on QOL, few patients report it as their chief symptom, and few physicians attempt to treat it.¹⁹

In order to target polypharmacy as a cause of dry mouth, the objectives for this study were to evaluate (1) the prevalence of xerostomia; (2) the relationship between xerostomia and other oral conditions; and (3) the impact of polypharmacy on dry mouth in a veteran population.

Methods

This is a retrospective cross-sectional study of all outpatient visits in fiscal year (FY) 2015 (October 1, 2014 to September 30, 2015) at the VA Palo Alto Health Care System (VAPAHCS), a tertiary care US Department of Veterans Affairs (VA) hospital. This study was approved by the Stanford University Institutional Review Board. All patients diagnosed with xerostomia in the 1-year study period were identified using ICD-9 diagnosis codes for dry mouth or disturbance of salivary gland secretion (527.7, 527.8, R68.2) and Systemized Nomenclature of Medicine Clinical Terms (SNOMED CT) codes covering dry mouth, xerostomia, aptyalism, absent salivary secretion, and disturbance of salivary secretion (87715008, 78948009). Data analysts in the VA Office of Business Analytics assisted in gathering data from the Veterans Information Systems and Technology Architecture (VistA) electronic health record.

The statistical analysis of that data was performed using Microsoft Excel. Age and gender distributions were determined for the patients. The relationship between xerostomia and the number and types of medications taken by patients also was examined. A previous Swedish study examining the link between dry mouth and quantities of medications used a scale ranging from 0 to ≥ 7 medications.¹⁶ The scale for this study was made wider to include the following groups: 0-2, 3-5, 6-8, 9-11, and ≥ 12 medications. Items that do not have xerogenic risks, such as medical supplies (eg, gloves, syringes, etc) and topical medications, were excluded from the analysis. Finally, the number of subjects with comorbid problems with speech, dentition, or swallowing (SDS) was recorded. Non-VA medications were included to capture any self- or externally prescribed xerogenic medications.

Results

Of the patients seen at VAPAHCS during FY 2015, 138 had a diagnostic code for xerostomia, including 129 men (93.5%) and 9 women (6.5%). The average (SD) age of this xerostomia cohort was 69.3 (12.6) years, and the 3 most common age groups were 60 to 69 years (37.7%), 70 to 79 years (28.3%), and 80 to 89 years (13.0%) (Table 1). Of those 138 patients with a xerostomia diagnosis, a majority (84; 60.9%) had at least 1 documented SDS problem (Table 2). Conversely, during FY 2015, although 4,971 patients seen at VAPAHCS had documented SDS problems, only 77 (1.5%) had a recorded diagnosis of xerostomia.

Of the 138 patients with xerostomia, 55 (39.9%) were taking ≥ 12 medications, more than twice as many patients as in any of the other groups studied (0-2, 3-5, 6-8, and 9-11 medications taken) (Table 3). On average, each patient with xerostomia filled prescriptions for 10.4 (SD, 7.2) different drugs. In this cohort of 138 patients diagnosed with xerostomia, antihypertensive medications or analgesics were taken by > 50% of patients, while statins, psychiatric medications, antibiotics, proton pump inhibitors, or drugs known to have anticholinergic activity were taken by > 40%. Antihistamines, anticonvulsants, diuretics, or inhaled respiratory agents were used by > 20% of the patients in this cohort (Table 4).

Data on each individual medication were split into 2 categories: the percentage of patients that filled ≥ 1 prescription for that drug, and the total number of prescriptions filled and/or refilled for that drug (ie, including all fills and refills made by individual patients). The 5 most widely used medications in this cohort were omeprazole (39.1%), docusate sodium (29.7%), gabapentin (29.7%), aspirin (27.5%), and hydrocodone/acetaminophen (26.1%) (Table 5). The 5 prescriptions that were cumulatively most filled and/or refilled were omeprazole (128), sildenafil citrate (108), gabapentin (101), hydrocodone/acetaminophen (100), and oxycodone (92) (Table 6). Though sildenafil citrate and oxycodone were among the most-filled prescriptions, these were not included in Table 5 as neither was taken by > 15% of the patients studied. These prescriptions were filled multiple times by a small subset of patients.

Regarding treatment for dry mouth, artificial saliva spray was one of the most widely used (23.2%) and the seventh most-filled prescription within this cohort (86). The only other medication taken by > 15% of patients in a formulation other than a tablet or capsule was chlorhexidine, a germicidal mouthwash used to improve oral care.

Discussion

Saliva is an essential component for the maintenance of normal oral health.^20,21 Decreased saliva production causes problems, including difficulties with speech, mastication, dysphagia, changes in taste, dental caries, impaired use of prostheses, recurrent infections, halitosis, deterioration of soft tissues, and compromised QOL.^22,23 Among patients with a diagnosed SDS abnormality who were seen at this facility during FY 2015, the prevalence of xerostomia was only 1.5%. However, the true prevalence and incidence of xerostomia among veterans is not known. Given the role of xerostomia as a common risk factor for SDS problems and the polypharmacy exhibited by those presented here with SDS problems, it is probable that xerostomia was underreported in this veteran cohort.

Additionally, although salivary acinar cells are known to atrophy with age, as is consistent with this xerostomia cohort’s average age (SD) of 69.3 (12.6) years, the development of dry mouth is a multifactorial process. The current scientific literature asserts that most salivary loss is due to local and systemic diseases, immunologic disorders, external radiation, and as was highlighted by this study, multiple prescription and nonprescription medications.^24-26

It has also been demonstrated previously that dry mouth complaints and low salivary flow rates are directly proportional to the number of medications taken by patients.^2,27-30 Polypharmacy is therefore an area of great interest, and ≥ 40 categories of xerogenic medications have been identified by investigators such as Sreebny and Schwartz.³¹ Among those, some of the most xerogenic medication classes include antihypertensives, antiulcer agents, anticholinergics, and antidepressants, are all very commonly consumed in this cohort of patients with dry mouth (58.7%, 42.0%, 47.1%, and 38.4%, respectively). The medication regimens within this cohort of veterans with xerostomia were prime examples of polypharmacy as each patient took an average (SD) of 10.4 (7.2) medications, 39.9% took ≥ 12, and 72.5% of patients with xerostomia were taking ≥ 6 prescription drugs during a 1-year period.

Given the dangers of polypharmacy, a more conservative approach to prescribing medications could feasibly help with preventing xerostomia and SGH. In practice, while clinicians try to avoid prescribing anticholinergics, antimuscarinics, and antihistaminergic drugs for geriatric patients, they are tasked with the complex management of medication adverse effects (AEs) when dealing with multiple health conditions. The clinicians’ primary responsibilities are to follow the standard of care and not to introduce unnecessary harm when managing patients, but they also must push for, stay abreast of, and conduct more basic research and clinical trials to inform, adjust, and improve our current standard.

Research into polypharmacy and its role in inducing dry mouth is ongoing. Twenty years ago, Thomson and colleagues identified reduced salivary flow in patients who used antianginals, thyroxine, diuretics, antidepressants, and medications for asthma, while only 5 years earlier Loesche and colleagues reported the role of antiulcer medications, such as proton pump inhibitors, in the development of xerostomia.^2,32 Within the past 5 years, Viljakainen and colleagues and Ohara and colleagues have echoed some of those findings by identifying associations between xerostomia and agents that impact digestive organs.^33,34 A strong association recently was identified between the use of antipsychotic drugs and xerostomia.³⁵ Additionally, when attributing xerostomia to polypharmacy, the interaction between medications is often overlooked in favor of considering the raw number of prescriptions taken. Whereas 1 medication alone may not have drying properties, combinations of medications might be more likely to induce xerostomia. Thomson and colleagues suggested such a situation regarding the interaction between thyroxine and diuretics.³⁶ Future studies should focus on identifying viable substitutes for existing medications that reduce risk for xerostomia without compromising the management of other serious conditions.

Treatment

Another pressing question for clinicians concerns artificial saliva. Although 23.2% of patients with dry mouth in this xerostomia cohort used artificial saliva, the efficacy of this treatment is still unproven. Saliva substitutes are often used by patients who cannot produce sufficient amounts of natural saliva. In practice, artificial saliva produces, at best, modest temporary improvement in dry mouth symptoms in up to 40% of patients. At worst, as put forth by the Cochrane Review, artificial saliva may be no better than placebo in treating dry mouth.^37,38 The volumes needed for symptom relief are large, ranging between 40 mL and 150 mL per day depending on the substitute’s composition. Saliva substitutes also must be reapplied throughout each day. This is particularly bothersome when patients must wake up repeatedly to reapply the treatment at night.³⁷ In short, these substances do not seem wholly effective in managing dry mouth, and other options must be made available to patients with refractory xerostomia when artificial saliva and lifestyle modifications fail.

For now, few alternatives exist. Chewing gums and lozenges help to stimulate salivary flow, as do muscarinic agonists like pilocarpine. Unfortunately, muscarinic agonists are seldom used due to cholinergic AEs. Humidifiers are effective in increasing nighttime moisture but are contraindicated in patients with dust mite allergies.³⁹ Reservoir-based devices with automated pumps funnel water and/or salivary substitutes from a fanny pack into patients’ mouths for lubrication.³⁷ Other more esoteric pharmacologic treatments include D-limonene, yohimbine, and amifostine, which purportedly protect salivary progenitor cells, increase peripheral cholinergic activity, and protect salivary glands from free-radical damage during radiation treatment, respectively. Although these agents have shown some promise, D-limonene is difficult to administer given the high dosage required for treatment, yohimbine hasn’t been seriously investigated for improving salivary secretion since 1997, and amifostine isn’t used widely due to its AE profile despite its US Food and Drug Administration approval for prevention of xerostomia.³⁹

Substance Abuse

The impact of smoking on xerostomia remains controversial. Some studies report an association between active smoking and xerostomia; others suggest that the local irritant effect of tobacco smoke may increase salivary gland output.^40,41 The same may be true for chronic alcohol use as there are no epidemiologic studies showing a causal relationship between alcohol use and xerostomia. Studies with rats that are chronically exposed to ethanol have found increased salivary flow rates.⁴² In the xerostomia cohort presented here, 30 patients (21.7%) had a documented history of substance misuse. That percentage is likely underestimated, as substance misuse is often underreported, and frequent use may not always constitute misuse. Therefore, nicotine exposure, alcohol exposure, illicit drug use, and vaping all should be considered during the workup of a patient with xerostomia.

Limitations

It is common for medications to remain in a patient’s health record long after that patient stops taking them. Developing methods to track when patients discontinue their prescriptions will be essential for clearing up uncertainty in our data and in other similar studies. This study also did not account for patients’ medication adherence and the duration of exposure to medications and illicit substances. Furthermore, the results of this veteran study are not easily generalizable as this cohort is disproportionately male, of advanced age, and especially prone to exhibiting both substance use and psychiatric diagnoses relative to the general population. As described by Viljakainen and colleagues, risk factors for xerostomia include advanced age, female gender, low body mass index (BMI), malnutrition, and depressive symptoms, but because the demographic scope of this veteran population was narrow, it was not possible to discern the impact of, for example, gender.³³ Data on variables like BMI, malnutrition, and depressive symptoms were not available. For this study, xerostomia could only be considered as an all-or-nothing phenomenon because the dataset did not describe different levels of dry mouth severity (eg, mild, moderate, severe).

Additionally, past polypharmacy studies have acknowledged an inability to tell whether xerostomia is mainly due to medications or to underlying medical conditions. For example, for emphysema, ß-adrenergic stimulation from bronchodilators could cause dry mouth by thickening saliva and decreasing salivary volume, but the pathophysiology and/or cardinal symptoms of emphysema, including chronic obstructive pulmonary disease-associated tachypnea, might contribute independently to dryness.

Though we can make inferences based on the medications taken by this cohort (eg, those taking antihypertensives have high blood pressure), this dataset did not explicitly detail comorbid conditions and ICD codes for chronic diseases that commonly arise with xerostomia. Those conditions, however, are of great clinical importance. Diabetes mellitus, HIV/AIDS, and, classically, Sjögren syndrome, all are known to cause dry mouth.⁴³ Identifying new conditions that co-occur with xerostomia would allow clinicians to describe the root causes of and risk factors for dry mouth and SDS conditions in greater detail. Patients with dry mouth without SDS problems in this dataset are of particular interest as closer examination of their medications and comorbid conditions could help us understand why some individuals and not others develop SDS problems. The subjects of how comorbidities contribute to dry mouth and how their influences can be judged independently from the effects of medications are of great interest to us and will be investigated rigorously in our future studies.

Conclusions

In this cohort, few patients with SDS problems had documentation of a concomitant xerostomia diagnosis. This could represent a true infrequency of dry mouth or more likely, an underrecognition by clinicians. Heightened physician awareness regarding the signs and symptoms of and risk factors for xerostomia is needed to improve providers’ ability to diagnose this condition. 

In particular, polypharmacy should be a major consideration when evaluating patients for xerostomia. This continues to be an important area of research, and some of the latest data on polypharmacy among older patients were compiled in a recent meta-analysis by Tan and colleagues. The authors of that systematic review reiterated the significant association between salivary gland hypofunction and the number of medications taken by patients. They also advocated for the creation of a risk score for medication-induced dry mouth to aid in medication management.⁴⁴ Per their recommendations, it is now as crucial as ever to consider the numbers and types of medications taken by patients, to discontinue unnecessary prescriptions when possible, and to continue developing new strategies for preventing and treating xerostomia.

Xerostomia, the subjective sensation of dry mouth, is a common problem developed by geriatric patients. In practice, xerostomia can impair swallowing, speech, and oral hygiene, and if left unchecked, symptoms such as dysphagia and dysarthria can diminish patients’ quality of life (QOL). Salivary gland hypofunction (SGH) is the objective measure of decreased saliva production, determined by sialometry. Although xerostomia and SGH can coexist, the 2 conditions are not necessarily related.^1-4 For this discussion, the term xerostomia will denote dry mouth with or without a concomitant diagnosis of SGH.

Xerostomia is seen in a wide variety of patients with varied comorbidities. It is commonly associated with Sjögren syndrome and head and neck irradiation. The diagnosis and treatment of xerostomia often involves rheumatologists, dentists, otolaryngologists, and oncologists. Additionally, most of the scientific literature about this topic exists in dental journals, such as the Journal of the American Dental Association and the British Dental Journal. Rarer still are studies in the veteran population.⁵

Faced with increasing time pressure to treat the many chronic diseases affecting aging veterans, health care providers (HCPs) tend to deprioritize diagnosing dry mouth. To that point, saliva is often not considered in the same category as other bodily fluids. According to Mandel, “It lacks the drama of blood, the sincerity of sweat…[and] the emotional appeal of tears.”⁶ In reality, saliva plays a critical role in the oral-digestive tract and in swallowing. It contains the first digestive enzymes in the gastrointestinal tract and is key for maintaining homeostasis in the oral cavity.⁷ Decreased saliva production results in difficulties with speech and mastication as well as problems of dysphagia, esophageal dysfunction, dysgeusia, nutritional compromises, new and recurrent dental caries, candidiasis, glossitis, impaired use of dentures, halitosis, and susceptibility to mucosal injury.^7,8 Problems with the production of saliva may lead to loss of QOL, such as enjoying a meal or conversing with others.⁴

Although xerostomia is often associated with advanced age, it is more often explained by the diseases that afflict geriatric patients and the arsenal of medications used to treat them.^2,9-16 Polypharmacy, the simultaneous use of multiple drugs by a single patient for ≥ 1 conditions, is an independent risk factor for xerostomia regardless of the types of medication taken.¹⁶ From 2005 to 2011, older adults in the US significantly increased their prescription medication use and dietary supplements. More than one-third of older adults used ≥ 5 prescription medications concurrently, and two-thirds of older adults used combinations of prescribed medications, over-the-counter medications, and dietary supplements.¹⁷ Several drug classes have the capacity to induce xerostomia, such as antihypertensives, antiulcer agents, anticholinergics, and antidepressants.^2,5,12 Prevalence of dry mouth also can range from 10% to 46%, and women typically are more medicated and symptomatic.^{2,3,9,13,14,16} Xerostomia can also lead to depression and even reduce patients’ will to live.¹⁸ Despite xerostomia’s prevalence and impact on QOL, few patients report it as their chief symptom, and few physicians attempt to treat it.¹⁹

In order to target polypharmacy as a cause of dry mouth, the objectives for this study were to evaluate (1) the prevalence of xerostomia; (2) the relationship between xerostomia and other oral conditions; and (3) the impact of polypharmacy on dry mouth in a veteran population.

Methods

This is a retrospective cross-sectional study of all outpatient visits in fiscal year (FY) 2015 (October 1, 2014 to September 30, 2015) at the VA Palo Alto Health Care System (VAPAHCS), a tertiary care US Department of Veterans Affairs (VA) hospital. This study was approved by the Stanford University Institutional Review Board. All patients diagnosed with xerostomia in the 1-year study period were identified using ICD-9 diagnosis codes for dry mouth or disturbance of salivary gland secretion (527.7, 527.8, R68.2) and Systemized Nomenclature of Medicine Clinical Terms (SNOMED CT) codes covering dry mouth, xerostomia, aptyalism, absent salivary secretion, and disturbance of salivary secretion (87715008, 78948009). Data analysts in the VA Office of Business Analytics assisted in gathering data from the Veterans Information Systems and Technology Architecture (VistA) electronic health record.

The statistical analysis of that data was performed using Microsoft Excel. Age and gender distributions were determined for the patients. The relationship between xerostomia and the number and types of medications taken by patients also was examined. A previous Swedish study examining the link between dry mouth and quantities of medications used a scale ranging from 0 to ≥ 7 medications.¹⁶ The scale for this study was made wider to include the following groups: 0-2, 3-5, 6-8, 9-11, and ≥ 12 medications. Items that do not have xerogenic risks, such as medical supplies (eg, gloves, syringes, etc) and topical medications, were excluded from the analysis. Finally, the number of subjects with comorbid problems with speech, dentition, or swallowing (SDS) was recorded. Non-VA medications were included to capture any self- or externally prescribed xerogenic medications.

Results

Of the patients seen at VAPAHCS during FY 2015, 138 had a diagnostic code for xerostomia, including 129 men (93.5%) and 9 women (6.5%). The average (SD) age of this xerostomia cohort was 69.3 (12.6) years, and the 3 most common age groups were 60 to 69 years (37.7%), 70 to 79 years (28.3%), and 80 to 89 years (13.0%) (Table 1). Of those 138 patients with a xerostomia diagnosis, a majority (84; 60.9%) had at least 1 documented SDS problem (Table 2). Conversely, during FY 2015, although 4,971 patients seen at VAPAHCS had documented SDS problems, only 77 (1.5%) had a recorded diagnosis of xerostomia.

Of the 138 patients with xerostomia, 55 (39.9%) were taking ≥ 12 medications, more than twice as many patients as in any of the other groups studied (0-2, 3-5, 6-8, and 9-11 medications taken) (Table 3). On average, each patient with xerostomia filled prescriptions for 10.4 (SD, 7.2) different drugs. In this cohort of 138 patients diagnosed with xerostomia, antihypertensive medications or analgesics were taken by > 50% of patients, while statins, psychiatric medications, antibiotics, proton pump inhibitors, or drugs known to have anticholinergic activity were taken by > 40%. Antihistamines, anticonvulsants, diuretics, or inhaled respiratory agents were used by > 20% of the patients in this cohort (Table 4).

Data on each individual medication were split into 2 categories: the percentage of patients that filled ≥ 1 prescription for that drug, and the total number of prescriptions filled and/or refilled for that drug (ie, including all fills and refills made by individual patients). The 5 most widely used medications in this cohort were omeprazole (39.1%), docusate sodium (29.7%), gabapentin (29.7%), aspirin (27.5%), and hydrocodone/acetaminophen (26.1%) (Table 5). The 5 prescriptions that were cumulatively most filled and/or refilled were omeprazole (128), sildenafil citrate (108), gabapentin (101), hydrocodone/acetaminophen (100), and oxycodone (92) (Table 6). Though sildenafil citrate and oxycodone were among the most-filled prescriptions, these were not included in Table 5 as neither was taken by > 15% of the patients studied. These prescriptions were filled multiple times by a small subset of patients.

Regarding treatment for dry mouth, artificial saliva spray was one of the most widely used (23.2%) and the seventh most-filled prescription within this cohort (86). The only other medication taken by > 15% of patients in a formulation other than a tablet or capsule was chlorhexidine, a germicidal mouthwash used to improve oral care.

Discussion

Saliva is an essential component for the maintenance of normal oral health.^20,21 Decreased saliva production causes problems, including difficulties with speech, mastication, dysphagia, changes in taste, dental caries, impaired use of prostheses, recurrent infections, halitosis, deterioration of soft tissues, and compromised QOL.^22,23 Among patients with a diagnosed SDS abnormality who were seen at this facility during FY 2015, the prevalence of xerostomia was only 1.5%. However, the true prevalence and incidence of xerostomia among veterans is not known. Given the role of xerostomia as a common risk factor for SDS problems and the polypharmacy exhibited by those presented here with SDS problems, it is probable that xerostomia was underreported in this veteran cohort.

Additionally, although salivary acinar cells are known to atrophy with age, as is consistent with this xerostomia cohort’s average age (SD) of 69.3 (12.6) years, the development of dry mouth is a multifactorial process. The current scientific literature asserts that most salivary loss is due to local and systemic diseases, immunologic disorders, external radiation, and as was highlighted by this study, multiple prescription and nonprescription medications.^24-26

It has also been demonstrated previously that dry mouth complaints and low salivary flow rates are directly proportional to the number of medications taken by patients.^2,27-30 Polypharmacy is therefore an area of great interest, and ≥ 40 categories of xerogenic medications have been identified by investigators such as Sreebny and Schwartz.³¹ Among those, some of the most xerogenic medication classes include antihypertensives, antiulcer agents, anticholinergics, and antidepressants, are all very commonly consumed in this cohort of patients with dry mouth (58.7%, 42.0%, 47.1%, and 38.4%, respectively). The medication regimens within this cohort of veterans with xerostomia were prime examples of polypharmacy as each patient took an average (SD) of 10.4 (7.2) medications, 39.9% took ≥ 12, and 72.5% of patients with xerostomia were taking ≥ 6 prescription drugs during a 1-year period.

Given the dangers of polypharmacy, a more conservative approach to prescribing medications could feasibly help with preventing xerostomia and SGH. In practice, while clinicians try to avoid prescribing anticholinergics, antimuscarinics, and antihistaminergic drugs for geriatric patients, they are tasked with the complex management of medication adverse effects (AEs) when dealing with multiple health conditions. The clinicians’ primary responsibilities are to follow the standard of care and not to introduce unnecessary harm when managing patients, but they also must push for, stay abreast of, and conduct more basic research and clinical trials to inform, adjust, and improve our current standard.

Research into polypharmacy and its role in inducing dry mouth is ongoing. Twenty years ago, Thomson and colleagues identified reduced salivary flow in patients who used antianginals, thyroxine, diuretics, antidepressants, and medications for asthma, while only 5 years earlier Loesche and colleagues reported the role of antiulcer medications, such as proton pump inhibitors, in the development of xerostomia.^2,32 Within the past 5 years, Viljakainen and colleagues and Ohara and colleagues have echoed some of those findings by identifying associations between xerostomia and agents that impact digestive organs.^33,34 A strong association recently was identified between the use of antipsychotic drugs and xerostomia.³⁵ Additionally, when attributing xerostomia to polypharmacy, the interaction between medications is often overlooked in favor of considering the raw number of prescriptions taken. Whereas 1 medication alone may not have drying properties, combinations of medications might be more likely to induce xerostomia. Thomson and colleagues suggested such a situation regarding the interaction between thyroxine and diuretics.³⁶ Future studies should focus on identifying viable substitutes for existing medications that reduce risk for xerostomia without compromising the management of other serious conditions.

Treatment

Another pressing question for clinicians concerns artificial saliva. Although 23.2% of patients with dry mouth in this xerostomia cohort used artificial saliva, the efficacy of this treatment is still unproven. Saliva substitutes are often used by patients who cannot produce sufficient amounts of natural saliva. In practice, artificial saliva produces, at best, modest temporary improvement in dry mouth symptoms in up to 40% of patients. At worst, as put forth by the Cochrane Review, artificial saliva may be no better than placebo in treating dry mouth.^37,38 The volumes needed for symptom relief are large, ranging between 40 mL and 150 mL per day depending on the substitute’s composition. Saliva substitutes also must be reapplied throughout each day. This is particularly bothersome when patients must wake up repeatedly to reapply the treatment at night.³⁷ In short, these substances do not seem wholly effective in managing dry mouth, and other options must be made available to patients with refractory xerostomia when artificial saliva and lifestyle modifications fail.

For now, few alternatives exist. Chewing gums and lozenges help to stimulate salivary flow, as do muscarinic agonists like pilocarpine. Unfortunately, muscarinic agonists are seldom used due to cholinergic AEs. Humidifiers are effective in increasing nighttime moisture but are contraindicated in patients with dust mite allergies.³⁹ Reservoir-based devices with automated pumps funnel water and/or salivary substitutes from a fanny pack into patients’ mouths for lubrication.³⁷ Other more esoteric pharmacologic treatments include D-limonene, yohimbine, and amifostine, which purportedly protect salivary progenitor cells, increase peripheral cholinergic activity, and protect salivary glands from free-radical damage during radiation treatment, respectively. Although these agents have shown some promise, D-limonene is difficult to administer given the high dosage required for treatment, yohimbine hasn’t been seriously investigated for improving salivary secretion since 1997, and amifostine isn’t used widely due to its AE profile despite its US Food and Drug Administration approval for prevention of xerostomia.³⁹

Substance Abuse

The impact of smoking on xerostomia remains controversial. Some studies report an association between active smoking and xerostomia; others suggest that the local irritant effect of tobacco smoke may increase salivary gland output.^40,41 The same may be true for chronic alcohol use as there are no epidemiologic studies showing a causal relationship between alcohol use and xerostomia. Studies with rats that are chronically exposed to ethanol have found increased salivary flow rates.⁴² In the xerostomia cohort presented here, 30 patients (21.7%) had a documented history of substance misuse. That percentage is likely underestimated, as substance misuse is often underreported, and frequent use may not always constitute misuse. Therefore, nicotine exposure, alcohol exposure, illicit drug use, and vaping all should be considered during the workup of a patient with xerostomia.

Limitations

It is common for medications to remain in a patient’s health record long after that patient stops taking them. Developing methods to track when patients discontinue their prescriptions will be essential for clearing up uncertainty in our data and in other similar studies. This study also did not account for patients’ medication adherence and the duration of exposure to medications and illicit substances. Furthermore, the results of this veteran study are not easily generalizable as this cohort is disproportionately male, of advanced age, and especially prone to exhibiting both substance use and psychiatric diagnoses relative to the general population. As described by Viljakainen and colleagues, risk factors for xerostomia include advanced age, female gender, low body mass index (BMI), malnutrition, and depressive symptoms, but because the demographic scope of this veteran population was narrow, it was not possible to discern the impact of, for example, gender.³³ Data on variables like BMI, malnutrition, and depressive symptoms were not available. For this study, xerostomia could only be considered as an all-or-nothing phenomenon because the dataset did not describe different levels of dry mouth severity (eg, mild, moderate, severe).

Additionally, past polypharmacy studies have acknowledged an inability to tell whether xerostomia is mainly due to medications or to underlying medical conditions. For example, for emphysema, ß-adrenergic stimulation from bronchodilators could cause dry mouth by thickening saliva and decreasing salivary volume, but the pathophysiology and/or cardinal symptoms of emphysema, including chronic obstructive pulmonary disease-associated tachypnea, might contribute independently to dryness.

Though we can make inferences based on the medications taken by this cohort (eg, those taking antihypertensives have high blood pressure), this dataset did not explicitly detail comorbid conditions and ICD codes for chronic diseases that commonly arise with xerostomia. Those conditions, however, are of great clinical importance. Diabetes mellitus, HIV/AIDS, and, classically, Sjögren syndrome, all are known to cause dry mouth.⁴³ Identifying new conditions that co-occur with xerostomia would allow clinicians to describe the root causes of and risk factors for dry mouth and SDS conditions in greater detail. Patients with dry mouth without SDS problems in this dataset are of particular interest as closer examination of their medications and comorbid conditions could help us understand why some individuals and not others develop SDS problems. The subjects of how comorbidities contribute to dry mouth and how their influences can be judged independently from the effects of medications are of great interest to us and will be investigated rigorously in our future studies.

Conclusions

In this cohort, few patients with SDS problems had documentation of a concomitant xerostomia diagnosis. This could represent a true infrequency of dry mouth or more likely, an underrecognition by clinicians. Heightened physician awareness regarding the signs and symptoms of and risk factors for xerostomia is needed to improve providers’ ability to diagnose this condition. 

In particular, polypharmacy should be a major consideration when evaluating patients for xerostomia. This continues to be an important area of research, and some of the latest data on polypharmacy among older patients were compiled in a recent meta-analysis by Tan and colleagues. The authors of that systematic review reiterated the significant association between salivary gland hypofunction and the number of medications taken by patients. They also advocated for the creation of a risk score for medication-induced dry mouth to aid in medication management.⁴⁴ Per their recommendations, it is now as crucial as ever to consider the numbers and types of medications taken by patients, to discontinue unnecessary prescriptions when possible, and to continue developing new strategies for preventing and treating xerostomia.

COVID-19 is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and its symptoms range from mild to severe respiratory illness, fever, cough, fatigue, and shortness of breath.¹ Diarrhea is common early on with infection and loss of taste and smell have also been reported.¹ Follicular conjunctivitis has also been reported, either as an early sign of infection or during hospitalization for severe COVID-19 disease.^2-4 The incubation period of COVID-19 falls within 2 to 14 days according to the CDC.⁵

It has been confirmed that COVID-19 is transmitted through both respiratory droplets and direct contact. Another possible route of viral transmission is entry through aerosolized droplets into the tears, which then pass through the nasolacrimal ducts and into the respiratory tract.⁶

Preparations Prior to Office Visit

It is essential for the eye care provider to prioritize patient care in order of absolute necessity, such as sudden vision loss, sudden onset flashes and floaters, and eye trauma. In cases of potentially sight threatening pathology, it is in the best interest of the patient to conduct a face-to-face appointment. Therefore, it is important to implement new guidelines and protocols as we continue to see these patients (Figure 1).

Prior to the patient entering the medical facility, measures should be implemented to minimize exposure risk. This can be done over the telephone or at vehicle entrance screening stations. The triage technician answering the telephone should have a script of questions to ask. The patient should be instructed to come into the office alone unless, for physical or mental reasons, a caregiver is required.

SARS-CoV-2 Screening Questions

Preparedness through risk mitigation strategies are recommended with a targeted questionnaire and noncontact temperature check at the clinic or hospital entrance. Below are some general questions to further triage patients exposed to SARS-CoV-2.

• Do you have fever or any respiratory symptoms?
• Do you have new or worsening cough or shortness of breath?
• Do you have flulike symptoms?
• Have you been in close contact with someone, including health care workers, confirmed to have the COVID-19?

If the patient answers yes to any of the above questions, the CDC urges health care providers to immediately notify both infection control personnel at your health care facility and your local or state health department.^1,2 In regions currently managing significant outbreaks of COVID-19, the AAO recommends that eye care providers assume that any patient could be infected with SARS-CoV-2 and to proceed accordingly.² If urgent eye care is needed, a referral call should be made to a hospital or center equipped to deal with COVID-19 and urgent eye conditions. When calling the referral center, ensure adequate staffing and space and relay all pertinent information along with receiving approval from the treating physician.

Face-to-Face Office Visits

Once it has been determined that it is in the best interest of the patient to be seen in a face-to-face visit, the patient should be instructed to call the office when they arrive in the parking lot. The CDC recommends limiting points of entry upon arrival and during the visit.¹ As soon as an examination lane is ready, the patient can then be messaged to come into the office and escorted into the examination room.

An urgent or emergent ophthalmic examination for a patient with no respiratory symptoms, no fever, and no COVID-19 risk factors should include proper hand hygiene, use of personal protective equipment (PPE), and proper disinfection. Several studies have documented SARS-CoV-2 infection in asymptomatic and presymptomatic patients, making PPE of the up most importance.^2,7,8 PPE should include mask, face shield, and gloves. Currently, there are national and international shortages on PPE and a heightened topic of discussion concerning mask use, effectiveness with extended wear, and reuse. Please refer to the CDC and AAO websites for up-to-date guidelines (Table).^1,2 According to the CDC, N95 respirators are restricted to those performing or present for an aerosol-generating procedure.⁹

It is recommended that the eye care provider should only perform necessary tests and procedures. Noncontact tonometry should be avoided, as this might cause aerosolization of virus particles. The close proximity between eye care providers and their patients during slit-lamp examination may require further precautions to lower the risk of transmission via droplets or through hand to eye contact. The patient should be advised not to speak during the examination portion and the AAO also recommends a surgical mask or cloth face covering for the patient.² An additional protective device that may be used during the slit-lamp exam is a breath shield or a barrier shield (Figures 2 and 3).² Some manufacturers are offering clinicians free slit-lamp breath shields online.

Infection Prevention and Control Measures

Last, once the patient leaves the examination room, it should be properly disinfected. A disinfection checklist may be made to ensure uniform systematic cleaning. Alcohol and bleach-based disinfectants commonly used in health care settings are likely very effective against virus particles that cause COVID-19.¹⁰ During the disinfection process, gloves should be worn and careful attention paid to the contact time. Contact time is the amount of time the surface should appear visibly wet for proper disinfection. For example, Metrex CaviWipes have a recommended contact time of 3 minutes; however, this varies depending on type of virus and formulation, check labels or manufacturers’ websites for further directions.¹⁰ Also, the US Environmental Protection Agency has a database search available for disinfectants that meet their criteria for use against SARS-CoV-2.¹¹

In an ever-changing environment, we offer this article to help equip providers to deliver the best possible patient care when face-to-face encounters are necessary. Currently nonurgent eye care follow-up visits are being conducted by telephone or video clinics. It is our goal to inform fellow practitioners on options and strategies to elevate the safety of staff and patients while minimizing the risk of exposure.

COVID-19 is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and its symptoms range from mild to severe respiratory illness, fever, cough, fatigue, and shortness of breath.¹ Diarrhea is common early on with infection and loss of taste and smell have also been reported.¹ Follicular conjunctivitis has also been reported, either as an early sign of infection or during hospitalization for severe COVID-19 disease.^2-4 The incubation period of COVID-19 falls within 2 to 14 days according to the CDC.⁵

It has been confirmed that COVID-19 is transmitted through both respiratory droplets and direct contact. Another possible route of viral transmission is entry through aerosolized droplets into the tears, which then pass through the nasolacrimal ducts and into the respiratory tract.⁶

Preparations Prior to Office Visit

It is essential for the eye care provider to prioritize patient care in order of absolute necessity, such as sudden vision loss, sudden onset flashes and floaters, and eye trauma. In cases of potentially sight threatening pathology, it is in the best interest of the patient to conduct a face-to-face appointment. Therefore, it is important to implement new guidelines and protocols as we continue to see these patients (Figure 1).

Prior to the patient entering the medical facility, measures should be implemented to minimize exposure risk. This can be done over the telephone or at vehicle entrance screening stations. The triage technician answering the telephone should have a script of questions to ask. The patient should be instructed to come into the office alone unless, for physical or mental reasons, a caregiver is required.

SARS-CoV-2 Screening Questions

Preparedness through risk mitigation strategies are recommended with a targeted questionnaire and noncontact temperature check at the clinic or hospital entrance. Below are some general questions to further triage patients exposed to SARS-CoV-2.

• Do you have fever or any respiratory symptoms?
• Do you have new or worsening cough or shortness of breath?
• Do you have flulike symptoms?
• Have you been in close contact with someone, including health care workers, confirmed to have the COVID-19?

If the patient answers yes to any of the above questions, the CDC urges health care providers to immediately notify both infection control personnel at your health care facility and your local or state health department.^1,2 In regions currently managing significant outbreaks of COVID-19, the AAO recommends that eye care providers assume that any patient could be infected with SARS-CoV-2 and to proceed accordingly.² If urgent eye care is needed, a referral call should be made to a hospital or center equipped to deal with COVID-19 and urgent eye conditions. When calling the referral center, ensure adequate staffing and space and relay all pertinent information along with receiving approval from the treating physician.

Face-to-Face Office Visits

Once it has been determined that it is in the best interest of the patient to be seen in a face-to-face visit, the patient should be instructed to call the office when they arrive in the parking lot. The CDC recommends limiting points of entry upon arrival and during the visit.¹ As soon as an examination lane is ready, the patient can then be messaged to come into the office and escorted into the examination room.

An urgent or emergent ophthalmic examination for a patient with no respiratory symptoms, no fever, and no COVID-19 risk factors should include proper hand hygiene, use of personal protective equipment (PPE), and proper disinfection. Several studies have documented SARS-CoV-2 infection in asymptomatic and presymptomatic patients, making PPE of the up most importance.^2,7,8 PPE should include mask, face shield, and gloves. Currently, there are national and international shortages on PPE and a heightened topic of discussion concerning mask use, effectiveness with extended wear, and reuse. Please refer to the CDC and AAO websites for up-to-date guidelines (Table).^1,2 According to the CDC, N95 respirators are restricted to those performing or present for an aerosol-generating procedure.⁹

It is recommended that the eye care provider should only perform necessary tests and procedures. Noncontact tonometry should be avoided, as this might cause aerosolization of virus particles. The close proximity between eye care providers and their patients during slit-lamp examination may require further precautions to lower the risk of transmission via droplets or through hand to eye contact. The patient should be advised not to speak during the examination portion and the AAO also recommends a surgical mask or cloth face covering for the patient.² An additional protective device that may be used during the slit-lamp exam is a breath shield or a barrier shield (Figures 2 and 3).² Some manufacturers are offering clinicians free slit-lamp breath shields online.

Infection Prevention and Control Measures

Last, once the patient leaves the examination room, it should be properly disinfected. A disinfection checklist may be made to ensure uniform systematic cleaning. Alcohol and bleach-based disinfectants commonly used in health care settings are likely very effective against virus particles that cause COVID-19.¹⁰ During the disinfection process, gloves should be worn and careful attention paid to the contact time. Contact time is the amount of time the surface should appear visibly wet for proper disinfection. For example, Metrex CaviWipes have a recommended contact time of 3 minutes; however, this varies depending on type of virus and formulation, check labels or manufacturers’ websites for further directions.¹⁰ Also, the US Environmental Protection Agency has a database search available for disinfectants that meet their criteria for use against SARS-CoV-2.¹¹

In an ever-changing environment, we offer this article to help equip providers to deliver the best possible patient care when face-to-face encounters are necessary. Currently nonurgent eye care follow-up visits are being conducted by telephone or video clinics. It is our goal to inform fellow practitioners on options and strategies to elevate the safety of staff and patients while minimizing the risk of exposure.

COVID-19 is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and its symptoms range from mild to severe respiratory illness, fever, cough, fatigue, and shortness of breath.¹ Diarrhea is common early on with infection and loss of taste and smell have also been reported.¹ Follicular conjunctivitis has also been reported, either as an early sign of infection or during hospitalization for severe COVID-19 disease.^2-4 The incubation period of COVID-19 falls within 2 to 14 days according to the CDC.⁵

It has been confirmed that COVID-19 is transmitted through both respiratory droplets and direct contact. Another possible route of viral transmission is entry through aerosolized droplets into the tears, which then pass through the nasolacrimal ducts and into the respiratory tract.⁶

Preparations Prior to Office Visit

It is essential for the eye care provider to prioritize patient care in order of absolute necessity, such as sudden vision loss, sudden onset flashes and floaters, and eye trauma. In cases of potentially sight threatening pathology, it is in the best interest of the patient to conduct a face-to-face appointment. Therefore, it is important to implement new guidelines and protocols as we continue to see these patients (Figure 1).

Prior to the patient entering the medical facility, measures should be implemented to minimize exposure risk. This can be done over the telephone or at vehicle entrance screening stations. The triage technician answering the telephone should have a script of questions to ask. The patient should be instructed to come into the office alone unless, for physical or mental reasons, a caregiver is required.

SARS-CoV-2 Screening Questions

Preparedness through risk mitigation strategies are recommended with a targeted questionnaire and noncontact temperature check at the clinic or hospital entrance. Below are some general questions to further triage patients exposed to SARS-CoV-2.

• Do you have fever or any respiratory symptoms?
• Do you have new or worsening cough or shortness of breath?
• Do you have flulike symptoms?
• Have you been in close contact with someone, including health care workers, confirmed to have the COVID-19?

If the patient answers yes to any of the above questions, the CDC urges health care providers to immediately notify both infection control personnel at your health care facility and your local or state health department.^1,2 In regions currently managing significant outbreaks of COVID-19, the AAO recommends that eye care providers assume that any patient could be infected with SARS-CoV-2 and to proceed accordingly.² If urgent eye care is needed, a referral call should be made to a hospital or center equipped to deal with COVID-19 and urgent eye conditions. When calling the referral center, ensure adequate staffing and space and relay all pertinent information along with receiving approval from the treating physician.

Face-to-Face Office Visits

Once it has been determined that it is in the best interest of the patient to be seen in a face-to-face visit, the patient should be instructed to call the office when they arrive in the parking lot. The CDC recommends limiting points of entry upon arrival and during the visit.¹ As soon as an examination lane is ready, the patient can then be messaged to come into the office and escorted into the examination room.

An urgent or emergent ophthalmic examination for a patient with no respiratory symptoms, no fever, and no COVID-19 risk factors should include proper hand hygiene, use of personal protective equipment (PPE), and proper disinfection. Several studies have documented SARS-CoV-2 infection in asymptomatic and presymptomatic patients, making PPE of the up most importance.^2,7,8 PPE should include mask, face shield, and gloves. Currently, there are national and international shortages on PPE and a heightened topic of discussion concerning mask use, effectiveness with extended wear, and reuse. Please refer to the CDC and AAO websites for up-to-date guidelines (Table).^1,2 According to the CDC, N95 respirators are restricted to those performing or present for an aerosol-generating procedure.⁹

It is recommended that the eye care provider should only perform necessary tests and procedures. Noncontact tonometry should be avoided, as this might cause aerosolization of virus particles. The close proximity between eye care providers and their patients during slit-lamp examination may require further precautions to lower the risk of transmission via droplets or through hand to eye contact. The patient should be advised not to speak during the examination portion and the AAO also recommends a surgical mask or cloth face covering for the patient.² An additional protective device that may be used during the slit-lamp exam is a breath shield or a barrier shield (Figures 2 and 3).² Some manufacturers are offering clinicians free slit-lamp breath shields online.

Infection Prevention and Control Measures

Last, once the patient leaves the examination room, it should be properly disinfected. A disinfection checklist may be made to ensure uniform systematic cleaning. Alcohol and bleach-based disinfectants commonly used in health care settings are likely very effective against virus particles that cause COVID-19.¹⁰ During the disinfection process, gloves should be worn and careful attention paid to the contact time. Contact time is the amount of time the surface should appear visibly wet for proper disinfection. For example, Metrex CaviWipes have a recommended contact time of 3 minutes; however, this varies depending on type of virus and formulation, check labels or manufacturers’ websites for further directions.¹⁰ Also, the US Environmental Protection Agency has a database search available for disinfectants that meet their criteria for use against SARS-CoV-2.¹¹

In an ever-changing environment, we offer this article to help equip providers to deliver the best possible patient care when face-to-face encounters are necessary. Currently nonurgent eye care follow-up visits are being conducted by telephone or video clinics. It is our goal to inform fellow practitioners on options and strategies to elevate the safety of staff and patients while minimizing the risk of exposure.