Hospital Medicine

As of April 9, 2020, the Centers for Disease Control and Prevention (CDC) reported that 9,282 health care providers in the US had contracted COVID-19, and 27 had died of the virus.² Medscape reports the toll as much higher. Thousands more nurses, doctors, epidemiologists, social workers, physician assistants, dentists, pharmacists, and other health care workers from Italy, China, and dozens of other countries have died fighting this plague.³

The truth is no one knows how many health care workers are actually sick or even have died. State and federal governments have not been routinely and specifically tracking that data, making these already grim statistics likely a gross underestimation.⁴ While not all of these health care providers were exposed to COVID-19 in the line of duty, many were, and many more will be as the pandemic subsides in one epicenter only to erupt in another, and smolders for months until a vaccine quenches it.

Each of those lost lives of promise had a story of hard work and sacrifice to become a health care professional, of friends and family who loved and cared for them when ill, who need and grieve for them, now gone far too soon. Nor should we forget to mourn all of the administrative professionals, the line and support staff of health care facilities, who also perished fighting the pestilence. It is fitting then, that this second editorial in my pledge to write each month about COVID-19 until the pandemic ends, be about the duty to care and its limits.

The duty to care is among the most fundamental and ancient ethical obligations of health care providers. It is included even in modern codes of ethics like that of the American Medical Association and American Nurses Association. The obligation to not abandon patients is even more compelling for the Military Health System, Veterans Health Administration (VHA), and the US Public Health Service whose health care mission also is a public trust. The duty is rooted in the fiduciary nature of the health professions in which the interests of the patient should take priority over other considerations, including a risk to their own health and life. Prioritization though has exceptions. Physician and attorney David Orentlicher points out the unconditional obligation that bound physicians in the 14th century Black Death, or the 1918 Spanish influenza, now admits exceptions and qualifications.⁵

The exception that has become the object of greatest concern to health care workers is personal protective equipment (PPE). In modern public health ethics, health care systems and state and federal governments have a corresponding ethical obligation of reciprocity toward their employees whose work places them at elevated risk of harm—in this case, COVID-19 exposure. The principle of reciprocity encompasses the measures and materials that health care institutions need to provide to health care workers to reasonably minimize the risk of viral transmission. The reasonableness standard does not demand that there be zero risk. It does require that health care workers have adequate and appropriate PPE so that in fulfilling their duty to care they are not exposed to a disproportionate risk.

This last assertion has been the subject of controversy in the media and consternation on the part of health care professionals for several disconcerting reasons. First and foremost, a cascade failure on the part of government and industry has resulted in PPE being the scarcest health care resource in this pandemic.⁶ The shortage is as serious as that of the life-saving ventilators that are rightly at the center of most crisis standards resource allocation plans.⁷ Second, the guidance from the CDC and other authoritative sources continues to change. This is, in part, to adjust to the even more rapid pace of knowledge about the virus and its behavior and to adapt to the reality of insufficient PPE.⁸

Understandably, health care providers, especially those on the frontlines, may lose trust in the scientific experts and the leadership of their institutions, compounding the climate of moral distress in a public health crisis. Health care workers in the community, and even in federal service, have launched socially distanced protests and taken to social media to voice their concern and rally assistance.^9,10 In response, VHA Executive-in-Charge Richard Stone, MD, admitted that VHA does have a shortage of PPE in a Washington Post interview.¹¹ He outlined how the organization plans to address staff concerns. The article also reported only a 4% absentee rate of VHA staff as opposed to the 40% that plans predicted was possible. This demonstrates once more the dedication of VHA health care professionals and workers to fulfill their duty to care for veterans even amid fears about inadequate PPE.

In the epigraph, Albert Camus captures the uncertainty and fear that as humans all health care providers experience as they face the unpredictable but very real threat of COVID-19.¹ Camus expresses even more strongly the devotion to duty of health care providers to care for vulnerable ill patients in need despite the inherent threat in a highly transmissible and potentially deadly infection that is inextricably linked to that caring. Orentlicher wisely opines that the integrity of the health professions and their respected role in society benefit from a strong duty to care.⁵ The best way to promote that duty is to do all in our power to protect those who willingly brave the pestilence to treat, and hope and pray someday to cure COVID-19.

As of April 9, 2020, the Centers for Disease Control and Prevention (CDC) reported that 9,282 health care providers in the US had contracted COVID-19, and 27 had died of the virus.² Medscape reports the toll as much higher. Thousands more nurses, doctors, epidemiologists, social workers, physician assistants, dentists, pharmacists, and other health care workers from Italy, China, and dozens of other countries have died fighting this plague.³

The truth is no one knows how many health care workers are actually sick or even have died. State and federal governments have not been routinely and specifically tracking that data, making these already grim statistics likely a gross underestimation.⁴ While not all of these health care providers were exposed to COVID-19 in the line of duty, many were, and many more will be as the pandemic subsides in one epicenter only to erupt in another, and smolders for months until a vaccine quenches it.

Each of those lost lives of promise had a story of hard work and sacrifice to become a health care professional, of friends and family who loved and cared for them when ill, who need and grieve for them, now gone far too soon. Nor should we forget to mourn all of the administrative professionals, the line and support staff of health care facilities, who also perished fighting the pestilence. It is fitting then, that this second editorial in my pledge to write each month about COVID-19 until the pandemic ends, be about the duty to care and its limits.

The duty to care is among the most fundamental and ancient ethical obligations of health care providers. It is included even in modern codes of ethics like that of the American Medical Association and American Nurses Association. The obligation to not abandon patients is even more compelling for the Military Health System, Veterans Health Administration (VHA), and the US Public Health Service whose health care mission also is a public trust. The duty is rooted in the fiduciary nature of the health professions in which the interests of the patient should take priority over other considerations, including a risk to their own health and life. Prioritization though has exceptions. Physician and attorney David Orentlicher points out the unconditional obligation that bound physicians in the 14th century Black Death, or the 1918 Spanish influenza, now admits exceptions and qualifications.⁵

The exception that has become the object of greatest concern to health care workers is personal protective equipment (PPE). In modern public health ethics, health care systems and state and federal governments have a corresponding ethical obligation of reciprocity toward their employees whose work places them at elevated risk of harm—in this case, COVID-19 exposure. The principle of reciprocity encompasses the measures and materials that health care institutions need to provide to health care workers to reasonably minimize the risk of viral transmission. The reasonableness standard does not demand that there be zero risk. It does require that health care workers have adequate and appropriate PPE so that in fulfilling their duty to care they are not exposed to a disproportionate risk.

This last assertion has been the subject of controversy in the media and consternation on the part of health care professionals for several disconcerting reasons. First and foremost, a cascade failure on the part of government and industry has resulted in PPE being the scarcest health care resource in this pandemic.⁶ The shortage is as serious as that of the life-saving ventilators that are rightly at the center of most crisis standards resource allocation plans.⁷ Second, the guidance from the CDC and other authoritative sources continues to change. This is, in part, to adjust to the even more rapid pace of knowledge about the virus and its behavior and to adapt to the reality of insufficient PPE.⁸

Understandably, health care providers, especially those on the frontlines, may lose trust in the scientific experts and the leadership of their institutions, compounding the climate of moral distress in a public health crisis. Health care workers in the community, and even in federal service, have launched socially distanced protests and taken to social media to voice their concern and rally assistance.^9,10 In response, VHA Executive-in-Charge Richard Stone, MD, admitted that VHA does have a shortage of PPE in a Washington Post interview.¹¹ He outlined how the organization plans to address staff concerns. The article also reported only a 4% absentee rate of VHA staff as opposed to the 40% that plans predicted was possible. This demonstrates once more the dedication of VHA health care professionals and workers to fulfill their duty to care for veterans even amid fears about inadequate PPE.

In the epigraph, Albert Camus captures the uncertainty and fear that as humans all health care providers experience as they face the unpredictable but very real threat of COVID-19.¹ Camus expresses even more strongly the devotion to duty of health care providers to care for vulnerable ill patients in need despite the inherent threat in a highly transmissible and potentially deadly infection that is inextricably linked to that caring. Orentlicher wisely opines that the integrity of the health professions and their respected role in society benefit from a strong duty to care.⁵ The best way to promote that duty is to do all in our power to protect those who willingly brave the pestilence to treat, and hope and pray someday to cure COVID-19.

As of April 9, 2020, the Centers for Disease Control and Prevention (CDC) reported that 9,282 health care providers in the US had contracted COVID-19, and 27 had died of the virus.² Medscape reports the toll as much higher. Thousands more nurses, doctors, epidemiologists, social workers, physician assistants, dentists, pharmacists, and other health care workers from Italy, China, and dozens of other countries have died fighting this plague.³

The truth is no one knows how many health care workers are actually sick or even have died. State and federal governments have not been routinely and specifically tracking that data, making these already grim statistics likely a gross underestimation.⁴ While not all of these health care providers were exposed to COVID-19 in the line of duty, many were, and many more will be as the pandemic subsides in one epicenter only to erupt in another, and smolders for months until a vaccine quenches it.

Each of those lost lives of promise had a story of hard work and sacrifice to become a health care professional, of friends and family who loved and cared for them when ill, who need and grieve for them, now gone far too soon. Nor should we forget to mourn all of the administrative professionals, the line and support staff of health care facilities, who also perished fighting the pestilence. It is fitting then, that this second editorial in my pledge to write each month about COVID-19 until the pandemic ends, be about the duty to care and its limits.

The duty to care is among the most fundamental and ancient ethical obligations of health care providers. It is included even in modern codes of ethics like that of the American Medical Association and American Nurses Association. The obligation to not abandon patients is even more compelling for the Military Health System, Veterans Health Administration (VHA), and the US Public Health Service whose health care mission also is a public trust. The duty is rooted in the fiduciary nature of the health professions in which the interests of the patient should take priority over other considerations, including a risk to their own health and life. Prioritization though has exceptions. Physician and attorney David Orentlicher points out the unconditional obligation that bound physicians in the 14th century Black Death, or the 1918 Spanish influenza, now admits exceptions and qualifications.⁵

The exception that has become the object of greatest concern to health care workers is personal protective equipment (PPE). In modern public health ethics, health care systems and state and federal governments have a corresponding ethical obligation of reciprocity toward their employees whose work places them at elevated risk of harm—in this case, COVID-19 exposure. The principle of reciprocity encompasses the measures and materials that health care institutions need to provide to health care workers to reasonably minimize the risk of viral transmission. The reasonableness standard does not demand that there be zero risk. It does require that health care workers have adequate and appropriate PPE so that in fulfilling their duty to care they are not exposed to a disproportionate risk.

This last assertion has been the subject of controversy in the media and consternation on the part of health care professionals for several disconcerting reasons. First and foremost, a cascade failure on the part of government and industry has resulted in PPE being the scarcest health care resource in this pandemic.⁶ The shortage is as serious as that of the life-saving ventilators that are rightly at the center of most crisis standards resource allocation plans.⁷ Second, the guidance from the CDC and other authoritative sources continues to change. This is, in part, to adjust to the even more rapid pace of knowledge about the virus and its behavior and to adapt to the reality of insufficient PPE.⁸

Understandably, health care providers, especially those on the frontlines, may lose trust in the scientific experts and the leadership of their institutions, compounding the climate of moral distress in a public health crisis. Health care workers in the community, and even in federal service, have launched socially distanced protests and taken to social media to voice their concern and rally assistance.^9,10 In response, VHA Executive-in-Charge Richard Stone, MD, admitted that VHA does have a shortage of PPE in a Washington Post interview.¹¹ He outlined how the organization plans to address staff concerns. The article also reported only a 4% absentee rate of VHA staff as opposed to the 40% that plans predicted was possible. This demonstrates once more the dedication of VHA health care professionals and workers to fulfill their duty to care for veterans even amid fears about inadequate PPE.

In the epigraph, Albert Camus captures the uncertainty and fear that as humans all health care providers experience as they face the unpredictable but very real threat of COVID-19.¹ Camus expresses even more strongly the devotion to duty of health care providers to care for vulnerable ill patients in need despite the inherent threat in a highly transmissible and potentially deadly infection that is inextricably linked to that caring. Orentlicher wisely opines that the integrity of the health professions and their respected role in society benefit from a strong duty to care.⁵ The best way to promote that duty is to do all in our power to protect those who willingly brave the pestilence to treat, and hope and pray someday to cure COVID-19.

“Takers believe in a zero-sum world, and they end up creating one where bosses, colleagues and clients don’t trust them. Givers build deeper and broader relationships—people are rooting for them instead of gunning for them.”

—Adam Grant

To succeed in a hospital, leaders need a generous supply of social and political capital. House officers learn this very quickly, especially when they are relying on other members of the healthcare team to obtain tests and studies for their patients and calling for specialty consultations. To be successful and efficient, building relationships and trust is key. Such capital, unfortunately, takes time to develop. Therefore, healthcare leaders and clinicians at all levels of training need to make an everyday investment of goodwill and friendliness with those they encounter. The dividends may be slow in coming, but they are substantial and sustained. Friends give you the benefit of the doubt—and help you when you are most in need.

Having friends (or friendly colleagues) at work is beneficial both professionally and personally. The benefits of social interactions have been studied for years and even more so in recent times with the dramatic increase in the use of handheld devices. Eye contact between casual acquaintances passing each other in the hallway is replaced with eyes focused downward on smartphones. The result? We are becoming more socially isolated. Our personal solution? When we see professional colleagues (or patients and families in the hallways of our hospital), we nod in acknowledgement with appropriate eye contact and say “Good morning” or “Hello” even if we don’t know them—even if their eyes are focused on their devices as they walk past you in the hallway. You get a gold star if you remember the names of the professional colleagues you see frequently in the hallways or around the hospital.

This isn’t soft science; it’s backed by hard data. When we conduct site visits of different hospitals around the country to help them improve their care quality and performance, we informally divide hospitals into two groups: The “How ya doin’?” hospitals vs the “Rec-Ignore” hospitals (in which employees recognize a colleague in the hallway but choose to not acknowledge them). Most prefer to work at a “How ya doin’?” hospital. Being friendly has been linked to increased team spirit and morale, knowledge sharing, trust, prevention of burnout, and sense of a positive working environment. It also makes you feel better about yourself—and makes other people feel similarly as well.

We’ll share an example from a search for a new department chair. The dean went on reverse site visits to meet the two finalists in their home institutions and asked them for tours of their hospitals. Candidate A walked around and it seemed like everyone knew her. She smiled and said hello to the people she came in contact with during the tour. Not so for candidate B—just the opposite. Guess which candidate the dean hired?

Put away your phone, interact with your colleagues, and learn to make small talk, and not just with your supervisors or peers. Chitchat is an important “social lubricant,” fostering a sense of community and teamwork. It helps bring down the divides that come from organizational hierarchies. It helps endear you to your staff.

Developing a reputation as a nice person who is quick with a smile and even quicker with a “How ya doin’?” pays off in the end. This reputation also makes it easier to give bad news, something that all leaders must do at some point. So make a friend before you need one—it usually will pay dividends.

“Takers believe in a zero-sum world, and they end up creating one where bosses, colleagues and clients don’t trust them. Givers build deeper and broader relationships—people are rooting for them instead of gunning for them.”

—Adam Grant

To succeed in a hospital, leaders need a generous supply of social and political capital. House officers learn this very quickly, especially when they are relying on other members of the healthcare team to obtain tests and studies for their patients and calling for specialty consultations. To be successful and efficient, building relationships and trust is key. Such capital, unfortunately, takes time to develop. Therefore, healthcare leaders and clinicians at all levels of training need to make an everyday investment of goodwill and friendliness with those they encounter. The dividends may be slow in coming, but they are substantial and sustained. Friends give you the benefit of the doubt—and help you when you are most in need.

Having friends (or friendly colleagues) at work is beneficial both professionally and personally. The benefits of social interactions have been studied for years and even more so in recent times with the dramatic increase in the use of handheld devices. Eye contact between casual acquaintances passing each other in the hallway is replaced with eyes focused downward on smartphones. The result? We are becoming more socially isolated. Our personal solution? When we see professional colleagues (or patients and families in the hallways of our hospital), we nod in acknowledgement with appropriate eye contact and say “Good morning” or “Hello” even if we don’t know them—even if their eyes are focused on their devices as they walk past you in the hallway. You get a gold star if you remember the names of the professional colleagues you see frequently in the hallways or around the hospital.

This isn’t soft science; it’s backed by hard data. When we conduct site visits of different hospitals around the country to help them improve their care quality and performance, we informally divide hospitals into two groups: The “How ya doin’?” hospitals vs the “Rec-Ignore” hospitals (in which employees recognize a colleague in the hallway but choose to not acknowledge them). Most prefer to work at a “How ya doin’?” hospital. Being friendly has been linked to increased team spirit and morale, knowledge sharing, trust, prevention of burnout, and sense of a positive working environment. It also makes you feel better about yourself—and makes other people feel similarly as well.

We’ll share an example from a search for a new department chair. The dean went on reverse site visits to meet the two finalists in their home institutions and asked them for tours of their hospitals. Candidate A walked around and it seemed like everyone knew her. She smiled and said hello to the people she came in contact with during the tour. Not so for candidate B—just the opposite. Guess which candidate the dean hired?

Put away your phone, interact with your colleagues, and learn to make small talk, and not just with your supervisors or peers. Chitchat is an important “social lubricant,” fostering a sense of community and teamwork. It helps bring down the divides that come from organizational hierarchies. It helps endear you to your staff.

Developing a reputation as a nice person who is quick with a smile and even quicker with a “How ya doin’?” pays off in the end. This reputation also makes it easier to give bad news, something that all leaders must do at some point. So make a friend before you need one—it usually will pay dividends.

“Takers believe in a zero-sum world, and they end up creating one where bosses, colleagues and clients don’t trust them. Givers build deeper and broader relationships—people are rooting for them instead of gunning for them.”

—Adam Grant

To succeed in a hospital, leaders need a generous supply of social and political capital. House officers learn this very quickly, especially when they are relying on other members of the healthcare team to obtain tests and studies for their patients and calling for specialty consultations. To be successful and efficient, building relationships and trust is key. Such capital, unfortunately, takes time to develop. Therefore, healthcare leaders and clinicians at all levels of training need to make an everyday investment of goodwill and friendliness with those they encounter. The dividends may be slow in coming, but they are substantial and sustained. Friends give you the benefit of the doubt—and help you when you are most in need.

Having friends (or friendly colleagues) at work is beneficial both professionally and personally. The benefits of social interactions have been studied for years and even more so in recent times with the dramatic increase in the use of handheld devices. Eye contact between casual acquaintances passing each other in the hallway is replaced with eyes focused downward on smartphones. The result? We are becoming more socially isolated. Our personal solution? When we see professional colleagues (or patients and families in the hallways of our hospital), we nod in acknowledgement with appropriate eye contact and say “Good morning” or “Hello” even if we don’t know them—even if their eyes are focused on their devices as they walk past you in the hallway. You get a gold star if you remember the names of the professional colleagues you see frequently in the hallways or around the hospital.

This isn’t soft science; it’s backed by hard data. When we conduct site visits of different hospitals around the country to help them improve their care quality and performance, we informally divide hospitals into two groups: The “How ya doin’?” hospitals vs the “Rec-Ignore” hospitals (in which employees recognize a colleague in the hallway but choose to not acknowledge them). Most prefer to work at a “How ya doin’?” hospital. Being friendly has been linked to increased team spirit and morale, knowledge sharing, trust, prevention of burnout, and sense of a positive working environment. It also makes you feel better about yourself—and makes other people feel similarly as well.

We’ll share an example from a search for a new department chair. The dean went on reverse site visits to meet the two finalists in their home institutions and asked them for tours of their hospitals. Candidate A walked around and it seemed like everyone knew her. She smiled and said hello to the people she came in contact with during the tour. Not so for candidate B—just the opposite. Guess which candidate the dean hired?

Put away your phone, interact with your colleagues, and learn to make small talk, and not just with your supervisors or peers. Chitchat is an important “social lubricant,” fostering a sense of community and teamwork. It helps bring down the divides that come from organizational hierarchies. It helps endear you to your staff.

Developing a reputation as a nice person who is quick with a smile and even quicker with a “How ya doin’?” pays off in the end. This reputation also makes it easier to give bad news, something that all leaders must do at some point. So make a friend before you need one—it usually will pay dividends.

The first case of coronavirus disease 2019 (COVID-19) in the United States was identified in Washington state in late January 2020. As of mid-April 2020, the number of US cases has increased to more than 800,000 with over 40,000 deaths. The limited available knowledge to guide medical decision-making combined with rapid progression of the pandemic has resulted in an urgent need to better define clinical, radiologic, and laboratory features of the disease, predictors of disease progression, predominant modes of transmission, and effective treatments. This urgency has led to a flood of manuscript submissions, which strains the scientific vetting process and leads to the spread of medical misinformation and potential for serious harm. As an example, a small observational (noncontrolled) study that used an antimalarial drug to treat COVID-19 patients was touted by several national leaders as proof of its effectiveness, despite substantial methodologic limitations.^1,2 While the article has not yet been retracted, the International Society of Antimicrobial Chemotherapy, the publishing journal’s society sponsor, subsequently issued a statement that “the article does not meet the Society’s expected standard.”³

With these concerns in mind, we recognize the importance of addressing the current pandemic and identifying areas where we can advance the field responsibly in the face of limited evidence in a rapidly evolving situation. Hospitalists throughout the world are facing unprecedented leadership challenges, navigating ethical stressors, and redesigning their care systems while learning rapidly and adapting nimbly. In this issue, we share leadership strategies, explore ethical challenges and controversies, describe successful practices, and provide personal reflections from a diverse group of hospitalists and leaders. As a journal, we have intentionally avoided rapid publication of articles with substantial methodologic limitations that are unlikely to advance our knowledge of COVID-19 even though such articles may generate substantial media coverage. Different regions of the country are at different stages of the pandemic; some hospitals are experiencing high patient volumes and struggling with shortages of equipment and supplies, while others are weeks away from peak disease activity or have avoided periods of high prevalence altogether. These varied experiences offer an opportunity to share our learnings and perspectives as we wait for more definitive evidence on best management practices. As part of our commitment to our colleagues in healthcare and to the broader scientific community, all Journal of Hospital Medicine articles related to COVID-19 and published during the pandemic will be open access (ie, freely accessible).

The first case of coronavirus disease 2019 (COVID-19) in the United States was identified in Washington state in late January 2020. As of mid-April 2020, the number of US cases has increased to more than 800,000 with over 40,000 deaths. The limited available knowledge to guide medical decision-making combined with rapid progression of the pandemic has resulted in an urgent need to better define clinical, radiologic, and laboratory features of the disease, predictors of disease progression, predominant modes of transmission, and effective treatments. This urgency has led to a flood of manuscript submissions, which strains the scientific vetting process and leads to the spread of medical misinformation and potential for serious harm. As an example, a small observational (noncontrolled) study that used an antimalarial drug to treat COVID-19 patients was touted by several national leaders as proof of its effectiveness, despite substantial methodologic limitations.^1,2 While the article has not yet been retracted, the International Society of Antimicrobial Chemotherapy, the publishing journal’s society sponsor, subsequently issued a statement that “the article does not meet the Society’s expected standard.”³

With these concerns in mind, we recognize the importance of addressing the current pandemic and identifying areas where we can advance the field responsibly in the face of limited evidence in a rapidly evolving situation. Hospitalists throughout the world are facing unprecedented leadership challenges, navigating ethical stressors, and redesigning their care systems while learning rapidly and adapting nimbly. In this issue, we share leadership strategies, explore ethical challenges and controversies, describe successful practices, and provide personal reflections from a diverse group of hospitalists and leaders. As a journal, we have intentionally avoided rapid publication of articles with substantial methodologic limitations that are unlikely to advance our knowledge of COVID-19 even though such articles may generate substantial media coverage. Different regions of the country are at different stages of the pandemic; some hospitals are experiencing high patient volumes and struggling with shortages of equipment and supplies, while others are weeks away from peak disease activity or have avoided periods of high prevalence altogether. These varied experiences offer an opportunity to share our learnings and perspectives as we wait for more definitive evidence on best management practices. As part of our commitment to our colleagues in healthcare and to the broader scientific community, all Journal of Hospital Medicine articles related to COVID-19 and published during the pandemic will be open access (ie, freely accessible).

The first case of coronavirus disease 2019 (COVID-19) in the United States was identified in Washington state in late January 2020. As of mid-April 2020, the number of US cases has increased to more than 800,000 with over 40,000 deaths. The limited available knowledge to guide medical decision-making combined with rapid progression of the pandemic has resulted in an urgent need to better define clinical, radiologic, and laboratory features of the disease, predictors of disease progression, predominant modes of transmission, and effective treatments. This urgency has led to a flood of manuscript submissions, which strains the scientific vetting process and leads to the spread of medical misinformation and potential for serious harm. As an example, a small observational (noncontrolled) study that used an antimalarial drug to treat COVID-19 patients was touted by several national leaders as proof of its effectiveness, despite substantial methodologic limitations.^1,2 While the article has not yet been retracted, the International Society of Antimicrobial Chemotherapy, the publishing journal’s society sponsor, subsequently issued a statement that “the article does not meet the Society’s expected standard.”³

With these concerns in mind, we recognize the importance of addressing the current pandemic and identifying areas where we can advance the field responsibly in the face of limited evidence in a rapidly evolving situation. Hospitalists throughout the world are facing unprecedented leadership challenges, navigating ethical stressors, and redesigning their care systems while learning rapidly and adapting nimbly. In this issue, we share leadership strategies, explore ethical challenges and controversies, describe successful practices, and provide personal reflections from a diverse group of hospitalists and leaders. As a journal, we have intentionally avoided rapid publication of articles with substantial methodologic limitations that are unlikely to advance our knowledge of COVID-19 even though such articles may generate substantial media coverage. Different regions of the country are at different stages of the pandemic; some hospitals are experiencing high patient volumes and struggling with shortages of equipment and supplies, while others are weeks away from peak disease activity or have avoided periods of high prevalence altogether. These varied experiences offer an opportunity to share our learnings and perspectives as we wait for more definitive evidence on best management practices. As part of our commitment to our colleagues in healthcare and to the broader scientific community, all Journal of Hospital Medicine articles related to COVID-19 and published during the pandemic will be open access (ie, freely accessible).

The hospitalist admitted a 56-year-old man with hypertension and hyperlipidemia to the general medical unit for community-­acquired pneumonia and started him on appropriate antimicrobial therapy. On the evening of admission, the nurse woke the patient to take his vital signs and noted a fever of 39.1°C (102.4°F). The patient had a pulse of 90 beats per minute, normal blood pressure, and a stable supplemental oxygen requirement via nasal cannula. The nurse noted an oral acetaminophen “as needed” order for fever. She woke the patient again to administer acetaminophen and notified the hospitalist.

Hospitalists frequently encounter febrile patients. According to one large hospital survey, fever occurs in 25% of pediatric and 31% of adult medical patients.¹ Fever in hospitalized patients most commonly results from infection, but autoimmune disease, malignancy, and an array of other inflammatory conditions cause fevers as well.¹

Defined as an elevated body temperature resulting from a raised hypothalamic set point², hospitalists often treat fever with acetaminophen or nonsteroidal anti-inflammatory drugs (NSAIDs). These routinely administered medications act centrally to temporarily lower the hypothalamic set point and relieve fever.^2,3 Standard hospital admission order sets commonly include an as-needed antipyretic every 4 to 6 hours for treatment of fever, regardless of the presence of fever-related symptoms.

Fever is differentiated from hyperthermia, where temperature increases because of dysregulated peripheral processes despite a normal hypothalamic set point.² Examples of hyperthermia include heat stroke, malignant hyperthermia, and neuroleptic malignant syndrome. Notably, antipyretic medications have no effect on hyperthermia, but physical means, such as cooling blankets, can lead to temperature reduction.²

Hospitalists prescribe antipyretic medication to alleviate fever-­related symptoms, including headache, chills and sweats, and joint and muscle aches.³ While researchers have sparingly studied this practice, available evidence and experience suggest that fever-related symptoms decline in parallel with defervescence after administration of acetaminophen or NSAIDs in both adult and pediatric populations.^4,5 One randomized, controlled, double-blind study of nearly 400 adult outpatients in Germany with febrile upper respiratory tract infections showed that both aspirin and acetaminophen bested the placebo in reducing fever and associated headache, achiness, and discomfort over a span of 6 hours.⁴ In another study, this time with pediatric patients hospitalized with fever and uncomplicated respiratory tract infections, patients who received acetaminophen had statistically significant improvements in activity, alertness, mood, comfort, appetite, and fluid intake 6 hours after receiving that therapy.⁵

Physicians, nurses, and caregivers also commonly believe that fever is inherently noxious and that treatment of infection-­related fever contributes to fighting the infection itself.^2,3,6 The pediatric literature describes parents, caretakers, and clinicians who suffer from “fever phobia,” the worry that fevers contribute to long-term neurologic complications, recurrent febrile seizures, and death.^6,7

Finally, healthcare providers administer antipyretic medication to mitigate the demand fever places on the cardiovascular and pulmonary systems.³ An elevated temperature increases the body’s metabolic rate, oxygen consumption, and cardiac output that critically ill patients who have acute and/or chronic compromise to those systems may not tolerate. For example, patients requiring pressor support for hemodynamic shock or mechanical ventilation for respiratory failure may not tolerate an elevated temperature.⁸

Fever serves as an adaptive host response to infection, boosting innate and adaptive immunity in a multitude of ways.⁸ In animal models, fever slows the replication of pathogenic bacteria and enhances the activity of antibiotic agents.⁸ In vitro studies demonstrate that fever increases mobility of leukocytes, phagocytic activity, and proliferation of T cells.⁸ Retrospective case-control studies of patients hospitalized with severe bacterial illnesses, including gram-negative bacteremia, spontaneous bacterial peritonitis, and community-­acquired pneumonia, found that patients with a documented febrile response had increased survival compared with those who remained afebrile during the infection.⁹ In addition, a large retrospective cohort study of septic ICU patients found a progressive decline in mortality in association with increasing peak temperature on the day of ICU admission.¹⁰

In addition to the above studies supporting the important role of fever in fighting infection, recent evidence definitively demonstrates no mortality or morbidity benefit of using antipyretic medications in infected patients. A 2017 meta-analysis that included eight observational and eight randomized studies, totaling 18,939 adult septic ICU patients, demonstrated no difference in hospital and 28-day mortality in patients treated with antipyretics vs those who were not.¹¹ The authors again found no mortality benefit with antipyretic use when separately analyzing data from only the randomized controlled trials (1,507 patients) or when stratifying patients based on the type of antipyretic received (acetaminophen, NSAIDs, or physical cooling).¹¹ They reported no differences in predefined secondary outcomes of shock reversal or nosocomial infections. The authors commented that these robust results likely would not change even with more data from additional trials. In children, a recent meta-analysis of three randomized controlled trials (540 patients) did not find the use of acetaminophen, ibuprofen, or diclofenac effective in preventing febrile seizures.¹²Pediatric practice guidelines consistently recommend using antipyretic medication to alleviate discomfort caused by fever and not solely to reduce temperature.^13,14

Antipyretic agents interfere with the effectiveness of the body’s immune response, as demonstrated in a number of infectious diseases.^2,15-18 Two randomized controlled studies conducted in healthy adult volunteers challenged with rhinovirus reported increased viral shedding and decreased antibody response in those subjects who received aspirin or acetaminophen, compared with those given placebo.^15,16 In another randomized controlled trial conducted in African children with malaria, paracetamol use delayed parasite clearance by 16 hours.¹⁷ A large case-control study correlated the use of NSAIDs with an increased risk of severe skin and soft-tissue complications in children with varicella and in adults with varicella zoster. ¹⁸ The international scientific community has raised concerns about worse outcomes with NSAID use in patients with COVID-19, the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); NSAIDs should be avoided in stable patients with COVID-19 until more data are available. ¹⁹

Additional risks and potential harms accompany antipyretic fever therapy. First, NSAIDs or acetaminophen may adversely affect patients with renal or hepatic insufficiency.^2,3 Second, masking fevers may impair the clinician’s ability to diagnose or evaluate response to treatment. Third, unnecessarily waking a sleeping patient to check temperature or administer unneeded antipyretics can contribute to hospital-associated problems, including delirium, insomnia, and falls. Treating these iatrogenic problems in turn may require additional medications or interventions. These unintended consequences may potentially prolong hospital stays, increase medication errors and polypharmacy, and detract from a patient’s overall healing and recovery.

While the use of antipyretic medications improves fever-­related symptoms, it comes at the cost of blunting a protective host response and exposes patients to medication risks without providing a clinical benefit. In sleeping, asymptomatic, or minimally symptomatic hospitalized patients, the risks of administering antipyretic medications clearly outweigh the benefits.

Treatment with antipyretic medication can alleviate fever-­related symptoms in those patients who have significant headache, body aches, chills, or sweats and in pediatric patients with notable malaise, irritability, or poor oral intake. Debate continues on the use of antipyretics in the ICU setting when managing critically ill patients with severe cardiopulmonary compromise who may not tolerate the additional hemodynamic strain a fever produces (eg, patients with shock requiring vasopressor support or respiratory failure requiring mechanical ventilation). Remember, decrease body temperature in hyperthermia syndromes by physical means.

Withhold antipyretic medication (ie, allow permissive fever) in well-appearing general medical patients with asymptomatic infection-related fevers. In patients who tolerate fever with minimal or no symptoms, potential benefits of permissive fever include decreased time to infection resolution and/or decreased risk of hospital-acquired infections. This may result in shorter hospital stays and significant cost savings. If we do not treat patients with asymptomatic fevers, then it follows that we should not check overnight temperatures in hospitalized patients sleeping comfortably.

• Do not order as-needed antipyretic medication for stable patients on general medical units with infection solely to reduce temperature or achieve normothermia.
• Only treat infected febrile patients with antipyretic medications for fever-related symptoms (headache, chills, or body aches or, in pediatric patients, irritability, malaise, or poor oral intake).
• Treat pathologically elevated temperatures (ie, hyperthermia syndromes) with physical measures because antipyretic medications will be ineffective.

In the clinical scenario, the hospitalist admitted the patient in stable condition for treatment of a community-acquired pneumonia. He mounted a febrile response to infection, which suggests that his active immune system may aid in recovery. The nurse noted the fever while the patient slept comfortably without fever-related symptoms.

After discussing these facts with the patient’s concerned nurse, the clinician should discontinue the order for as-needed acetaminophen for fever and instead recommend permissive fever without administration of antipyretic medication. This may facilitate recovery, avoid unnecessary polypharmacy, and allow the medical care team to follow his fever curve to ensure that the infection is adequately treated. If the patient develops bothersome fever-related symptoms, the hospitalist can reasonably treat with a single-dose of acetaminophen or NSAID.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason^™”? Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to propose ideas for other “Things We Do for No Reason” topics by emailing TWDFNR@hospitalmedicine.org.

The hospitalist admitted a 56-year-old man with hypertension and hyperlipidemia to the general medical unit for community-­acquired pneumonia and started him on appropriate antimicrobial therapy. On the evening of admission, the nurse woke the patient to take his vital signs and noted a fever of 39.1°C (102.4°F). The patient had a pulse of 90 beats per minute, normal blood pressure, and a stable supplemental oxygen requirement via nasal cannula. The nurse noted an oral acetaminophen “as needed” order for fever. She woke the patient again to administer acetaminophen and notified the hospitalist.

Hospitalists frequently encounter febrile patients. According to one large hospital survey, fever occurs in 25% of pediatric and 31% of adult medical patients.¹ Fever in hospitalized patients most commonly results from infection, but autoimmune disease, malignancy, and an array of other inflammatory conditions cause fevers as well.¹

Defined as an elevated body temperature resulting from a raised hypothalamic set point², hospitalists often treat fever with acetaminophen or nonsteroidal anti-inflammatory drugs (NSAIDs). These routinely administered medications act centrally to temporarily lower the hypothalamic set point and relieve fever.^2,3 Standard hospital admission order sets commonly include an as-needed antipyretic every 4 to 6 hours for treatment of fever, regardless of the presence of fever-related symptoms.

Fever is differentiated from hyperthermia, where temperature increases because of dysregulated peripheral processes despite a normal hypothalamic set point.² Examples of hyperthermia include heat stroke, malignant hyperthermia, and neuroleptic malignant syndrome. Notably, antipyretic medications have no effect on hyperthermia, but physical means, such as cooling blankets, can lead to temperature reduction.²

Hospitalists prescribe antipyretic medication to alleviate fever-­related symptoms, including headache, chills and sweats, and joint and muscle aches.³ While researchers have sparingly studied this practice, available evidence and experience suggest that fever-related symptoms decline in parallel with defervescence after administration of acetaminophen or NSAIDs in both adult and pediatric populations.^4,5 One randomized, controlled, double-blind study of nearly 400 adult outpatients in Germany with febrile upper respiratory tract infections showed that both aspirin and acetaminophen bested the placebo in reducing fever and associated headache, achiness, and discomfort over a span of 6 hours.⁴ In another study, this time with pediatric patients hospitalized with fever and uncomplicated respiratory tract infections, patients who received acetaminophen had statistically significant improvements in activity, alertness, mood, comfort, appetite, and fluid intake 6 hours after receiving that therapy.⁵

Physicians, nurses, and caregivers also commonly believe that fever is inherently noxious and that treatment of infection-­related fever contributes to fighting the infection itself.^2,3,6 The pediatric literature describes parents, caretakers, and clinicians who suffer from “fever phobia,” the worry that fevers contribute to long-term neurologic complications, recurrent febrile seizures, and death.^6,7

Finally, healthcare providers administer antipyretic medication to mitigate the demand fever places on the cardiovascular and pulmonary systems.³ An elevated temperature increases the body’s metabolic rate, oxygen consumption, and cardiac output that critically ill patients who have acute and/or chronic compromise to those systems may not tolerate. For example, patients requiring pressor support for hemodynamic shock or mechanical ventilation for respiratory failure may not tolerate an elevated temperature.⁸

Fever serves as an adaptive host response to infection, boosting innate and adaptive immunity in a multitude of ways.⁸ In animal models, fever slows the replication of pathogenic bacteria and enhances the activity of antibiotic agents.⁸ In vitro studies demonstrate that fever increases mobility of leukocytes, phagocytic activity, and proliferation of T cells.⁸ Retrospective case-control studies of patients hospitalized with severe bacterial illnesses, including gram-negative bacteremia, spontaneous bacterial peritonitis, and community-­acquired pneumonia, found that patients with a documented febrile response had increased survival compared with those who remained afebrile during the infection.⁹ In addition, a large retrospective cohort study of septic ICU patients found a progressive decline in mortality in association with increasing peak temperature on the day of ICU admission.¹⁰

In addition to the above studies supporting the important role of fever in fighting infection, recent evidence definitively demonstrates no mortality or morbidity benefit of using antipyretic medications in infected patients. A 2017 meta-analysis that included eight observational and eight randomized studies, totaling 18,939 adult septic ICU patients, demonstrated no difference in hospital and 28-day mortality in patients treated with antipyretics vs those who were not.¹¹ The authors again found no mortality benefit with antipyretic use when separately analyzing data from only the randomized controlled trials (1,507 patients) or when stratifying patients based on the type of antipyretic received (acetaminophen, NSAIDs, or physical cooling).¹¹ They reported no differences in predefined secondary outcomes of shock reversal or nosocomial infections. The authors commented that these robust results likely would not change even with more data from additional trials. In children, a recent meta-analysis of three randomized controlled trials (540 patients) did not find the use of acetaminophen, ibuprofen, or diclofenac effective in preventing febrile seizures.¹²Pediatric practice guidelines consistently recommend using antipyretic medication to alleviate discomfort caused by fever and not solely to reduce temperature.^13,14

Antipyretic agents interfere with the effectiveness of the body’s immune response, as demonstrated in a number of infectious diseases.^2,15-18 Two randomized controlled studies conducted in healthy adult volunteers challenged with rhinovirus reported increased viral shedding and decreased antibody response in those subjects who received aspirin or acetaminophen, compared with those given placebo.^15,16 In another randomized controlled trial conducted in African children with malaria, paracetamol use delayed parasite clearance by 16 hours.¹⁷ A large case-control study correlated the use of NSAIDs with an increased risk of severe skin and soft-tissue complications in children with varicella and in adults with varicella zoster. ¹⁸ The international scientific community has raised concerns about worse outcomes with NSAID use in patients with COVID-19, the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); NSAIDs should be avoided in stable patients with COVID-19 until more data are available. ¹⁹

Additional risks and potential harms accompany antipyretic fever therapy. First, NSAIDs or acetaminophen may adversely affect patients with renal or hepatic insufficiency.^2,3 Second, masking fevers may impair the clinician’s ability to diagnose or evaluate response to treatment. Third, unnecessarily waking a sleeping patient to check temperature or administer unneeded antipyretics can contribute to hospital-associated problems, including delirium, insomnia, and falls. Treating these iatrogenic problems in turn may require additional medications or interventions. These unintended consequences may potentially prolong hospital stays, increase medication errors and polypharmacy, and detract from a patient’s overall healing and recovery.

While the use of antipyretic medications improves fever-­related symptoms, it comes at the cost of blunting a protective host response and exposes patients to medication risks without providing a clinical benefit. In sleeping, asymptomatic, or minimally symptomatic hospitalized patients, the risks of administering antipyretic medications clearly outweigh the benefits.

Treatment with antipyretic medication can alleviate fever-­related symptoms in those patients who have significant headache, body aches, chills, or sweats and in pediatric patients with notable malaise, irritability, or poor oral intake. Debate continues on the use of antipyretics in the ICU setting when managing critically ill patients with severe cardiopulmonary compromise who may not tolerate the additional hemodynamic strain a fever produces (eg, patients with shock requiring vasopressor support or respiratory failure requiring mechanical ventilation). Remember, decrease body temperature in hyperthermia syndromes by physical means.

Withhold antipyretic medication (ie, allow permissive fever) in well-appearing general medical patients with asymptomatic infection-related fevers. In patients who tolerate fever with minimal or no symptoms, potential benefits of permissive fever include decreased time to infection resolution and/or decreased risk of hospital-acquired infections. This may result in shorter hospital stays and significant cost savings. If we do not treat patients with asymptomatic fevers, then it follows that we should not check overnight temperatures in hospitalized patients sleeping comfortably.

• Do not order as-needed antipyretic medication for stable patients on general medical units with infection solely to reduce temperature or achieve normothermia.
• Only treat infected febrile patients with antipyretic medications for fever-related symptoms (headache, chills, or body aches or, in pediatric patients, irritability, malaise, or poor oral intake).
• Treat pathologically elevated temperatures (ie, hyperthermia syndromes) with physical measures because antipyretic medications will be ineffective.

In the clinical scenario, the hospitalist admitted the patient in stable condition for treatment of a community-acquired pneumonia. He mounted a febrile response to infection, which suggests that his active immune system may aid in recovery. The nurse noted the fever while the patient slept comfortably without fever-related symptoms.

After discussing these facts with the patient’s concerned nurse, the clinician should discontinue the order for as-needed acetaminophen for fever and instead recommend permissive fever without administration of antipyretic medication. This may facilitate recovery, avoid unnecessary polypharmacy, and allow the medical care team to follow his fever curve to ensure that the infection is adequately treated. If the patient develops bothersome fever-related symptoms, the hospitalist can reasonably treat with a single-dose of acetaminophen or NSAID.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason^™”? Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to propose ideas for other “Things We Do for No Reason” topics by emailing TWDFNR@hospitalmedicine.org.

The hospitalist admitted a 56-year-old man with hypertension and hyperlipidemia to the general medical unit for community-­acquired pneumonia and started him on appropriate antimicrobial therapy. On the evening of admission, the nurse woke the patient to take his vital signs and noted a fever of 39.1°C (102.4°F). The patient had a pulse of 90 beats per minute, normal blood pressure, and a stable supplemental oxygen requirement via nasal cannula. The nurse noted an oral acetaminophen “as needed” order for fever. She woke the patient again to administer acetaminophen and notified the hospitalist.

Hospitalists frequently encounter febrile patients. According to one large hospital survey, fever occurs in 25% of pediatric and 31% of adult medical patients.¹ Fever in hospitalized patients most commonly results from infection, but autoimmune disease, malignancy, and an array of other inflammatory conditions cause fevers as well.¹

Defined as an elevated body temperature resulting from a raised hypothalamic set point², hospitalists often treat fever with acetaminophen or nonsteroidal anti-inflammatory drugs (NSAIDs). These routinely administered medications act centrally to temporarily lower the hypothalamic set point and relieve fever.^2,3 Standard hospital admission order sets commonly include an as-needed antipyretic every 4 to 6 hours for treatment of fever, regardless of the presence of fever-related symptoms.

Fever is differentiated from hyperthermia, where temperature increases because of dysregulated peripheral processes despite a normal hypothalamic set point.² Examples of hyperthermia include heat stroke, malignant hyperthermia, and neuroleptic malignant syndrome. Notably, antipyretic medications have no effect on hyperthermia, but physical means, such as cooling blankets, can lead to temperature reduction.²

Hospitalists prescribe antipyretic medication to alleviate fever-­related symptoms, including headache, chills and sweats, and joint and muscle aches.³ While researchers have sparingly studied this practice, available evidence and experience suggest that fever-related symptoms decline in parallel with defervescence after administration of acetaminophen or NSAIDs in both adult and pediatric populations.^4,5 One randomized, controlled, double-blind study of nearly 400 adult outpatients in Germany with febrile upper respiratory tract infections showed that both aspirin and acetaminophen bested the placebo in reducing fever and associated headache, achiness, and discomfort over a span of 6 hours.⁴ In another study, this time with pediatric patients hospitalized with fever and uncomplicated respiratory tract infections, patients who received acetaminophen had statistically significant improvements in activity, alertness, mood, comfort, appetite, and fluid intake 6 hours after receiving that therapy.⁵

Physicians, nurses, and caregivers also commonly believe that fever is inherently noxious and that treatment of infection-­related fever contributes to fighting the infection itself.^2,3,6 The pediatric literature describes parents, caretakers, and clinicians who suffer from “fever phobia,” the worry that fevers contribute to long-term neurologic complications, recurrent febrile seizures, and death.^6,7

Finally, healthcare providers administer antipyretic medication to mitigate the demand fever places on the cardiovascular and pulmonary systems.³ An elevated temperature increases the body’s metabolic rate, oxygen consumption, and cardiac output that critically ill patients who have acute and/or chronic compromise to those systems may not tolerate. For example, patients requiring pressor support for hemodynamic shock or mechanical ventilation for respiratory failure may not tolerate an elevated temperature.⁸

Fever serves as an adaptive host response to infection, boosting innate and adaptive immunity in a multitude of ways.⁸ In animal models, fever slows the replication of pathogenic bacteria and enhances the activity of antibiotic agents.⁸ In vitro studies demonstrate that fever increases mobility of leukocytes, phagocytic activity, and proliferation of T cells.⁸ Retrospective case-control studies of patients hospitalized with severe bacterial illnesses, including gram-negative bacteremia, spontaneous bacterial peritonitis, and community-­acquired pneumonia, found that patients with a documented febrile response had increased survival compared with those who remained afebrile during the infection.⁹ In addition, a large retrospective cohort study of septic ICU patients found a progressive decline in mortality in association with increasing peak temperature on the day of ICU admission.¹⁰

In addition to the above studies supporting the important role of fever in fighting infection, recent evidence definitively demonstrates no mortality or morbidity benefit of using antipyretic medications in infected patients. A 2017 meta-analysis that included eight observational and eight randomized studies, totaling 18,939 adult septic ICU patients, demonstrated no difference in hospital and 28-day mortality in patients treated with antipyretics vs those who were not.¹¹ The authors again found no mortality benefit with antipyretic use when separately analyzing data from only the randomized controlled trials (1,507 patients) or when stratifying patients based on the type of antipyretic received (acetaminophen, NSAIDs, or physical cooling).¹¹ They reported no differences in predefined secondary outcomes of shock reversal or nosocomial infections. The authors commented that these robust results likely would not change even with more data from additional trials. In children, a recent meta-analysis of three randomized controlled trials (540 patients) did not find the use of acetaminophen, ibuprofen, or diclofenac effective in preventing febrile seizures.¹²Pediatric practice guidelines consistently recommend using antipyretic medication to alleviate discomfort caused by fever and not solely to reduce temperature.^13,14

Antipyretic agents interfere with the effectiveness of the body’s immune response, as demonstrated in a number of infectious diseases.^2,15-18 Two randomized controlled studies conducted in healthy adult volunteers challenged with rhinovirus reported increased viral shedding and decreased antibody response in those subjects who received aspirin or acetaminophen, compared with those given placebo.^15,16 In another randomized controlled trial conducted in African children with malaria, paracetamol use delayed parasite clearance by 16 hours.¹⁷ A large case-control study correlated the use of NSAIDs with an increased risk of severe skin and soft-tissue complications in children with varicella and in adults with varicella zoster. ¹⁸ The international scientific community has raised concerns about worse outcomes with NSAID use in patients with COVID-19, the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); NSAIDs should be avoided in stable patients with COVID-19 until more data are available. ¹⁹

Additional risks and potential harms accompany antipyretic fever therapy. First, NSAIDs or acetaminophen may adversely affect patients with renal or hepatic insufficiency.^2,3 Second, masking fevers may impair the clinician’s ability to diagnose or evaluate response to treatment. Third, unnecessarily waking a sleeping patient to check temperature or administer unneeded antipyretics can contribute to hospital-associated problems, including delirium, insomnia, and falls. Treating these iatrogenic problems in turn may require additional medications or interventions. These unintended consequences may potentially prolong hospital stays, increase medication errors and polypharmacy, and detract from a patient’s overall healing and recovery.

While the use of antipyretic medications improves fever-­related symptoms, it comes at the cost of blunting a protective host response and exposes patients to medication risks without providing a clinical benefit. In sleeping, asymptomatic, or minimally symptomatic hospitalized patients, the risks of administering antipyretic medications clearly outweigh the benefits.

Treatment with antipyretic medication can alleviate fever-­related symptoms in those patients who have significant headache, body aches, chills, or sweats and in pediatric patients with notable malaise, irritability, or poor oral intake. Debate continues on the use of antipyretics in the ICU setting when managing critically ill patients with severe cardiopulmonary compromise who may not tolerate the additional hemodynamic strain a fever produces (eg, patients with shock requiring vasopressor support or respiratory failure requiring mechanical ventilation). Remember, decrease body temperature in hyperthermia syndromes by physical means.

Withhold antipyretic medication (ie, allow permissive fever) in well-appearing general medical patients with asymptomatic infection-related fevers. In patients who tolerate fever with minimal or no symptoms, potential benefits of permissive fever include decreased time to infection resolution and/or decreased risk of hospital-acquired infections. This may result in shorter hospital stays and significant cost savings. If we do not treat patients with asymptomatic fevers, then it follows that we should not check overnight temperatures in hospitalized patients sleeping comfortably.

• Do not order as-needed antipyretic medication for stable patients on general medical units with infection solely to reduce temperature or achieve normothermia.
• Only treat infected febrile patients with antipyretic medications for fever-related symptoms (headache, chills, or body aches or, in pediatric patients, irritability, malaise, or poor oral intake).
• Treat pathologically elevated temperatures (ie, hyperthermia syndromes) with physical measures because antipyretic medications will be ineffective.

In the clinical scenario, the hospitalist admitted the patient in stable condition for treatment of a community-acquired pneumonia. He mounted a febrile response to infection, which suggests that his active immune system may aid in recovery. The nurse noted the fever while the patient slept comfortably without fever-related symptoms.

After discussing these facts with the patient’s concerned nurse, the clinician should discontinue the order for as-needed acetaminophen for fever and instead recommend permissive fever without administration of antipyretic medication. This may facilitate recovery, avoid unnecessary polypharmacy, and allow the medical care team to follow his fever curve to ensure that the infection is adequately treated. If the patient develops bothersome fever-related symptoms, the hospitalist can reasonably treat with a single-dose of acetaminophen or NSAID.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason^™”? Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to propose ideas for other “Things We Do for No Reason” topics by emailing TWDFNR@hospitalmedicine.org.

In the era of multidrug resistant organisms spreading to healthcare facilities, as well as in the community, prevention of healthcare-associated infections (HAIs) has become one of the most important issues in the world. HAIs impact morbidity and mortality of patients, increase healthcare costs,^1,2 and are associated with a longer length of stay in the hospital.^3,4 In Japan, HAIs are a salient problem; more than 9% of patients admitted to the intensive care unit (ICU) developed an infection during their ICU stay,⁵ and the numbers of multidrug resistant organism isolates causing HAIs have been increasing annually.⁶

Hand hygiene is the most important strategy for preventing the spread of MDROs and reducing HAIs.⁷ Heightened attention to hand hygiene has occurred because of the recent global outbreak of coronavirus disease 2019 (COVID-19), which first appeared in Wuhan, China.⁸ Because no proven antiviral or vaccine is currently available for the disease, hand hygiene, appropriate cough etiquette, and physical distancing, including school closures, are the only way to prevent spread of the illness.^9,10 The virus appears to be highly contagious and spread by droplet or contact routes. The spread of COVID-19 in healthcare facilities has been significant,¹¹ and it could be a source of further spread of the disease in the community.

Unfortunately, hand hygiene adherence remains low in most settings.¹² The World Health Organization (WHO) created a strategy to improve hand hygiene adherence,¹³ which has been implemented in many countries.¹⁴ This strategy consists of five key components: (1) system change, (2) training/education, (3) evaluation and feedback, (4) reminders in the workplace, and (5) institutional safety climate.¹³ Implementing a multimodal intervention including these five elements has increased hand hygiene adherence among healthcare workers (HCWs) and appears to reduce HAIs in different locations.^15-17 Improving hand hygiene practice among HCWs is considered one of the most important ways to decrease the incidence of HAIs.^15,18,19

There are two types of practice for hand hygiene: either hand washing with soap and water or using alcohol-based hand rub (AHR). The former requires water, soap, a sink, and paper towels, whereas the latter requires only hand rub, which is easy to use and requires one-third the length of time as the former.²⁰ Therefore, AHR is strongly recommended, especially in acute and intensive care settings in hospitals, which require urgent care of patients. Importantly, previous studies demonstrated that greater use of AHR resulted in significant reductions in HAIs.^7,14

In Japan, the data related to hand hygiene adherence is limited. Previous studies at four hospitals in different regions of Japan demonstrated that hand hygiene rates were suboptimal²¹ and lower than reported adherence rates from other international studies.¹⁴ One study at three hospitals showed rates could be improved by a multimodal intervention tailored by each institution.²² A 5-year follow-up study demonstrated the sustainability of the multimodal intervention²³; however, hand hygiene adherence rates remained low at approximately 32%.

We hypothesized that perhaps focusing attention on just one single region (or prefecture) could boost hand hygiene rates. Niigata prefecture is located 200 miles north of Tokyo and is the largest prefecture facing the Japan Sea. There are five major tertiary hospitals in Niigata, and they communicate frequently and discuss infection control issues as a group. To investigate hand hygiene adherence before touching patients, and to evaluate the improvement of hand hygiene adherence induced by a multimodal intervention, we performed a pre- and postintervention study among HCWs at four of these tertiary care hospitals in Niigata.

Participating hospitals

Four tertiary care hospitals in Niigata, Japan, volunteered to participate in the study. The characteristics of the four participating hospitals are summarized in Table 1. All hospitals are public or community based. Hospital A included two units, consisting of a cardiovascular-cerebral ICU and an emergency department (ED), and Hospitals B, C, and D included various units containing surgical or medical wards, an ICU, or an ED. All four hospitals have at least one designated infection-­prevention nurse and an infection-prevention department. In addition, there is an infection control network system among the hospitals, and they communicate well to update the information related to local, domestic, or global infectious diseases through regular seminars and by distributing and exchanging electronic communication.

Preintervention

The preintervention infrastructure and existing activities to improve HCW hand hygiene in each hospital are summarized in Table 1. These activities were developed by each individual hospital and had been in place for at least 6 months before the study intervention. All hospitals used AHR and did direct observation for hand washing in designated wards or units and monitoring of AHR consumption; however, Hospital B did not have a wash basin in each room and no use of portable AHR. Preintervention hand hygiene data were collected from June to August 2018.

Intervention

To improve hand hygiene adherence, we initiated a multimodal intervention from September 2018 to February 2019 based on WHO recommendations¹³ and the findings from prior hand hygiene studies.²² Each facility was provided the same guidance on how to improve hand hygiene adherence and was asked to tailor their intervention to their settings (Table 2 and Appendix Figure). Suggested interventions included feedback regarding hand hygiene adherence observed during the preintervention period, interventions related to AHR, direct observation of and feedback regarding hand hygiene, new posters promoting hand hygiene in the workplace, a 1-month campaign for hand hygiene, seminars for HCWs related to hand hygiene, creation of a handbook for education/training, feedback regarding hand hygiene adherence during the intervention period, and others. The infection control team at each hospital designed the plans and strategies to improve hand hygiene adherence. Postintervention data were collected from February 2019 to March 2019.

Observation of Hand Hygiene Adherence

Hand hygiene adherence before patient contact was evaluated by board-certified infection control nurses. To reduce observation bias, external nurses from other participating hospitals conducted the observations. To minimize intraobserver variation, the same training as the previous study in Japan²¹ was provided. Hand hygiene observations were usually performed during the day Monday to Friday from 8 am to 1 pm because of observers’ availability.

Use of either AHR or soap and water before patient contact was defined as appropriate hand hygiene.^24,25 Hand hygiene adherence before patient contact for each provider-patient encounter was observed and recorded using a data collection form used in the previous studies.^19,26 The following information was obtained: unit name, time of initiation and completion of observations, HCW type (physician or nurse), and the type of hand hygiene (ie, AHR, hand washing with soap and water, or none). The observers kept an appropriate distance from the observed HCWs to avoid interfering with their regular clinical practice. In addition, we informed HCWs in the hospital that their clinical practices were going to be observed; however, they were not informed their hand hygiene adherence was going to be monitored.

Statistical Analysis

Overall hand hygiene adherence rates from the pre- and postintervention periods were compared based on hospitals and HCW subgroups. The Pearson’s chi-square test was used for the comparison of hand hygiene adherence rates between pre- and postintervention periods, and 95% CIs were estimated using binomial distribution. Poisson regression was used to look at changes in hand hygiene adherence with adjustment for HCW type. A two-tailed P value of <.05 was considered statistically significant. The study protocol was reviewed and approved by the ethics committees at all participating hospitals.

Overall Changes

In total, there were 2,018 and 1,630 observations of hand hygiene during the preintervention and postintervention periods, respectively. Most observations were of nurses: 1,643 of the 2,018 preintervention observations (81.4%) and 1,245 of the 1,630 postintervention observations (76.4%).

Findings from the HCW observations are summarized in Figure A. The overall postintervention hand hygiene adherence rate (548 of 1,630 observations; 33.6%; 95% CI, 31.3%-35.9%) was significantly higher than the preintervention rate (453 of 2,018 observations; 22.4%; 95% CI, 20.6%-24.3%; P < .001). This finding persisted after adjustment for the type of HCW (nurse vs physician), with proper hand hygiene adherence occurring 1.55 times more often after the intervention than before (95% CI, 1.37-1.76; P < .001). The overall improvement in hand hygiene adherence rates in the postintervention period was seen in all four hospitals (Figure B). However, the hand hygiene adherence rates of nurses in Hospitals C and D were lower than those in Hospitals A and B both before and after the intervention.

Changes by HCW Type

The rates of hand hygiene adherence in both physicians and nurses were higher in the postintervention period than in the preintervention period. However, the improvement of hand hygiene adherence among nurses—from 415 of 1,643 (25.2%) to 487 of 1,245 (39.1%) for an increase of 13.9 percentage points (95% CI,10.4-17.3)—was greater than that in physicians—from 38 of 375 (10.1%) to 61 of 385 (15.8%) for an increase of 5.7 percentage points (95% CI, 1.0-8.1; P < .001; Figure B). In general, nurse hand hygiene adherence was higher than that in physicians both in the preintervention period, with nurses at 25.2% (95% CI, 23.2%-27.4%) vs physicians at 10.1% (95% CI, 7.1%-13.2%; P < .001), and in the postintervention period, with nurses at 39.1% (95% CI, 36.4%-41.8%) vs physicians at 15.8% (95% CI, 12.2%-19.5%; P < .001).

Changes by Hospital

Overall, improvement of hand hygiene adherence was observed in all hospitals. However, the improvement rates differed in each hospital: They were 6.5 percentage points in Hospital A, 11.3 percentage points in Hospital C, 11.4 percentage points in Hospital D, and 18.4 percentage points in Hospital B. Hospital B achieved the highest postintervention adherence rates (42.6%), along with the highest improvement. The improvements of hand hygiene adherence in physicians were higher in Hospitals B (8.4 percentage points) and D (8.3 percentage points) than they were in Hospitals A (4.1 percentage points) and C (4.0 percentage points).

Interventions performed at each hospital to improve hand hygiene adherence are summarized in Table 2 and the Appendix Figure. All hospitals performed feedback of hand hygiene adherence after the preintervention period. Interventions related to AHR were frequently initiated; self-carry AHR was provided in two hospitals (Hospitals C and D), and location of AHR was moved (Hospitals B and D). In addition, new AHR products that caused less skin irritation were introduced in Hospital B. Direct observation by hospital staff (separate from our study observers) was also done as part of Hospital A and D’s improvement efforts. Other interventions included a 1-month campaign for hand hygiene including a contest for senryu (humorous 17-syllable poems; Table 2; Appendix Table), posters, seminars, and creation of a handbook related to hand hygiene. Posters emphasizing the importance of hand hygiene created by the local hospital infection control teams were put on the wall in several locations near wash basins. Seminars (1-hour lectures to emphasize the importance of hand hygiene) were provided to nurses. A 10-page hand hygiene handbook was created by one local infection control team and provided to nurses.

Our study demonstrated that the overall rate of hand hygiene adherence improved from 22.4% to 33.6% after multimodal intervention; however, the adherence rates even after intervention were suboptimal. The results were comparable with those of a previous study in Japan,²² which underscores how suboptimal HCW hand hygiene in Japan threatens patient safety. Hand hygiene among HCWs is one of the most important methods to prevent HAIs and to reduce spread of multidrug resistant organisms. High adherence has proven challenging because it requires behavior modification. We implemented WHO hand hygiene adherence strategies²⁷ and evaluated the efficacy of a multimodal intervention in hopes of finding the specific factors that could be related to behavior modification for HCWs.

We observed several important relationships between the intervention components and their improvement in hand hygiene adherence. Among the four participating hospitals, Hospital B was the most successful with improvement of hand hygiene adherence from 24.2% to 42.6%. One unique intervention for Hospital B was the introduction of new AHR products for the people who had felt uncomfortable with current products. Frequent hand washing or the use of certain AHR products could irritate skin causing dry or rough hands, which could reduce hand hygiene practices. In Japan, there are several AHR products available. Among them, a few products contain skin moisturizing elements; these products are 10%-20% higher in cost than nonmoisturizing products. The HCWs in our study stated that the new products were more comfortable to use, and they requested to introduce them as daily use products. Thus, use of a product containing a hand moisturizer may reduce some factors negatively affecting hand hygiene practice and improve adherence rates.

Although this study was unable to determine which components are definitively associated with improving hand hygiene adherence, the findings suggest initiation of multiple intervention components simultaneously may provide more motivation for change than initiating only one or two components at a time. It is also possible that certain intervention components were more beneficial than others. Consistent with a previous study, improving hand hygiene adherence cannot be simply achieved by improving infrastructure (eg, introducing portable AHR) alone, but rather depends on altering the behavior of physicians and nurses.

This study was performed at four tertiary care hospitals in Niigata that are affiliated with Niigata University. They are located closely in the region, within 100 km, have quarterly conferences, and use a mutual monitoring system related to infection prevention. The members of infection control communicate regularly, which we thought would optimize improvements in hand hygiene adherence, compared with the circumstances of previous studies. In this setting, HCWs have similar education and share knowledge related to infection control, and the effects of interventions in each hospital were equally evaluated if similar interventions were implemented. In the current study, the interventions at each hospital were similar, and there was limited variety; therefore, specific, novel interventions that could affect hand hygiene adherence significantly were difficult to find.

There are a few possible reasons why hand hygiene adherence rates were low in the current study. First, part of this study was conducted during the summer so that the consciousness and caution for hand hygiene might be lower, compared with that in winter. In general, HCWs become more cautious for hand hygiene practice when they take care of patients diagnosed with influenza or respiratory syncytial virus infection. Second, the infrastructure for hand hygiene practice in the hospitals in Japan is inadequate and not well designed. Because of safety reasons, a single dispenser of AHR is placed at the entrance of each room in general and not at each bedside. The number of private rooms is limited, and most of the rooms in wards have multiple beds per room, with no access to AHR within the room. In fact, the interventions at all four hospitals included a change in the location and/or access of AHR. Easier access to AHR is likely a key step to improving hand hygiene adherence rates. Finally, there was not an active intervention to include hospital or unit leaders. This is important given the involvement of leaders in hand hygiene practice significantly changed the hand adherence rates in a previous study.¹⁹

Given the suboptimal hand hygiene adherence rates in Japan noted in this and previous Japanese studies,^21,22 the spread of COVID-19 within the hospital setting is a concern. Transmission of COVID-19 by asymptomatic carriers has been suggested,¹¹ which emphasizes the importance of regular standard precautions with good hand hygiene practice to prevent further transmission.

Although the hand hygiene rate was suboptimal, we were able to achieve a few sustainable, structural modifications in the clinical environment after the intervention. These include adding AHR in new locations, changing the location of existing AHR to more appropriate locations, and introducing new products. These will remain in the clinical environment and will contribute to hand hygiene adherence in the future.

This study has several limitations. First, the presence of external observers in their clinical settings might have affected the behavior of HCWs.²⁸ Although they were not informed that their hand hygiene adherence was going to be monitored, the existence of an external observer in their clinical setting might have changed normal behavior. Second, the infrastructure and interventions for hand hygiene adherence before the intervention were different in each hospital, so there is a possibility that hospitals with less infrastructure for hand hygiene adherence had more room for improvement with the interventions. Third, we included observations at different units at each hospital, which might affect the results of the study because of the inclusion of different medical settings and HCWs. Fourth, the number of physician hand hygiene observations was limited: We conducted our observations between 8 am and 1 pm on weekdays because of observer availability, and many physicians visited their patients during other times of the day. Finally, we were unable to determine whether the improvements seen in each hospital were caused by specific intervention components. However, it is known that recognizing the importance of hand hygiene varies in different regions and countries in the world, and the goal for hand hygiene interventions is to establish a culture of hand hygiene practice.¹³ Further evaluation is necessary to assess sustainability.

In conclusion, a multimodal intervention to improve hand hygiene adherence successfully improved HCWs’ hand hygiene adherence in Niigata, Japan; however, the adherence rates are still relatively low compared with those reported from other countries. Further intervention is required to improve hand hygiene adherence.

In the era of multidrug resistant organisms spreading to healthcare facilities, as well as in the community, prevention of healthcare-associated infections (HAIs) has become one of the most important issues in the world. HAIs impact morbidity and mortality of patients, increase healthcare costs,^1,2 and are associated with a longer length of stay in the hospital.^3,4 In Japan, HAIs are a salient problem; more than 9% of patients admitted to the intensive care unit (ICU) developed an infection during their ICU stay,⁵ and the numbers of multidrug resistant organism isolates causing HAIs have been increasing annually.⁶

Hand hygiene is the most important strategy for preventing the spread of MDROs and reducing HAIs.⁷ Heightened attention to hand hygiene has occurred because of the recent global outbreak of coronavirus disease 2019 (COVID-19), which first appeared in Wuhan, China.⁸ Because no proven antiviral or vaccine is currently available for the disease, hand hygiene, appropriate cough etiquette, and physical distancing, including school closures, are the only way to prevent spread of the illness.^9,10 The virus appears to be highly contagious and spread by droplet or contact routes. The spread of COVID-19 in healthcare facilities has been significant,¹¹ and it could be a source of further spread of the disease in the community.

Unfortunately, hand hygiene adherence remains low in most settings.¹² The World Health Organization (WHO) created a strategy to improve hand hygiene adherence,¹³ which has been implemented in many countries.¹⁴ This strategy consists of five key components: (1) system change, (2) training/education, (3) evaluation and feedback, (4) reminders in the workplace, and (5) institutional safety climate.¹³ Implementing a multimodal intervention including these five elements has increased hand hygiene adherence among healthcare workers (HCWs) and appears to reduce HAIs in different locations.^15-17 Improving hand hygiene practice among HCWs is considered one of the most important ways to decrease the incidence of HAIs.^15,18,19

There are two types of practice for hand hygiene: either hand washing with soap and water or using alcohol-based hand rub (AHR). The former requires water, soap, a sink, and paper towels, whereas the latter requires only hand rub, which is easy to use and requires one-third the length of time as the former.²⁰ Therefore, AHR is strongly recommended, especially in acute and intensive care settings in hospitals, which require urgent care of patients. Importantly, previous studies demonstrated that greater use of AHR resulted in significant reductions in HAIs.^7,14

In Japan, the data related to hand hygiene adherence is limited. Previous studies at four hospitals in different regions of Japan demonstrated that hand hygiene rates were suboptimal²¹ and lower than reported adherence rates from other international studies.¹⁴ One study at three hospitals showed rates could be improved by a multimodal intervention tailored by each institution.²² A 5-year follow-up study demonstrated the sustainability of the multimodal intervention²³; however, hand hygiene adherence rates remained low at approximately 32%.

We hypothesized that perhaps focusing attention on just one single region (or prefecture) could boost hand hygiene rates. Niigata prefecture is located 200 miles north of Tokyo and is the largest prefecture facing the Japan Sea. There are five major tertiary hospitals in Niigata, and they communicate frequently and discuss infection control issues as a group. To investigate hand hygiene adherence before touching patients, and to evaluate the improvement of hand hygiene adherence induced by a multimodal intervention, we performed a pre- and postintervention study among HCWs at four of these tertiary care hospitals in Niigata.

Participating hospitals

Four tertiary care hospitals in Niigata, Japan, volunteered to participate in the study. The characteristics of the four participating hospitals are summarized in Table 1. All hospitals are public or community based. Hospital A included two units, consisting of a cardiovascular-cerebral ICU and an emergency department (ED), and Hospitals B, C, and D included various units containing surgical or medical wards, an ICU, or an ED. All four hospitals have at least one designated infection-­prevention nurse and an infection-prevention department. In addition, there is an infection control network system among the hospitals, and they communicate well to update the information related to local, domestic, or global infectious diseases through regular seminars and by distributing and exchanging electronic communication.

Preintervention

The preintervention infrastructure and existing activities to improve HCW hand hygiene in each hospital are summarized in Table 1. These activities were developed by each individual hospital and had been in place for at least 6 months before the study intervention. All hospitals used AHR and did direct observation for hand washing in designated wards or units and monitoring of AHR consumption; however, Hospital B did not have a wash basin in each room and no use of portable AHR. Preintervention hand hygiene data were collected from June to August 2018.

Intervention

To improve hand hygiene adherence, we initiated a multimodal intervention from September 2018 to February 2019 based on WHO recommendations¹³ and the findings from prior hand hygiene studies.²² Each facility was provided the same guidance on how to improve hand hygiene adherence and was asked to tailor their intervention to their settings (Table 2 and Appendix Figure). Suggested interventions included feedback regarding hand hygiene adherence observed during the preintervention period, interventions related to AHR, direct observation of and feedback regarding hand hygiene, new posters promoting hand hygiene in the workplace, a 1-month campaign for hand hygiene, seminars for HCWs related to hand hygiene, creation of a handbook for education/training, feedback regarding hand hygiene adherence during the intervention period, and others. The infection control team at each hospital designed the plans and strategies to improve hand hygiene adherence. Postintervention data were collected from February 2019 to March 2019.

Observation of Hand Hygiene Adherence

Hand hygiene adherence before patient contact was evaluated by board-certified infection control nurses. To reduce observation bias, external nurses from other participating hospitals conducted the observations. To minimize intraobserver variation, the same training as the previous study in Japan²¹ was provided. Hand hygiene observations were usually performed during the day Monday to Friday from 8 am to 1 pm because of observers’ availability.

Use of either AHR or soap and water before patient contact was defined as appropriate hand hygiene.^24,25 Hand hygiene adherence before patient contact for each provider-patient encounter was observed and recorded using a data collection form used in the previous studies.^19,26 The following information was obtained: unit name, time of initiation and completion of observations, HCW type (physician or nurse), and the type of hand hygiene (ie, AHR, hand washing with soap and water, or none). The observers kept an appropriate distance from the observed HCWs to avoid interfering with their regular clinical practice. In addition, we informed HCWs in the hospital that their clinical practices were going to be observed; however, they were not informed their hand hygiene adherence was going to be monitored.

Statistical Analysis

Overall hand hygiene adherence rates from the pre- and postintervention periods were compared based on hospitals and HCW subgroups. The Pearson’s chi-square test was used for the comparison of hand hygiene adherence rates between pre- and postintervention periods, and 95% CIs were estimated using binomial distribution. Poisson regression was used to look at changes in hand hygiene adherence with adjustment for HCW type. A two-tailed P value of <.05 was considered statistically significant. The study protocol was reviewed and approved by the ethics committees at all participating hospitals.

Overall Changes

In total, there were 2,018 and 1,630 observations of hand hygiene during the preintervention and postintervention periods, respectively. Most observations were of nurses: 1,643 of the 2,018 preintervention observations (81.4%) and 1,245 of the 1,630 postintervention observations (76.4%).

Findings from the HCW observations are summarized in Figure A. The overall postintervention hand hygiene adherence rate (548 of 1,630 observations; 33.6%; 95% CI, 31.3%-35.9%) was significantly higher than the preintervention rate (453 of 2,018 observations; 22.4%; 95% CI, 20.6%-24.3%; P < .001). This finding persisted after adjustment for the type of HCW (nurse vs physician), with proper hand hygiene adherence occurring 1.55 times more often after the intervention than before (95% CI, 1.37-1.76; P < .001). The overall improvement in hand hygiene adherence rates in the postintervention period was seen in all four hospitals (Figure B). However, the hand hygiene adherence rates of nurses in Hospitals C and D were lower than those in Hospitals A and B both before and after the intervention.

Changes by HCW Type

The rates of hand hygiene adherence in both physicians and nurses were higher in the postintervention period than in the preintervention period. However, the improvement of hand hygiene adherence among nurses—from 415 of 1,643 (25.2%) to 487 of 1,245 (39.1%) for an increase of 13.9 percentage points (95% CI,10.4-17.3)—was greater than that in physicians—from 38 of 375 (10.1%) to 61 of 385 (15.8%) for an increase of 5.7 percentage points (95% CI, 1.0-8.1; P < .001; Figure B). In general, nurse hand hygiene adherence was higher than that in physicians both in the preintervention period, with nurses at 25.2% (95% CI, 23.2%-27.4%) vs physicians at 10.1% (95% CI, 7.1%-13.2%; P < .001), and in the postintervention period, with nurses at 39.1% (95% CI, 36.4%-41.8%) vs physicians at 15.8% (95% CI, 12.2%-19.5%; P < .001).

Changes by Hospital

Overall, improvement of hand hygiene adherence was observed in all hospitals. However, the improvement rates differed in each hospital: They were 6.5 percentage points in Hospital A, 11.3 percentage points in Hospital C, 11.4 percentage points in Hospital D, and 18.4 percentage points in Hospital B. Hospital B achieved the highest postintervention adherence rates (42.6%), along with the highest improvement. The improvements of hand hygiene adherence in physicians were higher in Hospitals B (8.4 percentage points) and D (8.3 percentage points) than they were in Hospitals A (4.1 percentage points) and C (4.0 percentage points).

Interventions performed at each hospital to improve hand hygiene adherence are summarized in Table 2 and the Appendix Figure. All hospitals performed feedback of hand hygiene adherence after the preintervention period. Interventions related to AHR were frequently initiated; self-carry AHR was provided in two hospitals (Hospitals C and D), and location of AHR was moved (Hospitals B and D). In addition, new AHR products that caused less skin irritation were introduced in Hospital B. Direct observation by hospital staff (separate from our study observers) was also done as part of Hospital A and D’s improvement efforts. Other interventions included a 1-month campaign for hand hygiene including a contest for senryu (humorous 17-syllable poems; Table 2; Appendix Table), posters, seminars, and creation of a handbook related to hand hygiene. Posters emphasizing the importance of hand hygiene created by the local hospital infection control teams were put on the wall in several locations near wash basins. Seminars (1-hour lectures to emphasize the importance of hand hygiene) were provided to nurses. A 10-page hand hygiene handbook was created by one local infection control team and provided to nurses.

Our study demonstrated that the overall rate of hand hygiene adherence improved from 22.4% to 33.6% after multimodal intervention; however, the adherence rates even after intervention were suboptimal. The results were comparable with those of a previous study in Japan,²² which underscores how suboptimal HCW hand hygiene in Japan threatens patient safety. Hand hygiene among HCWs is one of the most important methods to prevent HAIs and to reduce spread of multidrug resistant organisms. High adherence has proven challenging because it requires behavior modification. We implemented WHO hand hygiene adherence strategies²⁷ and evaluated the efficacy of a multimodal intervention in hopes of finding the specific factors that could be related to behavior modification for HCWs.

We observed several important relationships between the intervention components and their improvement in hand hygiene adherence. Among the four participating hospitals, Hospital B was the most successful with improvement of hand hygiene adherence from 24.2% to 42.6%. One unique intervention for Hospital B was the introduction of new AHR products for the people who had felt uncomfortable with current products. Frequent hand washing or the use of certain AHR products could irritate skin causing dry or rough hands, which could reduce hand hygiene practices. In Japan, there are several AHR products available. Among them, a few products contain skin moisturizing elements; these products are 10%-20% higher in cost than nonmoisturizing products. The HCWs in our study stated that the new products were more comfortable to use, and they requested to introduce them as daily use products. Thus, use of a product containing a hand moisturizer may reduce some factors negatively affecting hand hygiene practice and improve adherence rates.

Although this study was unable to determine which components are definitively associated with improving hand hygiene adherence, the findings suggest initiation of multiple intervention components simultaneously may provide more motivation for change than initiating only one or two components at a time. It is also possible that certain intervention components were more beneficial than others. Consistent with a previous study, improving hand hygiene adherence cannot be simply achieved by improving infrastructure (eg, introducing portable AHR) alone, but rather depends on altering the behavior of physicians and nurses.

This study was performed at four tertiary care hospitals in Niigata that are affiliated with Niigata University. They are located closely in the region, within 100 km, have quarterly conferences, and use a mutual monitoring system related to infection prevention. The members of infection control communicate regularly, which we thought would optimize improvements in hand hygiene adherence, compared with the circumstances of previous studies. In this setting, HCWs have similar education and share knowledge related to infection control, and the effects of interventions in each hospital were equally evaluated if similar interventions were implemented. In the current study, the interventions at each hospital were similar, and there was limited variety; therefore, specific, novel interventions that could affect hand hygiene adherence significantly were difficult to find.

There are a few possible reasons why hand hygiene adherence rates were low in the current study. First, part of this study was conducted during the summer so that the consciousness and caution for hand hygiene might be lower, compared with that in winter. In general, HCWs become more cautious for hand hygiene practice when they take care of patients diagnosed with influenza or respiratory syncytial virus infection. Second, the infrastructure for hand hygiene practice in the hospitals in Japan is inadequate and not well designed. Because of safety reasons, a single dispenser of AHR is placed at the entrance of each room in general and not at each bedside. The number of private rooms is limited, and most of the rooms in wards have multiple beds per room, with no access to AHR within the room. In fact, the interventions at all four hospitals included a change in the location and/or access of AHR. Easier access to AHR is likely a key step to improving hand hygiene adherence rates. Finally, there was not an active intervention to include hospital or unit leaders. This is important given the involvement of leaders in hand hygiene practice significantly changed the hand adherence rates in a previous study.¹⁹

Given the suboptimal hand hygiene adherence rates in Japan noted in this and previous Japanese studies,^21,22 the spread of COVID-19 within the hospital setting is a concern. Transmission of COVID-19 by asymptomatic carriers has been suggested,¹¹ which emphasizes the importance of regular standard precautions with good hand hygiene practice to prevent further transmission.

Although the hand hygiene rate was suboptimal, we were able to achieve a few sustainable, structural modifications in the clinical environment after the intervention. These include adding AHR in new locations, changing the location of existing AHR to more appropriate locations, and introducing new products. These will remain in the clinical environment and will contribute to hand hygiene adherence in the future.

This study has several limitations. First, the presence of external observers in their clinical settings might have affected the behavior of HCWs.²⁸ Although they were not informed that their hand hygiene adherence was going to be monitored, the existence of an external observer in their clinical setting might have changed normal behavior. Second, the infrastructure and interventions for hand hygiene adherence before the intervention were different in each hospital, so there is a possibility that hospitals with less infrastructure for hand hygiene adherence had more room for improvement with the interventions. Third, we included observations at different units at each hospital, which might affect the results of the study because of the inclusion of different medical settings and HCWs. Fourth, the number of physician hand hygiene observations was limited: We conducted our observations between 8 am and 1 pm on weekdays because of observer availability, and many physicians visited their patients during other times of the day. Finally, we were unable to determine whether the improvements seen in each hospital were caused by specific intervention components. However, it is known that recognizing the importance of hand hygiene varies in different regions and countries in the world, and the goal for hand hygiene interventions is to establish a culture of hand hygiene practice.¹³ Further evaluation is necessary to assess sustainability.

In conclusion, a multimodal intervention to improve hand hygiene adherence successfully improved HCWs’ hand hygiene adherence in Niigata, Japan; however, the adherence rates are still relatively low compared with those reported from other countries. Further intervention is required to improve hand hygiene adherence.

In the era of multidrug resistant organisms spreading to healthcare facilities, as well as in the community, prevention of healthcare-associated infections (HAIs) has become one of the most important issues in the world. HAIs impact morbidity and mortality of patients, increase healthcare costs,^1,2 and are associated with a longer length of stay in the hospital.^3,4 In Japan, HAIs are a salient problem; more than 9% of patients admitted to the intensive care unit (ICU) developed an infection during their ICU stay,⁵ and the numbers of multidrug resistant organism isolates causing HAIs have been increasing annually.⁶

Hand hygiene is the most important strategy for preventing the spread of MDROs and reducing HAIs.⁷ Heightened attention to hand hygiene has occurred because of the recent global outbreak of coronavirus disease 2019 (COVID-19), which first appeared in Wuhan, China.⁸ Because no proven antiviral or vaccine is currently available for the disease, hand hygiene, appropriate cough etiquette, and physical distancing, including school closures, are the only way to prevent spread of the illness.^9,10 The virus appears to be highly contagious and spread by droplet or contact routes. The spread of COVID-19 in healthcare facilities has been significant,¹¹ and it could be a source of further spread of the disease in the community.

Unfortunately, hand hygiene adherence remains low in most settings.¹² The World Health Organization (WHO) created a strategy to improve hand hygiene adherence,¹³ which has been implemented in many countries.¹⁴ This strategy consists of five key components: (1) system change, (2) training/education, (3) evaluation and feedback, (4) reminders in the workplace, and (5) institutional safety climate.¹³ Implementing a multimodal intervention including these five elements has increased hand hygiene adherence among healthcare workers (HCWs) and appears to reduce HAIs in different locations.^15-17 Improving hand hygiene practice among HCWs is considered one of the most important ways to decrease the incidence of HAIs.^15,18,19

There are two types of practice for hand hygiene: either hand washing with soap and water or using alcohol-based hand rub (AHR). The former requires water, soap, a sink, and paper towels, whereas the latter requires only hand rub, which is easy to use and requires one-third the length of time as the former.²⁰ Therefore, AHR is strongly recommended, especially in acute and intensive care settings in hospitals, which require urgent care of patients. Importantly, previous studies demonstrated that greater use of AHR resulted in significant reductions in HAIs.^7,14

In Japan, the data related to hand hygiene adherence is limited. Previous studies at four hospitals in different regions of Japan demonstrated that hand hygiene rates were suboptimal²¹ and lower than reported adherence rates from other international studies.¹⁴ One study at three hospitals showed rates could be improved by a multimodal intervention tailored by each institution.²² A 5-year follow-up study demonstrated the sustainability of the multimodal intervention²³; however, hand hygiene adherence rates remained low at approximately 32%.

We hypothesized that perhaps focusing attention on just one single region (or prefecture) could boost hand hygiene rates. Niigata prefecture is located 200 miles north of Tokyo and is the largest prefecture facing the Japan Sea. There are five major tertiary hospitals in Niigata, and they communicate frequently and discuss infection control issues as a group. To investigate hand hygiene adherence before touching patients, and to evaluate the improvement of hand hygiene adherence induced by a multimodal intervention, we performed a pre- and postintervention study among HCWs at four of these tertiary care hospitals in Niigata.

Participating hospitals

Four tertiary care hospitals in Niigata, Japan, volunteered to participate in the study. The characteristics of the four participating hospitals are summarized in Table 1. All hospitals are public or community based. Hospital A included two units, consisting of a cardiovascular-cerebral ICU and an emergency department (ED), and Hospitals B, C, and D included various units containing surgical or medical wards, an ICU, or an ED. All four hospitals have at least one designated infection-­prevention nurse and an infection-prevention department. In addition, there is an infection control network system among the hospitals, and they communicate well to update the information related to local, domestic, or global infectious diseases through regular seminars and by distributing and exchanging electronic communication.

Preintervention

The preintervention infrastructure and existing activities to improve HCW hand hygiene in each hospital are summarized in Table 1. These activities were developed by each individual hospital and had been in place for at least 6 months before the study intervention. All hospitals used AHR and did direct observation for hand washing in designated wards or units and monitoring of AHR consumption; however, Hospital B did not have a wash basin in each room and no use of portable AHR. Preintervention hand hygiene data were collected from June to August 2018.

Intervention

To improve hand hygiene adherence, we initiated a multimodal intervention from September 2018 to February 2019 based on WHO recommendations¹³ and the findings from prior hand hygiene studies.²² Each facility was provided the same guidance on how to improve hand hygiene adherence and was asked to tailor their intervention to their settings (Table 2 and Appendix Figure). Suggested interventions included feedback regarding hand hygiene adherence observed during the preintervention period, interventions related to AHR, direct observation of and feedback regarding hand hygiene, new posters promoting hand hygiene in the workplace, a 1-month campaign for hand hygiene, seminars for HCWs related to hand hygiene, creation of a handbook for education/training, feedback regarding hand hygiene adherence during the intervention period, and others. The infection control team at each hospital designed the plans and strategies to improve hand hygiene adherence. Postintervention data were collected from February 2019 to March 2019.

Observation of Hand Hygiene Adherence

Hand hygiene adherence before patient contact was evaluated by board-certified infection control nurses. To reduce observation bias, external nurses from other participating hospitals conducted the observations. To minimize intraobserver variation, the same training as the previous study in Japan²¹ was provided. Hand hygiene observations were usually performed during the day Monday to Friday from 8 am to 1 pm because of observers’ availability.

Use of either AHR or soap and water before patient contact was defined as appropriate hand hygiene.^24,25 Hand hygiene adherence before patient contact for each provider-patient encounter was observed and recorded using a data collection form used in the previous studies.^19,26 The following information was obtained: unit name, time of initiation and completion of observations, HCW type (physician or nurse), and the type of hand hygiene (ie, AHR, hand washing with soap and water, or none). The observers kept an appropriate distance from the observed HCWs to avoid interfering with their regular clinical practice. In addition, we informed HCWs in the hospital that their clinical practices were going to be observed; however, they were not informed their hand hygiene adherence was going to be monitored.

Statistical Analysis

Overall hand hygiene adherence rates from the pre- and postintervention periods were compared based on hospitals and HCW subgroups. The Pearson’s chi-square test was used for the comparison of hand hygiene adherence rates between pre- and postintervention periods, and 95% CIs were estimated using binomial distribution. Poisson regression was used to look at changes in hand hygiene adherence with adjustment for HCW type. A two-tailed P value of <.05 was considered statistically significant. The study protocol was reviewed and approved by the ethics committees at all participating hospitals.

Overall Changes

In total, there were 2,018 and 1,630 observations of hand hygiene during the preintervention and postintervention periods, respectively. Most observations were of nurses: 1,643 of the 2,018 preintervention observations (81.4%) and 1,245 of the 1,630 postintervention observations (76.4%).

Findings from the HCW observations are summarized in Figure A. The overall postintervention hand hygiene adherence rate (548 of 1,630 observations; 33.6%; 95% CI, 31.3%-35.9%) was significantly higher than the preintervention rate (453 of 2,018 observations; 22.4%; 95% CI, 20.6%-24.3%; P < .001). This finding persisted after adjustment for the type of HCW (nurse vs physician), with proper hand hygiene adherence occurring 1.55 times more often after the intervention than before (95% CI, 1.37-1.76; P < .001). The overall improvement in hand hygiene adherence rates in the postintervention period was seen in all four hospitals (Figure B). However, the hand hygiene adherence rates of nurses in Hospitals C and D were lower than those in Hospitals A and B both before and after the intervention.

Changes by HCW Type

The rates of hand hygiene adherence in both physicians and nurses were higher in the postintervention period than in the preintervention period. However, the improvement of hand hygiene adherence among nurses—from 415 of 1,643 (25.2%) to 487 of 1,245 (39.1%) for an increase of 13.9 percentage points (95% CI,10.4-17.3)—was greater than that in physicians—from 38 of 375 (10.1%) to 61 of 385 (15.8%) for an increase of 5.7 percentage points (95% CI, 1.0-8.1; P < .001; Figure B). In general, nurse hand hygiene adherence was higher than that in physicians both in the preintervention period, with nurses at 25.2% (95% CI, 23.2%-27.4%) vs physicians at 10.1% (95% CI, 7.1%-13.2%; P < .001), and in the postintervention period, with nurses at 39.1% (95% CI, 36.4%-41.8%) vs physicians at 15.8% (95% CI, 12.2%-19.5%; P < .001).

Changes by Hospital

Overall, improvement of hand hygiene adherence was observed in all hospitals. However, the improvement rates differed in each hospital: They were 6.5 percentage points in Hospital A, 11.3 percentage points in Hospital C, 11.4 percentage points in Hospital D, and 18.4 percentage points in Hospital B. Hospital B achieved the highest postintervention adherence rates (42.6%), along with the highest improvement. The improvements of hand hygiene adherence in physicians were higher in Hospitals B (8.4 percentage points) and D (8.3 percentage points) than they were in Hospitals A (4.1 percentage points) and C (4.0 percentage points).

Interventions performed at each hospital to improve hand hygiene adherence are summarized in Table 2 and the Appendix Figure. All hospitals performed feedback of hand hygiene adherence after the preintervention period. Interventions related to AHR were frequently initiated; self-carry AHR was provided in two hospitals (Hospitals C and D), and location of AHR was moved (Hospitals B and D). In addition, new AHR products that caused less skin irritation were introduced in Hospital B. Direct observation by hospital staff (separate from our study observers) was also done as part of Hospital A and D’s improvement efforts. Other interventions included a 1-month campaign for hand hygiene including a contest for senryu (humorous 17-syllable poems; Table 2; Appendix Table), posters, seminars, and creation of a handbook related to hand hygiene. Posters emphasizing the importance of hand hygiene created by the local hospital infection control teams were put on the wall in several locations near wash basins. Seminars (1-hour lectures to emphasize the importance of hand hygiene) were provided to nurses. A 10-page hand hygiene handbook was created by one local infection control team and provided to nurses.

Our study demonstrated that the overall rate of hand hygiene adherence improved from 22.4% to 33.6% after multimodal intervention; however, the adherence rates even after intervention were suboptimal. The results were comparable with those of a previous study in Japan,²² which underscores how suboptimal HCW hand hygiene in Japan threatens patient safety. Hand hygiene among HCWs is one of the most important methods to prevent HAIs and to reduce spread of multidrug resistant organisms. High adherence has proven challenging because it requires behavior modification. We implemented WHO hand hygiene adherence strategies²⁷ and evaluated the efficacy of a multimodal intervention in hopes of finding the specific factors that could be related to behavior modification for HCWs.

We observed several important relationships between the intervention components and their improvement in hand hygiene adherence. Among the four participating hospitals, Hospital B was the most successful with improvement of hand hygiene adherence from 24.2% to 42.6%. One unique intervention for Hospital B was the introduction of new AHR products for the people who had felt uncomfortable with current products. Frequent hand washing or the use of certain AHR products could irritate skin causing dry or rough hands, which could reduce hand hygiene practices. In Japan, there are several AHR products available. Among them, a few products contain skin moisturizing elements; these products are 10%-20% higher in cost than nonmoisturizing products. The HCWs in our study stated that the new products were more comfortable to use, and they requested to introduce them as daily use products. Thus, use of a product containing a hand moisturizer may reduce some factors negatively affecting hand hygiene practice and improve adherence rates.

Although this study was unable to determine which components are definitively associated with improving hand hygiene adherence, the findings suggest initiation of multiple intervention components simultaneously may provide more motivation for change than initiating only one or two components at a time. It is also possible that certain intervention components were more beneficial than others. Consistent with a previous study, improving hand hygiene adherence cannot be simply achieved by improving infrastructure (eg, introducing portable AHR) alone, but rather depends on altering the behavior of physicians and nurses.

This study was performed at four tertiary care hospitals in Niigata that are affiliated with Niigata University. They are located closely in the region, within 100 km, have quarterly conferences, and use a mutual monitoring system related to infection prevention. The members of infection control communicate regularly, which we thought would optimize improvements in hand hygiene adherence, compared with the circumstances of previous studies. In this setting, HCWs have similar education and share knowledge related to infection control, and the effects of interventions in each hospital were equally evaluated if similar interventions were implemented. In the current study, the interventions at each hospital were similar, and there was limited variety; therefore, specific, novel interventions that could affect hand hygiene adherence significantly were difficult to find.

There are a few possible reasons why hand hygiene adherence rates were low in the current study. First, part of this study was conducted during the summer so that the consciousness and caution for hand hygiene might be lower, compared with that in winter. In general, HCWs become more cautious for hand hygiene practice when they take care of patients diagnosed with influenza or respiratory syncytial virus infection. Second, the infrastructure for hand hygiene practice in the hospitals in Japan is inadequate and not well designed. Because of safety reasons, a single dispenser of AHR is placed at the entrance of each room in general and not at each bedside. The number of private rooms is limited, and most of the rooms in wards have multiple beds per room, with no access to AHR within the room. In fact, the interventions at all four hospitals included a change in the location and/or access of AHR. Easier access to AHR is likely a key step to improving hand hygiene adherence rates. Finally, there was not an active intervention to include hospital or unit leaders. This is important given the involvement of leaders in hand hygiene practice significantly changed the hand adherence rates in a previous study.¹⁹

Given the suboptimal hand hygiene adherence rates in Japan noted in this and previous Japanese studies,^21,22 the spread of COVID-19 within the hospital setting is a concern. Transmission of COVID-19 by asymptomatic carriers has been suggested,¹¹ which emphasizes the importance of regular standard precautions with good hand hygiene practice to prevent further transmission.

Although the hand hygiene rate was suboptimal, we were able to achieve a few sustainable, structural modifications in the clinical environment after the intervention. These include adding AHR in new locations, changing the location of existing AHR to more appropriate locations, and introducing new products. These will remain in the clinical environment and will contribute to hand hygiene adherence in the future.

This study has several limitations. First, the presence of external observers in their clinical settings might have affected the behavior of HCWs.²⁸ Although they were not informed that their hand hygiene adherence was going to be monitored, the existence of an external observer in their clinical setting might have changed normal behavior. Second, the infrastructure and interventions for hand hygiene adherence before the intervention were different in each hospital, so there is a possibility that hospitals with less infrastructure for hand hygiene adherence had more room for improvement with the interventions. Third, we included observations at different units at each hospital, which might affect the results of the study because of the inclusion of different medical settings and HCWs. Fourth, the number of physician hand hygiene observations was limited: We conducted our observations between 8 am and 1 pm on weekdays because of observer availability, and many physicians visited their patients during other times of the day. Finally, we were unable to determine whether the improvements seen in each hospital were caused by specific intervention components. However, it is known that recognizing the importance of hand hygiene varies in different regions and countries in the world, and the goal for hand hygiene interventions is to establish a culture of hand hygiene practice.¹³ Further evaluation is necessary to assess sustainability.

In conclusion, a multimodal intervention to improve hand hygiene adherence successfully improved HCWs’ hand hygiene adherence in Niigata, Japan; however, the adherence rates are still relatively low compared with those reported from other countries. Further intervention is required to improve hand hygiene adherence.

Inspired by the ABIM Foundation’s Choosing Wisely^® campaign, the “Things We Do for No Reason^™”(TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent clear-cut conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion.

The hospitalist admits a 73-year-old man with non–insulin dependent diabetes and essential hypertension to the general medicine ward for lower extremity cellulitis. The hospitalist uses standard admission orders, encourages him to elevate his leg above his heart, starts intravenous antibiotics, and monitors him throughout the day and night with regular vital signs. On his second day of admission, the patient’s cellulitis clinically improves, and the team prepares for discharge. However, the nurse notes that the patient did not sleep well and has not slept since his 4 am vitals were taken. Now a lethargic and confused patient, the team adds delirium to his problem list.

General medicine floors commonly default frequency for measuring vital signs to every 4 hours (Q4), a practice that dates back more than a century to the time of Florence Nightingale.This custom remains in place to ensure the ability to identify and intervene for those at risk for clinical deterioration and preventable death. Research supports the notion that frequent and consistent vital sign checks can minimize mortality and morbidity in the hospital. In fact, validated scoring systems incorporate vital signs with other clinical findings as a way of quickly identifying a patient with worsening clinical status.¹ Further, trends and trajectories in vital signs may enable us to identify those with impending decompensation.² A 2008 consensus statement made by experts in patient safety encouraged hospitals to use frequent vital sign monitoring of patients when available and affordable.³ These interventions aim to help identify and treat patients with early clinical deterioration to prevent poor outcomes.

The practice of checking vital signs every 4 hours throughout the night dates to long before the modern era of evidence-­based medicine. Research thus far has not focused on the necessity of vital sign checks every 4 hours throughout the night, despite affecting almost every hospitalized patient. Further, patient acuity or need for monitoring does not drive the frequency of overnight vital signs; instead habit and defaults do. We often monitor patients at high risk for clinical deterioration just as frequently as patients at low risk.⁴

While evidence-based medicine influences much of clinical care, “real-world” needs encountered at the bedside often drive early adapters to innovate. Nurses, who spend the most time at the bedside and conduct the most regular patient assessments, have recognized that not all patients need vital signs checked every 4 hours throughout the night. In 2013, Hands et al conducted a chart review of hospital patterns and found that nurses obtained complete vital sign checks on patients less frequently throughout the night than during the day.⁵ Their work further showed that nurses used their clinical judgment to make decisions about risk: Those patients deemed low risk by the nurses received fewer vital sign checks while the sicker patients received monitoring every 4 hours throughout the night.

Few researchers have quantitatively identified reasons why nurses may choose to not conduct frequent observations for some patients, beyond the providers’ own experience and judgment. In one study, Hope et al conducted a qualitative analysis of nurses to better understand their reasoning behind who should and should not receive overnight monitoring.⁶ The results of the analysis revealed that nurses recognize the importance of sleep in support of health and healing and use their clinical judgement when deciding which patients and conditions can forgo frequent observations.Stiver et al conducted trailblazing work that examines the outcomes of decreasing overnight vital sign checks for low-risk hospitalized patients through a randomized pilot study.⁷ In order to ensure patient safety, their group employed regular nurse observations throughout the night without waking the patient. Those patients assigned to less monitoring overnight reported a trend toward better sleep during hospitalization without the occurrence of any adverse events or escalation in care.

Most important, evidence indicates that sleep disruptions in the hospital worsen health and impede healing; further supporting nurses’ instincts and practices. Hospitalized adults without comorbidities who experience inadequate sleep during hospitalization have a higher perception of pain.⁸ Similarly, research has associated hospital-induced sleep deprivation and a higher odds of elevated blood glucose in those without diabetes, or “hyperglycemia of hospitalization.” ⁹ Furthermore, national organizations have recognized the importance of sleep. The American Academy of Nursing, as part of its Choosing Wisely™ campaign, states that, in the hospital, nurses should not disturb a patient’s sleep “unless the patient’s condition or care specifically requires it.”¹⁰

Finally, in the era of COVID-19, any opportunity to support physical distancing and to limit face-to-face interaction could protect our patients and staff from acquiring SARS-CoV-2.

While consistent vital sign checks allow for early identification of those trending toward clinical deterioration, risk stratification of ward patients can identify those who may benefit from overnight Q4 vital sign checks. While clinicians often use their judgment to identify a subset of low-risk patients for de-escalation of overnight care, artificial intelligence such as Modified Early Warning Score (MEWS) and Pediatric Early Warning Signs (PEWS) may have a role to play. These validated systems use physiologic symptoms that present prior to significant vital sign alterations to identify patients at risk for clinical deterioration.¹¹ As an example, one randomized, controlled trial used a risk stratification tool to eliminate overnight monitoring for low-risk patients. Patients slept more soundly and reported fewer noise disruptions and higher satisfaction with the nursing staff. No adverse events were reported for those who were electronically stratified as low risk.¹²Further, forcing clinicians to decide on the need for overnight vitals by removing the Q4 vital sign default in the electronic health records (EHR) may minimize overnight disruptions. The University of Chicago in Illinois has implemented “sleep-friendly” options for vital sign ordering in the EHR for both children and adults. Enhanced order sets force providers to consider whether patients qualify for fewer overnight interventions. This change, alongside staff education and empowerment, reduced interruptions overnight for both populations and improved patient experience.¹³ This patient-centered practice mirrors a recent recommendation from the American Academy of Nursing to minimize sleep disruptions for hospitalized patients by letting low-risk patients sleep.¹⁰

• Use clinical judgment or an existing risk stratification system, such as MEWS or PEWS, to identify patients who may benefit from more or less monitoring.
• Forgo overnight vital sign checks for low-risk patients.
• Check overnight vitals for low-risk patients at 10 pm and 6 am.
• Use pulse oximetry or regular nurse checks as a balancing measure, especially in the pediatric population.

Inspired by the ABIM Foundation’s Choosing Wisely^® campaign, the “Things We Do for No Reason^™”(TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent clear-cut conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion.

The hospitalist admits a 73-year-old man with non–insulin dependent diabetes and essential hypertension to the general medicine ward for lower extremity cellulitis. The hospitalist uses standard admission orders, encourages him to elevate his leg above his heart, starts intravenous antibiotics, and monitors him throughout the day and night with regular vital signs. On his second day of admission, the patient’s cellulitis clinically improves, and the team prepares for discharge. However, the nurse notes that the patient did not sleep well and has not slept since his 4 am vitals were taken. Now a lethargic and confused patient, the team adds delirium to his problem list.

General medicine floors commonly default frequency for measuring vital signs to every 4 hours (Q4), a practice that dates back more than a century to the time of Florence Nightingale.This custom remains in place to ensure the ability to identify and intervene for those at risk for clinical deterioration and preventable death. Research supports the notion that frequent and consistent vital sign checks can minimize mortality and morbidity in the hospital. In fact, validated scoring systems incorporate vital signs with other clinical findings as a way of quickly identifying a patient with worsening clinical status.¹ Further, trends and trajectories in vital signs may enable us to identify those with impending decompensation.² A 2008 consensus statement made by experts in patient safety encouraged hospitals to use frequent vital sign monitoring of patients when available and affordable.³ These interventions aim to help identify and treat patients with early clinical deterioration to prevent poor outcomes.

The practice of checking vital signs every 4 hours throughout the night dates to long before the modern era of evidence-­based medicine. Research thus far has not focused on the necessity of vital sign checks every 4 hours throughout the night, despite affecting almost every hospitalized patient. Further, patient acuity or need for monitoring does not drive the frequency of overnight vital signs; instead habit and defaults do. We often monitor patients at high risk for clinical deterioration just as frequently as patients at low risk.⁴

While evidence-based medicine influences much of clinical care, “real-world” needs encountered at the bedside often drive early adapters to innovate. Nurses, who spend the most time at the bedside and conduct the most regular patient assessments, have recognized that not all patients need vital signs checked every 4 hours throughout the night. In 2013, Hands et al conducted a chart review of hospital patterns and found that nurses obtained complete vital sign checks on patients less frequently throughout the night than during the day.⁵ Their work further showed that nurses used their clinical judgment to make decisions about risk: Those patients deemed low risk by the nurses received fewer vital sign checks while the sicker patients received monitoring every 4 hours throughout the night.

Few researchers have quantitatively identified reasons why nurses may choose to not conduct frequent observations for some patients, beyond the providers’ own experience and judgment. In one study, Hope et al conducted a qualitative analysis of nurses to better understand their reasoning behind who should and should not receive overnight monitoring.⁶ The results of the analysis revealed that nurses recognize the importance of sleep in support of health and healing and use their clinical judgement when deciding which patients and conditions can forgo frequent observations.Stiver et al conducted trailblazing work that examines the outcomes of decreasing overnight vital sign checks for low-risk hospitalized patients through a randomized pilot study.⁷ In order to ensure patient safety, their group employed regular nurse observations throughout the night without waking the patient. Those patients assigned to less monitoring overnight reported a trend toward better sleep during hospitalization without the occurrence of any adverse events or escalation in care.

Most important, evidence indicates that sleep disruptions in the hospital worsen health and impede healing; further supporting nurses’ instincts and practices. Hospitalized adults without comorbidities who experience inadequate sleep during hospitalization have a higher perception of pain.⁸ Similarly, research has associated hospital-induced sleep deprivation and a higher odds of elevated blood glucose in those without diabetes, or “hyperglycemia of hospitalization.” ⁹ Furthermore, national organizations have recognized the importance of sleep. The American Academy of Nursing, as part of its Choosing Wisely™ campaign, states that, in the hospital, nurses should not disturb a patient’s sleep “unless the patient’s condition or care specifically requires it.”¹⁰

Finally, in the era of COVID-19, any opportunity to support physical distancing and to limit face-to-face interaction could protect our patients and staff from acquiring SARS-CoV-2.

While consistent vital sign checks allow for early identification of those trending toward clinical deterioration, risk stratification of ward patients can identify those who may benefit from overnight Q4 vital sign checks. While clinicians often use their judgment to identify a subset of low-risk patients for de-escalation of overnight care, artificial intelligence such as Modified Early Warning Score (MEWS) and Pediatric Early Warning Signs (PEWS) may have a role to play. These validated systems use physiologic symptoms that present prior to significant vital sign alterations to identify patients at risk for clinical deterioration.¹¹ As an example, one randomized, controlled trial used a risk stratification tool to eliminate overnight monitoring for low-risk patients. Patients slept more soundly and reported fewer noise disruptions and higher satisfaction with the nursing staff. No adverse events were reported for those who were electronically stratified as low risk.¹²Further, forcing clinicians to decide on the need for overnight vitals by removing the Q4 vital sign default in the electronic health records (EHR) may minimize overnight disruptions. The University of Chicago in Illinois has implemented “sleep-friendly” options for vital sign ordering in the EHR for both children and adults. Enhanced order sets force providers to consider whether patients qualify for fewer overnight interventions. This change, alongside staff education and empowerment, reduced interruptions overnight for both populations and improved patient experience.¹³ This patient-centered practice mirrors a recent recommendation from the American Academy of Nursing to minimize sleep disruptions for hospitalized patients by letting low-risk patients sleep.¹⁰

• Use clinical judgment or an existing risk stratification system, such as MEWS or PEWS, to identify patients who may benefit from more or less monitoring.
• Forgo overnight vital sign checks for low-risk patients.
• Check overnight vitals for low-risk patients at 10 pm and 6 am.
• Use pulse oximetry or regular nurse checks as a balancing measure, especially in the pediatric population.

Inspired by the ABIM Foundation’s Choosing Wisely^® campaign, the “Things We Do for No Reason^™”(TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent clear-cut conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion.

The hospitalist admits a 73-year-old man with non–insulin dependent diabetes and essential hypertension to the general medicine ward for lower extremity cellulitis. The hospitalist uses standard admission orders, encourages him to elevate his leg above his heart, starts intravenous antibiotics, and monitors him throughout the day and night with regular vital signs. On his second day of admission, the patient’s cellulitis clinically improves, and the team prepares for discharge. However, the nurse notes that the patient did not sleep well and has not slept since his 4 am vitals were taken. Now a lethargic and confused patient, the team adds delirium to his problem list.

General medicine floors commonly default frequency for measuring vital signs to every 4 hours (Q4), a practice that dates back more than a century to the time of Florence Nightingale.This custom remains in place to ensure the ability to identify and intervene for those at risk for clinical deterioration and preventable death. Research supports the notion that frequent and consistent vital sign checks can minimize mortality and morbidity in the hospital. In fact, validated scoring systems incorporate vital signs with other clinical findings as a way of quickly identifying a patient with worsening clinical status.¹ Further, trends and trajectories in vital signs may enable us to identify those with impending decompensation.² A 2008 consensus statement made by experts in patient safety encouraged hospitals to use frequent vital sign monitoring of patients when available and affordable.³ These interventions aim to help identify and treat patients with early clinical deterioration to prevent poor outcomes.

The practice of checking vital signs every 4 hours throughout the night dates to long before the modern era of evidence-­based medicine. Research thus far has not focused on the necessity of vital sign checks every 4 hours throughout the night, despite affecting almost every hospitalized patient. Further, patient acuity or need for monitoring does not drive the frequency of overnight vital signs; instead habit and defaults do. We often monitor patients at high risk for clinical deterioration just as frequently as patients at low risk.⁴

While evidence-based medicine influences much of clinical care, “real-world” needs encountered at the bedside often drive early adapters to innovate. Nurses, who spend the most time at the bedside and conduct the most regular patient assessments, have recognized that not all patients need vital signs checked every 4 hours throughout the night. In 2013, Hands et al conducted a chart review of hospital patterns and found that nurses obtained complete vital sign checks on patients less frequently throughout the night than during the day.⁵ Their work further showed that nurses used their clinical judgment to make decisions about risk: Those patients deemed low risk by the nurses received fewer vital sign checks while the sicker patients received monitoring every 4 hours throughout the night.

Few researchers have quantitatively identified reasons why nurses may choose to not conduct frequent observations for some patients, beyond the providers’ own experience and judgment. In one study, Hope et al conducted a qualitative analysis of nurses to better understand their reasoning behind who should and should not receive overnight monitoring.⁶ The results of the analysis revealed that nurses recognize the importance of sleep in support of health and healing and use their clinical judgement when deciding which patients and conditions can forgo frequent observations.Stiver et al conducted trailblazing work that examines the outcomes of decreasing overnight vital sign checks for low-risk hospitalized patients through a randomized pilot study.⁷ In order to ensure patient safety, their group employed regular nurse observations throughout the night without waking the patient. Those patients assigned to less monitoring overnight reported a trend toward better sleep during hospitalization without the occurrence of any adverse events or escalation in care.

Most important, evidence indicates that sleep disruptions in the hospital worsen health and impede healing; further supporting nurses’ instincts and practices. Hospitalized adults without comorbidities who experience inadequate sleep during hospitalization have a higher perception of pain.⁸ Similarly, research has associated hospital-induced sleep deprivation and a higher odds of elevated blood glucose in those without diabetes, or “hyperglycemia of hospitalization.” ⁹ Furthermore, national organizations have recognized the importance of sleep. The American Academy of Nursing, as part of its Choosing Wisely™ campaign, states that, in the hospital, nurses should not disturb a patient’s sleep “unless the patient’s condition or care specifically requires it.”¹⁰

Finally, in the era of COVID-19, any opportunity to support physical distancing and to limit face-to-face interaction could protect our patients and staff from acquiring SARS-CoV-2.

While consistent vital sign checks allow for early identification of those trending toward clinical deterioration, risk stratification of ward patients can identify those who may benefit from overnight Q4 vital sign checks. While clinicians often use their judgment to identify a subset of low-risk patients for de-escalation of overnight care, artificial intelligence such as Modified Early Warning Score (MEWS) and Pediatric Early Warning Signs (PEWS) may have a role to play. These validated systems use physiologic symptoms that present prior to significant vital sign alterations to identify patients at risk for clinical deterioration.¹¹ As an example, one randomized, controlled trial used a risk stratification tool to eliminate overnight monitoring for low-risk patients. Patients slept more soundly and reported fewer noise disruptions and higher satisfaction with the nursing staff. No adverse events were reported for those who were electronically stratified as low risk.¹²Further, forcing clinicians to decide on the need for overnight vitals by removing the Q4 vital sign default in the electronic health records (EHR) may minimize overnight disruptions. The University of Chicago in Illinois has implemented “sleep-friendly” options for vital sign ordering in the EHR for both children and adults. Enhanced order sets force providers to consider whether patients qualify for fewer overnight interventions. This change, alongside staff education and empowerment, reduced interruptions overnight for both populations and improved patient experience.¹³ This patient-centered practice mirrors a recent recommendation from the American Academy of Nursing to minimize sleep disruptions for hospitalized patients by letting low-risk patients sleep.¹⁰

• Use clinical judgment or an existing risk stratification system, such as MEWS or PEWS, to identify patients who may benefit from more or less monitoring.
• Forgo overnight vital sign checks for low-risk patients.
• Check overnight vitals for low-risk patients at 10 pm and 6 am.
• Use pulse oximetry or regular nurse checks as a balancing measure, especially in the pediatric population.

One contentious issue during the COVID-19 crisis has been the appropriate selection and use of respiratory protective equipment (RPE) for healthcare workers (HCWs) in hospitals and long-term care settings. As of April 2020, discrepancies exist in the recommendations from health authorities such as the World Health Organization (WHO), Centers for Disease Control and Prevention (CDC), and Canadian Standards Association (CSA). The first of these recommends a surgical mask for routine care and a respirator for high-risk care such as aerosol-generating procedures, while the CDC recommends respirators for all aspects of patient care for these SARS-CoV-2–infected patients, and the CSA risk assessment tool would also result in selection of a respirator.^1-3

Given the contradictory guidance, we will discuss several important considerations for hospital leaders in the implementation of a healthcare respiratory protection program during the current pandemic, including a focused review of the empirical data on surgical mask vs face-fitted respirator (most commonly available in healthcare as N95 in North America), continuous use of the RPE throughout an entire shift vs targeted use when caring for patients, and key areas of uncertainty.

Surgical masks are traditionally used for protection against droplet transmission of respiratory infections, in which large droplets often fall to the ground within short distances; on the other hand, N95 respirators are used for much smaller airborne pathogens, which can remain suspended in the air for long periods of time. Although empiric studies have supported the superiority of respirators over surgical masks in simulated settings (frequently defined as a calculated concentration ratio outside vs inside the RPE), most clinical studies fail to demonstrate a difference in clinical outcomes such as the prevention of respiratory infection. For instance, an exposure study using saline aerosol to simulate viral particles showed that N95 respirators conferred up to 8 to 12 times greater protection against particulate penetration, compared with surgical masks.⁴ However, these advantages of respirators over surgical masks in carefully controlled laboratory studies do not seem to translate to decreased infection risk in real-world settings.

The effectiveness of N95 respirators vs surgical masks in preventing respiratory infections has been evaluated in a small number of clinical randomized, controlled trials (RCTs). We identified five systematic reviews and/or metanalyses published after 2010 and three RCTs published after 1990.^5-12 The RCTs used laboratory-confirmed respiratory virus or clinical infection in HCWs as a clinical outcome, but studies differed in the implementation of RPE use (ie, continuous or targeted use). In a systematic review and metanalysis, Long et al identified six RCTs (9,171 participants) and concluded that, with the exception of laboratory-confirmed bacterial colonization, N95 respirators did not reduce the rate of laboratory-­confirmed influenza, viral respiratory infections, or influenza-like illness among HCWs, compared with surgical masks.⁵ The authors noted risks of bias in these studies owing to the inability to blind and conceal allocation. In addition, the studies focused on infections that are known to transmit via droplet, such as influenza, so the results might not be applicable in the face of a new pandemic in which the important modes of transmission are not yet clear.

The evidence base evaluating continuous vs targeted use of RPE in healthcare settings is quite small. Continuous use refers to using the RPE during an entire shift, whereas targeted use involves using RPE only when caring for confirmed or suspected respiratory patients. In our literature review we identified only one RCT that included separate study arms for continuous and targeted N95 respirator use.¹³ The authors found a significantly lower rate of clinical respiratory illness among HCWs in the continuous-use group, compared with that in the targeted-use group. Limitations of the study included a relatively short follow-up of 4 weeks and uneven distribution of baseline characteristics, although the authors adjusted for these differences in their analysis. The study, however, did not compare continuous vs targeted use of surgical masks with regard to clinical outcomes. Based on the study results, we can only infer that continuous use of RPE, either surgical mask or N95 respirator, may provide additional benefit to HCWs vs targeted use only.

Given the lack of robust evidence informing continuous or targeted RPE use, we suggest some additional factors to guide decision making. In settings with high HCW compliance with universal RPE (above 50%), even noncompliant HCW are protected against clinical respiratory illness, which suggests a herd protective effect when universal RPE use is implemented, likely owing to the prevention of symptomatic or asymptomatic infectious spread among HCWs.¹⁴ It is important to note that the compliance rate may be limited by discomfort of prolonged wear of certain RPEs. One study reported that compliance rate is lower for continuous use (66%) than it is for targeted use (82%).¹³ Accumulated respiratory pathogen deposition on RPEs from an extended period of use that could result in self-­contamination to the wearer is a potential concern, although these risks must be balanced against the repeated donning and doffing required by targeted use. Pilot studies examining viral particles left on surgical masks after being worn for entire shifts (or as long as tolerated) found that there were significantly more viral particles detected after 6 hours of continuous wear, which may increase the risk of self-contamination.¹⁵

The current literature is applicable to infections that are known to spread via droplet contact, and this is a major limitation in generalizing the available evidence to the SARS-CoV-2 pandemic, in which debate persists regarding the exact mode of transmission. It is postulated that, even in infections traditionally considered to be spread by droplets, such as influenza, aerosol transmission may occur when HCWs are working in close proximity to the exposure source or when the droplet evaporates and becomes droplet nuclei. The United States National Academies of Science, Engineering, and Medicine expert consultation report, published in April 2020, concluded that current studies support the possibility of aerosolization of SARS-CoV-2 virus from normal breathing.¹⁶ As of April 2020, the WHO recommendation for SARS-CoV-2 is to use droplet contact precautions with a surgical mask for regular patient care and N95 respirator for aerosol-generating procedures.¹ Although we have not come across any studies specifically comparing the efficacy between surgical mask to N95 respirator protection while performing aerosol-generating procedures, a systematic review found that certain aerosol-generating procedures, such as endotracheal intubation and noninvasive ventilation, conferred a significantly higher risk of transmission of SARS-CoV-1 to HCWs in 2003.¹⁷ For the current crisis, the CDC is taking a cautious approach in which N95 respirators are recommended for HCWs caring for patients with confirmed or suspected SARS-CoV-2 infection if the supply chain is secure, with advice in place in times of RPE shortage, such as use of expired respirators, other types of equivalent respirators, or respirators not approved by the National Institute for Occupational Safety and Health, as well as optimization of administrative and engineering controls (eg, telemedicine, limiting patient and visitor numbers, physical barriers, optimizing ventilation systems).^2,18 This advice is unusual in terms of deviating from advising the most appropriate RPE, and we presume it reflects the present global supply problems.

RPEs are only one component of a necessary personal protective equipment ensemble. Although eye protection (goggles or face shields) is recommended by the WHO and CDC when caring for patients with SARS-CoV-2, there is considerable uncertainty regarding the incremental effectiveness of eye protection because such protection is usually worn in conjunction with RPE. A 2019 Cochrane review did not identify any good-quality studies that could inform judgments regarding the effectiveness of eye protective equipment,¹⁹ and a recent rapid review reporting on the efficacy of eye protection in primary care settings reached a similar conclusion.²⁰ A risk-based approach would be to include eye protection in a well-­designed personal protective equipment program.

In the absence of aerosol-generating procedures, N95 respirators confer no additional benefit in preventing HCW respiratory infections when droplet transmission is suspected. However, the applicability of the available evidence is limited given the uncertainties surrounding SARS-CoV-2 transmission. When RPE may become scarce during a pandemic, the risk of potential self-contamination must be weighed against RPE conservation strategies. RPE compliance, herd-protection effects of routine RPE use, and RPE contamination from prolonged use are therefore important elements to consider when implementing hospital policies regarding universal masking because they all impact the potential effectiveness of RPE.

At the present time we lack definitive evidence on the effectiveness of surgical masks vs respirators and continuous vs targeted RPE use in the hospital setting for SARS-CoV-2. If our goal is to minimize risk of HCW infection, continuous use of N95 respirator could be considered. However, a more pragmatic solution in the setting of a limited supply of N95 respirators would be continuous use of surgical masks while engaged in clinical care of patients under investigation or with confirmed COVID-19.

One contentious issue during the COVID-19 crisis has been the appropriate selection and use of respiratory protective equipment (RPE) for healthcare workers (HCWs) in hospitals and long-term care settings. As of April 2020, discrepancies exist in the recommendations from health authorities such as the World Health Organization (WHO), Centers for Disease Control and Prevention (CDC), and Canadian Standards Association (CSA). The first of these recommends a surgical mask for routine care and a respirator for high-risk care such as aerosol-generating procedures, while the CDC recommends respirators for all aspects of patient care for these SARS-CoV-2–infected patients, and the CSA risk assessment tool would also result in selection of a respirator.^1-3

Given the contradictory guidance, we will discuss several important considerations for hospital leaders in the implementation of a healthcare respiratory protection program during the current pandemic, including a focused review of the empirical data on surgical mask vs face-fitted respirator (most commonly available in healthcare as N95 in North America), continuous use of the RPE throughout an entire shift vs targeted use when caring for patients, and key areas of uncertainty.

Surgical masks are traditionally used for protection against droplet transmission of respiratory infections, in which large droplets often fall to the ground within short distances; on the other hand, N95 respirators are used for much smaller airborne pathogens, which can remain suspended in the air for long periods of time. Although empiric studies have supported the superiority of respirators over surgical masks in simulated settings (frequently defined as a calculated concentration ratio outside vs inside the RPE), most clinical studies fail to demonstrate a difference in clinical outcomes such as the prevention of respiratory infection. For instance, an exposure study using saline aerosol to simulate viral particles showed that N95 respirators conferred up to 8 to 12 times greater protection against particulate penetration, compared with surgical masks.⁴ However, these advantages of respirators over surgical masks in carefully controlled laboratory studies do not seem to translate to decreased infection risk in real-world settings.

The effectiveness of N95 respirators vs surgical masks in preventing respiratory infections has been evaluated in a small number of clinical randomized, controlled trials (RCTs). We identified five systematic reviews and/or metanalyses published after 2010 and three RCTs published after 1990.^5-12 The RCTs used laboratory-confirmed respiratory virus or clinical infection in HCWs as a clinical outcome, but studies differed in the implementation of RPE use (ie, continuous or targeted use). In a systematic review and metanalysis, Long et al identified six RCTs (9,171 participants) and concluded that, with the exception of laboratory-confirmed bacterial colonization, N95 respirators did not reduce the rate of laboratory-­confirmed influenza, viral respiratory infections, or influenza-like illness among HCWs, compared with surgical masks.⁵ The authors noted risks of bias in these studies owing to the inability to blind and conceal allocation. In addition, the studies focused on infections that are known to transmit via droplet, such as influenza, so the results might not be applicable in the face of a new pandemic in which the important modes of transmission are not yet clear.

The evidence base evaluating continuous vs targeted use of RPE in healthcare settings is quite small. Continuous use refers to using the RPE during an entire shift, whereas targeted use involves using RPE only when caring for confirmed or suspected respiratory patients. In our literature review we identified only one RCT that included separate study arms for continuous and targeted N95 respirator use.¹³ The authors found a significantly lower rate of clinical respiratory illness among HCWs in the continuous-use group, compared with that in the targeted-use group. Limitations of the study included a relatively short follow-up of 4 weeks and uneven distribution of baseline characteristics, although the authors adjusted for these differences in their analysis. The study, however, did not compare continuous vs targeted use of surgical masks with regard to clinical outcomes. Based on the study results, we can only infer that continuous use of RPE, either surgical mask or N95 respirator, may provide additional benefit to HCWs vs targeted use only.

Given the lack of robust evidence informing continuous or targeted RPE use, we suggest some additional factors to guide decision making. In settings with high HCW compliance with universal RPE (above 50%), even noncompliant HCW are protected against clinical respiratory illness, which suggests a herd protective effect when universal RPE use is implemented, likely owing to the prevention of symptomatic or asymptomatic infectious spread among HCWs.¹⁴ It is important to note that the compliance rate may be limited by discomfort of prolonged wear of certain RPEs. One study reported that compliance rate is lower for continuous use (66%) than it is for targeted use (82%).¹³ Accumulated respiratory pathogen deposition on RPEs from an extended period of use that could result in self-­contamination to the wearer is a potential concern, although these risks must be balanced against the repeated donning and doffing required by targeted use. Pilot studies examining viral particles left on surgical masks after being worn for entire shifts (or as long as tolerated) found that there were significantly more viral particles detected after 6 hours of continuous wear, which may increase the risk of self-contamination.¹⁵

The current literature is applicable to infections that are known to spread via droplet contact, and this is a major limitation in generalizing the available evidence to the SARS-CoV-2 pandemic, in which debate persists regarding the exact mode of transmission. It is postulated that, even in infections traditionally considered to be spread by droplets, such as influenza, aerosol transmission may occur when HCWs are working in close proximity to the exposure source or when the droplet evaporates and becomes droplet nuclei. The United States National Academies of Science, Engineering, and Medicine expert consultation report, published in April 2020, concluded that current studies support the possibility of aerosolization of SARS-CoV-2 virus from normal breathing.¹⁶ As of April 2020, the WHO recommendation for SARS-CoV-2 is to use droplet contact precautions with a surgical mask for regular patient care and N95 respirator for aerosol-generating procedures.¹ Although we have not come across any studies specifically comparing the efficacy between surgical mask to N95 respirator protection while performing aerosol-generating procedures, a systematic review found that certain aerosol-generating procedures, such as endotracheal intubation and noninvasive ventilation, conferred a significantly higher risk of transmission of SARS-CoV-1 to HCWs in 2003.¹⁷ For the current crisis, the CDC is taking a cautious approach in which N95 respirators are recommended for HCWs caring for patients with confirmed or suspected SARS-CoV-2 infection if the supply chain is secure, with advice in place in times of RPE shortage, such as use of expired respirators, other types of equivalent respirators, or respirators not approved by the National Institute for Occupational Safety and Health, as well as optimization of administrative and engineering controls (eg, telemedicine, limiting patient and visitor numbers, physical barriers, optimizing ventilation systems).^2,18 This advice is unusual in terms of deviating from advising the most appropriate RPE, and we presume it reflects the present global supply problems.

RPEs are only one component of a necessary personal protective equipment ensemble. Although eye protection (goggles or face shields) is recommended by the WHO and CDC when caring for patients with SARS-CoV-2, there is considerable uncertainty regarding the incremental effectiveness of eye protection because such protection is usually worn in conjunction with RPE. A 2019 Cochrane review did not identify any good-quality studies that could inform judgments regarding the effectiveness of eye protective equipment,¹⁹ and a recent rapid review reporting on the efficacy of eye protection in primary care settings reached a similar conclusion.²⁰ A risk-based approach would be to include eye protection in a well-­designed personal protective equipment program.

In the absence of aerosol-generating procedures, N95 respirators confer no additional benefit in preventing HCW respiratory infections when droplet transmission is suspected. However, the applicability of the available evidence is limited given the uncertainties surrounding SARS-CoV-2 transmission. When RPE may become scarce during a pandemic, the risk of potential self-contamination must be weighed against RPE conservation strategies. RPE compliance, herd-protection effects of routine RPE use, and RPE contamination from prolonged use are therefore important elements to consider when implementing hospital policies regarding universal masking because they all impact the potential effectiveness of RPE.

At the present time we lack definitive evidence on the effectiveness of surgical masks vs respirators and continuous vs targeted RPE use in the hospital setting for SARS-CoV-2. If our goal is to minimize risk of HCW infection, continuous use of N95 respirator could be considered. However, a more pragmatic solution in the setting of a limited supply of N95 respirators would be continuous use of surgical masks while engaged in clinical care of patients under investigation or with confirmed COVID-19.

One contentious issue during the COVID-19 crisis has been the appropriate selection and use of respiratory protective equipment (RPE) for healthcare workers (HCWs) in hospitals and long-term care settings. As of April 2020, discrepancies exist in the recommendations from health authorities such as the World Health Organization (WHO), Centers for Disease Control and Prevention (CDC), and Canadian Standards Association (CSA). The first of these recommends a surgical mask for routine care and a respirator for high-risk care such as aerosol-generating procedures, while the CDC recommends respirators for all aspects of patient care for these SARS-CoV-2–infected patients, and the CSA risk assessment tool would also result in selection of a respirator.^1-3

Given the contradictory guidance, we will discuss several important considerations for hospital leaders in the implementation of a healthcare respiratory protection program during the current pandemic, including a focused review of the empirical data on surgical mask vs face-fitted respirator (most commonly available in healthcare as N95 in North America), continuous use of the RPE throughout an entire shift vs targeted use when caring for patients, and key areas of uncertainty.

Surgical masks are traditionally used for protection against droplet transmission of respiratory infections, in which large droplets often fall to the ground within short distances; on the other hand, N95 respirators are used for much smaller airborne pathogens, which can remain suspended in the air for long periods of time. Although empiric studies have supported the superiority of respirators over surgical masks in simulated settings (frequently defined as a calculated concentration ratio outside vs inside the RPE), most clinical studies fail to demonstrate a difference in clinical outcomes such as the prevention of respiratory infection. For instance, an exposure study using saline aerosol to simulate viral particles showed that N95 respirators conferred up to 8 to 12 times greater protection against particulate penetration, compared with surgical masks.⁴ However, these advantages of respirators over surgical masks in carefully controlled laboratory studies do not seem to translate to decreased infection risk in real-world settings.

The effectiveness of N95 respirators vs surgical masks in preventing respiratory infections has been evaluated in a small number of clinical randomized, controlled trials (RCTs). We identified five systematic reviews and/or metanalyses published after 2010 and three RCTs published after 1990.^5-12 The RCTs used laboratory-confirmed respiratory virus or clinical infection in HCWs as a clinical outcome, but studies differed in the implementation of RPE use (ie, continuous or targeted use). In a systematic review and metanalysis, Long et al identified six RCTs (9,171 participants) and concluded that, with the exception of laboratory-confirmed bacterial colonization, N95 respirators did not reduce the rate of laboratory-­confirmed influenza, viral respiratory infections, or influenza-like illness among HCWs, compared with surgical masks.⁵ The authors noted risks of bias in these studies owing to the inability to blind and conceal allocation. In addition, the studies focused on infections that are known to transmit via droplet, such as influenza, so the results might not be applicable in the face of a new pandemic in which the important modes of transmission are not yet clear.

The evidence base evaluating continuous vs targeted use of RPE in healthcare settings is quite small. Continuous use refers to using the RPE during an entire shift, whereas targeted use involves using RPE only when caring for confirmed or suspected respiratory patients. In our literature review we identified only one RCT that included separate study arms for continuous and targeted N95 respirator use.¹³ The authors found a significantly lower rate of clinical respiratory illness among HCWs in the continuous-use group, compared with that in the targeted-use group. Limitations of the study included a relatively short follow-up of 4 weeks and uneven distribution of baseline characteristics, although the authors adjusted for these differences in their analysis. The study, however, did not compare continuous vs targeted use of surgical masks with regard to clinical outcomes. Based on the study results, we can only infer that continuous use of RPE, either surgical mask or N95 respirator, may provide additional benefit to HCWs vs targeted use only.

Given the lack of robust evidence informing continuous or targeted RPE use, we suggest some additional factors to guide decision making. In settings with high HCW compliance with universal RPE (above 50%), even noncompliant HCW are protected against clinical respiratory illness, which suggests a herd protective effect when universal RPE use is implemented, likely owing to the prevention of symptomatic or asymptomatic infectious spread among HCWs.¹⁴ It is important to note that the compliance rate may be limited by discomfort of prolonged wear of certain RPEs. One study reported that compliance rate is lower for continuous use (66%) than it is for targeted use (82%).¹³ Accumulated respiratory pathogen deposition on RPEs from an extended period of use that could result in self-­contamination to the wearer is a potential concern, although these risks must be balanced against the repeated donning and doffing required by targeted use. Pilot studies examining viral particles left on surgical masks after being worn for entire shifts (or as long as tolerated) found that there were significantly more viral particles detected after 6 hours of continuous wear, which may increase the risk of self-contamination.¹⁵

The current literature is applicable to infections that are known to spread via droplet contact, and this is a major limitation in generalizing the available evidence to the SARS-CoV-2 pandemic, in which debate persists regarding the exact mode of transmission. It is postulated that, even in infections traditionally considered to be spread by droplets, such as influenza, aerosol transmission may occur when HCWs are working in close proximity to the exposure source or when the droplet evaporates and becomes droplet nuclei. The United States National Academies of Science, Engineering, and Medicine expert consultation report, published in April 2020, concluded that current studies support the possibility of aerosolization of SARS-CoV-2 virus from normal breathing.¹⁶ As of April 2020, the WHO recommendation for SARS-CoV-2 is to use droplet contact precautions with a surgical mask for regular patient care and N95 respirator for aerosol-generating procedures.¹ Although we have not come across any studies specifically comparing the efficacy between surgical mask to N95 respirator protection while performing aerosol-generating procedures, a systematic review found that certain aerosol-generating procedures, such as endotracheal intubation and noninvasive ventilation, conferred a significantly higher risk of transmission of SARS-CoV-1 to HCWs in 2003.¹⁷ For the current crisis, the CDC is taking a cautious approach in which N95 respirators are recommended for HCWs caring for patients with confirmed or suspected SARS-CoV-2 infection if the supply chain is secure, with advice in place in times of RPE shortage, such as use of expired respirators, other types of equivalent respirators, or respirators not approved by the National Institute for Occupational Safety and Health, as well as optimization of administrative and engineering controls (eg, telemedicine, limiting patient and visitor numbers, physical barriers, optimizing ventilation systems).^2,18 This advice is unusual in terms of deviating from advising the most appropriate RPE, and we presume it reflects the present global supply problems.

RPEs are only one component of a necessary personal protective equipment ensemble. Although eye protection (goggles or face shields) is recommended by the WHO and CDC when caring for patients with SARS-CoV-2, there is considerable uncertainty regarding the incremental effectiveness of eye protection because such protection is usually worn in conjunction with RPE. A 2019 Cochrane review did not identify any good-quality studies that could inform judgments regarding the effectiveness of eye protective equipment,¹⁹ and a recent rapid review reporting on the efficacy of eye protection in primary care settings reached a similar conclusion.²⁰ A risk-based approach would be to include eye protection in a well-­designed personal protective equipment program.

In the absence of aerosol-generating procedures, N95 respirators confer no additional benefit in preventing HCW respiratory infections when droplet transmission is suspected. However, the applicability of the available evidence is limited given the uncertainties surrounding SARS-CoV-2 transmission. When RPE may become scarce during a pandemic, the risk of potential self-contamination must be weighed against RPE conservation strategies. RPE compliance, herd-protection effects of routine RPE use, and RPE contamination from prolonged use are therefore important elements to consider when implementing hospital policies regarding universal masking because they all impact the potential effectiveness of RPE.

At the present time we lack definitive evidence on the effectiveness of surgical masks vs respirators and continuous vs targeted RPE use in the hospital setting for SARS-CoV-2. If our goal is to minimize risk of HCW infection, continuous use of N95 respirator could be considered. However, a more pragmatic solution in the setting of a limited supply of N95 respirators would be continuous use of surgical masks while engaged in clinical care of patients under investigation or with confirmed COVID-19.

Emerging studies on coronavirus disease 2019 (COVID-19) confirm high rates of infection among healthcare workers (HCWs).¹ As widespread community transmission increases, frontline HCWs, such as hospitalists, are at particularly high risk of exposure to people with undiagnosed COVID-19. Although there is no known effective treatment for COVID-19, early detection is vital to decreasing ongoing transmission through contact tracing and quarantine. However, lack of adequate testing capacity prevented basic public health interventions from curbing the pandemic at an earlier stage. As a result, given high rates of presumed community transmission of COVID-19 and evidence for asymptomatic transmission, there have been moves toward the use of universal personal protective equipment (PPE). This strategy is challenging to implement because of the acute PPE shortage, which has resulted in an urgent need to embrace innovation in infection prevention.

The current pandemic has resulted in an unprecedented volume of data being generated and disseminated, with the potential to impact real-time responses in geographically disparate regions. Here, we focus on the potential for innovation and knowledge sharing from an infection prevention perspective, which could enhance frontline HCW safety in the current COVID-19 pandemic.

Every outbreak begins and ends with a diagnostic test. Widespread population testing coupled with intensive contact tracing had the potential to curb national epidemics if it had been implemented in time. In the United States, which now has the highest number of COVID-19 cases worldwide, there were technical difficulties with the first diagnostic test developed by the Centers for Disease Control and Prevention (CDC) and subsequent delays in scaling up access to COVID-19 diagnostic testing.² The strategy of initially reserving testing only for those who were critically ill meant that by the time patients with COVID-19 were being diagnosed, widespread community spread had occurred because of the lack of detection of individuals with less severe or asymptomatic infections.

In contrast, scaled-up testing in South Korea has helped limit the spread and consequences of COVID-19. The use of drive-through testing centers enabled safe and efficient testing, while minimizing the risk to HCWs and eliminating the possibility of cross infection among people being tested.³ Although outdoor testing is not feasible in all settings, this approach avoids resources and time typically needed for ventilation (typically a negative pressure room with 12 air changes per hour would be used) and cleaning of specimen collection rooms.

The other major diagnostic gap is the ability to identify individuals who have recovered from COVID-19 and are immune. There is an urgent need to develop and scale up a rapid serological test that avoids cross-reactivity with other coronaviruses. Ideally, this test would permit testing of HCWs to determine who is likely immune and can therefore return safely to work.

There has been a massive and rapid increase in the need for PPE globally because of overwhelmed health systems having to care for large numbers of patients with suspected or confirmed COVID-19. This has been exacerbated by public fear, which has led to panic buying of medical face masks (primarily used to protect others from infections with a droplet mode of transmission) and filtering facepiece half-mask respirators, which include N95 respirators (used to protect the wearer from infections with an airborne mode of transmission).

COVID-19 is thought to be predominantly spread by transmission of respiratory droplets (>5 and <10 μm in diameter), which occurs when people are in close contact (within 1 meter) with others who typically (but not always) have respiratory symptoms such as cough or sneeze or with fomites that have come into contact with an infected person. This is in contrast to infectious diseases such as tuberculosis (TB) or measles, which are spread by airborne transmission of virus suspended in droplet nuclei (<5 μm in diameter), which can remain in the air for prolonged periods of time and can be transmitted over distances greater than 1 meter.⁴

While World Health Organization (WHO) and CDC infection prevention guidance have cited droplet transmission as the primary mode of transmission for COVID-19, current CDC guidelines state that respirators are preferred for the care of patients with known or suspected COVID-19, given the potential for opportunistic airborne transmission.⁵ However, in the setting of respirator shortages, it is recommended that these should be prioritized for HCWs caring for patients with COVID-19 in the context of aerosol-generating procedures or other patients with infections spread by airborne transmission such as TB or varicella until the supply chain is restored. Of note, optimal use of respirators requires fit testing, which is often lacking in nursing homes and outpatient facilities, as well as more widely in resource-limited countries.

Universal masking (use of surgical mask) for HCWs caring for any patient irrespective of symptoms or presenting complaint has also been implemented by many large hospital systems in recent days. Although universal masking adds to the burden of the PPE shortage, in settings with widespread community transmission and given increasing evidence⁶ demonstrating transmission from people with asymptomatic and presymptomatic infection, universal masking may be useful to decrease transmission. However, particularly in the setting of PPE shortages, it is important to understand that surgical masks are designed to be single use and that dampness and frequent adjustment of the mask affects their effectiveness.

As urgent attempts to coordinate and increase PPE manufacture are being made by health systems, in conjunction with private partnerships, there has also been a burst of public campaigns to sew cloth masks to mitigate the real-time shortages. Although it is likely that cloth masks provide inadequate protection in comparison with surgical masks,⁷ evidence does suggest that cloth masks provide some degree of protection from the spread of respiratory viruses,⁸ particularly if these are replaced promptly when damp or damaged and if combined with other interventions such as hand hygiene. This has led to recommendations for the general public in various countries to wear cloth face coverings in public settings, particularly where social distancing may be harder to maintain, but these are not recommended for use by HCWs in healthcare settings.

Strategies to navigate the PPE shortage in the era of COVID-19 include importing, reclaiming, reusing, and repurposing PPE; generating and extending supply; eliminating nonessential services; reducing patient contact; and using nonhuman services such as drones to deliver equipment and undertake tasks such as decontamination.^9,10 Multidisciplinary teams are working on creative ways to use existing resources to make effective PPE, including alternatives to N95 respirators. An outbreak simulation study at Emory University in Atlanta, Georgia, and the University of Texas Health Science Center at Houston in Texas demonstrated that HCWs could be rapidly trained and fit tested to use elastomeric half-mask respirators, which are reusable.¹¹ A multidisciplinary team at Boston Children’s Hospital in Massachusetts has developed and completed a small pilot study of a reusable elastomeric respirator made using an anesthesia facemask, antimicrobial filter, and elastic straps.¹²

Given evidence that suggests that COVID-19 involves a component of airborne transmission,¹³ in addition to droplet spread and surface (fomite) contamination,¹⁴ using known infection prevention techniques that work to decrease airborne transmission of other respiratory infectious diseases should also be considered. Germicidal ultraviolet (GUV) air disinfection rapidly disinfects upper room air, which is then continually exchanged with contaminated lower-room air. GUV air disinfection has been demonstrated to be a safe and cost-effective intervention, with an efficacy of approximately 80% for decreasing TB transmission.¹⁵ GUV air disinfection is also effective against airborne influenza and measles and may play a role in surface decontamination by accelerating viral inactivation. Enabling GUV in high-risk areas such as the emergency department or intensive care unit could be a high-yield intervention to decrease transmission of COVID-19.

HCWs exposed to other respiratory infections such as influenza or TB may receive preventive therapy to reduce the risk of developing disease. High rates of COVID-19 in HCWs have prompted several initiatives to evaluate innovative approaches to decreasing this risk. Multiple studies are underway to determine whether hydroxychloroquine could be used for pre- or postexposure prophylaxis to prevent COVID-19. Another multisite trial will evaluate whether the BCG vaccine, primarily used to reduce the risk of TB, provides protection against COVID-19 in HCWs, driven by data suggesting a correlation between universal BCG vaccination policy and reduced morbidity and mortality for COVID-19.¹⁶

Infection prevention efforts can benefit from the unprecedented amount of data on COVID-19 that is being generated and shared. Successful examples of the rapid intensification of infection prevention measures to decrease transmission in healthcare facilities should be emulated. The hospital authority of Hong Kong implemented a bundle of measures focused on early recognition, isolation, notification, and molecular diagnostics for people being evaluated for COVID-19.¹⁷ They subsequently broadened the clinical and epidemiological criteria of surveillance as the outbreak evolved and intensified PPE recommendations to all HCWs (face masks for all and N95 respirators for those performing aerosol-generating procedures), which appears to have resulted in no cases of HCW infection or nosocomial transmission.

Data characterizing the extent of occupational infections in HCWs during acute and chronic epidemics is often lacking and subject to wide variability in reporting, which limits its impact. For example, HCWs in high TB incidence countries have at least twice the risk of developing TB, compared with the general population. Although there are still major gaps in national data collection regarding the incidence of occupational TB, recent attempts by WHO to systematically record this data have resulted in increasing prioritization of this group as an at-risk population who may benefit from TB preventive therapy. We strongly advocate that health systems systematically record and share longitudinal data on numbers of HCWs infected with COVID-19. This transparency will facilitate urgent action to replenish and sustain resources such as PPE and enable institutions to share and adapt successful infection prevention strategies. Examples such as the prevention of central line- associated bloodstream infections demonstrate the potential impact of national collaborative efforts to strengthen infection prevention, although further effort is needed to optimize knowledge sharing in the context of outbreaks.

The cost of not investing in public health pandemic preparedness including measures to protect HCWs is now widely apparent. HCWs have a right to safety in their workplaces as they fulfil their duty of care to patients.¹⁸ Rapid scale-up of diagnostic testing capacity, and bundles of infection prevention interventions including universal masking and drive-through testing, can safeguard HCWs and the patients they serve in the current COVID-19 pandemic. Re-establishing immediate access to quality-assured PPE is imperative to reduce the individual and workforce consequences of HCWs developing COVID-19 or other infectious diseases like TB that are continuously a threat to the workforce. Meanwhile, innovative approaches such as repurposing resources to develop PPE and use of GUV air disinfection may help to mitigate PPE shortages, and use of preventive therapies may also decrease COVID-19 risk in HCWs. Reliable surveillance data on HCW infection rates can help identify and track gaps in infection prevention, as well as identify strategies that impact this outcome. Ultimately greater top-down political commitment is urgently needed to ensure that frontline HCWs have the necessary resources to address the current pandemic and to sustain these interventions to protect HCWs in the future.

Emerging studies on coronavirus disease 2019 (COVID-19) confirm high rates of infection among healthcare workers (HCWs).¹ As widespread community transmission increases, frontline HCWs, such as hospitalists, are at particularly high risk of exposure to people with undiagnosed COVID-19. Although there is no known effective treatment for COVID-19, early detection is vital to decreasing ongoing transmission through contact tracing and quarantine. However, lack of adequate testing capacity prevented basic public health interventions from curbing the pandemic at an earlier stage. As a result, given high rates of presumed community transmission of COVID-19 and evidence for asymptomatic transmission, there have been moves toward the use of universal personal protective equipment (PPE). This strategy is challenging to implement because of the acute PPE shortage, which has resulted in an urgent need to embrace innovation in infection prevention.

The current pandemic has resulted in an unprecedented volume of data being generated and disseminated, with the potential to impact real-time responses in geographically disparate regions. Here, we focus on the potential for innovation and knowledge sharing from an infection prevention perspective, which could enhance frontline HCW safety in the current COVID-19 pandemic.

Every outbreak begins and ends with a diagnostic test. Widespread population testing coupled with intensive contact tracing had the potential to curb national epidemics if it had been implemented in time. In the United States, which now has the highest number of COVID-19 cases worldwide, there were technical difficulties with the first diagnostic test developed by the Centers for Disease Control and Prevention (CDC) and subsequent delays in scaling up access to COVID-19 diagnostic testing.² The strategy of initially reserving testing only for those who were critically ill meant that by the time patients with COVID-19 were being diagnosed, widespread community spread had occurred because of the lack of detection of individuals with less severe or asymptomatic infections.

In contrast, scaled-up testing in South Korea has helped limit the spread and consequences of COVID-19. The use of drive-through testing centers enabled safe and efficient testing, while minimizing the risk to HCWs and eliminating the possibility of cross infection among people being tested.³ Although outdoor testing is not feasible in all settings, this approach avoids resources and time typically needed for ventilation (typically a negative pressure room with 12 air changes per hour would be used) and cleaning of specimen collection rooms.

The other major diagnostic gap is the ability to identify individuals who have recovered from COVID-19 and are immune. There is an urgent need to develop and scale up a rapid serological test that avoids cross-reactivity with other coronaviruses. Ideally, this test would permit testing of HCWs to determine who is likely immune and can therefore return safely to work.

There has been a massive and rapid increase in the need for PPE globally because of overwhelmed health systems having to care for large numbers of patients with suspected or confirmed COVID-19. This has been exacerbated by public fear, which has led to panic buying of medical face masks (primarily used to protect others from infections with a droplet mode of transmission) and filtering facepiece half-mask respirators, which include N95 respirators (used to protect the wearer from infections with an airborne mode of transmission).

COVID-19 is thought to be predominantly spread by transmission of respiratory droplets (>5 and <10 μm in diameter), which occurs when people are in close contact (within 1 meter) with others who typically (but not always) have respiratory symptoms such as cough or sneeze or with fomites that have come into contact with an infected person. This is in contrast to infectious diseases such as tuberculosis (TB) or measles, which are spread by airborne transmission of virus suspended in droplet nuclei (<5 μm in diameter), which can remain in the air for prolonged periods of time and can be transmitted over distances greater than 1 meter.⁴

While World Health Organization (WHO) and CDC infection prevention guidance have cited droplet transmission as the primary mode of transmission for COVID-19, current CDC guidelines state that respirators are preferred for the care of patients with known or suspected COVID-19, given the potential for opportunistic airborne transmission.⁵ However, in the setting of respirator shortages, it is recommended that these should be prioritized for HCWs caring for patients with COVID-19 in the context of aerosol-generating procedures or other patients with infections spread by airborne transmission such as TB or varicella until the supply chain is restored. Of note, optimal use of respirators requires fit testing, which is often lacking in nursing homes and outpatient facilities, as well as more widely in resource-limited countries.

Universal masking (use of surgical mask) for HCWs caring for any patient irrespective of symptoms or presenting complaint has also been implemented by many large hospital systems in recent days. Although universal masking adds to the burden of the PPE shortage, in settings with widespread community transmission and given increasing evidence⁶ demonstrating transmission from people with asymptomatic and presymptomatic infection, universal masking may be useful to decrease transmission. However, particularly in the setting of PPE shortages, it is important to understand that surgical masks are designed to be single use and that dampness and frequent adjustment of the mask affects their effectiveness.

As urgent attempts to coordinate and increase PPE manufacture are being made by health systems, in conjunction with private partnerships, there has also been a burst of public campaigns to sew cloth masks to mitigate the real-time shortages. Although it is likely that cloth masks provide inadequate protection in comparison with surgical masks,⁷ evidence does suggest that cloth masks provide some degree of protection from the spread of respiratory viruses,⁸ particularly if these are replaced promptly when damp or damaged and if combined with other interventions such as hand hygiene. This has led to recommendations for the general public in various countries to wear cloth face coverings in public settings, particularly where social distancing may be harder to maintain, but these are not recommended for use by HCWs in healthcare settings.

Strategies to navigate the PPE shortage in the era of COVID-19 include importing, reclaiming, reusing, and repurposing PPE; generating and extending supply; eliminating nonessential services; reducing patient contact; and using nonhuman services such as drones to deliver equipment and undertake tasks such as decontamination.^9,10 Multidisciplinary teams are working on creative ways to use existing resources to make effective PPE, including alternatives to N95 respirators. An outbreak simulation study at Emory University in Atlanta, Georgia, and the University of Texas Health Science Center at Houston in Texas demonstrated that HCWs could be rapidly trained and fit tested to use elastomeric half-mask respirators, which are reusable.¹¹ A multidisciplinary team at Boston Children’s Hospital in Massachusetts has developed and completed a small pilot study of a reusable elastomeric respirator made using an anesthesia facemask, antimicrobial filter, and elastic straps.¹²

Given evidence that suggests that COVID-19 involves a component of airborne transmission,¹³ in addition to droplet spread and surface (fomite) contamination,¹⁴ using known infection prevention techniques that work to decrease airborne transmission of other respiratory infectious diseases should also be considered. Germicidal ultraviolet (GUV) air disinfection rapidly disinfects upper room air, which is then continually exchanged with contaminated lower-room air. GUV air disinfection has been demonstrated to be a safe and cost-effective intervention, with an efficacy of approximately 80% for decreasing TB transmission.¹⁵ GUV air disinfection is also effective against airborne influenza and measles and may play a role in surface decontamination by accelerating viral inactivation. Enabling GUV in high-risk areas such as the emergency department or intensive care unit could be a high-yield intervention to decrease transmission of COVID-19.

HCWs exposed to other respiratory infections such as influenza or TB may receive preventive therapy to reduce the risk of developing disease. High rates of COVID-19 in HCWs have prompted several initiatives to evaluate innovative approaches to decreasing this risk. Multiple studies are underway to determine whether hydroxychloroquine could be used for pre- or postexposure prophylaxis to prevent COVID-19. Another multisite trial will evaluate whether the BCG vaccine, primarily used to reduce the risk of TB, provides protection against COVID-19 in HCWs, driven by data suggesting a correlation between universal BCG vaccination policy and reduced morbidity and mortality for COVID-19.¹⁶

Infection prevention efforts can benefit from the unprecedented amount of data on COVID-19 that is being generated and shared. Successful examples of the rapid intensification of infection prevention measures to decrease transmission in healthcare facilities should be emulated. The hospital authority of Hong Kong implemented a bundle of measures focused on early recognition, isolation, notification, and molecular diagnostics for people being evaluated for COVID-19.¹⁷ They subsequently broadened the clinical and epidemiological criteria of surveillance as the outbreak evolved and intensified PPE recommendations to all HCWs (face masks for all and N95 respirators for those performing aerosol-generating procedures), which appears to have resulted in no cases of HCW infection or nosocomial transmission.

Data characterizing the extent of occupational infections in HCWs during acute and chronic epidemics is often lacking and subject to wide variability in reporting, which limits its impact. For example, HCWs in high TB incidence countries have at least twice the risk of developing TB, compared with the general population. Although there are still major gaps in national data collection regarding the incidence of occupational TB, recent attempts by WHO to systematically record this data have resulted in increasing prioritization of this group as an at-risk population who may benefit from TB preventive therapy. We strongly advocate that health systems systematically record and share longitudinal data on numbers of HCWs infected with COVID-19. This transparency will facilitate urgent action to replenish and sustain resources such as PPE and enable institutions to share and adapt successful infection prevention strategies. Examples such as the prevention of central line- associated bloodstream infections demonstrate the potential impact of national collaborative efforts to strengthen infection prevention, although further effort is needed to optimize knowledge sharing in the context of outbreaks.

The cost of not investing in public health pandemic preparedness including measures to protect HCWs is now widely apparent. HCWs have a right to safety in their workplaces as they fulfil their duty of care to patients.¹⁸ Rapid scale-up of diagnostic testing capacity, and bundles of infection prevention interventions including universal masking and drive-through testing, can safeguard HCWs and the patients they serve in the current COVID-19 pandemic. Re-establishing immediate access to quality-assured PPE is imperative to reduce the individual and workforce consequences of HCWs developing COVID-19 or other infectious diseases like TB that are continuously a threat to the workforce. Meanwhile, innovative approaches such as repurposing resources to develop PPE and use of GUV air disinfection may help to mitigate PPE shortages, and use of preventive therapies may also decrease COVID-19 risk in HCWs. Reliable surveillance data on HCW infection rates can help identify and track gaps in infection prevention, as well as identify strategies that impact this outcome. Ultimately greater top-down political commitment is urgently needed to ensure that frontline HCWs have the necessary resources to address the current pandemic and to sustain these interventions to protect HCWs in the future.

Emerging studies on coronavirus disease 2019 (COVID-19) confirm high rates of infection among healthcare workers (HCWs).¹ As widespread community transmission increases, frontline HCWs, such as hospitalists, are at particularly high risk of exposure to people with undiagnosed COVID-19. Although there is no known effective treatment for COVID-19, early detection is vital to decreasing ongoing transmission through contact tracing and quarantine. However, lack of adequate testing capacity prevented basic public health interventions from curbing the pandemic at an earlier stage. As a result, given high rates of presumed community transmission of COVID-19 and evidence for asymptomatic transmission, there have been moves toward the use of universal personal protective equipment (PPE). This strategy is challenging to implement because of the acute PPE shortage, which has resulted in an urgent need to embrace innovation in infection prevention.

The current pandemic has resulted in an unprecedented volume of data being generated and disseminated, with the potential to impact real-time responses in geographically disparate regions. Here, we focus on the potential for innovation and knowledge sharing from an infection prevention perspective, which could enhance frontline HCW safety in the current COVID-19 pandemic.

Every outbreak begins and ends with a diagnostic test. Widespread population testing coupled with intensive contact tracing had the potential to curb national epidemics if it had been implemented in time. In the United States, which now has the highest number of COVID-19 cases worldwide, there were technical difficulties with the first diagnostic test developed by the Centers for Disease Control and Prevention (CDC) and subsequent delays in scaling up access to COVID-19 diagnostic testing.² The strategy of initially reserving testing only for those who were critically ill meant that by the time patients with COVID-19 were being diagnosed, widespread community spread had occurred because of the lack of detection of individuals with less severe or asymptomatic infections.

In contrast, scaled-up testing in South Korea has helped limit the spread and consequences of COVID-19. The use of drive-through testing centers enabled safe and efficient testing, while minimizing the risk to HCWs and eliminating the possibility of cross infection among people being tested.³ Although outdoor testing is not feasible in all settings, this approach avoids resources and time typically needed for ventilation (typically a negative pressure room with 12 air changes per hour would be used) and cleaning of specimen collection rooms.

The other major diagnostic gap is the ability to identify individuals who have recovered from COVID-19 and are immune. There is an urgent need to develop and scale up a rapid serological test that avoids cross-reactivity with other coronaviruses. Ideally, this test would permit testing of HCWs to determine who is likely immune and can therefore return safely to work.

There has been a massive and rapid increase in the need for PPE globally because of overwhelmed health systems having to care for large numbers of patients with suspected or confirmed COVID-19. This has been exacerbated by public fear, which has led to panic buying of medical face masks (primarily used to protect others from infections with a droplet mode of transmission) and filtering facepiece half-mask respirators, which include N95 respirators (used to protect the wearer from infections with an airborne mode of transmission).

COVID-19 is thought to be predominantly spread by transmission of respiratory droplets (>5 and <10 μm in diameter), which occurs when people are in close contact (within 1 meter) with others who typically (but not always) have respiratory symptoms such as cough or sneeze or with fomites that have come into contact with an infected person. This is in contrast to infectious diseases such as tuberculosis (TB) or measles, which are spread by airborne transmission of virus suspended in droplet nuclei (<5 μm in diameter), which can remain in the air for prolonged periods of time and can be transmitted over distances greater than 1 meter.⁴

While World Health Organization (WHO) and CDC infection prevention guidance have cited droplet transmission as the primary mode of transmission for COVID-19, current CDC guidelines state that respirators are preferred for the care of patients with known or suspected COVID-19, given the potential for opportunistic airborne transmission.⁵ However, in the setting of respirator shortages, it is recommended that these should be prioritized for HCWs caring for patients with COVID-19 in the context of aerosol-generating procedures or other patients with infections spread by airborne transmission such as TB or varicella until the supply chain is restored. Of note, optimal use of respirators requires fit testing, which is often lacking in nursing homes and outpatient facilities, as well as more widely in resource-limited countries.

Universal masking (use of surgical mask) for HCWs caring for any patient irrespective of symptoms or presenting complaint has also been implemented by many large hospital systems in recent days. Although universal masking adds to the burden of the PPE shortage, in settings with widespread community transmission and given increasing evidence⁶ demonstrating transmission from people with asymptomatic and presymptomatic infection, universal masking may be useful to decrease transmission. However, particularly in the setting of PPE shortages, it is important to understand that surgical masks are designed to be single use and that dampness and frequent adjustment of the mask affects their effectiveness.

As urgent attempts to coordinate and increase PPE manufacture are being made by health systems, in conjunction with private partnerships, there has also been a burst of public campaigns to sew cloth masks to mitigate the real-time shortages. Although it is likely that cloth masks provide inadequate protection in comparison with surgical masks,⁷ evidence does suggest that cloth masks provide some degree of protection from the spread of respiratory viruses,⁸ particularly if these are replaced promptly when damp or damaged and if combined with other interventions such as hand hygiene. This has led to recommendations for the general public in various countries to wear cloth face coverings in public settings, particularly where social distancing may be harder to maintain, but these are not recommended for use by HCWs in healthcare settings.

Strategies to navigate the PPE shortage in the era of COVID-19 include importing, reclaiming, reusing, and repurposing PPE; generating and extending supply; eliminating nonessential services; reducing patient contact; and using nonhuman services such as drones to deliver equipment and undertake tasks such as decontamination.^9,10 Multidisciplinary teams are working on creative ways to use existing resources to make effective PPE, including alternatives to N95 respirators. An outbreak simulation study at Emory University in Atlanta, Georgia, and the University of Texas Health Science Center at Houston in Texas demonstrated that HCWs could be rapidly trained and fit tested to use elastomeric half-mask respirators, which are reusable.¹¹ A multidisciplinary team at Boston Children’s Hospital in Massachusetts has developed and completed a small pilot study of a reusable elastomeric respirator made using an anesthesia facemask, antimicrobial filter, and elastic straps.¹²

Given evidence that suggests that COVID-19 involves a component of airborne transmission,¹³ in addition to droplet spread and surface (fomite) contamination,¹⁴ using known infection prevention techniques that work to decrease airborne transmission of other respiratory infectious diseases should also be considered. Germicidal ultraviolet (GUV) air disinfection rapidly disinfects upper room air, which is then continually exchanged with contaminated lower-room air. GUV air disinfection has been demonstrated to be a safe and cost-effective intervention, with an efficacy of approximately 80% for decreasing TB transmission.¹⁵ GUV air disinfection is also effective against airborne influenza and measles and may play a role in surface decontamination by accelerating viral inactivation. Enabling GUV in high-risk areas such as the emergency department or intensive care unit could be a high-yield intervention to decrease transmission of COVID-19.

HCWs exposed to other respiratory infections such as influenza or TB may receive preventive therapy to reduce the risk of developing disease. High rates of COVID-19 in HCWs have prompted several initiatives to evaluate innovative approaches to decreasing this risk. Multiple studies are underway to determine whether hydroxychloroquine could be used for pre- or postexposure prophylaxis to prevent COVID-19. Another multisite trial will evaluate whether the BCG vaccine, primarily used to reduce the risk of TB, provides protection against COVID-19 in HCWs, driven by data suggesting a correlation between universal BCG vaccination policy and reduced morbidity and mortality for COVID-19.¹⁶

Infection prevention efforts can benefit from the unprecedented amount of data on COVID-19 that is being generated and shared. Successful examples of the rapid intensification of infection prevention measures to decrease transmission in healthcare facilities should be emulated. The hospital authority of Hong Kong implemented a bundle of measures focused on early recognition, isolation, notification, and molecular diagnostics for people being evaluated for COVID-19.¹⁷ They subsequently broadened the clinical and epidemiological criteria of surveillance as the outbreak evolved and intensified PPE recommendations to all HCWs (face masks for all and N95 respirators for those performing aerosol-generating procedures), which appears to have resulted in no cases of HCW infection or nosocomial transmission.

Data characterizing the extent of occupational infections in HCWs during acute and chronic epidemics is often lacking and subject to wide variability in reporting, which limits its impact. For example, HCWs in high TB incidence countries have at least twice the risk of developing TB, compared with the general population. Although there are still major gaps in national data collection regarding the incidence of occupational TB, recent attempts by WHO to systematically record this data have resulted in increasing prioritization of this group as an at-risk population who may benefit from TB preventive therapy. We strongly advocate that health systems systematically record and share longitudinal data on numbers of HCWs infected with COVID-19. This transparency will facilitate urgent action to replenish and sustain resources such as PPE and enable institutions to share and adapt successful infection prevention strategies. Examples such as the prevention of central line- associated bloodstream infections demonstrate the potential impact of national collaborative efforts to strengthen infection prevention, although further effort is needed to optimize knowledge sharing in the context of outbreaks.

The cost of not investing in public health pandemic preparedness including measures to protect HCWs is now widely apparent. HCWs have a right to safety in their workplaces as they fulfil their duty of care to patients.¹⁸ Rapid scale-up of diagnostic testing capacity, and bundles of infection prevention interventions including universal masking and drive-through testing, can safeguard HCWs and the patients they serve in the current COVID-19 pandemic. Re-establishing immediate access to quality-assured PPE is imperative to reduce the individual and workforce consequences of HCWs developing COVID-19 or other infectious diseases like TB that are continuously a threat to the workforce. Meanwhile, innovative approaches such as repurposing resources to develop PPE and use of GUV air disinfection may help to mitigate PPE shortages, and use of preventive therapies may also decrease COVID-19 risk in HCWs. Reliable surveillance data on HCW infection rates can help identify and track gaps in infection prevention, as well as identify strategies that impact this outcome. Ultimately greater top-down political commitment is urgently needed to ensure that frontline HCWs have the necessary resources to address the current pandemic and to sustain these interventions to protect HCWs in the future.

Hospitalists serve as frontline healthcare professionals caring for the increasing number of COVID-19 patients in the United States. The safety of hospitalists and other frontline healthcare workers is paramount to preventing high nosocomial transmission as has been reported in several other countries. Much effort to date has rightly focused on ensuring healthcare workers have appropriate personal protective equipment (PPE) given the known increased risk of nosocomial infection to healthcare workers. However, another important strategy to prevent nosocomial transmission is to implement “social distancing,” or avoiding close contact with others. While this approach has received considerable press with regards to implementation in communities, social, or physical, distancing in the hospital is also a critical way to prevent nosocomial transmission and ensure the health and welfare of our workforce to meet the challenge. The Centers for Disease Control and Prevention (CDC) defines close contact as less than 6 feet away for over 10 minutes.¹ Given the myriad clinical interactions that occur within teams in the hospital, such distancing can prove challenging.

At the University of Chicago Medicine in Illinois, our hospitalist group was an early adopter of implementing several strategies to facilitate physical distancing in the context of clinical care to minimize community transmission of COVID-19 among healthcare professionals. We describe how to implement physical distancing effectively in specific hospital settings, including some challenges and strategies to surmount them.

Educational conferences and administrative meetings need to be transitioned to virtual meetings. While it may be easy to broadcast a conference in lieu of meeting in a conference room, it is critical that hospital clinicians do not “huddle close together” in front of a computer, which would defeat the purpose of physical distancing. While “flipping the classroom” in preclinical and higher education is common, this method can be effective to deliver standard education followed by a virtual question and answer session or chat room.²

Educational discussions can also occur asynchronously through learning management systems, such as Canvas, or even closed social media channels, such as Slack, that enable discussions. These tools require training to work, so it is important to invest in education on the chosen platform to ensure that it functions smoothly. It is equally important that administrators become familiar with these tools while working remotely and can facilitate administrative meetings without difficulty. We created a one-page tip sheet to help ease the transition for department administrators. The tip sheet highlighted how to start a virtual meeting and meeting etiquette (eg, mute upon entry into the meeting, mute when not talking, announce yourself when talking) as well as ensuring that dial-ins could easily access the meeting by including one-touch options, when available, on calendar invites in addition to the weblink. A daily email update can be an important adjunct to administrative meetings to ensure critical updates are reaching all clinicians in a group and also preserves meeting time for clarifying questions.

Perhaps the biggest challenge is how many clinical workrooms in hospitals today are crowded with computers next to each other. Ventilation can also be poor, making conditions riskier. This makes implemention of social distancing extremely challenging, but also critical, given how much time hospital-based clinicians spend on computers and in their workrooms. The first step to achieving social distancing in the workroom is to take an inventory of how many people work there and get a log of the number of computers. Consider whether existing computers can be rearranged with a goal of keeping people 6 feet apart. For particularly cramped workrooms, this may require assigning computer spaces to physicians across a floor or several floors, using computers out on a unit, or using mobile computers to limit the number of people in the workroom at one time. We suggest working with physical plant leaders and Information Technology to reallocate mobile workstations, laptops, or desktops to conference rooms, patient visiting areas, and offices that are not being used. Because coronavirus can survive on surfaces for several hours, it is also important to stock work rooms with disinfectants to clean surfaces such as keyboards and desktops frequently. One other important thing to consider is whether computers can be assigned to specific teams or people to limit the use of a computer by multiple people.

Rounding

Perhaps one of the most fundamental hardships with physical distancing is how to conduct routine clinical care such as rounds, sign-out, or multidisciplinary rounds. Rounds on teaching services are particularly challenging given the number of people. At many teaching institutions, medical students are no longer on clinical rotations, which immediately reduces the number of people on teaching teams. The other thing to consider is how rounds are conducted. As opposed to a large team walking together, assign one person from the team as the liaison for the patient, which also has the added benefit of conserving precious PPE. Virtual rounding enables clinicians, including residents and attendings, to work together and decide the plan for the day without first crowding into a patient room. This is perhaps the most important cultural hurdle that one may face.

Another administrative hurdle and common concern is how to bill for such interactions. While federal guidance evolves, our institution created smartphrases for this type of virtual rounding whereby attendings attest to resident notes even if they did not physically see the patient. Additional information may be obtained from patients by calling them on their patient-room phones or by using telemedicine as some hospitals are implementing.³ For large “mega” teams, split the team into smaller groups to facilitate continuity and easier conversations.

Sign-out

When feasible, it is important to transition to phone sign-out supplemented with viewing an updated shared sign-out, ideally electronically, for shift change. When using phone sign-out, it is ideal to implement a verbal read-back to ensure understanding and to keep your sign-out updated. Because using the telephone is not the most effective communication channel for sign-out, it is key to be vigilant with other sign-out best practices, such as using a standard template like IPASS⁴ or another framework, prioritizing sick patients, and ensuring a focus on to-do and if/then items that are critical for the receiver to ensure understanding.⁵

Multidisciplinary Rounds

As multidisciplinary rounds typically occur either at the bedside or in a conference room, it is key to ensure that these occur virtually whenever possible. One option is to use conference calls or video chat (eg, Zoom) for multidisciplinary rounds whenever possible. Calendar invites or paging reminders can be used to prompt teams when to call in to discuss patients. Because multiple people are entering a virtual room at once, it is important to establish an order or have a leader orchestrate who is next. In addition, given the importance of multiple people contributing to the discussion, it is also equally important for those speaking always to announce who they are and their role (eg, social worker, case manager, physical therapist) since it may not be possible to recognize people’s voices alone. This is where visual recognition can be helpful through use of institutional video conferencing that enables hearing and seeing someone. Further, it is important to ensure that the platform being used is HIPAA compliant.

Call rooms in hospitals can be particularly challenging if they are shared. Finding additional call rooms may require use of cots or reallocation of patient rooms. It is also possible for hospitalists to consider air mattresses in their offices or other private spaces to avoid sharing call rooms. Consider assigning the same call room to the same few people over the course of a rotation or period to avoid many people sharing one room. If a hospital is converting units to group patients under investigation or those who are COVID-19 positive, reallocating call rooms may be necessary to accommodate new teams. Lastly, it is important to communicate proactively with environmental services staff to make sure all call rooms are equipped with cleaning supplies and hand sanitizer and are cleaned daily to avoid nosocomial transmission.

Hospitalists serve as frontline healthcare professionals caring for the increasing number of COVID-19 patients in the United States. The safety of hospitalists and other frontline healthcare workers is paramount to preventing high nosocomial transmission as has been reported in several other countries. Much effort to date has rightly focused on ensuring healthcare workers have appropriate personal protective equipment (PPE) given the known increased risk of nosocomial infection to healthcare workers. However, another important strategy to prevent nosocomial transmission is to implement “social distancing,” or avoiding close contact with others. While this approach has received considerable press with regards to implementation in communities, social, or physical, distancing in the hospital is also a critical way to prevent nosocomial transmission and ensure the health and welfare of our workforce to meet the challenge. The Centers for Disease Control and Prevention (CDC) defines close contact as less than 6 feet away for over 10 minutes.¹ Given the myriad clinical interactions that occur within teams in the hospital, such distancing can prove challenging.

At the University of Chicago Medicine in Illinois, our hospitalist group was an early adopter of implementing several strategies to facilitate physical distancing in the context of clinical care to minimize community transmission of COVID-19 among healthcare professionals. We describe how to implement physical distancing effectively in specific hospital settings, including some challenges and strategies to surmount them.

Educational conferences and administrative meetings need to be transitioned to virtual meetings. While it may be easy to broadcast a conference in lieu of meeting in a conference room, it is critical that hospital clinicians do not “huddle close together” in front of a computer, which would defeat the purpose of physical distancing. While “flipping the classroom” in preclinical and higher education is common, this method can be effective to deliver standard education followed by a virtual question and answer session or chat room.²

Educational discussions can also occur asynchronously through learning management systems, such as Canvas, or even closed social media channels, such as Slack, that enable discussions. These tools require training to work, so it is important to invest in education on the chosen platform to ensure that it functions smoothly. It is equally important that administrators become familiar with these tools while working remotely and can facilitate administrative meetings without difficulty. We created a one-page tip sheet to help ease the transition for department administrators. The tip sheet highlighted how to start a virtual meeting and meeting etiquette (eg, mute upon entry into the meeting, mute when not talking, announce yourself when talking) as well as ensuring that dial-ins could easily access the meeting by including one-touch options, when available, on calendar invites in addition to the weblink. A daily email update can be an important adjunct to administrative meetings to ensure critical updates are reaching all clinicians in a group and also preserves meeting time for clarifying questions.

Perhaps the biggest challenge is how many clinical workrooms in hospitals today are crowded with computers next to each other. Ventilation can also be poor, making conditions riskier. This makes implemention of social distancing extremely challenging, but also critical, given how much time hospital-based clinicians spend on computers and in their workrooms. The first step to achieving social distancing in the workroom is to take an inventory of how many people work there and get a log of the number of computers. Consider whether existing computers can be rearranged with a goal of keeping people 6 feet apart. For particularly cramped workrooms, this may require assigning computer spaces to physicians across a floor or several floors, using computers out on a unit, or using mobile computers to limit the number of people in the workroom at one time. We suggest working with physical plant leaders and Information Technology to reallocate mobile workstations, laptops, or desktops to conference rooms, patient visiting areas, and offices that are not being used. Because coronavirus can survive on surfaces for several hours, it is also important to stock work rooms with disinfectants to clean surfaces such as keyboards and desktops frequently. One other important thing to consider is whether computers can be assigned to specific teams or people to limit the use of a computer by multiple people.

Rounding

Perhaps one of the most fundamental hardships with physical distancing is how to conduct routine clinical care such as rounds, sign-out, or multidisciplinary rounds. Rounds on teaching services are particularly challenging given the number of people. At many teaching institutions, medical students are no longer on clinical rotations, which immediately reduces the number of people on teaching teams. The other thing to consider is how rounds are conducted. As opposed to a large team walking together, assign one person from the team as the liaison for the patient, which also has the added benefit of conserving precious PPE. Virtual rounding enables clinicians, including residents and attendings, to work together and decide the plan for the day without first crowding into a patient room. This is perhaps the most important cultural hurdle that one may face.

Another administrative hurdle and common concern is how to bill for such interactions. While federal guidance evolves, our institution created smartphrases for this type of virtual rounding whereby attendings attest to resident notes even if they did not physically see the patient. Additional information may be obtained from patients by calling them on their patient-room phones or by using telemedicine as some hospitals are implementing.³ For large “mega” teams, split the team into smaller groups to facilitate continuity and easier conversations.

Sign-out

When feasible, it is important to transition to phone sign-out supplemented with viewing an updated shared sign-out, ideally electronically, for shift change. When using phone sign-out, it is ideal to implement a verbal read-back to ensure understanding and to keep your sign-out updated. Because using the telephone is not the most effective communication channel for sign-out, it is key to be vigilant with other sign-out best practices, such as using a standard template like IPASS⁴ or another framework, prioritizing sick patients, and ensuring a focus on to-do and if/then items that are critical for the receiver to ensure understanding.⁵

Multidisciplinary Rounds

As multidisciplinary rounds typically occur either at the bedside or in a conference room, it is key to ensure that these occur virtually whenever possible. One option is to use conference calls or video chat (eg, Zoom) for multidisciplinary rounds whenever possible. Calendar invites or paging reminders can be used to prompt teams when to call in to discuss patients. Because multiple people are entering a virtual room at once, it is important to establish an order or have a leader orchestrate who is next. In addition, given the importance of multiple people contributing to the discussion, it is also equally important for those speaking always to announce who they are and their role (eg, social worker, case manager, physical therapist) since it may not be possible to recognize people’s voices alone. This is where visual recognition can be helpful through use of institutional video conferencing that enables hearing and seeing someone. Further, it is important to ensure that the platform being used is HIPAA compliant.

Call rooms in hospitals can be particularly challenging if they are shared. Finding additional call rooms may require use of cots or reallocation of patient rooms. It is also possible for hospitalists to consider air mattresses in their offices or other private spaces to avoid sharing call rooms. Consider assigning the same call room to the same few people over the course of a rotation or period to avoid many people sharing one room. If a hospital is converting units to group patients under investigation or those who are COVID-19 positive, reallocating call rooms may be necessary to accommodate new teams. Lastly, it is important to communicate proactively with environmental services staff to make sure all call rooms are equipped with cleaning supplies and hand sanitizer and are cleaned daily to avoid nosocomial transmission.

Hospitalists serve as frontline healthcare professionals caring for the increasing number of COVID-19 patients in the United States. The safety of hospitalists and other frontline healthcare workers is paramount to preventing high nosocomial transmission as has been reported in several other countries. Much effort to date has rightly focused on ensuring healthcare workers have appropriate personal protective equipment (PPE) given the known increased risk of nosocomial infection to healthcare workers. However, another important strategy to prevent nosocomial transmission is to implement “social distancing,” or avoiding close contact with others. While this approach has received considerable press with regards to implementation in communities, social, or physical, distancing in the hospital is also a critical way to prevent nosocomial transmission and ensure the health and welfare of our workforce to meet the challenge. The Centers for Disease Control and Prevention (CDC) defines close contact as less than 6 feet away for over 10 minutes.¹ Given the myriad clinical interactions that occur within teams in the hospital, such distancing can prove challenging.

At the University of Chicago Medicine in Illinois, our hospitalist group was an early adopter of implementing several strategies to facilitate physical distancing in the context of clinical care to minimize community transmission of COVID-19 among healthcare professionals. We describe how to implement physical distancing effectively in specific hospital settings, including some challenges and strategies to surmount them.

Educational conferences and administrative meetings need to be transitioned to virtual meetings. While it may be easy to broadcast a conference in lieu of meeting in a conference room, it is critical that hospital clinicians do not “huddle close together” in front of a computer, which would defeat the purpose of physical distancing. While “flipping the classroom” in preclinical and higher education is common, this method can be effective to deliver standard education followed by a virtual question and answer session or chat room.²

Educational discussions can also occur asynchronously through learning management systems, such as Canvas, or even closed social media channels, such as Slack, that enable discussions. These tools require training to work, so it is important to invest in education on the chosen platform to ensure that it functions smoothly. It is equally important that administrators become familiar with these tools while working remotely and can facilitate administrative meetings without difficulty. We created a one-page tip sheet to help ease the transition for department administrators. The tip sheet highlighted how to start a virtual meeting and meeting etiquette (eg, mute upon entry into the meeting, mute when not talking, announce yourself when talking) as well as ensuring that dial-ins could easily access the meeting by including one-touch options, when available, on calendar invites in addition to the weblink. A daily email update can be an important adjunct to administrative meetings to ensure critical updates are reaching all clinicians in a group and also preserves meeting time for clarifying questions.

Perhaps the biggest challenge is how many clinical workrooms in hospitals today are crowded with computers next to each other. Ventilation can also be poor, making conditions riskier. This makes implemention of social distancing extremely challenging, but also critical, given how much time hospital-based clinicians spend on computers and in their workrooms. The first step to achieving social distancing in the workroom is to take an inventory of how many people work there and get a log of the number of computers. Consider whether existing computers can be rearranged with a goal of keeping people 6 feet apart. For particularly cramped workrooms, this may require assigning computer spaces to physicians across a floor or several floors, using computers out on a unit, or using mobile computers to limit the number of people in the workroom at one time. We suggest working with physical plant leaders and Information Technology to reallocate mobile workstations, laptops, or desktops to conference rooms, patient visiting areas, and offices that are not being used. Because coronavirus can survive on surfaces for several hours, it is also important to stock work rooms with disinfectants to clean surfaces such as keyboards and desktops frequently. One other important thing to consider is whether computers can be assigned to specific teams or people to limit the use of a computer by multiple people.

Rounding

Perhaps one of the most fundamental hardships with physical distancing is how to conduct routine clinical care such as rounds, sign-out, or multidisciplinary rounds. Rounds on teaching services are particularly challenging given the number of people. At many teaching institutions, medical students are no longer on clinical rotations, which immediately reduces the number of people on teaching teams. The other thing to consider is how rounds are conducted. As opposed to a large team walking together, assign one person from the team as the liaison for the patient, which also has the added benefit of conserving precious PPE. Virtual rounding enables clinicians, including residents and attendings, to work together and decide the plan for the day without first crowding into a patient room. This is perhaps the most important cultural hurdle that one may face.

Another administrative hurdle and common concern is how to bill for such interactions. While federal guidance evolves, our institution created smartphrases for this type of virtual rounding whereby attendings attest to resident notes even if they did not physically see the patient. Additional information may be obtained from patients by calling them on their patient-room phones or by using telemedicine as some hospitals are implementing.³ For large “mega” teams, split the team into smaller groups to facilitate continuity and easier conversations.

Sign-out

When feasible, it is important to transition to phone sign-out supplemented with viewing an updated shared sign-out, ideally electronically, for shift change. When using phone sign-out, it is ideal to implement a verbal read-back to ensure understanding and to keep your sign-out updated. Because using the telephone is not the most effective communication channel for sign-out, it is key to be vigilant with other sign-out best practices, such as using a standard template like IPASS⁴ or another framework, prioritizing sick patients, and ensuring a focus on to-do and if/then items that are critical for the receiver to ensure understanding.⁵

Multidisciplinary Rounds

As multidisciplinary rounds typically occur either at the bedside or in a conference room, it is key to ensure that these occur virtually whenever possible. One option is to use conference calls or video chat (eg, Zoom) for multidisciplinary rounds whenever possible. Calendar invites or paging reminders can be used to prompt teams when to call in to discuss patients. Because multiple people are entering a virtual room at once, it is important to establish an order or have a leader orchestrate who is next. In addition, given the importance of multiple people contributing to the discussion, it is also equally important for those speaking always to announce who they are and their role (eg, social worker, case manager, physical therapist) since it may not be possible to recognize people’s voices alone. This is where visual recognition can be helpful through use of institutional video conferencing that enables hearing and seeing someone. Further, it is important to ensure that the platform being used is HIPAA compliant.

Call rooms in hospitals can be particularly challenging if they are shared. Finding additional call rooms may require use of cots or reallocation of patient rooms. It is also possible for hospitalists to consider air mattresses in their offices or other private spaces to avoid sharing call rooms. Consider assigning the same call room to the same few people over the course of a rotation or period to avoid many people sharing one room. If a hospital is converting units to group patients under investigation or those who are COVID-19 positive, reallocating call rooms may be necessary to accommodate new teams. Lastly, it is important to communicate proactively with environmental services staff to make sure all call rooms are equipped with cleaning supplies and hand sanitizer and are cleaned daily to avoid nosocomial transmission.

“To journey for the sake of saving our own lives is little by little to cease to live in any sense that really matters, even to ourselves, because it is only by journeying for the world’s sake—even when the world bores and sickens and scares you half to death—that little by little we start to come alive.”

—Frederick Buechner

On February 29, 2020, I find out by text from my intern when the first patient at our hospital in Seattle tests positive for COVID-19. He learns of it from his fellow intern who is caring for the patient. The news quickly spreads through the hospital like the virus itself, going from person to person while official communication channels remain initially silent. The news comes on the heels of a friend’s text that her daughter’s high school is closing for disinfection after a classmate also tested positive for COVID-19. I know the cataclysmic significance of these two events: Public health efforts to contain the SARS-CoV-2 coronavirus have failed, and there is ongoing community spread of the infection in Washington state. I text my intern back with the emoji of The Scream by Edvard Munch.

Could I be asymptomatically infected with the coronavirus? I work in close quarters with my colleagues who cared for the COVID-19–positive patient before he was placed in infection precautions. Social distancing has yet to enter our lexicon and our lives. In our crowded office, shared surfaces abound. Suddenly, every hard surface seems suspect—chairs, phones, dictaphone handsets, code pagers, printers, keypads, and door handles. All can be vectors of viral transmission. Normally insouciant about cleanliness, my coworkers and I start swabbing down every surface with disinfectant wipes. I ponder my likelihood of infection and decide it is possible but not probable.

In the next few days, I have a trip to Sedona, Arizona, planned with my extended family. Originally conceived as a celebration for my mom’s 80th birthday, we repurposed it as a time to grieve together after she unexpectedly passed away. I debate back and forth whether to go on the trip. If there is a chance I am infected with the coronavirus, it feels irresponsible to board an airplane with hundreds of other people. Yet the trip carries such high value for me. My family holds out hope I can get tested for the coronavirus, but I know just how limited testing capability is. It cuts me to the heart, but I cancel my flight. The deciding factor is that my sister has an autoimmune disease and is immunosuppressed. I don’t want to jeopardize her health. The world has truly gone topsy-turvy when the greatest thoughtfulness you can show to someone you love is to stay the hell away from her.

After my work stretch, I hunker down at home to monitor myself. I have a mild sore throat but convince myself it is psychosomatic. My plausible deniability of illness dies when I develop a cough and fatigue. Based on my symptoms, it is impossible to tell if I have the coronavirus or a common cold. I place myself on home quarantine. I don’t pursue coronavirus testing because there are hospitalized patients who need it much more than I do. I diligently monitor my temperature twice daily and it remains normal. My sore throat and fatigue go away, but my cough and some mild shortness of breath persists. I attribute it to my asthma, but the possibility of COVID-19 always lurks in the back of my mind. COVID-19 patients often don’t worsen until their second week of infection. Ordinarily, I would start using my steroid inhaler, but I hold off since steroids are thought to prolong viral replication. 

When I tire of staying in the house, I go outside to work in the yard. I get on a low ladder to pull down the English ivy climbing up and smothering a tree. The ivy strand I’m tugging on suddenly breaks and I fall hard onto my back. Like a slap in the face, the accident shocks me into a new state of mental clarity. As a hospitalist, I’m a precious resource to my community right now. I can’t knock myself out of commission for dumb reasons. I ban myself from climbing any more ladders. 

As my time in quarantine draws to a close, I put my legal and financial affairs in order and pack a just-in-case backpack. The emergency room doctor hospitalized at a nearby hospital with severe COVID-19 is about my age. I am still coughing so I check in with the head of Infection Control to see if I need to be tested before returning to work. He tells me no. As I start working, I realize that coughing is the new leprosy. Even though I wear a mask, I get tense looks from others who carefully keep their distance from me. I tell everyone I have cough variant asthma, but what they all want to know is if I have been tested for the coronavirus. I haven’t been.

When my hospital sets up a new dedicated Employee Health screening phone line, I call right away. The nurse tells me I don’t meet criteria for coronavirus testing even though I am working on the COVID-19 rule-out unit with patients who have tested positive. While I agree with her from a medical standpoint, I don’t from a social or psychological perspective. This is a particularly unpropitious time in history to be a Chinese American doctor who can’t stop coughing. A negative test will reassure my patients and coworkers I am not a risk to them. A positive test, which is a possibility because of known prolonged viral shedding of the coronavirus, will reassure me I’m likely on my way to developing serologic immunity. I don’t get a test. When I tell my colleague, he suggests I resort to lying, but I won’t do it. As I’ve watched how power, wealth, and privilege play out in access to testing, I refuse to manipulate the system. But my experience is pointed commentary on the abysmal failure of testing in the United States when a frontline symptomatic doctor taking care of COVID-19 patients in one of the epicenters of the pandemic can’t get a coronavirus test. During a meeting, the head of Infection Control bluntly states he hopes we know that all of us are going to get the coronavirus at some point, but hopefully it won’t take us out of commission all at once. I feel better hearing him acknowledge that because it confirms my own sense of reality. 

While the ongoing pandemic definitely increases stress and anxiety levels in the hospital, there also continues to be caring and kindness. As I don my personal protective equipment (PPE), a nurse notices an exposed gap in the back of my gown and fixes it for me. He has my back, literally and figuratively. Our unit clerk, hearing my persistent cough, braves the 6-foot danger zone to hand me cough drops. Another nurse asks when the last time was I drank anything, and I give her a blank look because I can’t remember. I’m trying to minimize my use of masks, so I have kept my current one on all day. The front of the mask may be contaminated, but as long as I don’t touch that surface, it is still protecting me. She hands me a cup of water and I consider the benefit of staying hydrated versus using up another mask. Having previously landed in the emergency room with a kidney stone from not properly hydrating, I take off the mask, throw it away, wash my hands, and drink the water. But in my mind, I’m acutely aware of our shrinking supply of PPE. 

On our time off work, my coworkers and I reach out to everyone we know to ask for mask donations. One friend tells me her husband is fashioning a mask for her from their furnace filter. She considers it the most romantic thing he’s ever done for her. The two of us agree that if we run out of PPE, we will go on caring for our patients anyway. We are doctors and caring for others is not only what we do, but an intrinsic part of who we are. Our journey amid the coronavirus pandemic may at times scare us half to death, but in caring for others “little by little we start to come alive.”

“To journey for the sake of saving our own lives is little by little to cease to live in any sense that really matters, even to ourselves, because it is only by journeying for the world’s sake—even when the world bores and sickens and scares you half to death—that little by little we start to come alive.”

—Frederick Buechner

On February 29, 2020, I find out by text from my intern when the first patient at our hospital in Seattle tests positive for COVID-19. He learns of it from his fellow intern who is caring for the patient. The news quickly spreads through the hospital like the virus itself, going from person to person while official communication channels remain initially silent. The news comes on the heels of a friend’s text that her daughter’s high school is closing for disinfection after a classmate also tested positive for COVID-19. I know the cataclysmic significance of these two events: Public health efforts to contain the SARS-CoV-2 coronavirus have failed, and there is ongoing community spread of the infection in Washington state. I text my intern back with the emoji of The Scream by Edvard Munch.

Could I be asymptomatically infected with the coronavirus? I work in close quarters with my colleagues who cared for the COVID-19–positive patient before he was placed in infection precautions. Social distancing has yet to enter our lexicon and our lives. In our crowded office, shared surfaces abound. Suddenly, every hard surface seems suspect—chairs, phones, dictaphone handsets, code pagers, printers, keypads, and door handles. All can be vectors of viral transmission. Normally insouciant about cleanliness, my coworkers and I start swabbing down every surface with disinfectant wipes. I ponder my likelihood of infection and decide it is possible but not probable.

In the next few days, I have a trip to Sedona, Arizona, planned with my extended family. Originally conceived as a celebration for my mom’s 80th birthday, we repurposed it as a time to grieve together after she unexpectedly passed away. I debate back and forth whether to go on the trip. If there is a chance I am infected with the coronavirus, it feels irresponsible to board an airplane with hundreds of other people. Yet the trip carries such high value for me. My family holds out hope I can get tested for the coronavirus, but I know just how limited testing capability is. It cuts me to the heart, but I cancel my flight. The deciding factor is that my sister has an autoimmune disease and is immunosuppressed. I don’t want to jeopardize her health. The world has truly gone topsy-turvy when the greatest thoughtfulness you can show to someone you love is to stay the hell away from her.

After my work stretch, I hunker down at home to monitor myself. I have a mild sore throat but convince myself it is psychosomatic. My plausible deniability of illness dies when I develop a cough and fatigue. Based on my symptoms, it is impossible to tell if I have the coronavirus or a common cold. I place myself on home quarantine. I don’t pursue coronavirus testing because there are hospitalized patients who need it much more than I do. I diligently monitor my temperature twice daily and it remains normal. My sore throat and fatigue go away, but my cough and some mild shortness of breath persists. I attribute it to my asthma, but the possibility of COVID-19 always lurks in the back of my mind. COVID-19 patients often don’t worsen until their second week of infection. Ordinarily, I would start using my steroid inhaler, but I hold off since steroids are thought to prolong viral replication. 

When I tire of staying in the house, I go outside to work in the yard. I get on a low ladder to pull down the English ivy climbing up and smothering a tree. The ivy strand I’m tugging on suddenly breaks and I fall hard onto my back. Like a slap in the face, the accident shocks me into a new state of mental clarity. As a hospitalist, I’m a precious resource to my community right now. I can’t knock myself out of commission for dumb reasons. I ban myself from climbing any more ladders. 

As my time in quarantine draws to a close, I put my legal and financial affairs in order and pack a just-in-case backpack. The emergency room doctor hospitalized at a nearby hospital with severe COVID-19 is about my age. I am still coughing so I check in with the head of Infection Control to see if I need to be tested before returning to work. He tells me no. As I start working, I realize that coughing is the new leprosy. Even though I wear a mask, I get tense looks from others who carefully keep their distance from me. I tell everyone I have cough variant asthma, but what they all want to know is if I have been tested for the coronavirus. I haven’t been.

When my hospital sets up a new dedicated Employee Health screening phone line, I call right away. The nurse tells me I don’t meet criteria for coronavirus testing even though I am working on the COVID-19 rule-out unit with patients who have tested positive. While I agree with her from a medical standpoint, I don’t from a social or psychological perspective. This is a particularly unpropitious time in history to be a Chinese American doctor who can’t stop coughing. A negative test will reassure my patients and coworkers I am not a risk to them. A positive test, which is a possibility because of known prolonged viral shedding of the coronavirus, will reassure me I’m likely on my way to developing serologic immunity. I don’t get a test. When I tell my colleague, he suggests I resort to lying, but I won’t do it. As I’ve watched how power, wealth, and privilege play out in access to testing, I refuse to manipulate the system. But my experience is pointed commentary on the abysmal failure of testing in the United States when a frontline symptomatic doctor taking care of COVID-19 patients in one of the epicenters of the pandemic can’t get a coronavirus test. During a meeting, the head of Infection Control bluntly states he hopes we know that all of us are going to get the coronavirus at some point, but hopefully it won’t take us out of commission all at once. I feel better hearing him acknowledge that because it confirms my own sense of reality. 

While the ongoing pandemic definitely increases stress and anxiety levels in the hospital, there also continues to be caring and kindness. As I don my personal protective equipment (PPE), a nurse notices an exposed gap in the back of my gown and fixes it for me. He has my back, literally and figuratively. Our unit clerk, hearing my persistent cough, braves the 6-foot danger zone to hand me cough drops. Another nurse asks when the last time was I drank anything, and I give her a blank look because I can’t remember. I’m trying to minimize my use of masks, so I have kept my current one on all day. The front of the mask may be contaminated, but as long as I don’t touch that surface, it is still protecting me. She hands me a cup of water and I consider the benefit of staying hydrated versus using up another mask. Having previously landed in the emergency room with a kidney stone from not properly hydrating, I take off the mask, throw it away, wash my hands, and drink the water. But in my mind, I’m acutely aware of our shrinking supply of PPE. 

On our time off work, my coworkers and I reach out to everyone we know to ask for mask donations. One friend tells me her husband is fashioning a mask for her from their furnace filter. She considers it the most romantic thing he’s ever done for her. The two of us agree that if we run out of PPE, we will go on caring for our patients anyway. We are doctors and caring for others is not only what we do, but an intrinsic part of who we are. Our journey amid the coronavirus pandemic may at times scare us half to death, but in caring for others “little by little we start to come alive.”

“To journey for the sake of saving our own lives is little by little to cease to live in any sense that really matters, even to ourselves, because it is only by journeying for the world’s sake—even when the world bores and sickens and scares you half to death—that little by little we start to come alive.”

—Frederick Buechner

On February 29, 2020, I find out by text from my intern when the first patient at our hospital in Seattle tests positive for COVID-19. He learns of it from his fellow intern who is caring for the patient. The news quickly spreads through the hospital like the virus itself, going from person to person while official communication channels remain initially silent. The news comes on the heels of a friend’s text that her daughter’s high school is closing for disinfection after a classmate also tested positive for COVID-19. I know the cataclysmic significance of these two events: Public health efforts to contain the SARS-CoV-2 coronavirus have failed, and there is ongoing community spread of the infection in Washington state. I text my intern back with the emoji of The Scream by Edvard Munch.

Could I be asymptomatically infected with the coronavirus? I work in close quarters with my colleagues who cared for the COVID-19–positive patient before he was placed in infection precautions. Social distancing has yet to enter our lexicon and our lives. In our crowded office, shared surfaces abound. Suddenly, every hard surface seems suspect—chairs, phones, dictaphone handsets, code pagers, printers, keypads, and door handles. All can be vectors of viral transmission. Normally insouciant about cleanliness, my coworkers and I start swabbing down every surface with disinfectant wipes. I ponder my likelihood of infection and decide it is possible but not probable.

In the next few days, I have a trip to Sedona, Arizona, planned with my extended family. Originally conceived as a celebration for my mom’s 80th birthday, we repurposed it as a time to grieve together after she unexpectedly passed away. I debate back and forth whether to go on the trip. If there is a chance I am infected with the coronavirus, it feels irresponsible to board an airplane with hundreds of other people. Yet the trip carries such high value for me. My family holds out hope I can get tested for the coronavirus, but I know just how limited testing capability is. It cuts me to the heart, but I cancel my flight. The deciding factor is that my sister has an autoimmune disease and is immunosuppressed. I don’t want to jeopardize her health. The world has truly gone topsy-turvy when the greatest thoughtfulness you can show to someone you love is to stay the hell away from her.

After my work stretch, I hunker down at home to monitor myself. I have a mild sore throat but convince myself it is psychosomatic. My plausible deniability of illness dies when I develop a cough and fatigue. Based on my symptoms, it is impossible to tell if I have the coronavirus or a common cold. I place myself on home quarantine. I don’t pursue coronavirus testing because there are hospitalized patients who need it much more than I do. I diligently monitor my temperature twice daily and it remains normal. My sore throat and fatigue go away, but my cough and some mild shortness of breath persists. I attribute it to my asthma, but the possibility of COVID-19 always lurks in the back of my mind. COVID-19 patients often don’t worsen until their second week of infection. Ordinarily, I would start using my steroid inhaler, but I hold off since steroids are thought to prolong viral replication. 

When I tire of staying in the house, I go outside to work in the yard. I get on a low ladder to pull down the English ivy climbing up and smothering a tree. The ivy strand I’m tugging on suddenly breaks and I fall hard onto my back. Like a slap in the face, the accident shocks me into a new state of mental clarity. As a hospitalist, I’m a precious resource to my community right now. I can’t knock myself out of commission for dumb reasons. I ban myself from climbing any more ladders. 

As my time in quarantine draws to a close, I put my legal and financial affairs in order and pack a just-in-case backpack. The emergency room doctor hospitalized at a nearby hospital with severe COVID-19 is about my age. I am still coughing so I check in with the head of Infection Control to see if I need to be tested before returning to work. He tells me no. As I start working, I realize that coughing is the new leprosy. Even though I wear a mask, I get tense looks from others who carefully keep their distance from me. I tell everyone I have cough variant asthma, but what they all want to know is if I have been tested for the coronavirus. I haven’t been.

When my hospital sets up a new dedicated Employee Health screening phone line, I call right away. The nurse tells me I don’t meet criteria for coronavirus testing even though I am working on the COVID-19 rule-out unit with patients who have tested positive. While I agree with her from a medical standpoint, I don’t from a social or psychological perspective. This is a particularly unpropitious time in history to be a Chinese American doctor who can’t stop coughing. A negative test will reassure my patients and coworkers I am not a risk to them. A positive test, which is a possibility because of known prolonged viral shedding of the coronavirus, will reassure me I’m likely on my way to developing serologic immunity. I don’t get a test. When I tell my colleague, he suggests I resort to lying, but I won’t do it. As I’ve watched how power, wealth, and privilege play out in access to testing, I refuse to manipulate the system. But my experience is pointed commentary on the abysmal failure of testing in the United States when a frontline symptomatic doctor taking care of COVID-19 patients in one of the epicenters of the pandemic can’t get a coronavirus test. During a meeting, the head of Infection Control bluntly states he hopes we know that all of us are going to get the coronavirus at some point, but hopefully it won’t take us out of commission all at once. I feel better hearing him acknowledge that because it confirms my own sense of reality. 

While the ongoing pandemic definitely increases stress and anxiety levels in the hospital, there also continues to be caring and kindness. As I don my personal protective equipment (PPE), a nurse notices an exposed gap in the back of my gown and fixes it for me. He has my back, literally and figuratively. Our unit clerk, hearing my persistent cough, braves the 6-foot danger zone to hand me cough drops. Another nurse asks when the last time was I drank anything, and I give her a blank look because I can’t remember. I’m trying to minimize my use of masks, so I have kept my current one on all day. The front of the mask may be contaminated, but as long as I don’t touch that surface, it is still protecting me. She hands me a cup of water and I consider the benefit of staying hydrated versus using up another mask. Having previously landed in the emergency room with a kidney stone from not properly hydrating, I take off the mask, throw it away, wash my hands, and drink the water. But in my mind, I’m acutely aware of our shrinking supply of PPE. 

On our time off work, my coworkers and I reach out to everyone we know to ask for mask donations. One friend tells me her husband is fashioning a mask for her from their furnace filter. She considers it the most romantic thing he’s ever done for her. The two of us agree that if we run out of PPE, we will go on caring for our patients anyway. We are doctors and caring for others is not only what we do, but an intrinsic part of who we are. Our journey amid the coronavirus pandemic may at times scare us half to death, but in caring for others “little by little we start to come alive.”