Original Research

Length of stay (LOS) is a significant driver of costs after hip fracture surgery.^1-3 Multiple studies have identified factors associated with increased LOS in hip fracture patients. These factors include admission time, delay to surgery, presence of comorbidities, and older age.^4-9

One significant and potentially modifiable factor affecting LOS is delayed transfer to a rehabilitation center after surgery.^8-11 Although patients after orthopedic surgeries require additional rehabilitation services or subacute care directly attributable to their injuries, specialized rehabilitation centers may not always have beds readily available.^6-11 Studies have shown that delays in transfer to skilled nursing facilities or rehabilitation centers are highly common among orthopedic patients.⁸ It is therefore imperative that orthopedists have a mechanism for predicting and identifying which patients require rehabilitation services early in the postoperative period. Identifying risk factors and stratifying patients who are most likely to require rehabilitation would facilitate the early transfer of these patients and thereby directly decrease LOS and hospitalization-related costs.

In this article, we report results from prospective, national, multicenter data to identify commonly measured risk factors for discharge to rehabilitation facilities for hip fracture patients. Through multivariate analysis of ACS-NSQIP (American College of Surgeons National Surgical Quality Improvement Program) data, we determined which risk factors significantly predispose patients to discharge to rehabilitation centers versus discharge home. Knowledge of these risk factors allows the practicing orthopedist to be better equipped to identify patients who require additional rehabilitation early in the postoperative course. By mobilizing case managers and social workers to help avoid delays in the transfers of these identified patients, LOS-associated costs may ultimately decrease.

Materials and Methods

After obtaining institutional review board approval for this study from the Office of Research at Vanderbilt University, we prospectively collected 2011 discharge data from the ACS-NSQIP database (these data are unavailable for earlier years). All patients who underwent hip fracture surgery in 2011 were identified by CPT (Current Procedural Terminology) codes. Cases of patients with unknown discharge information and of those who died during their hospitalizations were excluded from analysis. For the remaining patients, discharge information as categorized by ACS-NSQIP included skilled care (eg, subacute hospital, skilled nursing home), unskilled facility (eg, nursing home, assisted facility), separate acute care, and rehabilitation. All other patients were discharged home without additional assistance or to the previous home where they received chronic care, assisted living, or unskilled aid. Patients were dichotomized according to whether they were discharged home or to one of the rehabilitation facilities mentioned.

To determine which risk factors significantly contributed to a patient’s discharge to rehabilitation, we ran univariate analyses using Fisher exact tests for categorical variables and Student t tests for continuous variables on multiple patient factors, including demographics, preoperative comorbidities, and operative factors. Demographics included age and sex. Preoperative comorbidities included 32 conditions: diabetes mellitus, active smoking status, current alcohol use, dyspnea, history of chronic obstructive pulmonary disease, history of congestive heart failure, hypertension requiring medication, history of esophageal varices, history of myocardial infarction, current renal failure, current dialysis dependence, steroid use, recent weight loss, existing bleeding disorder, transfusion before discharge, presence of central nervous system tumor, recent chemotherapy, recent radiation therapy, previous percutaneous coronary intervention, previous percutaneous coronary stenting, history of angina, peripheral vascular disease, cerebrovascular accidents, recent surgery (within 30 days), rest pain, impaired sensorium, history of transient ischemic attacks, current hemiplegia status, current paraplegia status, current quadriplegia status, current ascites, hypertension, and disseminated cancer. Operative factors included wound infection, DNR (do not resuscitate) status, ventilator support, anesthesia type, wound class, ASA (American Society of Anesthesiologists) class, and operative time.

For the univariate analyses, significance was set at P < .05. Demographics, preoperative comorbidities, and operative factors that were significantly associated with discharge to a rehabilitation facility in the univariate analysis were selected as covariates for a multivariate analysis. We incorporated a binary logistic regression to analyze which of these significant risk factors are correlated with a patient’s discharge to a rehabilitation facility after hip fracture surgery.

Results

A total of 4974 patients undergoing surgery for hip fractures in 2011 were identified. Of these patients, 4815 had complete information on discharge location and were included in the analysis.

Table 1 lists the results of the univariate analysis comparing demographics, preoperative comorbidities, and operative factors between the home and rehabilitation groups. Both age (P < .001) and sex (P = .012) were significantly different between groups; the rehabilitation group was older by about 10 years and included significantly more females. In addition to demographic factors, 16 preoperative comorbidities, and 5 surgical factors were significantly associated with discharge to rehabilitation.

Surgery type significantly affected discharge to rehabilitation (Figure). Patients who were undergoing open plating of a femoral neck fracture or intramedullary nailing of an intertrochanteric, peritrochanteric, or subtrochanteric femoral fracture constituted 30% of all patients discharged to rehabilitation centers. In contrast, patients undergoing percutaneous skeletal fixation of a proximal femoral fracture constituted only 5.5% of all patients discharged to rehabilitation. Based on surgery type, we broke down discharge location further, into categories of skilled nursing facility, unskilled facility (not patient’s previous home), separate acute-care facility, dedicated rehabilitation center, and home. Of all 4815 patients combined, 2102 (43.6%) were discharged to a skilled nursing facility, 31 (0.6%) to an unskilled facility (not home), 106 (2.2%) to separate acute care, 1312 (27.2%) to a dedicated rehabilitation center, and 950 (19.7%) home.

Table 2 lists the significant results from the multivariate logistical analysis comparing discharge to a rehabilitation center and discharge home after controlling for the significant risk factors (Table 1). Current diabetes, history of dyspnea, previous myocardial infarction, history of ischemic attacks, current bleeding disorder, transfusion during hospitalization, previous percutaneous cardiac stenting, chemotherapy, past cerebrovascular accident, presence of cancer, surgery type based on CPT code, history of chronic obstructive pulmonary disease or congestive heart failure, current smoking status, and operative time longer than 90 minutes were not significantly correlated with discharge to rehabilitation in the multivariate analysis. All significant factors were associated with higher odds of discharge to rehabilitation except for DNR status. DNR patients were 2.04 times more likely (95% CI, 1.49-2.78; P < .001) to be discharged home than to rehabilitation centers.

Applying these adjusted odds ratios, we see that an elderly woman (age, >65 years) who underwent general anesthesia with an ASA class higher than 2 was 17.63 times more likely than a patient without these risk factors to be discharged to rehabilitation. If this patient were also dialysis-dependent, she would be 61.52 times more likely than a similar patient without dialysis needs to be discharged to rehabilitation.

Even when controlling for all significant and nonsignificant variables in multivariate logistical analysis, age over 65 years (β = 1.05; P < .001), female sex (β = 1.76; P = .004), dialysis dependence (β = 12.98; P = .036), hypertension requiring medication (β = 1.53; P = .032), and ASA class higher than 2 (β = 1.98; P = .001) were found to be significant risk factors for discharge to rehabilitation.

Discussion

This study was the first to investigate the issue of which patient risk factors allow the practicing orthopedist to identify patients who require rehabilitation after hip fracture surgery. Through our multivariate analysis, which controlled for demographics, comorbidities, and operative factors, we found that older age, female sex, history of percutaneous coronary intervention, dialysis dependence, general anesthesia, and ASA class higher than 2 significantly increased the odds of discharge to a rehabilitation center versus home.

Using our study’s results, we can create a risk stratification model for patients and thereby a means of targeting patients who need rehabilitation and starting the process of finding a rehabilitation bed early in the postoperative course. Our study’s variables are easily measured metrics that may be collected in any hospital setting. Especially for hip fracture patients, early planning and discharge to the appropriate rehabilitation center are important in decreasing LOS and associated hospitalization costs. According to one report,³ about 85% of all hip fracture costs are directly related to LOS, given the unnecessarily long rehabilitation periods in hospitals. Hollingworth and colleagues² compared costs for patients who remained in the hospital with costs for those discharged with rehabilitation services. Overall costs were significantly lower for patients discharged home with rehabilitation. The authors concluded that 40% of hip fracture patients may be suitable for early discharge.² In an analysis of Medicare payments for hip fracture treatment, hospital costs including LOS accounted for 60% of all payments.¹² The results of these 2 studies suggest that the overall driver of hip fracture costs is prolonged LOS and that, if patients are discharged to rehabilitation, then overall costs may be lowered through a direct reduction in hospital LOS. Given that hip fractures account for almost 350,000 hospital admissions in the United States each year, and using our institution’s average hospital charge per day ($4500), about $1.6 billion may be saved if each patient’s LOS decreased by 1 day.¹³ Although multiple factors affect LOS, discharge planning is under orthopedists’ direct control. Therefore, early identification of patients who will require rehabilitation may help reduce LOS-associated costs in our health care system.

The patient variables that were significantly associated with discharge to rehabilitation are also associated with increased morbidity and mortality in hip fracture patients, according to the literature,^14-20 which provides some external validation of using these risk factors as predictors for rehabilitation. A patient with one of these risk factors may require rehabilitation, given that rehabilitation services are specifically linked to lower morbidity and mortality rates among hip fracture patients. For example, patients with dialysis needs were 3.49 times more likely to be discharged to a rehabilitation center in our study. In a 2000 study by Coco and Rush,¹⁶ hip fracture patients on dialysis had a 1-year mortality rate 2.5 times higher than that of patients who were not dialysis-dependent. In 2010, Cameron and colleagues¹⁷ found that cardiovascular disease was associated with a 2.68 times higher risk of mortality in hip fracture patients. Similarly in our study, both hypertension and history of percutaneous coronary intervention were associated with discharge to rehabilitation. We found higher odds of discharge to rehabilitation with higher ASA classes, which mirror results from a study by Michel and colleagues,¹⁵ who found that higher (vs lower) preoperative ASA classes were associated with higher 1-year mortality in hip fracture patients. Interestingly, DNR status was associated with higher odds of discharge home, which may reflect patients’ desires to forgo noninvasive or lifesaving procedures that may be performed at rehabilitation facilities. Although general anesthesia predisposed patients to discharge to a rehabilitation center, multiple studies have found no association between anesthesia type and postoperative mortality rates for hip fracture patients.^18,19 Last, Marcantonio and colleagues²⁰ found delirium specifically had a higher odds ratio for discharge, but our univariate analysis did not find a significant association between impaired sensorium and discharge location. Given the correlation of our risk factors with increased morbidity and mortality in the literature, our study’s results provide the initial groundwork for creating a risk calculator that orthopedists can use to predict discharge to rehabilitation.

Our study had some limitations. Although we analyzed a large number of demographics, preoperative comorbidities, and surgical factors, our univariate analysis was limited to information in the ACS-NSQIP database. We did not incorporate other clinically relevant factors (eg, social factors, including patients’ support networks) that may influence discharge decisions. Furthermore, ACS-NSQIP records patient data only up to 30 days after surgery. Discharge information for the time after that was missing for a subset of hip fracture patients, and these patients had to be excluded, potentially skewing our data. ACS-NSQIP also does not collect cost data for patients based on hospitalization or LOS, so we could not determine whether patients discharged to rehabilitation incurred higher costs because of longer hospitalizations.

Nevertheless, our study identified significant patient and operative variables that are associated with discharge to a rehabilitation center. By identifying hip fracture patients with these risk factors early and mobilizing the appropriate resources, practicing orthopedists should be better equipped to help facilitate the discharge of patients to the appropriate location after surgery. Validation of these risk factors should be prospectively determined with an analysis of LOS and cost implications. Use of a risk calculator may in fact result in decreased LOS and hospital-related costs. Furthermore, using these risk factors in a prospective patient cohort would help validate their use and determine whether there is clinical correlation. The orthopedists in our institution are becoming more aware of these risk factors, but validation is necessary.

Length of stay (LOS) is a significant driver of costs after hip fracture surgery.^1-3 Multiple studies have identified factors associated with increased LOS in hip fracture patients. These factors include admission time, delay to surgery, presence of comorbidities, and older age.^4-9

One significant and potentially modifiable factor affecting LOS is delayed transfer to a rehabilitation center after surgery.^8-11 Although patients after orthopedic surgeries require additional rehabilitation services or subacute care directly attributable to their injuries, specialized rehabilitation centers may not always have beds readily available.^6-11 Studies have shown that delays in transfer to skilled nursing facilities or rehabilitation centers are highly common among orthopedic patients.⁸ It is therefore imperative that orthopedists have a mechanism for predicting and identifying which patients require rehabilitation services early in the postoperative period. Identifying risk factors and stratifying patients who are most likely to require rehabilitation would facilitate the early transfer of these patients and thereby directly decrease LOS and hospitalization-related costs.

In this article, we report results from prospective, national, multicenter data to identify commonly measured risk factors for discharge to rehabilitation facilities for hip fracture patients. Through multivariate analysis of ACS-NSQIP (American College of Surgeons National Surgical Quality Improvement Program) data, we determined which risk factors significantly predispose patients to discharge to rehabilitation centers versus discharge home. Knowledge of these risk factors allows the practicing orthopedist to be better equipped to identify patients who require additional rehabilitation early in the postoperative course. By mobilizing case managers and social workers to help avoid delays in the transfers of these identified patients, LOS-associated costs may ultimately decrease.

Materials and Methods

After obtaining institutional review board approval for this study from the Office of Research at Vanderbilt University, we prospectively collected 2011 discharge data from the ACS-NSQIP database (these data are unavailable for earlier years). All patients who underwent hip fracture surgery in 2011 were identified by CPT (Current Procedural Terminology) codes. Cases of patients with unknown discharge information and of those who died during their hospitalizations were excluded from analysis. For the remaining patients, discharge information as categorized by ACS-NSQIP included skilled care (eg, subacute hospital, skilled nursing home), unskilled facility (eg, nursing home, assisted facility), separate acute care, and rehabilitation. All other patients were discharged home without additional assistance or to the previous home where they received chronic care, assisted living, or unskilled aid. Patients were dichotomized according to whether they were discharged home or to one of the rehabilitation facilities mentioned.

To determine which risk factors significantly contributed to a patient’s discharge to rehabilitation, we ran univariate analyses using Fisher exact tests for categorical variables and Student t tests for continuous variables on multiple patient factors, including demographics, preoperative comorbidities, and operative factors. Demographics included age and sex. Preoperative comorbidities included 32 conditions: diabetes mellitus, active smoking status, current alcohol use, dyspnea, history of chronic obstructive pulmonary disease, history of congestive heart failure, hypertension requiring medication, history of esophageal varices, history of myocardial infarction, current renal failure, current dialysis dependence, steroid use, recent weight loss, existing bleeding disorder, transfusion before discharge, presence of central nervous system tumor, recent chemotherapy, recent radiation therapy, previous percutaneous coronary intervention, previous percutaneous coronary stenting, history of angina, peripheral vascular disease, cerebrovascular accidents, recent surgery (within 30 days), rest pain, impaired sensorium, history of transient ischemic attacks, current hemiplegia status, current paraplegia status, current quadriplegia status, current ascites, hypertension, and disseminated cancer. Operative factors included wound infection, DNR (do not resuscitate) status, ventilator support, anesthesia type, wound class, ASA (American Society of Anesthesiologists) class, and operative time.

For the univariate analyses, significance was set at P < .05. Demographics, preoperative comorbidities, and operative factors that were significantly associated with discharge to a rehabilitation facility in the univariate analysis were selected as covariates for a multivariate analysis. We incorporated a binary logistic regression to analyze which of these significant risk factors are correlated with a patient’s discharge to a rehabilitation facility after hip fracture surgery.

Results

A total of 4974 patients undergoing surgery for hip fractures in 2011 were identified. Of these patients, 4815 had complete information on discharge location and were included in the analysis.

Table 1 lists the results of the univariate analysis comparing demographics, preoperative comorbidities, and operative factors between the home and rehabilitation groups. Both age (P < .001) and sex (P = .012) were significantly different between groups; the rehabilitation group was older by about 10 years and included significantly more females. In addition to demographic factors, 16 preoperative comorbidities, and 5 surgical factors were significantly associated with discharge to rehabilitation.

Surgery type significantly affected discharge to rehabilitation (Figure). Patients who were undergoing open plating of a femoral neck fracture or intramedullary nailing of an intertrochanteric, peritrochanteric, or subtrochanteric femoral fracture constituted 30% of all patients discharged to rehabilitation centers. In contrast, patients undergoing percutaneous skeletal fixation of a proximal femoral fracture constituted only 5.5% of all patients discharged to rehabilitation. Based on surgery type, we broke down discharge location further, into categories of skilled nursing facility, unskilled facility (not patient’s previous home), separate acute-care facility, dedicated rehabilitation center, and home. Of all 4815 patients combined, 2102 (43.6%) were discharged to a skilled nursing facility, 31 (0.6%) to an unskilled facility (not home), 106 (2.2%) to separate acute care, 1312 (27.2%) to a dedicated rehabilitation center, and 950 (19.7%) home.

Table 2 lists the significant results from the multivariate logistical analysis comparing discharge to a rehabilitation center and discharge home after controlling for the significant risk factors (Table 1). Current diabetes, history of dyspnea, previous myocardial infarction, history of ischemic attacks, current bleeding disorder, transfusion during hospitalization, previous percutaneous cardiac stenting, chemotherapy, past cerebrovascular accident, presence of cancer, surgery type based on CPT code, history of chronic obstructive pulmonary disease or congestive heart failure, current smoking status, and operative time longer than 90 minutes were not significantly correlated with discharge to rehabilitation in the multivariate analysis. All significant factors were associated with higher odds of discharge to rehabilitation except for DNR status. DNR patients were 2.04 times more likely (95% CI, 1.49-2.78; P < .001) to be discharged home than to rehabilitation centers.

Applying these adjusted odds ratios, we see that an elderly woman (age, >65 years) who underwent general anesthesia with an ASA class higher than 2 was 17.63 times more likely than a patient without these risk factors to be discharged to rehabilitation. If this patient were also dialysis-dependent, she would be 61.52 times more likely than a similar patient without dialysis needs to be discharged to rehabilitation.

Even when controlling for all significant and nonsignificant variables in multivariate logistical analysis, age over 65 years (β = 1.05; P < .001), female sex (β = 1.76; P = .004), dialysis dependence (β = 12.98; P = .036), hypertension requiring medication (β = 1.53; P = .032), and ASA class higher than 2 (β = 1.98; P = .001) were found to be significant risk factors for discharge to rehabilitation.

Discussion

This study was the first to investigate the issue of which patient risk factors allow the practicing orthopedist to identify patients who require rehabilitation after hip fracture surgery. Through our multivariate analysis, which controlled for demographics, comorbidities, and operative factors, we found that older age, female sex, history of percutaneous coronary intervention, dialysis dependence, general anesthesia, and ASA class higher than 2 significantly increased the odds of discharge to a rehabilitation center versus home.

Using our study’s results, we can create a risk stratification model for patients and thereby a means of targeting patients who need rehabilitation and starting the process of finding a rehabilitation bed early in the postoperative course. Our study’s variables are easily measured metrics that may be collected in any hospital setting. Especially for hip fracture patients, early planning and discharge to the appropriate rehabilitation center are important in decreasing LOS and associated hospitalization costs. According to one report,³ about 85% of all hip fracture costs are directly related to LOS, given the unnecessarily long rehabilitation periods in hospitals. Hollingworth and colleagues² compared costs for patients who remained in the hospital with costs for those discharged with rehabilitation services. Overall costs were significantly lower for patients discharged home with rehabilitation. The authors concluded that 40% of hip fracture patients may be suitable for early discharge.² In an analysis of Medicare payments for hip fracture treatment, hospital costs including LOS accounted for 60% of all payments.¹² The results of these 2 studies suggest that the overall driver of hip fracture costs is prolonged LOS and that, if patients are discharged to rehabilitation, then overall costs may be lowered through a direct reduction in hospital LOS. Given that hip fractures account for almost 350,000 hospital admissions in the United States each year, and using our institution’s average hospital charge per day ($4500), about $1.6 billion may be saved if each patient’s LOS decreased by 1 day.¹³ Although multiple factors affect LOS, discharge planning is under orthopedists’ direct control. Therefore, early identification of patients who will require rehabilitation may help reduce LOS-associated costs in our health care system.

The patient variables that were significantly associated with discharge to rehabilitation are also associated with increased morbidity and mortality in hip fracture patients, according to the literature,^14-20 which provides some external validation of using these risk factors as predictors for rehabilitation. A patient with one of these risk factors may require rehabilitation, given that rehabilitation services are specifically linked to lower morbidity and mortality rates among hip fracture patients. For example, patients with dialysis needs were 3.49 times more likely to be discharged to a rehabilitation center in our study. In a 2000 study by Coco and Rush,¹⁶ hip fracture patients on dialysis had a 1-year mortality rate 2.5 times higher than that of patients who were not dialysis-dependent. In 2010, Cameron and colleagues¹⁷ found that cardiovascular disease was associated with a 2.68 times higher risk of mortality in hip fracture patients. Similarly in our study, both hypertension and history of percutaneous coronary intervention were associated with discharge to rehabilitation. We found higher odds of discharge to rehabilitation with higher ASA classes, which mirror results from a study by Michel and colleagues,¹⁵ who found that higher (vs lower) preoperative ASA classes were associated with higher 1-year mortality in hip fracture patients. Interestingly, DNR status was associated with higher odds of discharge home, which may reflect patients’ desires to forgo noninvasive or lifesaving procedures that may be performed at rehabilitation facilities. Although general anesthesia predisposed patients to discharge to a rehabilitation center, multiple studies have found no association between anesthesia type and postoperative mortality rates for hip fracture patients.^18,19 Last, Marcantonio and colleagues²⁰ found delirium specifically had a higher odds ratio for discharge, but our univariate analysis did not find a significant association between impaired sensorium and discharge location. Given the correlation of our risk factors with increased morbidity and mortality in the literature, our study’s results provide the initial groundwork for creating a risk calculator that orthopedists can use to predict discharge to rehabilitation.

Our study had some limitations. Although we analyzed a large number of demographics, preoperative comorbidities, and surgical factors, our univariate analysis was limited to information in the ACS-NSQIP database. We did not incorporate other clinically relevant factors (eg, social factors, including patients’ support networks) that may influence discharge decisions. Furthermore, ACS-NSQIP records patient data only up to 30 days after surgery. Discharge information for the time after that was missing for a subset of hip fracture patients, and these patients had to be excluded, potentially skewing our data. ACS-NSQIP also does not collect cost data for patients based on hospitalization or LOS, so we could not determine whether patients discharged to rehabilitation incurred higher costs because of longer hospitalizations.

Nevertheless, our study identified significant patient and operative variables that are associated with discharge to a rehabilitation center. By identifying hip fracture patients with these risk factors early and mobilizing the appropriate resources, practicing orthopedists should be better equipped to help facilitate the discharge of patients to the appropriate location after surgery. Validation of these risk factors should be prospectively determined with an analysis of LOS and cost implications. Use of a risk calculator may in fact result in decreased LOS and hospital-related costs. Furthermore, using these risk factors in a prospective patient cohort would help validate their use and determine whether there is clinical correlation. The orthopedists in our institution are becoming more aware of these risk factors, but validation is necessary.

Length of stay (LOS) is a significant driver of costs after hip fracture surgery.^1-3 Multiple studies have identified factors associated with increased LOS in hip fracture patients. These factors include admission time, delay to surgery, presence of comorbidities, and older age.^4-9

One significant and potentially modifiable factor affecting LOS is delayed transfer to a rehabilitation center after surgery.^8-11 Although patients after orthopedic surgeries require additional rehabilitation services or subacute care directly attributable to their injuries, specialized rehabilitation centers may not always have beds readily available.^6-11 Studies have shown that delays in transfer to skilled nursing facilities or rehabilitation centers are highly common among orthopedic patients.⁸ It is therefore imperative that orthopedists have a mechanism for predicting and identifying which patients require rehabilitation services early in the postoperative period. Identifying risk factors and stratifying patients who are most likely to require rehabilitation would facilitate the early transfer of these patients and thereby directly decrease LOS and hospitalization-related costs.

In this article, we report results from prospective, national, multicenter data to identify commonly measured risk factors for discharge to rehabilitation facilities for hip fracture patients. Through multivariate analysis of ACS-NSQIP (American College of Surgeons National Surgical Quality Improvement Program) data, we determined which risk factors significantly predispose patients to discharge to rehabilitation centers versus discharge home. Knowledge of these risk factors allows the practicing orthopedist to be better equipped to identify patients who require additional rehabilitation early in the postoperative course. By mobilizing case managers and social workers to help avoid delays in the transfers of these identified patients, LOS-associated costs may ultimately decrease.

Materials and Methods

After obtaining institutional review board approval for this study from the Office of Research at Vanderbilt University, we prospectively collected 2011 discharge data from the ACS-NSQIP database (these data are unavailable for earlier years). All patients who underwent hip fracture surgery in 2011 were identified by CPT (Current Procedural Terminology) codes. Cases of patients with unknown discharge information and of those who died during their hospitalizations were excluded from analysis. For the remaining patients, discharge information as categorized by ACS-NSQIP included skilled care (eg, subacute hospital, skilled nursing home), unskilled facility (eg, nursing home, assisted facility), separate acute care, and rehabilitation. All other patients were discharged home without additional assistance or to the previous home where they received chronic care, assisted living, or unskilled aid. Patients were dichotomized according to whether they were discharged home or to one of the rehabilitation facilities mentioned.

To determine which risk factors significantly contributed to a patient’s discharge to rehabilitation, we ran univariate analyses using Fisher exact tests for categorical variables and Student t tests for continuous variables on multiple patient factors, including demographics, preoperative comorbidities, and operative factors. Demographics included age and sex. Preoperative comorbidities included 32 conditions: diabetes mellitus, active smoking status, current alcohol use, dyspnea, history of chronic obstructive pulmonary disease, history of congestive heart failure, hypertension requiring medication, history of esophageal varices, history of myocardial infarction, current renal failure, current dialysis dependence, steroid use, recent weight loss, existing bleeding disorder, transfusion before discharge, presence of central nervous system tumor, recent chemotherapy, recent radiation therapy, previous percutaneous coronary intervention, previous percutaneous coronary stenting, history of angina, peripheral vascular disease, cerebrovascular accidents, recent surgery (within 30 days), rest pain, impaired sensorium, history of transient ischemic attacks, current hemiplegia status, current paraplegia status, current quadriplegia status, current ascites, hypertension, and disseminated cancer. Operative factors included wound infection, DNR (do not resuscitate) status, ventilator support, anesthesia type, wound class, ASA (American Society of Anesthesiologists) class, and operative time.

For the univariate analyses, significance was set at P < .05. Demographics, preoperative comorbidities, and operative factors that were significantly associated with discharge to a rehabilitation facility in the univariate analysis were selected as covariates for a multivariate analysis. We incorporated a binary logistic regression to analyze which of these significant risk factors are correlated with a patient’s discharge to a rehabilitation facility after hip fracture surgery.

Results

A total of 4974 patients undergoing surgery for hip fractures in 2011 were identified. Of these patients, 4815 had complete information on discharge location and were included in the analysis.

Table 1 lists the results of the univariate analysis comparing demographics, preoperative comorbidities, and operative factors between the home and rehabilitation groups. Both age (P < .001) and sex (P = .012) were significantly different between groups; the rehabilitation group was older by about 10 years and included significantly more females. In addition to demographic factors, 16 preoperative comorbidities, and 5 surgical factors were significantly associated with discharge to rehabilitation.

Surgery type significantly affected discharge to rehabilitation (Figure). Patients who were undergoing open plating of a femoral neck fracture or intramedullary nailing of an intertrochanteric, peritrochanteric, or subtrochanteric femoral fracture constituted 30% of all patients discharged to rehabilitation centers. In contrast, patients undergoing percutaneous skeletal fixation of a proximal femoral fracture constituted only 5.5% of all patients discharged to rehabilitation. Based on surgery type, we broke down discharge location further, into categories of skilled nursing facility, unskilled facility (not patient’s previous home), separate acute-care facility, dedicated rehabilitation center, and home. Of all 4815 patients combined, 2102 (43.6%) were discharged to a skilled nursing facility, 31 (0.6%) to an unskilled facility (not home), 106 (2.2%) to separate acute care, 1312 (27.2%) to a dedicated rehabilitation center, and 950 (19.7%) home.

Table 2 lists the significant results from the multivariate logistical analysis comparing discharge to a rehabilitation center and discharge home after controlling for the significant risk factors (Table 1). Current diabetes, history of dyspnea, previous myocardial infarction, history of ischemic attacks, current bleeding disorder, transfusion during hospitalization, previous percutaneous cardiac stenting, chemotherapy, past cerebrovascular accident, presence of cancer, surgery type based on CPT code, history of chronic obstructive pulmonary disease or congestive heart failure, current smoking status, and operative time longer than 90 minutes were not significantly correlated with discharge to rehabilitation in the multivariate analysis. All significant factors were associated with higher odds of discharge to rehabilitation except for DNR status. DNR patients were 2.04 times more likely (95% CI, 1.49-2.78; P < .001) to be discharged home than to rehabilitation centers.

Applying these adjusted odds ratios, we see that an elderly woman (age, >65 years) who underwent general anesthesia with an ASA class higher than 2 was 17.63 times more likely than a patient without these risk factors to be discharged to rehabilitation. If this patient were also dialysis-dependent, she would be 61.52 times more likely than a similar patient without dialysis needs to be discharged to rehabilitation.

Even when controlling for all significant and nonsignificant variables in multivariate logistical analysis, age over 65 years (β = 1.05; P < .001), female sex (β = 1.76; P = .004), dialysis dependence (β = 12.98; P = .036), hypertension requiring medication (β = 1.53; P = .032), and ASA class higher than 2 (β = 1.98; P = .001) were found to be significant risk factors for discharge to rehabilitation.

Discussion

This study was the first to investigate the issue of which patient risk factors allow the practicing orthopedist to identify patients who require rehabilitation after hip fracture surgery. Through our multivariate analysis, which controlled for demographics, comorbidities, and operative factors, we found that older age, female sex, history of percutaneous coronary intervention, dialysis dependence, general anesthesia, and ASA class higher than 2 significantly increased the odds of discharge to a rehabilitation center versus home.

Using our study’s results, we can create a risk stratification model for patients and thereby a means of targeting patients who need rehabilitation and starting the process of finding a rehabilitation bed early in the postoperative course. Our study’s variables are easily measured metrics that may be collected in any hospital setting. Especially for hip fracture patients, early planning and discharge to the appropriate rehabilitation center are important in decreasing LOS and associated hospitalization costs. According to one report,³ about 85% of all hip fracture costs are directly related to LOS, given the unnecessarily long rehabilitation periods in hospitals. Hollingworth and colleagues² compared costs for patients who remained in the hospital with costs for those discharged with rehabilitation services. Overall costs were significantly lower for patients discharged home with rehabilitation. The authors concluded that 40% of hip fracture patients may be suitable for early discharge.² In an analysis of Medicare payments for hip fracture treatment, hospital costs including LOS accounted for 60% of all payments.¹² The results of these 2 studies suggest that the overall driver of hip fracture costs is prolonged LOS and that, if patients are discharged to rehabilitation, then overall costs may be lowered through a direct reduction in hospital LOS. Given that hip fractures account for almost 350,000 hospital admissions in the United States each year, and using our institution’s average hospital charge per day ($4500), about $1.6 billion may be saved if each patient’s LOS decreased by 1 day.¹³ Although multiple factors affect LOS, discharge planning is under orthopedists’ direct control. Therefore, early identification of patients who will require rehabilitation may help reduce LOS-associated costs in our health care system.

The patient variables that were significantly associated with discharge to rehabilitation are also associated with increased morbidity and mortality in hip fracture patients, according to the literature,^14-20 which provides some external validation of using these risk factors as predictors for rehabilitation. A patient with one of these risk factors may require rehabilitation, given that rehabilitation services are specifically linked to lower morbidity and mortality rates among hip fracture patients. For example, patients with dialysis needs were 3.49 times more likely to be discharged to a rehabilitation center in our study. In a 2000 study by Coco and Rush,¹⁶ hip fracture patients on dialysis had a 1-year mortality rate 2.5 times higher than that of patients who were not dialysis-dependent. In 2010, Cameron and colleagues¹⁷ found that cardiovascular disease was associated with a 2.68 times higher risk of mortality in hip fracture patients. Similarly in our study, both hypertension and history of percutaneous coronary intervention were associated with discharge to rehabilitation. We found higher odds of discharge to rehabilitation with higher ASA classes, which mirror results from a study by Michel and colleagues,¹⁵ who found that higher (vs lower) preoperative ASA classes were associated with higher 1-year mortality in hip fracture patients. Interestingly, DNR status was associated with higher odds of discharge home, which may reflect patients’ desires to forgo noninvasive or lifesaving procedures that may be performed at rehabilitation facilities. Although general anesthesia predisposed patients to discharge to a rehabilitation center, multiple studies have found no association between anesthesia type and postoperative mortality rates for hip fracture patients.^18,19 Last, Marcantonio and colleagues²⁰ found delirium specifically had a higher odds ratio for discharge, but our univariate analysis did not find a significant association between impaired sensorium and discharge location. Given the correlation of our risk factors with increased morbidity and mortality in the literature, our study’s results provide the initial groundwork for creating a risk calculator that orthopedists can use to predict discharge to rehabilitation.

Our study had some limitations. Although we analyzed a large number of demographics, preoperative comorbidities, and surgical factors, our univariate analysis was limited to information in the ACS-NSQIP database. We did not incorporate other clinically relevant factors (eg, social factors, including patients’ support networks) that may influence discharge decisions. Furthermore, ACS-NSQIP records patient data only up to 30 days after surgery. Discharge information for the time after that was missing for a subset of hip fracture patients, and these patients had to be excluded, potentially skewing our data. ACS-NSQIP also does not collect cost data for patients based on hospitalization or LOS, so we could not determine whether patients discharged to rehabilitation incurred higher costs because of longer hospitalizations.

Nevertheless, our study identified significant patient and operative variables that are associated with discharge to a rehabilitation center. By identifying hip fracture patients with these risk factors early and mobilizing the appropriate resources, practicing orthopedists should be better equipped to help facilitate the discharge of patients to the appropriate location after surgery. Validation of these risk factors should be prospectively determined with an analysis of LOS and cost implications. Use of a risk calculator may in fact result in decreased LOS and hospital-related costs. Furthermore, using these risk factors in a prospective patient cohort would help validate their use and determine whether there is clinical correlation. The orthopedists in our institution are becoming more aware of these risk factors, but validation is necessary.

Femur fractures (Orthopaedic Trauma Association classes 31, 32, 33)¹ are common childhood injuries, occurring at a rate of 19 per 100,000 children in the United States.² Peak occurrence is bimodal at ages 2 and 17 years. The most common mechanism of injury in children under 6 years is a fall, and hip spica casting is the preferred treatment modality in this group.^3-5

A bar connecting the legs of the spica cast has been shown to facilitate patient transport⁵ and significantly decrease mechanical failure of the spica cast.⁶ This bar often consists of a broom handle or pipe that must be cut to size during the case and subsequently incorporated into the cast—tasks that are often inconvenient and time-consuming for on-call or emergency department staff unfamiliar with orthopedic tools.

In this article, we review a spica cast application that incorporates a low-cost, lightweight technique for fabricating a connecting bar from existing fiberglass casting material. The Institutional Review Board at Connecticut Children’s Medical Center approved this work.

Technique of Double-Leg Spica Casting With Fiberglass Bar

A spica casting table (Orthopedic Systems) with a well-padded post is placed on the operating room table and adjusted to the length of the patient from perineum to just below the shoulders. With the patient under general anesthesia, folded towels are used to provide 2 to 4 cm of padding on the anterior torso, atop which a waterproof pantaloon is applied. The patient is transferred to the spica table, and the patient’s arms are gently secured to the casting table with cast padding or tape in an abducted position at the shoulders. A surgeon controls the legs by holding the feet with the long fingers just above the heels, the index fingers on the anterior ankle, and the thumbs on the soles of the feet. Cast padding is wrapped from the nipple line to the supramalleolar region on each leg. The bony prominences of the malleoli, patella, fibular head, femoral condyles, iliac crests, and coccyx are well padded.

Fiberglass is then rolled without compression onto the patient, beginning with the torso and perineal areas. The injured leg is wrapped to its final length above the malleoli while the uninjured leg is kept free. Maintaining the position of the injured leg with simultaneous molding at the fracture site, typically to promote valgus, allows fracture reduction. The fracture position is then checked under image intensification. For femur fractures, hip abduction and flexion are set to 45° and 90°, respectively, while knee flexion is between 50° and 90°. The uninjured leg is then wrapped with fiberglass. Additional strips of fiberglass can be used to reinforce weak junctional regions between the torso and the legs, posteriorly over the “intern’s triangle” and anteriorly along the hip crease.

A connecting fiberglass bar is then created using a fiberglass roll once the cast is hardened. A 2-inch fiberglass roll is wrapped around one leg to secure its position (Figure 1A) and then rolled around the second limb (Figure 1B). Fiberglass is then pulled taut and rolled around the bridge that has been created in order to thicken the bar (Figure 2). The roll is again brought around the closest limb, wrapped back across the bridge to the other limb, and rolled out to its full length. Last, the legs are abducted 1 to 2 cm to tension the bar (Figure 3). Although this does not produce enough movement to cause a crease and a resultant ulcer, careful inspection of common pressure points (eg, popliteal fossa) should be performed after the cast is complete.

The chest towels are removed, and the final cast is inspected clinically and fluoroscopically at the fracture site before extubation. The cast is trimmed as needed to ensure room for perineal care, as well as full ankle flexion and extension without impingement. Cast edges are further petaled with plastic tape (Hy-Tape International) to provide padding and prevent the waterproof lining from tearing.

Postoperative care involves overnight observation and caregiver practice in perineal care. Frequent rotation from supine to prone is encouraged. Nurses confirm car-seat fit before discharge. If needed, radiographs are obtained 7 to 10 days later to help with wedging adjustment. The cast is removed in the clinic when adequate callus is appreciated on subsequent radiographs.

Case Series

Our experience with this technique in 16 unilateral femur fractures has been favorable (Table). Patient age ranged from 5 months to 3 years. Mean pretreatment angulation was 13° varus and 11° procurvatum. The majority of fractures were femoral shaft fractures; 1 was proximal, 2 distal.

All fractures united without cast revision. Mean cast time was 4.5 weeks (range, 16 days–6 weeks). Immediate postoperative alignment was 2.5° varus (range, 11° valgus to 16° varus) and 7° procurvatum (range, 1° recurvatum to 22° procurvatum). Mean shortening was 1.5 cm (range, 0-2.7 cm). Final alignment was 1° valgus (range, 9° valgus to 12° varus) and 5° procurvatum (range, 0° to 22°). Mean follow-up was 8 months. There were no cases of skin maceration or cast failure. No casts precluded use of a spica car-seat. Figure 4 shows a typical case with a midshaft fracture treated with closed reduction and casting for 4 weeks with good remodeling at final follow-up, 19 months after injury.

Discussion

Although single-leg walking spica casts have been shown to safely treat low-energy femur fractures in children 1 to 6 years old,⁷ length-unstable femur fractures, bilateral femur fractures, and patients with hip dysplasia continue to be managed with a double-leg hip spica construct. Cast integrity remains fundamental to the control of most fractures and prevention of cast-related complications, such as skin maceration and ulceration. Surgeons typically use spica cast reinforcement schemes—such as cast augments of the torso–limb junction, with multiple layers of casting material or incorporation of a connecting bar between the legs, typically constructed by overwrapping a wooden dowel in casting material—to improve the mechanical stability of casts.⁶ The present technique of creating a connecting bar from fiberglass casting material significantly simplifies the standard wooden dowel approach and provided excellent results in our treatment group in terms of cast integrity and fracture alignment. In addition, at our institution, a roll of fiberglass costs $2.10, whereas a wooden dowel costs $3 to $10 and can be difficult to locate if not frequently used. Other tube-shaped materials, such as the disposable material used to package implants and tubes, carry an even lower cost. However, we have found that a single fiberglass roll is most readily available and easiest to apply.

Although proper spica cast application remains important in managing pediatric trauma, it lacks a good technical description in the literature. In this technical report, we have presented our standard spica cast application method, which minimizes the range of cast complications that have been reported, from minor skin irritation to superior mesenteric artery syndrome. Two salient technical highlights are use of waterproof pantaloon liners and cast petaling, which we have found almost eliminate the morbidity of potential skin complications, reported to occur at a rate of 28%.⁸ In addition, we forgo applying the cast on the injured leg in segments. Application of a short-leg cast on the injured leg to allow traction on the leg during cast application is of dubious utility and may be potentially harmful, with described complications of peroneal nerve palsy and compartment syndrome.^9-11 Further, it is important to use an abdominal spacer (eg, a stack of towels) under the cast padding to create room for abdominal expansion and minimize pressure thought to induce superior mesenteric artery syndrome. Plastic or rubber abdominal spacers have also been described.^12,13 Last, leg position is important for reduction and maintenance of the fracture, as well as patient care. Literature advocates minimizing hip abduction to just that needed for perineal care and maximizing hip flexion and knee extension to optimize car-seat fit and safety.¹⁴

Conclusion

Construction of a spica cast lower limb connecting bar from readily available fiberglass casting material allows a facile and rapid addition to the mechanical stability of a spica cast in the treatment of pediatric femur fractures. The technique is low-cost and obviates the need for additional extraneous materials.

Femur fractures (Orthopaedic Trauma Association classes 31, 32, 33)¹ are common childhood injuries, occurring at a rate of 19 per 100,000 children in the United States.² Peak occurrence is bimodal at ages 2 and 17 years. The most common mechanism of injury in children under 6 years is a fall, and hip spica casting is the preferred treatment modality in this group.^3-5

A bar connecting the legs of the spica cast has been shown to facilitate patient transport⁵ and significantly decrease mechanical failure of the spica cast.⁶ This bar often consists of a broom handle or pipe that must be cut to size during the case and subsequently incorporated into the cast—tasks that are often inconvenient and time-consuming for on-call or emergency department staff unfamiliar with orthopedic tools.

In this article, we review a spica cast application that incorporates a low-cost, lightweight technique for fabricating a connecting bar from existing fiberglass casting material. The Institutional Review Board at Connecticut Children’s Medical Center approved this work.

Technique of Double-Leg Spica Casting With Fiberglass Bar

A spica casting table (Orthopedic Systems) with a well-padded post is placed on the operating room table and adjusted to the length of the patient from perineum to just below the shoulders. With the patient under general anesthesia, folded towels are used to provide 2 to 4 cm of padding on the anterior torso, atop which a waterproof pantaloon is applied. The patient is transferred to the spica table, and the patient’s arms are gently secured to the casting table with cast padding or tape in an abducted position at the shoulders. A surgeon controls the legs by holding the feet with the long fingers just above the heels, the index fingers on the anterior ankle, and the thumbs on the soles of the feet. Cast padding is wrapped from the nipple line to the supramalleolar region on each leg. The bony prominences of the malleoli, patella, fibular head, femoral condyles, iliac crests, and coccyx are well padded.

Fiberglass is then rolled without compression onto the patient, beginning with the torso and perineal areas. The injured leg is wrapped to its final length above the malleoli while the uninjured leg is kept free. Maintaining the position of the injured leg with simultaneous molding at the fracture site, typically to promote valgus, allows fracture reduction. The fracture position is then checked under image intensification. For femur fractures, hip abduction and flexion are set to 45° and 90°, respectively, while knee flexion is between 50° and 90°. The uninjured leg is then wrapped with fiberglass. Additional strips of fiberglass can be used to reinforce weak junctional regions between the torso and the legs, posteriorly over the “intern’s triangle” and anteriorly along the hip crease.

A connecting fiberglass bar is then created using a fiberglass roll once the cast is hardened. A 2-inch fiberglass roll is wrapped around one leg to secure its position (Figure 1A) and then rolled around the second limb (Figure 1B). Fiberglass is then pulled taut and rolled around the bridge that has been created in order to thicken the bar (Figure 2). The roll is again brought around the closest limb, wrapped back across the bridge to the other limb, and rolled out to its full length. Last, the legs are abducted 1 to 2 cm to tension the bar (Figure 3). Although this does not produce enough movement to cause a crease and a resultant ulcer, careful inspection of common pressure points (eg, popliteal fossa) should be performed after the cast is complete.

The chest towels are removed, and the final cast is inspected clinically and fluoroscopically at the fracture site before extubation. The cast is trimmed as needed to ensure room for perineal care, as well as full ankle flexion and extension without impingement. Cast edges are further petaled with plastic tape (Hy-Tape International) to provide padding and prevent the waterproof lining from tearing.

Postoperative care involves overnight observation and caregiver practice in perineal care. Frequent rotation from supine to prone is encouraged. Nurses confirm car-seat fit before discharge. If needed, radiographs are obtained 7 to 10 days later to help with wedging adjustment. The cast is removed in the clinic when adequate callus is appreciated on subsequent radiographs.

Case Series

Our experience with this technique in 16 unilateral femur fractures has been favorable (Table). Patient age ranged from 5 months to 3 years. Mean pretreatment angulation was 13° varus and 11° procurvatum. The majority of fractures were femoral shaft fractures; 1 was proximal, 2 distal.

All fractures united without cast revision. Mean cast time was 4.5 weeks (range, 16 days–6 weeks). Immediate postoperative alignment was 2.5° varus (range, 11° valgus to 16° varus) and 7° procurvatum (range, 1° recurvatum to 22° procurvatum). Mean shortening was 1.5 cm (range, 0-2.7 cm). Final alignment was 1° valgus (range, 9° valgus to 12° varus) and 5° procurvatum (range, 0° to 22°). Mean follow-up was 8 months. There were no cases of skin maceration or cast failure. No casts precluded use of a spica car-seat. Figure 4 shows a typical case with a midshaft fracture treated with closed reduction and casting for 4 weeks with good remodeling at final follow-up, 19 months after injury.

Discussion

Although single-leg walking spica casts have been shown to safely treat low-energy femur fractures in children 1 to 6 years old,⁷ length-unstable femur fractures, bilateral femur fractures, and patients with hip dysplasia continue to be managed with a double-leg hip spica construct. Cast integrity remains fundamental to the control of most fractures and prevention of cast-related complications, such as skin maceration and ulceration. Surgeons typically use spica cast reinforcement schemes—such as cast augments of the torso–limb junction, with multiple layers of casting material or incorporation of a connecting bar between the legs, typically constructed by overwrapping a wooden dowel in casting material—to improve the mechanical stability of casts.⁶ The present technique of creating a connecting bar from fiberglass casting material significantly simplifies the standard wooden dowel approach and provided excellent results in our treatment group in terms of cast integrity and fracture alignment. In addition, at our institution, a roll of fiberglass costs $2.10, whereas a wooden dowel costs $3 to $10 and can be difficult to locate if not frequently used. Other tube-shaped materials, such as the disposable material used to package implants and tubes, carry an even lower cost. However, we have found that a single fiberglass roll is most readily available and easiest to apply.

Although proper spica cast application remains important in managing pediatric trauma, it lacks a good technical description in the literature. In this technical report, we have presented our standard spica cast application method, which minimizes the range of cast complications that have been reported, from minor skin irritation to superior mesenteric artery syndrome. Two salient technical highlights are use of waterproof pantaloon liners and cast petaling, which we have found almost eliminate the morbidity of potential skin complications, reported to occur at a rate of 28%.⁸ In addition, we forgo applying the cast on the injured leg in segments. Application of a short-leg cast on the injured leg to allow traction on the leg during cast application is of dubious utility and may be potentially harmful, with described complications of peroneal nerve palsy and compartment syndrome.^9-11 Further, it is important to use an abdominal spacer (eg, a stack of towels) under the cast padding to create room for abdominal expansion and minimize pressure thought to induce superior mesenteric artery syndrome. Plastic or rubber abdominal spacers have also been described.^12,13 Last, leg position is important for reduction and maintenance of the fracture, as well as patient care. Literature advocates minimizing hip abduction to just that needed for perineal care and maximizing hip flexion and knee extension to optimize car-seat fit and safety.¹⁴

Conclusion

Construction of a spica cast lower limb connecting bar from readily available fiberglass casting material allows a facile and rapid addition to the mechanical stability of a spica cast in the treatment of pediatric femur fractures. The technique is low-cost and obviates the need for additional extraneous materials.

Femur fractures (Orthopaedic Trauma Association classes 31, 32, 33)¹ are common childhood injuries, occurring at a rate of 19 per 100,000 children in the United States.² Peak occurrence is bimodal at ages 2 and 17 years. The most common mechanism of injury in children under 6 years is a fall, and hip spica casting is the preferred treatment modality in this group.^3-5

A bar connecting the legs of the spica cast has been shown to facilitate patient transport⁵ and significantly decrease mechanical failure of the spica cast.⁶ This bar often consists of a broom handle or pipe that must be cut to size during the case and subsequently incorporated into the cast—tasks that are often inconvenient and time-consuming for on-call or emergency department staff unfamiliar with orthopedic tools.

In this article, we review a spica cast application that incorporates a low-cost, lightweight technique for fabricating a connecting bar from existing fiberglass casting material. The Institutional Review Board at Connecticut Children’s Medical Center approved this work.

Technique of Double-Leg Spica Casting With Fiberglass Bar

A spica casting table (Orthopedic Systems) with a well-padded post is placed on the operating room table and adjusted to the length of the patient from perineum to just below the shoulders. With the patient under general anesthesia, folded towels are used to provide 2 to 4 cm of padding on the anterior torso, atop which a waterproof pantaloon is applied. The patient is transferred to the spica table, and the patient’s arms are gently secured to the casting table with cast padding or tape in an abducted position at the shoulders. A surgeon controls the legs by holding the feet with the long fingers just above the heels, the index fingers on the anterior ankle, and the thumbs on the soles of the feet. Cast padding is wrapped from the nipple line to the supramalleolar region on each leg. The bony prominences of the malleoli, patella, fibular head, femoral condyles, iliac crests, and coccyx are well padded.

Fiberglass is then rolled without compression onto the patient, beginning with the torso and perineal areas. The injured leg is wrapped to its final length above the malleoli while the uninjured leg is kept free. Maintaining the position of the injured leg with simultaneous molding at the fracture site, typically to promote valgus, allows fracture reduction. The fracture position is then checked under image intensification. For femur fractures, hip abduction and flexion are set to 45° and 90°, respectively, while knee flexion is between 50° and 90°. The uninjured leg is then wrapped with fiberglass. Additional strips of fiberglass can be used to reinforce weak junctional regions between the torso and the legs, posteriorly over the “intern’s triangle” and anteriorly along the hip crease.

A connecting fiberglass bar is then created using a fiberglass roll once the cast is hardened. A 2-inch fiberglass roll is wrapped around one leg to secure its position (Figure 1A) and then rolled around the second limb (Figure 1B). Fiberglass is then pulled taut and rolled around the bridge that has been created in order to thicken the bar (Figure 2). The roll is again brought around the closest limb, wrapped back across the bridge to the other limb, and rolled out to its full length. Last, the legs are abducted 1 to 2 cm to tension the bar (Figure 3). Although this does not produce enough movement to cause a crease and a resultant ulcer, careful inspection of common pressure points (eg, popliteal fossa) should be performed after the cast is complete.

The chest towels are removed, and the final cast is inspected clinically and fluoroscopically at the fracture site before extubation. The cast is trimmed as needed to ensure room for perineal care, as well as full ankle flexion and extension without impingement. Cast edges are further petaled with plastic tape (Hy-Tape International) to provide padding and prevent the waterproof lining from tearing.

Postoperative care involves overnight observation and caregiver practice in perineal care. Frequent rotation from supine to prone is encouraged. Nurses confirm car-seat fit before discharge. If needed, radiographs are obtained 7 to 10 days later to help with wedging adjustment. The cast is removed in the clinic when adequate callus is appreciated on subsequent radiographs.

Case Series

Our experience with this technique in 16 unilateral femur fractures has been favorable (Table). Patient age ranged from 5 months to 3 years. Mean pretreatment angulation was 13° varus and 11° procurvatum. The majority of fractures were femoral shaft fractures; 1 was proximal, 2 distal.

All fractures united without cast revision. Mean cast time was 4.5 weeks (range, 16 days–6 weeks). Immediate postoperative alignment was 2.5° varus (range, 11° valgus to 16° varus) and 7° procurvatum (range, 1° recurvatum to 22° procurvatum). Mean shortening was 1.5 cm (range, 0-2.7 cm). Final alignment was 1° valgus (range, 9° valgus to 12° varus) and 5° procurvatum (range, 0° to 22°). Mean follow-up was 8 months. There were no cases of skin maceration or cast failure. No casts precluded use of a spica car-seat. Figure 4 shows a typical case with a midshaft fracture treated with closed reduction and casting for 4 weeks with good remodeling at final follow-up, 19 months after injury.

Discussion

Although single-leg walking spica casts have been shown to safely treat low-energy femur fractures in children 1 to 6 years old,⁷ length-unstable femur fractures, bilateral femur fractures, and patients with hip dysplasia continue to be managed with a double-leg hip spica construct. Cast integrity remains fundamental to the control of most fractures and prevention of cast-related complications, such as skin maceration and ulceration. Surgeons typically use spica cast reinforcement schemes—such as cast augments of the torso–limb junction, with multiple layers of casting material or incorporation of a connecting bar between the legs, typically constructed by overwrapping a wooden dowel in casting material—to improve the mechanical stability of casts.⁶ The present technique of creating a connecting bar from fiberglass casting material significantly simplifies the standard wooden dowel approach and provided excellent results in our treatment group in terms of cast integrity and fracture alignment. In addition, at our institution, a roll of fiberglass costs $2.10, whereas a wooden dowel costs $3 to $10 and can be difficult to locate if not frequently used. Other tube-shaped materials, such as the disposable material used to package implants and tubes, carry an even lower cost. However, we have found that a single fiberglass roll is most readily available and easiest to apply.

Although proper spica cast application remains important in managing pediatric trauma, it lacks a good technical description in the literature. In this technical report, we have presented our standard spica cast application method, which minimizes the range of cast complications that have been reported, from minor skin irritation to superior mesenteric artery syndrome. Two salient technical highlights are use of waterproof pantaloon liners and cast petaling, which we have found almost eliminate the morbidity of potential skin complications, reported to occur at a rate of 28%.⁸ In addition, we forgo applying the cast on the injured leg in segments. Application of a short-leg cast on the injured leg to allow traction on the leg during cast application is of dubious utility and may be potentially harmful, with described complications of peroneal nerve palsy and compartment syndrome.^9-11 Further, it is important to use an abdominal spacer (eg, a stack of towels) under the cast padding to create room for abdominal expansion and minimize pressure thought to induce superior mesenteric artery syndrome. Plastic or rubber abdominal spacers have also been described.^12,13 Last, leg position is important for reduction and maintenance of the fracture, as well as patient care. Literature advocates minimizing hip abduction to just that needed for perineal care and maximizing hip flexion and knee extension to optimize car-seat fit and safety.¹⁴

Conclusion

Construction of a spica cast lower limb connecting bar from readily available fiberglass casting material allows a facile and rapid addition to the mechanical stability of a spica cast in the treatment of pediatric femur fractures. The technique is low-cost and obviates the need for additional extraneous materials.

The incidence of Dupuytren disease increases with advancing age,¹ as do the medical comorbidities of patients seeking treatment for disabling hand contractures. For patients with significant comorbidities, open surgical fasciectomy, the current standard of treatment for Dupuytren disease,^2,3 may be associated with increased perioperative risks.

Collagenase enzymatic fasciotomy has become an accepted nonsurgical treatment alternative to traditional fasciectomy or surgical fasciotomy for significant digital contractures caused by Dupuytren disease.^4-6 Clostridium histolyticum collagenase (CHC) is a foreign protein, made up of 2 collagenases isolated from the bacteria C histolyticum.⁷ The collagenases are zinc-dependent matrix metalloproteinases that cleave the triple helical structure of collagen molecules.⁸ Also known as Xiaflex (Auxilium Pharmaceuticals), CHC was approved by the US Food and Drug Administration (FDA) in February 2010 for use in patients with Dupuytren contractures.

 Enzymatic rupture is safe and efficacious at midterm follow-up and offers the theoretical advantage of avoiding palmar and digital fasciectomy and the associated risks of surgical-site infection and wound-healing complications.⁶ The risks of surgical wound complications are magnified in immunosuppressed patients, particularly those on chronic steroid therapy; wound-healing complication rates may be increased 2 to 5 times compared with controls.⁹ In a pooled literature review, wound-healing complications were reported after 22.9% of open primary fasciectomies, with infection occurring in 2.4%.¹⁰ A nonsurgical alternative is therefore particularly appealing for a patient cohort that may be at higher risk for a frequently described complication of surgery for Dupuytren contracture.

The exclusion criteria in the trials for FDA approval were extensive and included breast-feeding, pregnancy, bleeding disorder, recent stroke, use of tetracycline derivative within 14 days before start of study, use of anticoagulant within 7 days before start of study, allergy to collagenase, and chronic muscular, neurologic, or neuromuscular disorder affecting the hands.⁶ Safety and efficacy of collagenase in patients requiring chronic immunosuppressive therapy for medical comorbidities have not been previously documented. Furthermore, although skin tears were reported in 11% of patients after manual cord rupture in the CORD (Collagenase Option for the Reduction of Dupuytren’s) I trial,⁶ the likelihood of deep and superficial infection and delayed wound healing has not been quantitated.

In this article, we report on outcomes of 13 collagenase enzymatic fasciotomies performed in 8 patients who were on chronic immunosuppressive therapy.

Methods

Institutional review board approval was obtained at both academic hand surgery institutions. We retrospectively reviewed prospectively collected clinical data within our 2 centers’ databases of patients with Dupuytren disease. Eight patients on chronic immunosuppressive therapies treated with collagenase for metacarpophalangeal (MP) or proximal interphalangeal (PIP) joint contractures between February 2010 and December 2011 were identified. Three of these patients received collagenase injections into 2 or more separate Dupuytren cords at different encounters, resulting in a total of 13 individual collagenase enzymatic fasciotomies.

Collagenase injections were administered following CORD I trial protocol,⁶ except we injected Dupuytren cords crossing the PIP joint using a lateral approach to minimize risk of flexor tendon rupture. Manipulation of the treated joint was performed between 24 and 48 hours after collagenase injection under local anesthesia with 3 mL of 1% mepivacaine or lidocaine without epinephrine. After manipulation and cord rupture, patients were placed in a hand-based extension splint to wear at night for up to 3 months. Patients were followed at 1 and 12 months.

Results

Patients’ baseline characteristics are summarized in Table 1. Four patients were maintained on chronic prednisone therapy, 3 on methotrexate, and 1 on azathioprine. Therapy duration, medication dose, and diagnoses requiring immunosuppressant therapy varied among patients.

Outcomes and adverse events are summarized in Table 2. Mean number of joint contractures per hand treated was 2.8 (MP, 1.4; PIP, 1.4). However, not all joints met the intervention criteria. Of the 13 joints treated, 7 were MP joints, and 6 were PIP joints. Mean preinjection contracture of the treated joints was 53.0° (range, 20°-90°). Twelve of the 13 joint contractures improved. At mean follow-up of 6.7 months (range, 1-22 months), mean magnitude of contracture improved to 12.9° (range, 0°-45°). Mean MP joint contracture improved from 42.0° to 4.2° (range, 0°-10°), and mean PIP joint contracture improved from 65.8° to 21.7° (range, 0°-45°).

All 13 collagenase injections were well tolerated, and there were no systemic reactions. Injection-site pain was common. Mild injection-site bruising and edema were reported in all cases. Enzymatic fasciotomy was performed in all patients, and immediate improvement in contracture after manipulation 24 to 48 hours after injection was recorded.

Three of the 13 injections were complicated by skin tears during manipulation and cord rupture. All 3 skin tears were treated with local wound care, which included use of povidone-iodine and wet-to-dry dressings. There was no evidence of subsequent superficial or deep, local or regional infection. In 2 cases, the wound healed within 1 week; in the third case, wound healing was present by 2 weeks. Once the wounds showed early re-epithelialization, hand-based extension splinting in a position of comfort was used at night for up to 3 months after injection. Two of the 13 injections were complicated by small blood blisters. These were treated with observation and resolved spontaneously.

Discussion

Collagenase enzymatic fasciotomy appeared to be a safe and efficacious alternative to surgical treatment of Dupuytren contractures in this cohort of patients maintained on chronic immunosuppressive agents. MP contractures responded more substantially than PIP contractures did, as expected.⁶ No previously undescribed adverse outcomes were noted in these 8 patients on chronic immunosuppressive therapy beyond those reported in the CORD I trial. Three (23%) of the 13 collagenase injections in our series were complicated by skin tears after manipulation. Skins tears were reported in 22 (11%) of 204 patients after manual cord rupture in the CORD I trial.⁶ Given the limited numbers in this series, it remains unclear if chronic immunosuppression truly increases the risk of skin tears in this subset of patients. Other common treatment-related adverse events seen in the CORD I trial—injection-site hemorrhage (37%), pruritis (11%) and lymphadenopathy (10%)—were not seen after the 13 injections in our case series. We are prospectively following all patients with Dupuytren disease, and this is an area of ongoing research at our centers.

The immunosuppressive actions of prednisone, azathioprine, and methotrexate are well documented. Prednisone is a glucocorticoid, converted in the liver to prednisolone, which suppresses inflammation and immune responses by regulation of gene expression. Its immunosuppressive actions are multifactorial, relating to inhibition of lymphocytes, neutrophils, and monocytes. These effects are dose- and time-dependent¹¹ and may become evident in patients receiving low doses over prolonged periods. Skin atrophy¹² and delayed wound healing⁹ are side effects of long-term prednisone use. Skin atrophy may make the prednisone-treated patient more susceptible to skin tears after collagenase injection and manipulation. Azathioprine inhibits purine synthesis, which is especially important in the proliferation of immune cells.¹³ It has been shown to inhibit both cellular immunity at low doses and humoral immunity at higher doses.¹⁴ Methotrexate inhibits lymphocyte folic acid metabolism. The immunosuppressive properties of low-dose methotrexate have been linked to the induction of apoptosis in activated T cells.¹⁵

A more complex process in immunosuppressed patients is the immunogenicity of injected collagenase. As CHC in current use is a mixture of 2 foreign proteins, an immunologic response is expected in the host after injection. It has been shown that, after 3 injections of CHC into Dupuytren cords, 100% of patients developed antibodies to both enzymes in their serum.⁶ More than 85% demonstrated anti-CHC antibodies after a single injection. However, no patients showed signs of anaphylaxis or allergic reaction, and there was no correlation between serum levels of anti-CHC and adverse events. It has been hypothesized that there is a potential for cross-reactivity of the anti-CHC antibodies with human matrix metalloproteinases, causing enzymatic dysfunction within the host.¹⁶ This has yet to be reported clinically, and Xiaflex is currently under postmarketing surveillance. Immunocompromised people, with suppressed humoral and cellular immune responses, may produce less of an antibody response to the foreign CHC proteins. Whether this conclusively leads to a change in the side effect profile of the medication in these individuals is beyond the scope of this article. However, we identified no new side effects in this small but higher risk cohort. The issue should be continually monitored as collagenase is used in wider clinical settings.

Collagenase enzymatic fasciotomy is a new nonsurgical therapeutic option for Dupuytren disease. Indications and guidelines for use continue to evolve. This case series highlights the use of collagenase in 8 patients who were on long-term immunosuppressive therapy. This study has the limitations inherent to retrospective analyses. It is difficult to generalize results across broader immunosuppressed populations. A larger cohort, with long-term follow-up assessing recurrence of contracture, is needed to make definitive conclusions about use of collagenase in this challenging subset of patients. Based on our observations in this limited cohort, it appears appropriate to pursue further studies on use of collagenase enzymatic fasciotomy. A randomized, prospective or case–control series comparing surgical fasciectomy with enzymatic fasciotomy would yield further meaningful data. As more patients seek nonsurgical treatment for Dupuytren disease, its safety and efficacy in select cohorts of patients should continue to be evaluated.

The incidence of Dupuytren disease increases with advancing age,¹ as do the medical comorbidities of patients seeking treatment for disabling hand contractures. For patients with significant comorbidities, open surgical fasciectomy, the current standard of treatment for Dupuytren disease,^2,3 may be associated with increased perioperative risks.

Collagenase enzymatic fasciotomy has become an accepted nonsurgical treatment alternative to traditional fasciectomy or surgical fasciotomy for significant digital contractures caused by Dupuytren disease.^4-6 Clostridium histolyticum collagenase (CHC) is a foreign protein, made up of 2 collagenases isolated from the bacteria C histolyticum.⁷ The collagenases are zinc-dependent matrix metalloproteinases that cleave the triple helical structure of collagen molecules.⁸ Also known as Xiaflex (Auxilium Pharmaceuticals), CHC was approved by the US Food and Drug Administration (FDA) in February 2010 for use in patients with Dupuytren contractures.

 Enzymatic rupture is safe and efficacious at midterm follow-up and offers the theoretical advantage of avoiding palmar and digital fasciectomy and the associated risks of surgical-site infection and wound-healing complications.⁶ The risks of surgical wound complications are magnified in immunosuppressed patients, particularly those on chronic steroid therapy; wound-healing complication rates may be increased 2 to 5 times compared with controls.⁹ In a pooled literature review, wound-healing complications were reported after 22.9% of open primary fasciectomies, with infection occurring in 2.4%.¹⁰ A nonsurgical alternative is therefore particularly appealing for a patient cohort that may be at higher risk for a frequently described complication of surgery for Dupuytren contracture.

The exclusion criteria in the trials for FDA approval were extensive and included breast-feeding, pregnancy, bleeding disorder, recent stroke, use of tetracycline derivative within 14 days before start of study, use of anticoagulant within 7 days before start of study, allergy to collagenase, and chronic muscular, neurologic, or neuromuscular disorder affecting the hands.⁶ Safety and efficacy of collagenase in patients requiring chronic immunosuppressive therapy for medical comorbidities have not been previously documented. Furthermore, although skin tears were reported in 11% of patients after manual cord rupture in the CORD (Collagenase Option for the Reduction of Dupuytren’s) I trial,⁶ the likelihood of deep and superficial infection and delayed wound healing has not been quantitated.

In this article, we report on outcomes of 13 collagenase enzymatic fasciotomies performed in 8 patients who were on chronic immunosuppressive therapy.

Methods

Institutional review board approval was obtained at both academic hand surgery institutions. We retrospectively reviewed prospectively collected clinical data within our 2 centers’ databases of patients with Dupuytren disease. Eight patients on chronic immunosuppressive therapies treated with collagenase for metacarpophalangeal (MP) or proximal interphalangeal (PIP) joint contractures between February 2010 and December 2011 were identified. Three of these patients received collagenase injections into 2 or more separate Dupuytren cords at different encounters, resulting in a total of 13 individual collagenase enzymatic fasciotomies.

Collagenase injections were administered following CORD I trial protocol,⁶ except we injected Dupuytren cords crossing the PIP joint using a lateral approach to minimize risk of flexor tendon rupture. Manipulation of the treated joint was performed between 24 and 48 hours after collagenase injection under local anesthesia with 3 mL of 1% mepivacaine or lidocaine without epinephrine. After manipulation and cord rupture, patients were placed in a hand-based extension splint to wear at night for up to 3 months. Patients were followed at 1 and 12 months.

Results

Patients’ baseline characteristics are summarized in Table 1. Four patients were maintained on chronic prednisone therapy, 3 on methotrexate, and 1 on azathioprine. Therapy duration, medication dose, and diagnoses requiring immunosuppressant therapy varied among patients.

Outcomes and adverse events are summarized in Table 2. Mean number of joint contractures per hand treated was 2.8 (MP, 1.4; PIP, 1.4). However, not all joints met the intervention criteria. Of the 13 joints treated, 7 were MP joints, and 6 were PIP joints. Mean preinjection contracture of the treated joints was 53.0° (range, 20°-90°). Twelve of the 13 joint contractures improved. At mean follow-up of 6.7 months (range, 1-22 months), mean magnitude of contracture improved to 12.9° (range, 0°-45°). Mean MP joint contracture improved from 42.0° to 4.2° (range, 0°-10°), and mean PIP joint contracture improved from 65.8° to 21.7° (range, 0°-45°).

All 13 collagenase injections were well tolerated, and there were no systemic reactions. Injection-site pain was common. Mild injection-site bruising and edema were reported in all cases. Enzymatic fasciotomy was performed in all patients, and immediate improvement in contracture after manipulation 24 to 48 hours after injection was recorded.

Three of the 13 injections were complicated by skin tears during manipulation and cord rupture. All 3 skin tears were treated with local wound care, which included use of povidone-iodine and wet-to-dry dressings. There was no evidence of subsequent superficial or deep, local or regional infection. In 2 cases, the wound healed within 1 week; in the third case, wound healing was present by 2 weeks. Once the wounds showed early re-epithelialization, hand-based extension splinting in a position of comfort was used at night for up to 3 months after injection. Two of the 13 injections were complicated by small blood blisters. These were treated with observation and resolved spontaneously.

Discussion

Collagenase enzymatic fasciotomy appeared to be a safe and efficacious alternative to surgical treatment of Dupuytren contractures in this cohort of patients maintained on chronic immunosuppressive agents. MP contractures responded more substantially than PIP contractures did, as expected.⁶ No previously undescribed adverse outcomes were noted in these 8 patients on chronic immunosuppressive therapy beyond those reported in the CORD I trial. Three (23%) of the 13 collagenase injections in our series were complicated by skin tears after manipulation. Skins tears were reported in 22 (11%) of 204 patients after manual cord rupture in the CORD I trial.⁶ Given the limited numbers in this series, it remains unclear if chronic immunosuppression truly increases the risk of skin tears in this subset of patients. Other common treatment-related adverse events seen in the CORD I trial—injection-site hemorrhage (37%), pruritis (11%) and lymphadenopathy (10%)—were not seen after the 13 injections in our case series. We are prospectively following all patients with Dupuytren disease, and this is an area of ongoing research at our centers.

The immunosuppressive actions of prednisone, azathioprine, and methotrexate are well documented. Prednisone is a glucocorticoid, converted in the liver to prednisolone, which suppresses inflammation and immune responses by regulation of gene expression. Its immunosuppressive actions are multifactorial, relating to inhibition of lymphocytes, neutrophils, and monocytes. These effects are dose- and time-dependent¹¹ and may become evident in patients receiving low doses over prolonged periods. Skin atrophy¹² and delayed wound healing⁹ are side effects of long-term prednisone use. Skin atrophy may make the prednisone-treated patient more susceptible to skin tears after collagenase injection and manipulation. Azathioprine inhibits purine synthesis, which is especially important in the proliferation of immune cells.¹³ It has been shown to inhibit both cellular immunity at low doses and humoral immunity at higher doses.¹⁴ Methotrexate inhibits lymphocyte folic acid metabolism. The immunosuppressive properties of low-dose methotrexate have been linked to the induction of apoptosis in activated T cells.¹⁵

A more complex process in immunosuppressed patients is the immunogenicity of injected collagenase. As CHC in current use is a mixture of 2 foreign proteins, an immunologic response is expected in the host after injection. It has been shown that, after 3 injections of CHC into Dupuytren cords, 100% of patients developed antibodies to both enzymes in their serum.⁶ More than 85% demonstrated anti-CHC antibodies after a single injection. However, no patients showed signs of anaphylaxis or allergic reaction, and there was no correlation between serum levels of anti-CHC and adverse events. It has been hypothesized that there is a potential for cross-reactivity of the anti-CHC antibodies with human matrix metalloproteinases, causing enzymatic dysfunction within the host.¹⁶ This has yet to be reported clinically, and Xiaflex is currently under postmarketing surveillance. Immunocompromised people, with suppressed humoral and cellular immune responses, may produce less of an antibody response to the foreign CHC proteins. Whether this conclusively leads to a change in the side effect profile of the medication in these individuals is beyond the scope of this article. However, we identified no new side effects in this small but higher risk cohort. The issue should be continually monitored as collagenase is used in wider clinical settings.

Collagenase enzymatic fasciotomy is a new nonsurgical therapeutic option for Dupuytren disease. Indications and guidelines for use continue to evolve. This case series highlights the use of collagenase in 8 patients who were on long-term immunosuppressive therapy. This study has the limitations inherent to retrospective analyses. It is difficult to generalize results across broader immunosuppressed populations. A larger cohort, with long-term follow-up assessing recurrence of contracture, is needed to make definitive conclusions about use of collagenase in this challenging subset of patients. Based on our observations in this limited cohort, it appears appropriate to pursue further studies on use of collagenase enzymatic fasciotomy. A randomized, prospective or case–control series comparing surgical fasciectomy with enzymatic fasciotomy would yield further meaningful data. As more patients seek nonsurgical treatment for Dupuytren disease, its safety and efficacy in select cohorts of patients should continue to be evaluated.

The incidence of Dupuytren disease increases with advancing age,¹ as do the medical comorbidities of patients seeking treatment for disabling hand contractures. For patients with significant comorbidities, open surgical fasciectomy, the current standard of treatment for Dupuytren disease,^2,3 may be associated with increased perioperative risks.

Collagenase enzymatic fasciotomy has become an accepted nonsurgical treatment alternative to traditional fasciectomy or surgical fasciotomy for significant digital contractures caused by Dupuytren disease.^4-6 Clostridium histolyticum collagenase (CHC) is a foreign protein, made up of 2 collagenases isolated from the bacteria C histolyticum.⁷ The collagenases are zinc-dependent matrix metalloproteinases that cleave the triple helical structure of collagen molecules.⁸ Also known as Xiaflex (Auxilium Pharmaceuticals), CHC was approved by the US Food and Drug Administration (FDA) in February 2010 for use in patients with Dupuytren contractures.

 Enzymatic rupture is safe and efficacious at midterm follow-up and offers the theoretical advantage of avoiding palmar and digital fasciectomy and the associated risks of surgical-site infection and wound-healing complications.⁶ The risks of surgical wound complications are magnified in immunosuppressed patients, particularly those on chronic steroid therapy; wound-healing complication rates may be increased 2 to 5 times compared with controls.⁹ In a pooled literature review, wound-healing complications were reported after 22.9% of open primary fasciectomies, with infection occurring in 2.4%.¹⁰ A nonsurgical alternative is therefore particularly appealing for a patient cohort that may be at higher risk for a frequently described complication of surgery for Dupuytren contracture.

The exclusion criteria in the trials for FDA approval were extensive and included breast-feeding, pregnancy, bleeding disorder, recent stroke, use of tetracycline derivative within 14 days before start of study, use of anticoagulant within 7 days before start of study, allergy to collagenase, and chronic muscular, neurologic, or neuromuscular disorder affecting the hands.⁶ Safety and efficacy of collagenase in patients requiring chronic immunosuppressive therapy for medical comorbidities have not been previously documented. Furthermore, although skin tears were reported in 11% of patients after manual cord rupture in the CORD (Collagenase Option for the Reduction of Dupuytren’s) I trial,⁶ the likelihood of deep and superficial infection and delayed wound healing has not been quantitated.

In this article, we report on outcomes of 13 collagenase enzymatic fasciotomies performed in 8 patients who were on chronic immunosuppressive therapy.

Methods

Institutional review board approval was obtained at both academic hand surgery institutions. We retrospectively reviewed prospectively collected clinical data within our 2 centers’ databases of patients with Dupuytren disease. Eight patients on chronic immunosuppressive therapies treated with collagenase for metacarpophalangeal (MP) or proximal interphalangeal (PIP) joint contractures between February 2010 and December 2011 were identified. Three of these patients received collagenase injections into 2 or more separate Dupuytren cords at different encounters, resulting in a total of 13 individual collagenase enzymatic fasciotomies.

Collagenase injections were administered following CORD I trial protocol,⁶ except we injected Dupuytren cords crossing the PIP joint using a lateral approach to minimize risk of flexor tendon rupture. Manipulation of the treated joint was performed between 24 and 48 hours after collagenase injection under local anesthesia with 3 mL of 1% mepivacaine or lidocaine without epinephrine. After manipulation and cord rupture, patients were placed in a hand-based extension splint to wear at night for up to 3 months. Patients were followed at 1 and 12 months.

Results

Patients’ baseline characteristics are summarized in Table 1. Four patients were maintained on chronic prednisone therapy, 3 on methotrexate, and 1 on azathioprine. Therapy duration, medication dose, and diagnoses requiring immunosuppressant therapy varied among patients.

Outcomes and adverse events are summarized in Table 2. Mean number of joint contractures per hand treated was 2.8 (MP, 1.4; PIP, 1.4). However, not all joints met the intervention criteria. Of the 13 joints treated, 7 were MP joints, and 6 were PIP joints. Mean preinjection contracture of the treated joints was 53.0° (range, 20°-90°). Twelve of the 13 joint contractures improved. At mean follow-up of 6.7 months (range, 1-22 months), mean magnitude of contracture improved to 12.9° (range, 0°-45°). Mean MP joint contracture improved from 42.0° to 4.2° (range, 0°-10°), and mean PIP joint contracture improved from 65.8° to 21.7° (range, 0°-45°).

All 13 collagenase injections were well tolerated, and there were no systemic reactions. Injection-site pain was common. Mild injection-site bruising and edema were reported in all cases. Enzymatic fasciotomy was performed in all patients, and immediate improvement in contracture after manipulation 24 to 48 hours after injection was recorded.

Three of the 13 injections were complicated by skin tears during manipulation and cord rupture. All 3 skin tears were treated with local wound care, which included use of povidone-iodine and wet-to-dry dressings. There was no evidence of subsequent superficial or deep, local or regional infection. In 2 cases, the wound healed within 1 week; in the third case, wound healing was present by 2 weeks. Once the wounds showed early re-epithelialization, hand-based extension splinting in a position of comfort was used at night for up to 3 months after injection. Two of the 13 injections were complicated by small blood blisters. These were treated with observation and resolved spontaneously.

Discussion

Collagenase enzymatic fasciotomy appeared to be a safe and efficacious alternative to surgical treatment of Dupuytren contractures in this cohort of patients maintained on chronic immunosuppressive agents. MP contractures responded more substantially than PIP contractures did, as expected.⁶ No previously undescribed adverse outcomes were noted in these 8 patients on chronic immunosuppressive therapy beyond those reported in the CORD I trial. Three (23%) of the 13 collagenase injections in our series were complicated by skin tears after manipulation. Skins tears were reported in 22 (11%) of 204 patients after manual cord rupture in the CORD I trial.⁶ Given the limited numbers in this series, it remains unclear if chronic immunosuppression truly increases the risk of skin tears in this subset of patients. Other common treatment-related adverse events seen in the CORD I trial—injection-site hemorrhage (37%), pruritis (11%) and lymphadenopathy (10%)—were not seen after the 13 injections in our case series. We are prospectively following all patients with Dupuytren disease, and this is an area of ongoing research at our centers.

The immunosuppressive actions of prednisone, azathioprine, and methotrexate are well documented. Prednisone is a glucocorticoid, converted in the liver to prednisolone, which suppresses inflammation and immune responses by regulation of gene expression. Its immunosuppressive actions are multifactorial, relating to inhibition of lymphocytes, neutrophils, and monocytes. These effects are dose- and time-dependent¹¹ and may become evident in patients receiving low doses over prolonged periods. Skin atrophy¹² and delayed wound healing⁹ are side effects of long-term prednisone use. Skin atrophy may make the prednisone-treated patient more susceptible to skin tears after collagenase injection and manipulation. Azathioprine inhibits purine synthesis, which is especially important in the proliferation of immune cells.¹³ It has been shown to inhibit both cellular immunity at low doses and humoral immunity at higher doses.¹⁴ Methotrexate inhibits lymphocyte folic acid metabolism. The immunosuppressive properties of low-dose methotrexate have been linked to the induction of apoptosis in activated T cells.¹⁵

A more complex process in immunosuppressed patients is the immunogenicity of injected collagenase. As CHC in current use is a mixture of 2 foreign proteins, an immunologic response is expected in the host after injection. It has been shown that, after 3 injections of CHC into Dupuytren cords, 100% of patients developed antibodies to both enzymes in their serum.⁶ More than 85% demonstrated anti-CHC antibodies after a single injection. However, no patients showed signs of anaphylaxis or allergic reaction, and there was no correlation between serum levels of anti-CHC and adverse events. It has been hypothesized that there is a potential for cross-reactivity of the anti-CHC antibodies with human matrix metalloproteinases, causing enzymatic dysfunction within the host.¹⁶ This has yet to be reported clinically, and Xiaflex is currently under postmarketing surveillance. Immunocompromised people, with suppressed humoral and cellular immune responses, may produce less of an antibody response to the foreign CHC proteins. Whether this conclusively leads to a change in the side effect profile of the medication in these individuals is beyond the scope of this article. However, we identified no new side effects in this small but higher risk cohort. The issue should be continually monitored as collagenase is used in wider clinical settings.

Collagenase enzymatic fasciotomy is a new nonsurgical therapeutic option for Dupuytren disease. Indications and guidelines for use continue to evolve. This case series highlights the use of collagenase in 8 patients who were on long-term immunosuppressive therapy. This study has the limitations inherent to retrospective analyses. It is difficult to generalize results across broader immunosuppressed populations. A larger cohort, with long-term follow-up assessing recurrence of contracture, is needed to make definitive conclusions about use of collagenase in this challenging subset of patients. Based on our observations in this limited cohort, it appears appropriate to pursue further studies on use of collagenase enzymatic fasciotomy. A randomized, prospective or case–control series comparing surgical fasciectomy with enzymatic fasciotomy would yield further meaningful data. As more patients seek nonsurgical treatment for Dupuytren disease, its safety and efficacy in select cohorts of patients should continue to be evaluated.

Carpal tunnel syndrome (CTS) is a disorder characterized by entrapment of the median nerve at the wrist, which may lead to symptoms of pain, paresthesia, and, ultimately, thenar muscle atrophy. Surgical intervention is indicated with persistent or progressive symptoms despite nonoperative management. Timely surgical decompression aims to halt progression of this disorder and prevent permanent peripheral nerve injury.

Carpal tunnel release (CTR) is the most common hand and wrist surgery in the United States, with about 400,000 operations performed annually.^1,2 Several methods of decompressing the carpal tunnel have been described.³ These include standard open CTR (OCTR), mini-open approaches, and various endoscopic techniques. OCTR was initially described by Sir James Learmonth in 1933,⁴ and it remains the gold-standard surgical treatment for patients with symptomatic CTS. Uniform excellent results with high patient satisfaction and low complication rates have been reported in several series.^5-9 Common to all techniques is complete proximal-to-distal division of the transverse carpal ligament (TCL). Magnetic resonance imaging studies have shown that TCL transection and the resulting diastasis between the radial and ulnar leaflets cause a significant increase in the volume of the carpal tunnel, leading to decreased pressure.^10,11

Endoscopic CTR (ECTR) techniques were developed in an effort to reduce complications, scar sensitivity, and pillar pain and facilitate more rapid return to work.^12-17 Outcome studies have demonstrated that both open and endoscopic releases yield patient-reported subjective improvements over preoperative symptoms.^18-22 A randomized, controlled trial by Trumble and colleagues²³ in 2002 found that ECTR led to improved patient outcomes in the early postoperative period (first 3 months), though differences in outcomes were reduced at final follow-up. More recently (2007), a Cochrane review of 33 trials concluded there was no strong evidence favoring use of alternative techniques over OCTR.³ Further, OCTR has been found to be technically less demanding and associated with decreased complications and costs.²⁴

Indications

The benefit of median nerve decompression at the wrist for CTS is clear.^6,7 Indications for surgery in patients with CTS include persistent symptoms despite nonoperative treatment, objective sensory disturbance or motor weakness, and thenar atrophy. Symptomatic response to corticosteroid injection is predictive of success after carpal tunnel surgery.²⁵ More than 87% of patients who gain symptomatic relief from corticosteroid injection have an excellent surgical outcome.

Technique

OCTR allows direct visualization of the TCL and the distal volar forearm fascia (DVFF) and evaluation for the presence of anomalous branching patterns of the median nerve. OCTR traditionally was performed through a 4- to 5-cm longitudinal incision extending from the wrist crease proximally to the Kaplan cardinal line distally. The mini-open technique is identical with the exception of incision length. We routinely use a 2.5- to 3-cm incision. Regardless of incision length, each OCTR should proceed through the same reproducible steps.

We perform OCTR under tourniquet control. Choice of anesthesia is surgeon and patient preference. We prefer local anesthesia with conscious sedation. After conscious sedation is administered, we infiltrate the carpal tunnel and surrounding subcutaneous tissue with 10 mL of a 50:50 mixture of 0.5% bupivacaine and 1% lidocaine without epinephrine.

A 2.5- to 3-cm longitudinal incision is made along the axis of the radial border of the ring finger from the Kaplan cardinal line²⁶ and extending about 3 cm proximally toward the wrist flexion crease ulnar to the palmaris longus if present (Figure 1).

After the skin is incised longitudinally, the subcutaneous fat is mobilized and cutaneous sensory branches identified and protected. The underlying superficial palmar fascia is incised in line with the skin incision. The underlying midportion of the TCL is now visualized.

Transverse Carpal Ligament Release

Occasionally, the investing fascia along the ulnar edge of the thenar musculature is mobilized radialward (if the thenar musculature is well developed) to visualize the proximal limb of the TCL. Injury to any anomalous motor branch of the median nerve is avoided by directly visualizing and then incising the TCL (Figure 2). The TCL is incised along its ulnar border just radial to the hook of hamate from distal to proximal in line with the radial border of the ring finger. Staying near the ulnar attachment of the TCL keeps the plane of ligament division farther away from the median nerve and its recurrent motor branches. Although the ulnar neurovascular bundle typically resides ulnar to the hook of hamate in the canal of Guyon, the surgeon must be aware that it can be located radial to the hook in some instances.^27,28 In the elderly, the ulnar artery may be tortuous and enter the field and require retraction. The TCL is incised distally until the sentinel fat pad, which marks the superficial palmar arterial arch, is visualized. This bed of adipose tissue marks the distal edge of the TCL.²⁹

Proximally, subcutaneous tissues above the proximal limb of the TCL and DVFF are mobilized to about 2 cm proximal to the wrist flexion crease to create a plane for the fine long nasal turbinate speculum. The nasal turbinate speculum is then inserted into this plane above the proximal limb of the TCL and DVFF (Figure 3). Once inserted to the level of the confluence of the TCL and the DVFF, the speculum is opened.

Topside visualization is now encountered with the ulnar neurovascular bundle protected by the ulnar blade of the speculum. A long-handle scalpel is used to incise the TCL and the DVFF under direct visualization from proximal to distal in line with the previously completed distal release (Figure 4). As the nasal turbinate speculum is stretching the TCL and putting it under tension, the TCL can be heard splitting as it is being incised. Once the TCL and the DVFF are divided, the speculum is slowly closed and removed. Wide diastasis of the radial and ulnar leaflets of the TCL and the DVFF is directly visualized. Complete decompression of the median nerve from the distal forearm fascia to the superficial palmar arch is confirmed.

Adhesions between the undersurface of the radial leaflet and the flexor tendons and median nerve are mobilized. The median nerve is assessed for “hourglass” morphology or atrophy. The flexor tendons can be swept radialward with a free elevator to inspect the floor of the carpal tunnel. Flexor tenosynovectomy is not routinely performed. The incision is closed with interrupted simple sutures using 4-0 nylon.

Study Results

This study was conducted at Hand Surgery PC, Newton-Wellesley Hospital, Tufts University School of Medicine. Over a 10-month interval, 101 consecutive mini-OCTRs (63 right hands, 38 left hands) were performed with this proximal release modification in 88 patients (51 females, 37 males) by Dr. Ruchelsman and Dr. Belsky (Table). CTRs performed in the setting of wrist and/or carpal trauma were excluded. Mean age was 62.8 years. Mean follow-up was 11.3 weeks (~3 months). For isolated cases of CTR, mean tourniquet time was 16 minutes. CTS symptoms were relieved in all patients with a high degree of satisfaction as measured with history and examination findings at follow-up visits. There were no major complications (eg, infection, neural or vascular damage, severe residual pain). Four patients reported minor residual numbness in the fingers at latest follow-up but nevertheless had major improvement over preoperative baseline. These 4 patients had preoperative electromyograms or nerve conduction studies documenting the extent of their disease. There was 1 case of minor wound complication. Three weeks after surgery, the patient had a 1-cm wound opening, which closed with local wound care. The patient did not develop any drainage, infection, bleeding, or neurologic symptoms.

Discussion

Open release of the TCL—the gold standard of surgical treatment for CTS—produces reliable symptom relief in the vast majority of patients.^25,30 Given that the most common complication of carpal tunnel surgery is incomplete release of the TCL,^31,32 this technique, which uses a nasal turbinate speculum to better visualize the median nerve, could potentially reduce the reoperation rate. The nasal turbinate speculum allows the surgeon to see the confluence of the TCL and the DVFF. In addition, as the complete release can be visualized, there is minimal chance of injury.

The 2007 Cochrane review³ found no strong evidence supporting replacing OCTR with endoscopic techniques. Previous investigators have questioned the utility of ECTR given that it is higher in cost and more resource-intensive than OCTR^1,33,34 and is associated with higher rates of certain complications.^5,22,35-37 A 2004 meta-analysis of 13 randomized, controlled trials found a higher rate of reversible nerve damage with an odds ratio of 3.1 for ECTR versus OCTR.³⁵ A more recent (2006) review of more than 80 studies found transient neurapraxias in 1.45% of ECTR cases and 0.25% of OCTR cases.⁵ The same study reported overall complication rates (reversible and major neurovascular structural injuries) of 0.74% for OCTR and 1.63% for ECTR (P < .005). Another limitation of ECTR is that endoscopic techniques require a higher degree of surgical skill, which makes teaching residents and fellows more challenging.

The novel nasal turbinate speculum technique presented here is easily reproducible and allows first-time surgeons to visualize all important structures. Given that this technique does not require an endoscope or an endoscope-viewing tower, it is likely more cost-effective and requires less time for turnover between cases. Patients obtain good relief of their CTS symptoms with this technique, and most return to their daily activities within weeks after operation.

Carpal tunnel syndrome (CTS) is a disorder characterized by entrapment of the median nerve at the wrist, which may lead to symptoms of pain, paresthesia, and, ultimately, thenar muscle atrophy. Surgical intervention is indicated with persistent or progressive symptoms despite nonoperative management. Timely surgical decompression aims to halt progression of this disorder and prevent permanent peripheral nerve injury.

Carpal tunnel release (CTR) is the most common hand and wrist surgery in the United States, with about 400,000 operations performed annually.^1,2 Several methods of decompressing the carpal tunnel have been described.³ These include standard open CTR (OCTR), mini-open approaches, and various endoscopic techniques. OCTR was initially described by Sir James Learmonth in 1933,⁴ and it remains the gold-standard surgical treatment for patients with symptomatic CTS. Uniform excellent results with high patient satisfaction and low complication rates have been reported in several series.^5-9 Common to all techniques is complete proximal-to-distal division of the transverse carpal ligament (TCL). Magnetic resonance imaging studies have shown that TCL transection and the resulting diastasis between the radial and ulnar leaflets cause a significant increase in the volume of the carpal tunnel, leading to decreased pressure.^10,11

Endoscopic CTR (ECTR) techniques were developed in an effort to reduce complications, scar sensitivity, and pillar pain and facilitate more rapid return to work.^12-17 Outcome studies have demonstrated that both open and endoscopic releases yield patient-reported subjective improvements over preoperative symptoms.^18-22 A randomized, controlled trial by Trumble and colleagues²³ in 2002 found that ECTR led to improved patient outcomes in the early postoperative period (first 3 months), though differences in outcomes were reduced at final follow-up. More recently (2007), a Cochrane review of 33 trials concluded there was no strong evidence favoring use of alternative techniques over OCTR.³ Further, OCTR has been found to be technically less demanding and associated with decreased complications and costs.²⁴

Indications

The benefit of median nerve decompression at the wrist for CTS is clear.^6,7 Indications for surgery in patients with CTS include persistent symptoms despite nonoperative treatment, objective sensory disturbance or motor weakness, and thenar atrophy. Symptomatic response to corticosteroid injection is predictive of success after carpal tunnel surgery.²⁵ More than 87% of patients who gain symptomatic relief from corticosteroid injection have an excellent surgical outcome.

Technique

OCTR allows direct visualization of the TCL and the distal volar forearm fascia (DVFF) and evaluation for the presence of anomalous branching patterns of the median nerve. OCTR traditionally was performed through a 4- to 5-cm longitudinal incision extending from the wrist crease proximally to the Kaplan cardinal line distally. The mini-open technique is identical with the exception of incision length. We routinely use a 2.5- to 3-cm incision. Regardless of incision length, each OCTR should proceed through the same reproducible steps.

We perform OCTR under tourniquet control. Choice of anesthesia is surgeon and patient preference. We prefer local anesthesia with conscious sedation. After conscious sedation is administered, we infiltrate the carpal tunnel and surrounding subcutaneous tissue with 10 mL of a 50:50 mixture of 0.5% bupivacaine and 1% lidocaine without epinephrine.

A 2.5- to 3-cm longitudinal incision is made along the axis of the radial border of the ring finger from the Kaplan cardinal line²⁶ and extending about 3 cm proximally toward the wrist flexion crease ulnar to the palmaris longus if present (Figure 1).

After the skin is incised longitudinally, the subcutaneous fat is mobilized and cutaneous sensory branches identified and protected. The underlying superficial palmar fascia is incised in line with the skin incision. The underlying midportion of the TCL is now visualized.

Transverse Carpal Ligament Release

Occasionally, the investing fascia along the ulnar edge of the thenar musculature is mobilized radialward (if the thenar musculature is well developed) to visualize the proximal limb of the TCL. Injury to any anomalous motor branch of the median nerve is avoided by directly visualizing and then incising the TCL (Figure 2). The TCL is incised along its ulnar border just radial to the hook of hamate from distal to proximal in line with the radial border of the ring finger. Staying near the ulnar attachment of the TCL keeps the plane of ligament division farther away from the median nerve and its recurrent motor branches. Although the ulnar neurovascular bundle typically resides ulnar to the hook of hamate in the canal of Guyon, the surgeon must be aware that it can be located radial to the hook in some instances.^27,28 In the elderly, the ulnar artery may be tortuous and enter the field and require retraction. The TCL is incised distally until the sentinel fat pad, which marks the superficial palmar arterial arch, is visualized. This bed of adipose tissue marks the distal edge of the TCL.²⁹

Proximally, subcutaneous tissues above the proximal limb of the TCL and DVFF are mobilized to about 2 cm proximal to the wrist flexion crease to create a plane for the fine long nasal turbinate speculum. The nasal turbinate speculum is then inserted into this plane above the proximal limb of the TCL and DVFF (Figure 3). Once inserted to the level of the confluence of the TCL and the DVFF, the speculum is opened.

Topside visualization is now encountered with the ulnar neurovascular bundle protected by the ulnar blade of the speculum. A long-handle scalpel is used to incise the TCL and the DVFF under direct visualization from proximal to distal in line with the previously completed distal release (Figure 4). As the nasal turbinate speculum is stretching the TCL and putting it under tension, the TCL can be heard splitting as it is being incised. Once the TCL and the DVFF are divided, the speculum is slowly closed and removed. Wide diastasis of the radial and ulnar leaflets of the TCL and the DVFF is directly visualized. Complete decompression of the median nerve from the distal forearm fascia to the superficial palmar arch is confirmed.

Adhesions between the undersurface of the radial leaflet and the flexor tendons and median nerve are mobilized. The median nerve is assessed for “hourglass” morphology or atrophy. The flexor tendons can be swept radialward with a free elevator to inspect the floor of the carpal tunnel. Flexor tenosynovectomy is not routinely performed. The incision is closed with interrupted simple sutures using 4-0 nylon.

Study Results

This study was conducted at Hand Surgery PC, Newton-Wellesley Hospital, Tufts University School of Medicine. Over a 10-month interval, 101 consecutive mini-OCTRs (63 right hands, 38 left hands) were performed with this proximal release modification in 88 patients (51 females, 37 males) by Dr. Ruchelsman and Dr. Belsky (Table). CTRs performed in the setting of wrist and/or carpal trauma were excluded. Mean age was 62.8 years. Mean follow-up was 11.3 weeks (~3 months). For isolated cases of CTR, mean tourniquet time was 16 minutes. CTS symptoms were relieved in all patients with a high degree of satisfaction as measured with history and examination findings at follow-up visits. There were no major complications (eg, infection, neural or vascular damage, severe residual pain). Four patients reported minor residual numbness in the fingers at latest follow-up but nevertheless had major improvement over preoperative baseline. These 4 patients had preoperative electromyograms or nerve conduction studies documenting the extent of their disease. There was 1 case of minor wound complication. Three weeks after surgery, the patient had a 1-cm wound opening, which closed with local wound care. The patient did not develop any drainage, infection, bleeding, or neurologic symptoms.

Discussion

Open release of the TCL—the gold standard of surgical treatment for CTS—produces reliable symptom relief in the vast majority of patients.^25,30 Given that the most common complication of carpal tunnel surgery is incomplete release of the TCL,^31,32 this technique, which uses a nasal turbinate speculum to better visualize the median nerve, could potentially reduce the reoperation rate. The nasal turbinate speculum allows the surgeon to see the confluence of the TCL and the DVFF. In addition, as the complete release can be visualized, there is minimal chance of injury.

The 2007 Cochrane review³ found no strong evidence supporting replacing OCTR with endoscopic techniques. Previous investigators have questioned the utility of ECTR given that it is higher in cost and more resource-intensive than OCTR^1,33,34 and is associated with higher rates of certain complications.^5,22,35-37 A 2004 meta-analysis of 13 randomized, controlled trials found a higher rate of reversible nerve damage with an odds ratio of 3.1 for ECTR versus OCTR.³⁵ A more recent (2006) review of more than 80 studies found transient neurapraxias in 1.45% of ECTR cases and 0.25% of OCTR cases.⁵ The same study reported overall complication rates (reversible and major neurovascular structural injuries) of 0.74% for OCTR and 1.63% for ECTR (P < .005). Another limitation of ECTR is that endoscopic techniques require a higher degree of surgical skill, which makes teaching residents and fellows more challenging.

The novel nasal turbinate speculum technique presented here is easily reproducible and allows first-time surgeons to visualize all important structures. Given that this technique does not require an endoscope or an endoscope-viewing tower, it is likely more cost-effective and requires less time for turnover between cases. Patients obtain good relief of their CTS symptoms with this technique, and most return to their daily activities within weeks after operation.

Carpal tunnel syndrome (CTS) is a disorder characterized by entrapment of the median nerve at the wrist, which may lead to symptoms of pain, paresthesia, and, ultimately, thenar muscle atrophy. Surgical intervention is indicated with persistent or progressive symptoms despite nonoperative management. Timely surgical decompression aims to halt progression of this disorder and prevent permanent peripheral nerve injury.

Carpal tunnel release (CTR) is the most common hand and wrist surgery in the United States, with about 400,000 operations performed annually.^1,2 Several methods of decompressing the carpal tunnel have been described.³ These include standard open CTR (OCTR), mini-open approaches, and various endoscopic techniques. OCTR was initially described by Sir James Learmonth in 1933,⁴ and it remains the gold-standard surgical treatment for patients with symptomatic CTS. Uniform excellent results with high patient satisfaction and low complication rates have been reported in several series.^5-9 Common to all techniques is complete proximal-to-distal division of the transverse carpal ligament (TCL). Magnetic resonance imaging studies have shown that TCL transection and the resulting diastasis between the radial and ulnar leaflets cause a significant increase in the volume of the carpal tunnel, leading to decreased pressure.^10,11

Endoscopic CTR (ECTR) techniques were developed in an effort to reduce complications, scar sensitivity, and pillar pain and facilitate more rapid return to work.^12-17 Outcome studies have demonstrated that both open and endoscopic releases yield patient-reported subjective improvements over preoperative symptoms.^18-22 A randomized, controlled trial by Trumble and colleagues²³ in 2002 found that ECTR led to improved patient outcomes in the early postoperative period (first 3 months), though differences in outcomes were reduced at final follow-up. More recently (2007), a Cochrane review of 33 trials concluded there was no strong evidence favoring use of alternative techniques over OCTR.³ Further, OCTR has been found to be technically less demanding and associated with decreased complications and costs.²⁴

Indications

The benefit of median nerve decompression at the wrist for CTS is clear.^6,7 Indications for surgery in patients with CTS include persistent symptoms despite nonoperative treatment, objective sensory disturbance or motor weakness, and thenar atrophy. Symptomatic response to corticosteroid injection is predictive of success after carpal tunnel surgery.²⁵ More than 87% of patients who gain symptomatic relief from corticosteroid injection have an excellent surgical outcome.

Technique

OCTR allows direct visualization of the TCL and the distal volar forearm fascia (DVFF) and evaluation for the presence of anomalous branching patterns of the median nerve. OCTR traditionally was performed through a 4- to 5-cm longitudinal incision extending from the wrist crease proximally to the Kaplan cardinal line distally. The mini-open technique is identical with the exception of incision length. We routinely use a 2.5- to 3-cm incision. Regardless of incision length, each OCTR should proceed through the same reproducible steps.

We perform OCTR under tourniquet control. Choice of anesthesia is surgeon and patient preference. We prefer local anesthesia with conscious sedation. After conscious sedation is administered, we infiltrate the carpal tunnel and surrounding subcutaneous tissue with 10 mL of a 50:50 mixture of 0.5% bupivacaine and 1% lidocaine without epinephrine.

A 2.5- to 3-cm longitudinal incision is made along the axis of the radial border of the ring finger from the Kaplan cardinal line²⁶ and extending about 3 cm proximally toward the wrist flexion crease ulnar to the palmaris longus if present (Figure 1).

After the skin is incised longitudinally, the subcutaneous fat is mobilized and cutaneous sensory branches identified and protected. The underlying superficial palmar fascia is incised in line with the skin incision. The underlying midportion of the TCL is now visualized.

Transverse Carpal Ligament Release

Occasionally, the investing fascia along the ulnar edge of the thenar musculature is mobilized radialward (if the thenar musculature is well developed) to visualize the proximal limb of the TCL. Injury to any anomalous motor branch of the median nerve is avoided by directly visualizing and then incising the TCL (Figure 2). The TCL is incised along its ulnar border just radial to the hook of hamate from distal to proximal in line with the radial border of the ring finger. Staying near the ulnar attachment of the TCL keeps the plane of ligament division farther away from the median nerve and its recurrent motor branches. Although the ulnar neurovascular bundle typically resides ulnar to the hook of hamate in the canal of Guyon, the surgeon must be aware that it can be located radial to the hook in some instances.^27,28 In the elderly, the ulnar artery may be tortuous and enter the field and require retraction. The TCL is incised distally until the sentinel fat pad, which marks the superficial palmar arterial arch, is visualized. This bed of adipose tissue marks the distal edge of the TCL.²⁹

Proximally, subcutaneous tissues above the proximal limb of the TCL and DVFF are mobilized to about 2 cm proximal to the wrist flexion crease to create a plane for the fine long nasal turbinate speculum. The nasal turbinate speculum is then inserted into this plane above the proximal limb of the TCL and DVFF (Figure 3). Once inserted to the level of the confluence of the TCL and the DVFF, the speculum is opened.

Topside visualization is now encountered with the ulnar neurovascular bundle protected by the ulnar blade of the speculum. A long-handle scalpel is used to incise the TCL and the DVFF under direct visualization from proximal to distal in line with the previously completed distal release (Figure 4). As the nasal turbinate speculum is stretching the TCL and putting it under tension, the TCL can be heard splitting as it is being incised. Once the TCL and the DVFF are divided, the speculum is slowly closed and removed. Wide diastasis of the radial and ulnar leaflets of the TCL and the DVFF is directly visualized. Complete decompression of the median nerve from the distal forearm fascia to the superficial palmar arch is confirmed.

Adhesions between the undersurface of the radial leaflet and the flexor tendons and median nerve are mobilized. The median nerve is assessed for “hourglass” morphology or atrophy. The flexor tendons can be swept radialward with a free elevator to inspect the floor of the carpal tunnel. Flexor tenosynovectomy is not routinely performed. The incision is closed with interrupted simple sutures using 4-0 nylon.

Study Results

This study was conducted at Hand Surgery PC, Newton-Wellesley Hospital, Tufts University School of Medicine. Over a 10-month interval, 101 consecutive mini-OCTRs (63 right hands, 38 left hands) were performed with this proximal release modification in 88 patients (51 females, 37 males) by Dr. Ruchelsman and Dr. Belsky (Table). CTRs performed in the setting of wrist and/or carpal trauma were excluded. Mean age was 62.8 years. Mean follow-up was 11.3 weeks (~3 months). For isolated cases of CTR, mean tourniquet time was 16 minutes. CTS symptoms were relieved in all patients with a high degree of satisfaction as measured with history and examination findings at follow-up visits. There were no major complications (eg, infection, neural or vascular damage, severe residual pain). Four patients reported minor residual numbness in the fingers at latest follow-up but nevertheless had major improvement over preoperative baseline. These 4 patients had preoperative electromyograms or nerve conduction studies documenting the extent of their disease. There was 1 case of minor wound complication. Three weeks after surgery, the patient had a 1-cm wound opening, which closed with local wound care. The patient did not develop any drainage, infection, bleeding, or neurologic symptoms.

Discussion

Open release of the TCL—the gold standard of surgical treatment for CTS—produces reliable symptom relief in the vast majority of patients.^25,30 Given that the most common complication of carpal tunnel surgery is incomplete release of the TCL,^31,32 this technique, which uses a nasal turbinate speculum to better visualize the median nerve, could potentially reduce the reoperation rate. The nasal turbinate speculum allows the surgeon to see the confluence of the TCL and the DVFF. In addition, as the complete release can be visualized, there is minimal chance of injury.

The 2007 Cochrane review³ found no strong evidence supporting replacing OCTR with endoscopic techniques. Previous investigators have questioned the utility of ECTR given that it is higher in cost and more resource-intensive than OCTR^1,33,34 and is associated with higher rates of certain complications.^5,22,35-37 A 2004 meta-analysis of 13 randomized, controlled trials found a higher rate of reversible nerve damage with an odds ratio of 3.1 for ECTR versus OCTR.³⁵ A more recent (2006) review of more than 80 studies found transient neurapraxias in 1.45% of ECTR cases and 0.25% of OCTR cases.⁵ The same study reported overall complication rates (reversible and major neurovascular structural injuries) of 0.74% for OCTR and 1.63% for ECTR (P < .005). Another limitation of ECTR is that endoscopic techniques require a higher degree of surgical skill, which makes teaching residents and fellows more challenging.

The novel nasal turbinate speculum technique presented here is easily reproducible and allows first-time surgeons to visualize all important structures. Given that this technique does not require an endoscope or an endoscope-viewing tower, it is likely more cost-effective and requires less time for turnover between cases. Patients obtain good relief of their CTS symptoms with this technique, and most return to their daily activities within weeks after operation.

Fractures of the spinous process of the lower cervical spine or upper thoracic spine are frequently referred to as clay-shoveler’s fractures. Originally reported by Hall¹ in 1940, these fractures were described in workers in Australia who dug drains in clay soil and threw the clay overhead with long shovels. Occasionally, the mud would not release from the shovel, causing excess force to be transmitted to the supraspinous ligaments and resulting in a forceful avulsion fracture of one or multiple spinous processes. The few reports following the earliest description in the literature frequently describe the mechanism of injury as being athletic in nature.^2-4 The forceful contraction of the paraspinal and trapezius muscles on the supraspinous ligaments and the resultant attachment to the spinous processes make this a not uncommon injury during athletics, especially with a flexed position of the neck and shoulders. The resultant fracture or apophyseal avulsion is painful and often necessitates a visit to the physician, with plain films, computed tomography (CT) scans, or magnetic resonance imaging (MRI) confirming the diagnosis.⁵

Treatment of these fractures has not been well described, but frequently a period of rest followed by physical therapy will allow a return to activity. We present a series of adolescent athletes who developed nonunion of the fracture of the T1 spinous process with continued symptoms, despite rest and conservative therapy, and who underwent surgical excision of the ununited fragment.

Materials and Methods

We obtained institutional review board permission for this study and searched the surgical database between 2006 and 2013 for patients who had undergone resection of a spinous process nonunion. We collected demographic data on the patients, evaluated the radiographic studies, and reviewed operative reports and follow-up patient data.

Results

Dr. Hedequist operated on 3 patients with a spinous process nonunion over the study time period. The average age of the patients was 14 years; the location of the spinous process fracture was the T1 vertebra in all patients. Two patients sustained the injury while playing hockey and 1 during wrestling. The average duration of symptoms prior to operation was 10 months; all patients had seen physicians without a diagnosis prior to evaluation at out institution. All patients had a trial of physical therapy before surgery, and all had been unable to return to sport after injury secondary to pain.

Examination of all patients revealed pain directly over the fracture site and accentuated by forward flexion of the neck and shoulders. Evaluation of injury plain films revealed a fracture fragment in 2 patients (Figure 1). All 3 patients underwent MRI and CT scans confirming the diagnosis. MRI confirmed areas of increased signal at the tip of the T1 spinous process, with inflammation in the supraspinous ligament directly at that region (Figure 2). The CT scans confirmed the presence of a bony fragment correlating with the tip of the T1 spinous process (Figure 3).

Surgery was performed under general endotracheal anesthesia via a midline incision over the affected area down to the spinous process. The supraspinous ligament was opened revealing an easily identified and definable ununited ossicle, which was removed without taking down the interspinous ligament. All 3 nonunions were noted to be atrophic with no evidence of surrounding inflammatory tissue or bursa. The residual end of the spinous process was smoothed down with a rongeur. Standard closure was performed. There were no surgical complications.

All patients had complete relief of pain at follow-up; 1 patient returned to full sports activity at 6 weeks and the other 2 returned to full sports activity at 3 months. There was no loss of cervical motion or trapezial strength at follow-up. All patients voiced satisfaction with the decision for surgical intervention.

Discussion

Clay-shoveler’s fracture is an injury well known to orthopedists. This fracture is thought to be caused by a forceful contraction of the thoracic paraspinal and trapezial muscles, causing an avulsion fracture with pain and frequently a “pop” experienced by the patient.¹ Usually considered self-limiting injuries, treatment involves a period of rest and activity modification with occasional physical therapy. Return to sports has been reported with occasional pain but with patient satisfaction.^3,5,6

Our series of patients represent a group of adolescent athletes who sustained spinous process fractures of the T1 vertebra and, despite a significant period of rest and activity modification, were unable to return to sports given their pain. The examination of these patients revealed focal tenderness at the tip of the spinous process. The diagnosis is made clinically, with radiographic studies confirming the diagnosis. In our series of patients, MRI was the original modality used to confirm injury to the area, with hyperintensity seen in the area of the supraspinous ligament and tip of the spinous process. CT confirmed the nonunion and presence of an ossicle in all patients. Surgical exposure of that area easily exposed the ununited ossicle, which was removed in all patients.

To our knowledge, this is the first report in the literature describing surgical excision of an ununited spinous process fracture in adolescent athletes. The original descriptive case series by Hall¹ states “in the minds of surgeons who have seen many of these cases that early operative removal of the fragments is the proper routine treatment.” Since that original series, we have not found articles in the literature that support surgical removal; however, persistent symptoms after fracture are described.⁵ It is not surprising that these patients developed pain at the site of the fracture given the forces acting in that area. The trapezial and paraspinal muscles acting on that area are forceful and repetitive during activities, especially sports. All our patients had pain with attempts at activity and all had had a significant period of rest. In a recent article, this injury was described in adolescents without the patients having clear relief of symptoms despite a period of inactivity.⁵ While physical therapy is therapeutic in some patients experiencing pain, it can be a source of aggravation due to neck and shoulder motion and muscle contraction. It is not surprising that therapy would not help in most cases, as neck and shoulder motion and muscle contraction are the sources of continuing discomfort.

Clinical practice suggests that most patients with spinous process fractures will become pain-free; however, that is not universal. This series demonstrates that a small subset of patients with this injury will continue to have significant symptoms despite a period of rest. In those patients who desire a pain-free return to sports, we recommend consideration of surgical excision after confirmation of nonunion with radiographic studies. The inherent risks of surgical treatment are minimal with this procedure, and the benefits include return to pain-free sports activity, with the resultant physical and psychosocial benefits for adolescent athletes.

Fractures of the spinous process of the lower cervical spine or upper thoracic spine are frequently referred to as clay-shoveler’s fractures. Originally reported by Hall¹ in 1940, these fractures were described in workers in Australia who dug drains in clay soil and threw the clay overhead with long shovels. Occasionally, the mud would not release from the shovel, causing excess force to be transmitted to the supraspinous ligaments and resulting in a forceful avulsion fracture of one or multiple spinous processes. The few reports following the earliest description in the literature frequently describe the mechanism of injury as being athletic in nature.^2-4 The forceful contraction of the paraspinal and trapezius muscles on the supraspinous ligaments and the resultant attachment to the spinous processes make this a not uncommon injury during athletics, especially with a flexed position of the neck and shoulders. The resultant fracture or apophyseal avulsion is painful and often necessitates a visit to the physician, with plain films, computed tomography (CT) scans, or magnetic resonance imaging (MRI) confirming the diagnosis.⁵

Treatment of these fractures has not been well described, but frequently a period of rest followed by physical therapy will allow a return to activity. We present a series of adolescent athletes who developed nonunion of the fracture of the T1 spinous process with continued symptoms, despite rest and conservative therapy, and who underwent surgical excision of the ununited fragment.

Materials and Methods

We obtained institutional review board permission for this study and searched the surgical database between 2006 and 2013 for patients who had undergone resection of a spinous process nonunion. We collected demographic data on the patients, evaluated the radiographic studies, and reviewed operative reports and follow-up patient data.

Results

Dr. Hedequist operated on 3 patients with a spinous process nonunion over the study time period. The average age of the patients was 14 years; the location of the spinous process fracture was the T1 vertebra in all patients. Two patients sustained the injury while playing hockey and 1 during wrestling. The average duration of symptoms prior to operation was 10 months; all patients had seen physicians without a diagnosis prior to evaluation at out institution. All patients had a trial of physical therapy before surgery, and all had been unable to return to sport after injury secondary to pain.

Examination of all patients revealed pain directly over the fracture site and accentuated by forward flexion of the neck and shoulders. Evaluation of injury plain films revealed a fracture fragment in 2 patients (Figure 1). All 3 patients underwent MRI and CT scans confirming the diagnosis. MRI confirmed areas of increased signal at the tip of the T1 spinous process, with inflammation in the supraspinous ligament directly at that region (Figure 2). The CT scans confirmed the presence of a bony fragment correlating with the tip of the T1 spinous process (Figure 3).

Surgery was performed under general endotracheal anesthesia via a midline incision over the affected area down to the spinous process. The supraspinous ligament was opened revealing an easily identified and definable ununited ossicle, which was removed without taking down the interspinous ligament. All 3 nonunions were noted to be atrophic with no evidence of surrounding inflammatory tissue or bursa. The residual end of the spinous process was smoothed down with a rongeur. Standard closure was performed. There were no surgical complications.

All patients had complete relief of pain at follow-up; 1 patient returned to full sports activity at 6 weeks and the other 2 returned to full sports activity at 3 months. There was no loss of cervical motion or trapezial strength at follow-up. All patients voiced satisfaction with the decision for surgical intervention.

Discussion

Clay-shoveler’s fracture is an injury well known to orthopedists. This fracture is thought to be caused by a forceful contraction of the thoracic paraspinal and trapezial muscles, causing an avulsion fracture with pain and frequently a “pop” experienced by the patient.¹ Usually considered self-limiting injuries, treatment involves a period of rest and activity modification with occasional physical therapy. Return to sports has been reported with occasional pain but with patient satisfaction.^3,5,6

Our series of patients represent a group of adolescent athletes who sustained spinous process fractures of the T1 vertebra and, despite a significant period of rest and activity modification, were unable to return to sports given their pain. The examination of these patients revealed focal tenderness at the tip of the spinous process. The diagnosis is made clinically, with radiographic studies confirming the diagnosis. In our series of patients, MRI was the original modality used to confirm injury to the area, with hyperintensity seen in the area of the supraspinous ligament and tip of the spinous process. CT confirmed the nonunion and presence of an ossicle in all patients. Surgical exposure of that area easily exposed the ununited ossicle, which was removed in all patients.

To our knowledge, this is the first report in the literature describing surgical excision of an ununited spinous process fracture in adolescent athletes. The original descriptive case series by Hall¹ states “in the minds of surgeons who have seen many of these cases that early operative removal of the fragments is the proper routine treatment.” Since that original series, we have not found articles in the literature that support surgical removal; however, persistent symptoms after fracture are described.⁵ It is not surprising that these patients developed pain at the site of the fracture given the forces acting in that area. The trapezial and paraspinal muscles acting on that area are forceful and repetitive during activities, especially sports. All our patients had pain with attempts at activity and all had had a significant period of rest. In a recent article, this injury was described in adolescents without the patients having clear relief of symptoms despite a period of inactivity.⁵ While physical therapy is therapeutic in some patients experiencing pain, it can be a source of aggravation due to neck and shoulder motion and muscle contraction. It is not surprising that therapy would not help in most cases, as neck and shoulder motion and muscle contraction are the sources of continuing discomfort.

Clinical practice suggests that most patients with spinous process fractures will become pain-free; however, that is not universal. This series demonstrates that a small subset of patients with this injury will continue to have significant symptoms despite a period of rest. In those patients who desire a pain-free return to sports, we recommend consideration of surgical excision after confirmation of nonunion with radiographic studies. The inherent risks of surgical treatment are minimal with this procedure, and the benefits include return to pain-free sports activity, with the resultant physical and psychosocial benefits for adolescent athletes.

Fractures of the spinous process of the lower cervical spine or upper thoracic spine are frequently referred to as clay-shoveler’s fractures. Originally reported by Hall¹ in 1940, these fractures were described in workers in Australia who dug drains in clay soil and threw the clay overhead with long shovels. Occasionally, the mud would not release from the shovel, causing excess force to be transmitted to the supraspinous ligaments and resulting in a forceful avulsion fracture of one or multiple spinous processes. The few reports following the earliest description in the literature frequently describe the mechanism of injury as being athletic in nature.^2-4 The forceful contraction of the paraspinal and trapezius muscles on the supraspinous ligaments and the resultant attachment to the spinous processes make this a not uncommon injury during athletics, especially with a flexed position of the neck and shoulders. The resultant fracture or apophyseal avulsion is painful and often necessitates a visit to the physician, with plain films, computed tomography (CT) scans, or magnetic resonance imaging (MRI) confirming the diagnosis.⁵

Treatment of these fractures has not been well described, but frequently a period of rest followed by physical therapy will allow a return to activity. We present a series of adolescent athletes who developed nonunion of the fracture of the T1 spinous process with continued symptoms, despite rest and conservative therapy, and who underwent surgical excision of the ununited fragment.

Materials and Methods

We obtained institutional review board permission for this study and searched the surgical database between 2006 and 2013 for patients who had undergone resection of a spinous process nonunion. We collected demographic data on the patients, evaluated the radiographic studies, and reviewed operative reports and follow-up patient data.

Results

Dr. Hedequist operated on 3 patients with a spinous process nonunion over the study time period. The average age of the patients was 14 years; the location of the spinous process fracture was the T1 vertebra in all patients. Two patients sustained the injury while playing hockey and 1 during wrestling. The average duration of symptoms prior to operation was 10 months; all patients had seen physicians without a diagnosis prior to evaluation at out institution. All patients had a trial of physical therapy before surgery, and all had been unable to return to sport after injury secondary to pain.

Examination of all patients revealed pain directly over the fracture site and accentuated by forward flexion of the neck and shoulders. Evaluation of injury plain films revealed a fracture fragment in 2 patients (Figure 1). All 3 patients underwent MRI and CT scans confirming the diagnosis. MRI confirmed areas of increased signal at the tip of the T1 spinous process, with inflammation in the supraspinous ligament directly at that region (Figure 2). The CT scans confirmed the presence of a bony fragment correlating with the tip of the T1 spinous process (Figure 3).

Surgery was performed under general endotracheal anesthesia via a midline incision over the affected area down to the spinous process. The supraspinous ligament was opened revealing an easily identified and definable ununited ossicle, which was removed without taking down the interspinous ligament. All 3 nonunions were noted to be atrophic with no evidence of surrounding inflammatory tissue or bursa. The residual end of the spinous process was smoothed down with a rongeur. Standard closure was performed. There were no surgical complications.

All patients had complete relief of pain at follow-up; 1 patient returned to full sports activity at 6 weeks and the other 2 returned to full sports activity at 3 months. There was no loss of cervical motion or trapezial strength at follow-up. All patients voiced satisfaction with the decision for surgical intervention.

Discussion

Clay-shoveler’s fracture is an injury well known to orthopedists. This fracture is thought to be caused by a forceful contraction of the thoracic paraspinal and trapezial muscles, causing an avulsion fracture with pain and frequently a “pop” experienced by the patient.¹ Usually considered self-limiting injuries, treatment involves a period of rest and activity modification with occasional physical therapy. Return to sports has been reported with occasional pain but with patient satisfaction.^3,5,6

Our series of patients represent a group of adolescent athletes who sustained spinous process fractures of the T1 vertebra and, despite a significant period of rest and activity modification, were unable to return to sports given their pain. The examination of these patients revealed focal tenderness at the tip of the spinous process. The diagnosis is made clinically, with radiographic studies confirming the diagnosis. In our series of patients, MRI was the original modality used to confirm injury to the area, with hyperintensity seen in the area of the supraspinous ligament and tip of the spinous process. CT confirmed the nonunion and presence of an ossicle in all patients. Surgical exposure of that area easily exposed the ununited ossicle, which was removed in all patients.

To our knowledge, this is the first report in the literature describing surgical excision of an ununited spinous process fracture in adolescent athletes. The original descriptive case series by Hall¹ states “in the minds of surgeons who have seen many of these cases that early operative removal of the fragments is the proper routine treatment.” Since that original series, we have not found articles in the literature that support surgical removal; however, persistent symptoms after fracture are described.⁵ It is not surprising that these patients developed pain at the site of the fracture given the forces acting in that area. The trapezial and paraspinal muscles acting on that area are forceful and repetitive during activities, especially sports. All our patients had pain with attempts at activity and all had had a significant period of rest. In a recent article, this injury was described in adolescents without the patients having clear relief of symptoms despite a period of inactivity.⁵ While physical therapy is therapeutic in some patients experiencing pain, it can be a source of aggravation due to neck and shoulder motion and muscle contraction. It is not surprising that therapy would not help in most cases, as neck and shoulder motion and muscle contraction are the sources of continuing discomfort.

Clinical practice suggests that most patients with spinous process fractures will become pain-free; however, that is not universal. This series demonstrates that a small subset of patients with this injury will continue to have significant symptoms despite a period of rest. In those patients who desire a pain-free return to sports, we recommend consideration of surgical excision after confirmation of nonunion with radiographic studies. The inherent risks of surgical treatment are minimal with this procedure, and the benefits include return to pain-free sports activity, with the resultant physical and psychosocial benefits for adolescent athletes.

The responsibilities of team physicians have increased dramatically since the early 19th century, when these physicians first appeared on the sidelines during football games.¹ Although the primary role of the team physician is to care for the athlete, other responsibilities include administrative and legal duties, equipment- and environment-related duties, teaching, and communication with parents, coaches, and other physicians.^2-4 These responsibilities differ greatly by the level of the athlete and the team being covered. For example, compared with high school and collegiate sport physicians, physicians caring for professional athletes may have increased interaction with the media.⁵

Despite the increasing demands and responsibilities of team physicians, it is important that they continue to advance the field of sports medicine through teaching and research.^3,6 Team physicians have direct access to athletes at multiple levels of competition, from novice to professional, and therefore have a unique understanding of the injuries that commonly affect these athletes. Efforts to both teach and study the prevention, diagnosis, and treatment of these injuries have dramatically advanced the field of sports medicine. In fact, several advancements in sports medicine have come from team physicians, including advancements in anterior cruciate ligament reconstruction,^7,8 shoulder arthroscopy,⁹ and “Tommy John” surgery,¹⁰ to name a few.

Given the important role of team physicians (particularly orthopedic team physicians) in advancing sports medicine, it is important to understand the degree to which team physicians at all levels of sport contribute to teaching and research.

We conducted a study to determine the overall academic involvement of orthopedic team physicians at all levels of sport, including the degree to which these physicians are affiliated with academic medical centers (by level of sport and by professional sport) and the quantity and impact of these physicians’ scientific publications. We hypothesized that orthopedic physician academic involvement would be higher at the professional level of sport than at the collegiate or high school level and that the degree of physician academic involvement would differ between professional sporting leagues.

Materials and Methods

In August 2012, we performed a comprehensive telephone- and Internet-based search to identify a sample of team physicians caring for athletes at the high school, collegiate, and professional levels of sport. Data were collected on all team physicians, regardless of medical specialty. We defined a physician as any person listed as having either a doctor of medicine (MD) or a doctor of osteopathic medicine (DO) degree. A physician listed as a team physician at 2 different levels of competition (high school, college, professional) was included in both cohorts. A physician listed as a team physician in 2 different professional sports leagues was included independently for both leagues. All other medical personnel, including athletic trainers, therapists, and nursing staff, were excluded. Data on our sample population were collected as follows:

1. High school. Performing a comprehensive database search through the US Department of Education, we generated a list of all 20,989 US schools that include grades 9 to 12.¹¹ We then used a random number generator (random.org) to randomly select a sample of 120 high schools. These schools were contacted by telephone and asked to identify the team physician(s) for their sports teams. Twenty of these schools reported not having an athletic team, so we randomly generated a list of 20 additional high schools. High schools that had an athletic team but denied having a team physician were included in the analysis.

2. College. We used the National Collegiate Athletic Association (NCAA) website (ncaa.org) to generate a list of all colleges affiliated with the NCAA. Of these colleges, 347 were Division I, 316 were Division II, and 443 were Division III. The random.org random number generator was used to generate a list of 40 schools for each division, for a total of 120 schools. An Internet-based search was then performed to identify any and all team physicians caring for athletes at that particular school. In select cases, telephone calls were made to determine all the team physicians involved in the care of athletes at that institution.

3. Professional. Team physician data were collected for 4 of the most popular professional sporting leagues¹²: Major League Baseball (MLB), National Basketball Association (NBA), National Football League (NFL), and National Hockey League (NHL). Each team’s official website was identified through its league website (mlb.com, nba.com, nfl.com, nhl.com), and the roster or directory listing of all team physicians was recorded. In 2 cases, the team’s medical personnel listing could not be retrieved through the Internet, and a telephone call had to be made to identify all team physicians. Team physicians were identified for 122 professional teams: 30 MLB, 30 NBA, 32 NFL, and 30 NHL.

For this study, all physicians were classified as either orthopedic or nonorthopedic. Orthopedic surgeons—the focus of this study—were defined as those who completed residency training in orthopedic surgery. Median number of orthopedic and nonorthopedic surgeons per team was calculated at the high school, collegiate, and professional levels.

After identifying all orthopedic team physicians, we performed additional Internet searches to determine any affiliation between each physician and an applicable academic medical center. Physicians were placed in 1 of 3 different categories based on “level” of academic affiliation. Orthopedists with no identifiable connection to an academic medical center were listed under none. The first 100 search results were studied before this determination was made. Orthopedists with any academic affiliation below the level of full professorship were placed in the category associate/assistant/adjunct professor, which included any physician who was an associate professor, adjunct professor, clinical instructor, or volunteer instructor at an academic medical center. Last, orthopedists listed as full professors were placed in the professor category.

Number of publications written by each orthopedic team physician was then calculated using SciVerse Scopus (scopus.com), a comprehensive abstract and citation database of research literature that offers complete coverage of the Medline and Embase databases.¹³ Scopus offers a Scopus Author Identifier, which assigns each author in Scopus a unique identification number.¹⁴ This number is based on an “algorithm that matches author names based on their affiliation, address, subject area, source title, dates of publication citations, and co-authors.”¹⁴ Authors whose names did not appear in Scopus were assumed to have no publications, and this was reported after cross-referencing with Medline to ensure no documents were missed. This study included all publications: original research articles, reviews, letters, and commentaries. Any level of authorship (first, second, etc) was included. All publications were scanned, and duplicate listings were not included. Median number of publications per orthopedic team physician was calculated at the high school, college, and professional levels.

We also determined the h-index for each orthopedic team physician. The h-index is used to measure the impact of the published work of a scholar: “A scientist has index h if h of his/her papers have at least h citations each, and the other papers have no more than h citations each.”¹⁵ For example, an h-index of 12 means that, out of an author’s total number of publications, 12 have been cited at least 12 times, and all of his or her other publications have been cited fewer than 12 times. All authors in Scopus are automatically assigned h-indexes, and we collected these numbers.¹⁶ Of note, citations for articles published before 1996 are not included in the h-index calculation. Median h-index score per orthopedic team physician was calculated at the high school, college, and professional levels.

Analysis of variance was used to compare continuous data (eg, number of publications per surgeon) across different groups (eg, physicians from respective sports). Chi-square tests were used to detect whole-number differences between groups (eg, difference in number of physicians per team across the various professional sports leagues). Statistical significance was set at P < .05.

Results

We identified 1054 team physicians among the 362 total high schools, colleges, and professional sports teams included in this study. Of the 1054 physicians, 678 (64%) were orthopedic surgeons (Table 1). Seventy-two (60%) of the 120 high schools did not have a team physician, whereas all the colleges and professional teams did. Number of orthopedic surgeons per team was higher at the collegiate level (2.29; range, 0-11) and professional level (2.21; range, 1-9) than at the high school level (1.11; range, 0-24) (Table 1). Median number of nonorthopedic surgeons was highest in professional sports (1.88; range, 0-9) followed by college sports (1.06; range, 0-9) and high school sports (0.16; range, 0-2) (Table 1).

Of the 678 orthopedic team physicians, 298 (44%) were officially affiliated with an academic medical center, either as clinical instructor, associate/adjunct professor, or full professor. Percentage of orthopedists affiliated with an academic medical center was highest in professional sports (173/270, 64%) followed by collegiate sports (98/275, 36%) and high school sports (27/133, 20%) (P < .001, Table 2). Percentage of orthopedists identified as full professors was highest at the professional level (42/270, 16%) followed by the collegiate level (14/275, 5.1%) and the high school level (3/133, 2.3%) (P < .001, Table 2).

We found 12,036 publications written by the 678 orthopedic team physicians included in this study. Median number of publications per orthopedist was significantly higher in professional sports (30.6; range, 0-460) than in collegiate sports (10.7; range, 0-581) and high school sports (6.0; range, 0-220) (P < .001). Number of authors with more than 25 publications was highest at the professional level (82) followed by the collegiate level (27) and the high school level (7) (Table 3). Median number of publications per orthopedist was also higher at the professional level (12) than at the collegiate level (2) and high school level (1). Median h-index was higher among orthopedists in professional sports (7.1; range, 0-50) than at colleges (2.7; range, 0-63) and high schools (1.8; range, 0-32) (P < .001). Median h-index was also significantly higher at the professional level (5) than at the collegiate level (1) and high school level (0).

At the professional level of sports, we identified 499 team physicians (270 orthopedic, 54%; 229 nonorthopedic, 46%). Median number of orthopedic team physicians varied by sport, with MLB (2.8; range, 1-8) and the NFL (2.4; range, 1-4) having relatively more of these physicians than the NHL (2.0; range, 1-6) and the NBA (1.7; range, 1-9) (Table 4). Percentage of orthopedic team physicians affiliated with academic medical centers was highest in MLB (58/83, 69.9%) followed by the NFL (47/76, 61.8%), the NHL (37/60, 61.7%), and the NBA (31/51, 60.8%) (Table 5). Median number of publications by orthopedists also varied by sport, with the highest number in MLB (37.9; range, 0-225) followed by the NBA (32.0; range, 0-227) and the NFL (30.4; range, 0-460), with the lowest number in the NHL (20.7; range, 0-144) (Table 6). Median number of publications was the same (17.5) in MLB and the NFL and lower in the NBA (11) and the NHL (7.5). Median h-index was highest in the NFL (8.2; range, 0-50) and MLB (7.9; range, 0-32) followed by the NBA (6.6; range, 0-35) and the NHL (4.9; range, 0-20) (Table 7) Median h-index was the same (6) in MLB and the NFL and lower (3) in the NBA and the NHL.

Discussion

To our knowledge, this is the first study of academic involvement and the research activities of orthopedic team physicians at the high school, college, and professional levels of sport. We found that, on average, there were almost twice as many orthopedists at the collegiate and professional levels than at the high school level—likely because 72 of the 120 high schools randomly selected did not have a team physician, despite having sports teams. We can attribute this to the organizational structure of teams in a high school setting, where it is fairly common that no medically educated health care provider is readily available for the student athletes.⁵ Although the median number of orthopedists was similar at the collegiate and professional levels, the number of nonorthopedic team physicians was higher at the professional level than at the collegiate level. Although most collegiate and professional teams have an internist and an orthopedist on staff, medical staff at the professional level may also include several subspecialists from a variety of medical fields (eg, dental medicine, ophthalmology, neurology).¹⁷

We found that a significantly larger proportion of orthopedists at the professional level (64%) were affiliated with academic medical centers as associate/adjunct professors and full professors compared with orthopedists at the collegiate level (36%) and high school level (20%). The academic relationship with collegiate teams was much lower than expected. Regarding professional sports, however, this finding confirmed our hypothesis, and the explanation is likely multifactorial and historical. Moreover, the median number of publications was higher for orthopedists at the professional level (30.8) than at the collegiate level (10.7) and high school level (6). In the late 1940s and early 1950s, many orthopedic team physicians entered into contracts with major universities.⁴ For many physicians, this contractual relationship increased their prestige, and some orthopedic groups were alleged to have endorsed scholarships at those schools.⁴ Given the high level of publicity and scrutiny surrounding medical decisions at the professional level of sports, it is possible that professional sports teams specifically seek orthopedists who are well respected within academia. Moreover, contracts between universities/academic medical centers and professional teams may mandate that a faculty member from that organization provide the orthopedic/medical care for the team. This may also increase the likelihood of professional teams being paired with academic orthopedic physicians. However, such contractual agreements are made between professional teams and large private medical groups as well.

In addition to measuring quantity of publications, we used the h-index to measure their quality. Following the same pattern as the publication rate, median h-index per orthopedic team physician was significantly higher at the professional level (7.1) than at the collegiate level (2.7) and high school level (1.8). As with publication volume, this is not entirely surprising, as h-index has been shown to correlate with academic rank in other surgical specialties,¹⁸ and there was a higher percentage of academic physicians at the professional level than at the collegiate and high school levels.

At the professional level of sports, 56% of all team physicians were orthopedic surgeons. Orthopedists caring for MLB teams had the highest median number of publications (37.9), followed by the NBA (32.0), the NFL (30.4), and the NHL (20.7). One likely explanation is the higher percentage of MLB physicians affiliated with academic medical centers. Regarding the h-index, MLB and NFL physicians had the highest values (7.9 and 8.2, respectively).

Our study had several limitations. First, we may not have captured data on all the team physicians at the high school, college, and professional levels. By following a detailed protocol in identifying surgeons, however, we tried to minimize the impact of any such omissions. In addition, teams may have had many unofficial consultants acting as team physicians, whether orthopedic or nonorthopedic, and, if these physicians were not listed in an official capacity, they may have been omitted from this study. We further realize that a true measure of academic productivity should also include book chapters and books published, research grants awarded, and patents registered. By including only peer-reviewed articles, we omitted these other criteria.

To our knowledge, the data presented here represent the first attempt to quantify the academic involvement and research productivity of orthopedic team physicians at the high school, college, and professional levels of sport. These data help us understand how research productivity varies by orthopedic team physicians at different levels of sport and may be useful to those considering a career as a team physician, as they can better evaluate their own productivity in the context of team physicians across different levels of competition.

The responsibilities of team physicians have increased dramatically since the early 19th century, when these physicians first appeared on the sidelines during football games.¹ Although the primary role of the team physician is to care for the athlete, other responsibilities include administrative and legal duties, equipment- and environment-related duties, teaching, and communication with parents, coaches, and other physicians.^2-4 These responsibilities differ greatly by the level of the athlete and the team being covered. For example, compared with high school and collegiate sport physicians, physicians caring for professional athletes may have increased interaction with the media.⁵

Despite the increasing demands and responsibilities of team physicians, it is important that they continue to advance the field of sports medicine through teaching and research.^3,6 Team physicians have direct access to athletes at multiple levels of competition, from novice to professional, and therefore have a unique understanding of the injuries that commonly affect these athletes. Efforts to both teach and study the prevention, diagnosis, and treatment of these injuries have dramatically advanced the field of sports medicine. In fact, several advancements in sports medicine have come from team physicians, including advancements in anterior cruciate ligament reconstruction,^7,8 shoulder arthroscopy,⁹ and “Tommy John” surgery,¹⁰ to name a few.

Given the important role of team physicians (particularly orthopedic team physicians) in advancing sports medicine, it is important to understand the degree to which team physicians at all levels of sport contribute to teaching and research.

We conducted a study to determine the overall academic involvement of orthopedic team physicians at all levels of sport, including the degree to which these physicians are affiliated with academic medical centers (by level of sport and by professional sport) and the quantity and impact of these physicians’ scientific publications. We hypothesized that orthopedic physician academic involvement would be higher at the professional level of sport than at the collegiate or high school level and that the degree of physician academic involvement would differ between professional sporting leagues.

Materials and Methods

In August 2012, we performed a comprehensive telephone- and Internet-based search to identify a sample of team physicians caring for athletes at the high school, collegiate, and professional levels of sport. Data were collected on all team physicians, regardless of medical specialty. We defined a physician as any person listed as having either a doctor of medicine (MD) or a doctor of osteopathic medicine (DO) degree. A physician listed as a team physician at 2 different levels of competition (high school, college, professional) was included in both cohorts. A physician listed as a team physician in 2 different professional sports leagues was included independently for both leagues. All other medical personnel, including athletic trainers, therapists, and nursing staff, were excluded. Data on our sample population were collected as follows:

1. High school. Performing a comprehensive database search through the US Department of Education, we generated a list of all 20,989 US schools that include grades 9 to 12.¹¹ We then used a random number generator (random.org) to randomly select a sample of 120 high schools. These schools were contacted by telephone and asked to identify the team physician(s) for their sports teams. Twenty of these schools reported not having an athletic team, so we randomly generated a list of 20 additional high schools. High schools that had an athletic team but denied having a team physician were included in the analysis.

2. College. We used the National Collegiate Athletic Association (NCAA) website (ncaa.org) to generate a list of all colleges affiliated with the NCAA. Of these colleges, 347 were Division I, 316 were Division II, and 443 were Division III. The random.org random number generator was used to generate a list of 40 schools for each division, for a total of 120 schools. An Internet-based search was then performed to identify any and all team physicians caring for athletes at that particular school. In select cases, telephone calls were made to determine all the team physicians involved in the care of athletes at that institution.

3. Professional. Team physician data were collected for 4 of the most popular professional sporting leagues¹²: Major League Baseball (MLB), National Basketball Association (NBA), National Football League (NFL), and National Hockey League (NHL). Each team’s official website was identified through its league website (mlb.com, nba.com, nfl.com, nhl.com), and the roster or directory listing of all team physicians was recorded. In 2 cases, the team’s medical personnel listing could not be retrieved through the Internet, and a telephone call had to be made to identify all team physicians. Team physicians were identified for 122 professional teams: 30 MLB, 30 NBA, 32 NFL, and 30 NHL.

For this study, all physicians were classified as either orthopedic or nonorthopedic. Orthopedic surgeons—the focus of this study—were defined as those who completed residency training in orthopedic surgery. Median number of orthopedic and nonorthopedic surgeons per team was calculated at the high school, collegiate, and professional levels.

After identifying all orthopedic team physicians, we performed additional Internet searches to determine any affiliation between each physician and an applicable academic medical center. Physicians were placed in 1 of 3 different categories based on “level” of academic affiliation. Orthopedists with no identifiable connection to an academic medical center were listed under none. The first 100 search results were studied before this determination was made. Orthopedists with any academic affiliation below the level of full professorship were placed in the category associate/assistant/adjunct professor, which included any physician who was an associate professor, adjunct professor, clinical instructor, or volunteer instructor at an academic medical center. Last, orthopedists listed as full professors were placed in the professor category.

Number of publications written by each orthopedic team physician was then calculated using SciVerse Scopus (scopus.com), a comprehensive abstract and citation database of research literature that offers complete coverage of the Medline and Embase databases.¹³ Scopus offers a Scopus Author Identifier, which assigns each author in Scopus a unique identification number.¹⁴ This number is based on an “algorithm that matches author names based on their affiliation, address, subject area, source title, dates of publication citations, and co-authors.”¹⁴ Authors whose names did not appear in Scopus were assumed to have no publications, and this was reported after cross-referencing with Medline to ensure no documents were missed. This study included all publications: original research articles, reviews, letters, and commentaries. Any level of authorship (first, second, etc) was included. All publications were scanned, and duplicate listings were not included. Median number of publications per orthopedic team physician was calculated at the high school, college, and professional levels.

We also determined the h-index for each orthopedic team physician. The h-index is used to measure the impact of the published work of a scholar: “A scientist has index h if h of his/her papers have at least h citations each, and the other papers have no more than h citations each.”¹⁵ For example, an h-index of 12 means that, out of an author’s total number of publications, 12 have been cited at least 12 times, and all of his or her other publications have been cited fewer than 12 times. All authors in Scopus are automatically assigned h-indexes, and we collected these numbers.¹⁶ Of note, citations for articles published before 1996 are not included in the h-index calculation. Median h-index score per orthopedic team physician was calculated at the high school, college, and professional levels.

Analysis of variance was used to compare continuous data (eg, number of publications per surgeon) across different groups (eg, physicians from respective sports). Chi-square tests were used to detect whole-number differences between groups (eg, difference in number of physicians per team across the various professional sports leagues). Statistical significance was set at P < .05.

Results

We identified 1054 team physicians among the 362 total high schools, colleges, and professional sports teams included in this study. Of the 1054 physicians, 678 (64%) were orthopedic surgeons (Table 1). Seventy-two (60%) of the 120 high schools did not have a team physician, whereas all the colleges and professional teams did. Number of orthopedic surgeons per team was higher at the collegiate level (2.29; range, 0-11) and professional level (2.21; range, 1-9) than at the high school level (1.11; range, 0-24) (Table 1). Median number of nonorthopedic surgeons was highest in professional sports (1.88; range, 0-9) followed by college sports (1.06; range, 0-9) and high school sports (0.16; range, 0-2) (Table 1).

Of the 678 orthopedic team physicians, 298 (44%) were officially affiliated with an academic medical center, either as clinical instructor, associate/adjunct professor, or full professor. Percentage of orthopedists affiliated with an academic medical center was highest in professional sports (173/270, 64%) followed by collegiate sports (98/275, 36%) and high school sports (27/133, 20%) (P < .001, Table 2). Percentage of orthopedists identified as full professors was highest at the professional level (42/270, 16%) followed by the collegiate level (14/275, 5.1%) and the high school level (3/133, 2.3%) (P < .001, Table 2).

We found 12,036 publications written by the 678 orthopedic team physicians included in this study. Median number of publications per orthopedist was significantly higher in professional sports (30.6; range, 0-460) than in collegiate sports (10.7; range, 0-581) and high school sports (6.0; range, 0-220) (P < .001). Number of authors with more than 25 publications was highest at the professional level (82) followed by the collegiate level (27) and the high school level (7) (Table 3). Median number of publications per orthopedist was also higher at the professional level (12) than at the collegiate level (2) and high school level (1). Median h-index was higher among orthopedists in professional sports (7.1; range, 0-50) than at colleges (2.7; range, 0-63) and high schools (1.8; range, 0-32) (P < .001). Median h-index was also significantly higher at the professional level (5) than at the collegiate level (1) and high school level (0).

At the professional level of sports, we identified 499 team physicians (270 orthopedic, 54%; 229 nonorthopedic, 46%). Median number of orthopedic team physicians varied by sport, with MLB (2.8; range, 1-8) and the NFL (2.4; range, 1-4) having relatively more of these physicians than the NHL (2.0; range, 1-6) and the NBA (1.7; range, 1-9) (Table 4). Percentage of orthopedic team physicians affiliated with academic medical centers was highest in MLB (58/83, 69.9%) followed by the NFL (47/76, 61.8%), the NHL (37/60, 61.7%), and the NBA (31/51, 60.8%) (Table 5). Median number of publications by orthopedists also varied by sport, with the highest number in MLB (37.9; range, 0-225) followed by the NBA (32.0; range, 0-227) and the NFL (30.4; range, 0-460), with the lowest number in the NHL (20.7; range, 0-144) (Table 6). Median number of publications was the same (17.5) in MLB and the NFL and lower in the NBA (11) and the NHL (7.5). Median h-index was highest in the NFL (8.2; range, 0-50) and MLB (7.9; range, 0-32) followed by the NBA (6.6; range, 0-35) and the NHL (4.9; range, 0-20) (Table 7) Median h-index was the same (6) in MLB and the NFL and lower (3) in the NBA and the NHL.

Discussion

To our knowledge, this is the first study of academic involvement and the research activities of orthopedic team physicians at the high school, college, and professional levels of sport. We found that, on average, there were almost twice as many orthopedists at the collegiate and professional levels than at the high school level—likely because 72 of the 120 high schools randomly selected did not have a team physician, despite having sports teams. We can attribute this to the organizational structure of teams in a high school setting, where it is fairly common that no medically educated health care provider is readily available for the student athletes.⁵ Although the median number of orthopedists was similar at the collegiate and professional levels, the number of nonorthopedic team physicians was higher at the professional level than at the collegiate level. Although most collegiate and professional teams have an internist and an orthopedist on staff, medical staff at the professional level may also include several subspecialists from a variety of medical fields (eg, dental medicine, ophthalmology, neurology).¹⁷

We found that a significantly larger proportion of orthopedists at the professional level (64%) were affiliated with academic medical centers as associate/adjunct professors and full professors compared with orthopedists at the collegiate level (36%) and high school level (20%). The academic relationship with collegiate teams was much lower than expected. Regarding professional sports, however, this finding confirmed our hypothesis, and the explanation is likely multifactorial and historical. Moreover, the median number of publications was higher for orthopedists at the professional level (30.8) than at the collegiate level (10.7) and high school level (6). In the late 1940s and early 1950s, many orthopedic team physicians entered into contracts with major universities.⁴ For many physicians, this contractual relationship increased their prestige, and some orthopedic groups were alleged to have endorsed scholarships at those schools.⁴ Given the high level of publicity and scrutiny surrounding medical decisions at the professional level of sports, it is possible that professional sports teams specifically seek orthopedists who are well respected within academia. Moreover, contracts between universities/academic medical centers and professional teams may mandate that a faculty member from that organization provide the orthopedic/medical care for the team. This may also increase the likelihood of professional teams being paired with academic orthopedic physicians. However, such contractual agreements are made between professional teams and large private medical groups as well.

In addition to measuring quantity of publications, we used the h-index to measure their quality. Following the same pattern as the publication rate, median h-index per orthopedic team physician was significantly higher at the professional level (7.1) than at the collegiate level (2.7) and high school level (1.8). As with publication volume, this is not entirely surprising, as h-index has been shown to correlate with academic rank in other surgical specialties,¹⁸ and there was a higher percentage of academic physicians at the professional level than at the collegiate and high school levels.

At the professional level of sports, 56% of all team physicians were orthopedic surgeons. Orthopedists caring for MLB teams had the highest median number of publications (37.9), followed by the NBA (32.0), the NFL (30.4), and the NHL (20.7). One likely explanation is the higher percentage of MLB physicians affiliated with academic medical centers. Regarding the h-index, MLB and NFL physicians had the highest values (7.9 and 8.2, respectively).

Our study had several limitations. First, we may not have captured data on all the team physicians at the high school, college, and professional levels. By following a detailed protocol in identifying surgeons, however, we tried to minimize the impact of any such omissions. In addition, teams may have had many unofficial consultants acting as team physicians, whether orthopedic or nonorthopedic, and, if these physicians were not listed in an official capacity, they may have been omitted from this study. We further realize that a true measure of academic productivity should also include book chapters and books published, research grants awarded, and patents registered. By including only peer-reviewed articles, we omitted these other criteria.

To our knowledge, the data presented here represent the first attempt to quantify the academic involvement and research productivity of orthopedic team physicians at the high school, college, and professional levels of sport. These data help us understand how research productivity varies by orthopedic team physicians at different levels of sport and may be useful to those considering a career as a team physician, as they can better evaluate their own productivity in the context of team physicians across different levels of competition.

The responsibilities of team physicians have increased dramatically since the early 19th century, when these physicians first appeared on the sidelines during football games.¹ Although the primary role of the team physician is to care for the athlete, other responsibilities include administrative and legal duties, equipment- and environment-related duties, teaching, and communication with parents, coaches, and other physicians.^2-4 These responsibilities differ greatly by the level of the athlete and the team being covered. For example, compared with high school and collegiate sport physicians, physicians caring for professional athletes may have increased interaction with the media.⁵

Despite the increasing demands and responsibilities of team physicians, it is important that they continue to advance the field of sports medicine through teaching and research.^3,6 Team physicians have direct access to athletes at multiple levels of competition, from novice to professional, and therefore have a unique understanding of the injuries that commonly affect these athletes. Efforts to both teach and study the prevention, diagnosis, and treatment of these injuries have dramatically advanced the field of sports medicine. In fact, several advancements in sports medicine have come from team physicians, including advancements in anterior cruciate ligament reconstruction,^7,8 shoulder arthroscopy,⁹ and “Tommy John” surgery,¹⁰ to name a few.

Given the important role of team physicians (particularly orthopedic team physicians) in advancing sports medicine, it is important to understand the degree to which team physicians at all levels of sport contribute to teaching and research.

We conducted a study to determine the overall academic involvement of orthopedic team physicians at all levels of sport, including the degree to which these physicians are affiliated with academic medical centers (by level of sport and by professional sport) and the quantity and impact of these physicians’ scientific publications. We hypothesized that orthopedic physician academic involvement would be higher at the professional level of sport than at the collegiate or high school level and that the degree of physician academic involvement would differ between professional sporting leagues.

Materials and Methods

In August 2012, we performed a comprehensive telephone- and Internet-based search to identify a sample of team physicians caring for athletes at the high school, collegiate, and professional levels of sport. Data were collected on all team physicians, regardless of medical specialty. We defined a physician as any person listed as having either a doctor of medicine (MD) or a doctor of osteopathic medicine (DO) degree. A physician listed as a team physician at 2 different levels of competition (high school, college, professional) was included in both cohorts. A physician listed as a team physician in 2 different professional sports leagues was included independently for both leagues. All other medical personnel, including athletic trainers, therapists, and nursing staff, were excluded. Data on our sample population were collected as follows:

1. High school. Performing a comprehensive database search through the US Department of Education, we generated a list of all 20,989 US schools that include grades 9 to 12.¹¹ We then used a random number generator (random.org) to randomly select a sample of 120 high schools. These schools were contacted by telephone and asked to identify the team physician(s) for their sports teams. Twenty of these schools reported not having an athletic team, so we randomly generated a list of 20 additional high schools. High schools that had an athletic team but denied having a team physician were included in the analysis.

2. College. We used the National Collegiate Athletic Association (NCAA) website (ncaa.org) to generate a list of all colleges affiliated with the NCAA. Of these colleges, 347 were Division I, 316 were Division II, and 443 were Division III. The random.org random number generator was used to generate a list of 40 schools for each division, for a total of 120 schools. An Internet-based search was then performed to identify any and all team physicians caring for athletes at that particular school. In select cases, telephone calls were made to determine all the team physicians involved in the care of athletes at that institution.

3. Professional. Team physician data were collected for 4 of the most popular professional sporting leagues¹²: Major League Baseball (MLB), National Basketball Association (NBA), National Football League (NFL), and National Hockey League (NHL). Each team’s official website was identified through its league website (mlb.com, nba.com, nfl.com, nhl.com), and the roster or directory listing of all team physicians was recorded. In 2 cases, the team’s medical personnel listing could not be retrieved through the Internet, and a telephone call had to be made to identify all team physicians. Team physicians were identified for 122 professional teams: 30 MLB, 30 NBA, 32 NFL, and 30 NHL.

For this study, all physicians were classified as either orthopedic or nonorthopedic. Orthopedic surgeons—the focus of this study—were defined as those who completed residency training in orthopedic surgery. Median number of orthopedic and nonorthopedic surgeons per team was calculated at the high school, collegiate, and professional levels.

After identifying all orthopedic team physicians, we performed additional Internet searches to determine any affiliation between each physician and an applicable academic medical center. Physicians were placed in 1 of 3 different categories based on “level” of academic affiliation. Orthopedists with no identifiable connection to an academic medical center were listed under none. The first 100 search results were studied before this determination was made. Orthopedists with any academic affiliation below the level of full professorship were placed in the category associate/assistant/adjunct professor, which included any physician who was an associate professor, adjunct professor, clinical instructor, or volunteer instructor at an academic medical center. Last, orthopedists listed as full professors were placed in the professor category.

Number of publications written by each orthopedic team physician was then calculated using SciVerse Scopus (scopus.com), a comprehensive abstract and citation database of research literature that offers complete coverage of the Medline and Embase databases.¹³ Scopus offers a Scopus Author Identifier, which assigns each author in Scopus a unique identification number.¹⁴ This number is based on an “algorithm that matches author names based on their affiliation, address, subject area, source title, dates of publication citations, and co-authors.”¹⁴ Authors whose names did not appear in Scopus were assumed to have no publications, and this was reported after cross-referencing with Medline to ensure no documents were missed. This study included all publications: original research articles, reviews, letters, and commentaries. Any level of authorship (first, second, etc) was included. All publications were scanned, and duplicate listings were not included. Median number of publications per orthopedic team physician was calculated at the high school, college, and professional levels.

We also determined the h-index for each orthopedic team physician. The h-index is used to measure the impact of the published work of a scholar: “A scientist has index h if h of his/her papers have at least h citations each, and the other papers have no more than h citations each.”¹⁵ For example, an h-index of 12 means that, out of an author’s total number of publications, 12 have been cited at least 12 times, and all of his or her other publications have been cited fewer than 12 times. All authors in Scopus are automatically assigned h-indexes, and we collected these numbers.¹⁶ Of note, citations for articles published before 1996 are not included in the h-index calculation. Median h-index score per orthopedic team physician was calculated at the high school, college, and professional levels.

Analysis of variance was used to compare continuous data (eg, number of publications per surgeon) across different groups (eg, physicians from respective sports). Chi-square tests were used to detect whole-number differences between groups (eg, difference in number of physicians per team across the various professional sports leagues). Statistical significance was set at P < .05.

Results

We identified 1054 team physicians among the 362 total high schools, colleges, and professional sports teams included in this study. Of the 1054 physicians, 678 (64%) were orthopedic surgeons (Table 1). Seventy-two (60%) of the 120 high schools did not have a team physician, whereas all the colleges and professional teams did. Number of orthopedic surgeons per team was higher at the collegiate level (2.29; range, 0-11) and professional level (2.21; range, 1-9) than at the high school level (1.11; range, 0-24) (Table 1). Median number of nonorthopedic surgeons was highest in professional sports (1.88; range, 0-9) followed by college sports (1.06; range, 0-9) and high school sports (0.16; range, 0-2) (Table 1).

Of the 678 orthopedic team physicians, 298 (44%) were officially affiliated with an academic medical center, either as clinical instructor, associate/adjunct professor, or full professor. Percentage of orthopedists affiliated with an academic medical center was highest in professional sports (173/270, 64%) followed by collegiate sports (98/275, 36%) and high school sports (27/133, 20%) (P < .001, Table 2). Percentage of orthopedists identified as full professors was highest at the professional level (42/270, 16%) followed by the collegiate level (14/275, 5.1%) and the high school level (3/133, 2.3%) (P < .001, Table 2).

We found 12,036 publications written by the 678 orthopedic team physicians included in this study. Median number of publications per orthopedist was significantly higher in professional sports (30.6; range, 0-460) than in collegiate sports (10.7; range, 0-581) and high school sports (6.0; range, 0-220) (P < .001). Number of authors with more than 25 publications was highest at the professional level (82) followed by the collegiate level (27) and the high school level (7) (Table 3). Median number of publications per orthopedist was also higher at the professional level (12) than at the collegiate level (2) and high school level (1). Median h-index was higher among orthopedists in professional sports (7.1; range, 0-50) than at colleges (2.7; range, 0-63) and high schools (1.8; range, 0-32) (P < .001). Median h-index was also significantly higher at the professional level (5) than at the collegiate level (1) and high school level (0).

At the professional level of sports, we identified 499 team physicians (270 orthopedic, 54%; 229 nonorthopedic, 46%). Median number of orthopedic team physicians varied by sport, with MLB (2.8; range, 1-8) and the NFL (2.4; range, 1-4) having relatively more of these physicians than the NHL (2.0; range, 1-6) and the NBA (1.7; range, 1-9) (Table 4). Percentage of orthopedic team physicians affiliated with academic medical centers was highest in MLB (58/83, 69.9%) followed by the NFL (47/76, 61.8%), the NHL (37/60, 61.7%), and the NBA (31/51, 60.8%) (Table 5). Median number of publications by orthopedists also varied by sport, with the highest number in MLB (37.9; range, 0-225) followed by the NBA (32.0; range, 0-227) and the NFL (30.4; range, 0-460), with the lowest number in the NHL (20.7; range, 0-144) (Table 6). Median number of publications was the same (17.5) in MLB and the NFL and lower in the NBA (11) and the NHL (7.5). Median h-index was highest in the NFL (8.2; range, 0-50) and MLB (7.9; range, 0-32) followed by the NBA (6.6; range, 0-35) and the NHL (4.9; range, 0-20) (Table 7) Median h-index was the same (6) in MLB and the NFL and lower (3) in the NBA and the NHL.

Discussion

To our knowledge, this is the first study of academic involvement and the research activities of orthopedic team physicians at the high school, college, and professional levels of sport. We found that, on average, there were almost twice as many orthopedists at the collegiate and professional levels than at the high school level—likely because 72 of the 120 high schools randomly selected did not have a team physician, despite having sports teams. We can attribute this to the organizational structure of teams in a high school setting, where it is fairly common that no medically educated health care provider is readily available for the student athletes.⁵ Although the median number of orthopedists was similar at the collegiate and professional levels, the number of nonorthopedic team physicians was higher at the professional level than at the collegiate level. Although most collegiate and professional teams have an internist and an orthopedist on staff, medical staff at the professional level may also include several subspecialists from a variety of medical fields (eg, dental medicine, ophthalmology, neurology).¹⁷

We found that a significantly larger proportion of orthopedists at the professional level (64%) were affiliated with academic medical centers as associate/adjunct professors and full professors compared with orthopedists at the collegiate level (36%) and high school level (20%). The academic relationship with collegiate teams was much lower than expected. Regarding professional sports, however, this finding confirmed our hypothesis, and the explanation is likely multifactorial and historical. Moreover, the median number of publications was higher for orthopedists at the professional level (30.8) than at the collegiate level (10.7) and high school level (6). In the late 1940s and early 1950s, many orthopedic team physicians entered into contracts with major universities.⁴ For many physicians, this contractual relationship increased their prestige, and some orthopedic groups were alleged to have endorsed scholarships at those schools.⁴ Given the high level of publicity and scrutiny surrounding medical decisions at the professional level of sports, it is possible that professional sports teams specifically seek orthopedists who are well respected within academia. Moreover, contracts between universities/academic medical centers and professional teams may mandate that a faculty member from that organization provide the orthopedic/medical care for the team. This may also increase the likelihood of professional teams being paired with academic orthopedic physicians. However, such contractual agreements are made between professional teams and large private medical groups as well.

In addition to measuring quantity of publications, we used the h-index to measure their quality. Following the same pattern as the publication rate, median h-index per orthopedic team physician was significantly higher at the professional level (7.1) than at the collegiate level (2.7) and high school level (1.8). As with publication volume, this is not entirely surprising, as h-index has been shown to correlate with academic rank in other surgical specialties,¹⁸ and there was a higher percentage of academic physicians at the professional level than at the collegiate and high school levels.

At the professional level of sports, 56% of all team physicians were orthopedic surgeons. Orthopedists caring for MLB teams had the highest median number of publications (37.9), followed by the NBA (32.0), the NFL (30.4), and the NHL (20.7). One likely explanation is the higher percentage of MLB physicians affiliated with academic medical centers. Regarding the h-index, MLB and NFL physicians had the highest values (7.9 and 8.2, respectively).

Our study had several limitations. First, we may not have captured data on all the team physicians at the high school, college, and professional levels. By following a detailed protocol in identifying surgeons, however, we tried to minimize the impact of any such omissions. In addition, teams may have had many unofficial consultants acting as team physicians, whether orthopedic or nonorthopedic, and, if these physicians were not listed in an official capacity, they may have been omitted from this study. We further realize that a true measure of academic productivity should also include book chapters and books published, research grants awarded, and patents registered. By including only peer-reviewed articles, we omitted these other criteria.

To our knowledge, the data presented here represent the first attempt to quantify the academic involvement and research productivity of orthopedic team physicians at the high school, college, and professional levels of sport. These data help us understand how research productivity varies by orthopedic team physicians at different levels of sport and may be useful to those considering a career as a team physician, as they can better evaluate their own productivity in the context of team physicians across different levels of competition.

As defined by the American Academy of Orthopaedic Surgeons (AAOS) in 1996, conflict of interest (COI) is the “circumstance that exists when, because of personal financial gain, an individual has the potential to be less than objective when called on to reach a judgment or interpret a result.”¹ In medical research, COIs often occur in relationships between physician-researchers and pharmaceutical, medical device, and biotechnology companies. These relationships usually take the form of research grants but can also arise when the researcher has a financial interest in the product being tested or in the company that manufactures the product.

 Although constructive collaboration between academic medicine and industry has worked to improve health care and ultimately benefit patients, potential drawbacks of such relationships include sequestration and suppression of results that may be disadvantageous to the industry sponsor,² increased likelihood of reporting positive results (pro-industry),^3-7 and biased study designs.⁸ The nature of such relationships may threaten the integrity of scientific studies, the objectivity of medical education, the quality of patient care, and the public’s trust in medicine.⁹

Financial relationships and affiliations are increasing as we seek to answer a growing number of clinical questions—with funding often being a limiting factor. At national scientific meetings, the number of presentations reporting COIs reflects this trend. Paper and poster presentations accepted for annual meetings of the Orthopaedic Trauma Association (OTA) and reporting a COI increased from 7.6% in 1993 to 12.6% in 2002 (P = .0129).²

Medical subspecialties outside of orthopedics are experiencing similar trends. Most notable is the American Psychiatric Association (APA). After the APA published a mandatory financial COI disclosure policy in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), the percentage of task force members reporting industry relationships increased by 12%.¹⁰ Analysis of the DSM-5 panels demonstrated that the panels with the largest percentage of reported COIs are those for which pharmacological treatment is the first-line intervention, including the panels for mood disorders (67%), psychotic disorders (83%) and sleep/wake disorders (100%).¹⁰ Moreover, the industry ties reported are to the pharmaceutical companies that manufacture the medications used to treat these disorders or to companies that service the pharmaceutical industry.¹⁰

The degree to which financial COIs affect the interpretation of the orthopedic literature has never been quantified. Although it is clear that COIs can confound the results and reporting of data, how the medical community uses disclosures when interpreting the literature and when formulating opinions that may or may not affect their practice patterns is largely unknown.

We conducted a study to evaluate how a hypothetical financial COI disclosure would influence the interpretation of data by orthopedic clinicians. We also wanted to determine the reliability of the data as perceived in association with different study designs, levels of evidence, research institutional settings, and reporting of positive or negative results.

Methods

We asked members of the Arthroscopy Association of North America (AANA) and the American Orthopaedic Society for Sports Medicine (AOSSM) to complete a multiple-choice situational questionnaire (Table). The questionnaire assesses the degree to which respondents use COI disclosures when interpreting the literature. It further explores the perceived clinical value of a study with a given reported COI, assuming variations in study design, research institutional setting, and significance of results. The fictional research team disclosed the project was funded by a pharmaceutical company and all team members received consulting compensation. The survey and study were reviewed and approved by our institutional review board. The survey consisted of 14 multiple-choice questions that allowed for only 1 answer selection per person and allowed survey takers to skip questions they did not wish to answer. The survey questions and associated response options appear in edited form in the Table. A link to the questionnaire (https://www.surveymonkey.com/s/MPCCLCX) was sent with a message explaining the study. The responses to the questionnaire constituted the data.

Results

We sent a request to participate in the survey to 750 physicians and received 522 responses (overall response rate, 70%). The response rate for each question equaled or exceeded 98%.

The majority of respondents (95.6%) were male. Ninety-nine percent of respondents were orthopedic surgeons. The Northeast (US) was the most common geographical practice location of respondents (32%), followed by the Midwest (19.1%) and the Southeast (16.6%). Most respondents (40%) had been in practice for more than 20 years; 67% had been in practice a minimum of 10 years. The majority (68.8%) were employed by private practice groups, either single specialty (57.8%) or multispecialty (11%).

Eighty percent of respondents strongly agreed that COI disclosure is important when interpreting study results, 62% reported always reading the disclosure slide during academy or other meeting presentations, and 41% reported always using this information when deciding how to interpret scientific data.

Seventy-five percent of respondents thought the study—an academic-center case series with significant results in favor of the pharmaceutical company funding the study—was biased (42% indicated biased with merit, 33% biased without merit). Twenty-three percent thought the study was possibly biased, but likely trustworthy given the academic institutional affiliation. When the study setting was changed to community hospital, 95% thought the study was biased (51% biased with merit, 44% biased without merit). With the same study performed at an academic center, and no statistically significant results (not in favor of the pharmaceutical company funding the study), 88% thought the study had merit (46% biased with merit, 42% unbiased with merit).

When the study design was changed to a randomized controlled trial (level I evidence) conducted at an academic center with negative results, an overwhelming 95% of respondents thought the study had merit (33% biased with merit, 62% unbiased with merit). Given the same study design at an academic center, with positive results, 78% still thought the study had merit (39% biased with merit, 39% unbiased with merit). An additional 18% thought the study was biased, but still likely trustworthy given the academic institutional affiliation. Finally, given a randomized controlled trial and positive results, but with the research setting a small community practice, 90% thought the study had merit (51% biased with merit, 39% unbiased with merit). The percentage of respondents who found the study biased and likely without merit increased from 3.7% to 9.5% when the institutional affiliation changed from academic to community.

Discussion

As governmental funding sources become increasingly limited, the role of industry sponsorship of orthopedic research has grown. Potential drawbacks and biases of such research support have been well described—most notably, increased positive result reporting, suppression of results that may be disadvantageous to the industry sponsor, and biased study designs.^2-8 However, the extent to which financial COIs affect the orthopedic medical community’s interpretation of the literature has never been quantified. To our knowledge, the present study is the first to quantify the impact of reported COI on study interpretation.

Our goal was to examine how reported financial COIs influence the interpretation of the literature by the orthopedic medical community. Moreover, we wanted to determine the perceived reliability of the data when variables (study design, institutional affiliation, positive vs negative results) were changed. The results of our survey indicate that, when a financial COI is reported, study reliability is perceived as highest when negative results were found.

Our survey noted a discrepancy between the documented importance of the hypothetical research team’s COI disclosure and the use of such disclosures when interpreting study results. Eighty percent of respondents agreed that COI disclosure is important when interpreting study results, but only 62% reported always reading disclosures, and even fewer (41%) reported always using the information when interpreting results. It is unclear exactly why this trend exists, as one would expect the percentages to be more similar. These particular survey questions were formed around using COI disclosures when interpreting study results during academic presentations at national meetings and not during the review of published literature. It is possible that positioning the COI disclosure at the beginning of a presentation has an effect, but only 3.7% of respondents indicated they seldom remembered the disclosure by the end of the presentation. The results of our survey may have varied if the questions had targeted reading and interpreting the literature.

Interestingly, the results of these survey questions tended to be more consistent with rates of reported financial COI by presenters at national orthopedic meetings. A study published in the New England Journal of Medicine found that less than 80% of orthopedic surgeons reported their disclosures at a large annual meeting (AAOS), even when the disclosure involved payments pertinent to the research they were presenting.⁵ When the payments were indirectly related to the research, the percentage of surgeons reporting disclosures was 50%, almost the same as the disclosure rate for unrelated payments.⁵

When the study was changed to a level I randomized controlled trial, more survey respondents found it to be less biased and have more merit. Although it would seem intuitive for a study with a higher level of evidence to carry more clinical value during interpretation, this may not hold true in the setting of industry-sponsored clinical trials. Several studies have documented a significant association between the reporting of positive results and industry-sponsored randomized clinical trials. In 2008, Khan and colleagues³ examined 100 orthopedic randomized clinical trials reported in 5 major orthopedic subspecialty journals over a 2-year period. The association between industry funding and favorable outcome in all original randomized clinical trials was strong and significant (P < .001). This is not surprising, given the amount of time and money required for a well-designed clinical study. Commercial products with preclinical promise are pushed to testing in a clinical trial, whereas resources would not be wasted on products lacking preclinical merit.

The most important variable affecting interpretation of study merit by survey respondents was the reporting of negative results. As more researchers are developing COIs, many studies are discovering a relationship between COIs and outcomes of research studies. Reviewing the adult total joint literature, Ezzet⁸ found an industry funding rate of 50%. Positive results were reported in 93% of cases in commercially funded studies versus 37% of cases in independently funded studies. Furthermore, no negative results were reported by investigators who were receiving royalties from the respective companies.

Studies across the medical literature have also found this association between industry sponsorship and reporting of positive results. One such study, reported by Valachis and colleagues⁷ in the Journal of Clinical Oncology, examined more than 80 economic analyses of targeted oncologic therapies and found the studies funded by pharmaceutical companies were more likely to report favorable qualitative cost estimates. In addition, when studies with a COI disclosure were examined, those reporting any financial relationship with a manufacturer (eg, author affiliation, funding) were more likely than those without such a relationship to report favorable results.

Our study had several limitations. First, as most of the survey respondents were orthopedic surgeons, extrapolating their data to the medical community at large may not be appropriate, as each specialty may view industry affiliations differently. In addition, respondents were asked to base their interpretations of a study on conclusions we predetermined—no direct visualization of the data set or statistical testing methods. It is possible that these responses may have been different had the respondents had the opportunity to further evaluate the study in question. In a recent study, Altwairgi and colleagues¹¹ found that 10% of randomized clinical trials involving lung cancer treatment were reported with different conclusions in their full manuscripts relative to their abstracts. We think our survey design perhaps best mimics an annual meeting environment in which participants have very limited ability to interpret studies and may rely more heavily on the factors we investigated—study design, significance of findings, and setting, all similar to information presented in an abstract—when making informed decisions. Although our response rate was only 70%, this is comparable to or better than the rates in similar survey studies that used email-based questionnaires.^12,13

Another limitation was that our survey may have forced respondents into answers they did not entirely agree with, given the limited options of the multiple-choice response format and the specific wording of the questions. Our conclusions may have been more dramatic when we were evaluating whether the study was deemed meritorious or not. However, there is no adopted standard for evaluating the extent of bias perceived by a clinician. We thought it was important to include answer options indicating a study had merit despite obvious bias in design and execution. That a study had merit can mean different things. It may change clinical practice, may require further study and reproducibility, or may not be significant enough to matter. Asking follow-up questions to evaluate this perception among the respondents could have provided validity to the term merit. Further studies in this field are needed to determine how studies are interpreted and translated into clinical practice by various clinicians.

Conclusion

Although the present study is not a quantitative analysis of the determination of bias in the orthopedic community, it is the first to evaluate orthopedic surgeons’ perceptions on the basis of key fundamentals of orthopedic research relative to COI. It is clear from our study results that introducing levels of evidence to the orthopedic milieu has had a significant impact both on the quality of research and on the foundational use of deductive reasoning when interpreting the literature. Reporting negative outcomes is perhaps the most important factor in eliminating the perception of bias among orthopedic surgeons. To what extent a perceived COI plays into medical decision-making and the ultimate treatment of patients is still relatively unknown.

As defined by the American Academy of Orthopaedic Surgeons (AAOS) in 1996, conflict of interest (COI) is the “circumstance that exists when, because of personal financial gain, an individual has the potential to be less than objective when called on to reach a judgment or interpret a result.”¹ In medical research, COIs often occur in relationships between physician-researchers and pharmaceutical, medical device, and biotechnology companies. These relationships usually take the form of research grants but can also arise when the researcher has a financial interest in the product being tested or in the company that manufactures the product.

 Although constructive collaboration between academic medicine and industry has worked to improve health care and ultimately benefit patients, potential drawbacks of such relationships include sequestration and suppression of results that may be disadvantageous to the industry sponsor,² increased likelihood of reporting positive results (pro-industry),^3-7 and biased study designs.⁸ The nature of such relationships may threaten the integrity of scientific studies, the objectivity of medical education, the quality of patient care, and the public’s trust in medicine.⁹

Financial relationships and affiliations are increasing as we seek to answer a growing number of clinical questions—with funding often being a limiting factor. At national scientific meetings, the number of presentations reporting COIs reflects this trend. Paper and poster presentations accepted for annual meetings of the Orthopaedic Trauma Association (OTA) and reporting a COI increased from 7.6% in 1993 to 12.6% in 2002 (P = .0129).²

Medical subspecialties outside of orthopedics are experiencing similar trends. Most notable is the American Psychiatric Association (APA). After the APA published a mandatory financial COI disclosure policy in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), the percentage of task force members reporting industry relationships increased by 12%.¹⁰ Analysis of the DSM-5 panels demonstrated that the panels with the largest percentage of reported COIs are those for which pharmacological treatment is the first-line intervention, including the panels for mood disorders (67%), psychotic disorders (83%) and sleep/wake disorders (100%).¹⁰ Moreover, the industry ties reported are to the pharmaceutical companies that manufacture the medications used to treat these disorders or to companies that service the pharmaceutical industry.¹⁰

The degree to which financial COIs affect the interpretation of the orthopedic literature has never been quantified. Although it is clear that COIs can confound the results and reporting of data, how the medical community uses disclosures when interpreting the literature and when formulating opinions that may or may not affect their practice patterns is largely unknown.

We conducted a study to evaluate how a hypothetical financial COI disclosure would influence the interpretation of data by orthopedic clinicians. We also wanted to determine the reliability of the data as perceived in association with different study designs, levels of evidence, research institutional settings, and reporting of positive or negative results.

Methods

We asked members of the Arthroscopy Association of North America (AANA) and the American Orthopaedic Society for Sports Medicine (AOSSM) to complete a multiple-choice situational questionnaire (Table). The questionnaire assesses the degree to which respondents use COI disclosures when interpreting the literature. It further explores the perceived clinical value of a study with a given reported COI, assuming variations in study design, research institutional setting, and significance of results. The fictional research team disclosed the project was funded by a pharmaceutical company and all team members received consulting compensation. The survey and study were reviewed and approved by our institutional review board. The survey consisted of 14 multiple-choice questions that allowed for only 1 answer selection per person and allowed survey takers to skip questions they did not wish to answer. The survey questions and associated response options appear in edited form in the Table. A link to the questionnaire (https://www.surveymonkey.com/s/MPCCLCX) was sent with a message explaining the study. The responses to the questionnaire constituted the data.

Results

We sent a request to participate in the survey to 750 physicians and received 522 responses (overall response rate, 70%). The response rate for each question equaled or exceeded 98%.

The majority of respondents (95.6%) were male. Ninety-nine percent of respondents were orthopedic surgeons. The Northeast (US) was the most common geographical practice location of respondents (32%), followed by the Midwest (19.1%) and the Southeast (16.6%). Most respondents (40%) had been in practice for more than 20 years; 67% had been in practice a minimum of 10 years. The majority (68.8%) were employed by private practice groups, either single specialty (57.8%) or multispecialty (11%).

Eighty percent of respondents strongly agreed that COI disclosure is important when interpreting study results, 62% reported always reading the disclosure slide during academy or other meeting presentations, and 41% reported always using this information when deciding how to interpret scientific data.

Seventy-five percent of respondents thought the study—an academic-center case series with significant results in favor of the pharmaceutical company funding the study—was biased (42% indicated biased with merit, 33% biased without merit). Twenty-three percent thought the study was possibly biased, but likely trustworthy given the academic institutional affiliation. When the study setting was changed to community hospital, 95% thought the study was biased (51% biased with merit, 44% biased without merit). With the same study performed at an academic center, and no statistically significant results (not in favor of the pharmaceutical company funding the study), 88% thought the study had merit (46% biased with merit, 42% unbiased with merit).

When the study design was changed to a randomized controlled trial (level I evidence) conducted at an academic center with negative results, an overwhelming 95% of respondents thought the study had merit (33% biased with merit, 62% unbiased with merit). Given the same study design at an academic center, with positive results, 78% still thought the study had merit (39% biased with merit, 39% unbiased with merit). An additional 18% thought the study was biased, but still likely trustworthy given the academic institutional affiliation. Finally, given a randomized controlled trial and positive results, but with the research setting a small community practice, 90% thought the study had merit (51% biased with merit, 39% unbiased with merit). The percentage of respondents who found the study biased and likely without merit increased from 3.7% to 9.5% when the institutional affiliation changed from academic to community.

Discussion

As governmental funding sources become increasingly limited, the role of industry sponsorship of orthopedic research has grown. Potential drawbacks and biases of such research support have been well described—most notably, increased positive result reporting, suppression of results that may be disadvantageous to the industry sponsor, and biased study designs.^2-8 However, the extent to which financial COIs affect the orthopedic medical community’s interpretation of the literature has never been quantified. To our knowledge, the present study is the first to quantify the impact of reported COI on study interpretation.

Our goal was to examine how reported financial COIs influence the interpretation of the literature by the orthopedic medical community. Moreover, we wanted to determine the perceived reliability of the data when variables (study design, institutional affiliation, positive vs negative results) were changed. The results of our survey indicate that, when a financial COI is reported, study reliability is perceived as highest when negative results were found.

Our survey noted a discrepancy between the documented importance of the hypothetical research team’s COI disclosure and the use of such disclosures when interpreting study results. Eighty percent of respondents agreed that COI disclosure is important when interpreting study results, but only 62% reported always reading disclosures, and even fewer (41%) reported always using the information when interpreting results. It is unclear exactly why this trend exists, as one would expect the percentages to be more similar. These particular survey questions were formed around using COI disclosures when interpreting study results during academic presentations at national meetings and not during the review of published literature. It is possible that positioning the COI disclosure at the beginning of a presentation has an effect, but only 3.7% of respondents indicated they seldom remembered the disclosure by the end of the presentation. The results of our survey may have varied if the questions had targeted reading and interpreting the literature.

Interestingly, the results of these survey questions tended to be more consistent with rates of reported financial COI by presenters at national orthopedic meetings. A study published in the New England Journal of Medicine found that less than 80% of orthopedic surgeons reported their disclosures at a large annual meeting (AAOS), even when the disclosure involved payments pertinent to the research they were presenting.⁵ When the payments were indirectly related to the research, the percentage of surgeons reporting disclosures was 50%, almost the same as the disclosure rate for unrelated payments.⁵

When the study was changed to a level I randomized controlled trial, more survey respondents found it to be less biased and have more merit. Although it would seem intuitive for a study with a higher level of evidence to carry more clinical value during interpretation, this may not hold true in the setting of industry-sponsored clinical trials. Several studies have documented a significant association between the reporting of positive results and industry-sponsored randomized clinical trials. In 2008, Khan and colleagues³ examined 100 orthopedic randomized clinical trials reported in 5 major orthopedic subspecialty journals over a 2-year period. The association between industry funding and favorable outcome in all original randomized clinical trials was strong and significant (P < .001). This is not surprising, given the amount of time and money required for a well-designed clinical study. Commercial products with preclinical promise are pushed to testing in a clinical trial, whereas resources would not be wasted on products lacking preclinical merit.

The most important variable affecting interpretation of study merit by survey respondents was the reporting of negative results. As more researchers are developing COIs, many studies are discovering a relationship between COIs and outcomes of research studies. Reviewing the adult total joint literature, Ezzet⁸ found an industry funding rate of 50%. Positive results were reported in 93% of cases in commercially funded studies versus 37% of cases in independently funded studies. Furthermore, no negative results were reported by investigators who were receiving royalties from the respective companies.

Studies across the medical literature have also found this association between industry sponsorship and reporting of positive results. One such study, reported by Valachis and colleagues⁷ in the Journal of Clinical Oncology, examined more than 80 economic analyses of targeted oncologic therapies and found the studies funded by pharmaceutical companies were more likely to report favorable qualitative cost estimates. In addition, when studies with a COI disclosure were examined, those reporting any financial relationship with a manufacturer (eg, author affiliation, funding) were more likely than those without such a relationship to report favorable results.

Our study had several limitations. First, as most of the survey respondents were orthopedic surgeons, extrapolating their data to the medical community at large may not be appropriate, as each specialty may view industry affiliations differently. In addition, respondents were asked to base their interpretations of a study on conclusions we predetermined—no direct visualization of the data set or statistical testing methods. It is possible that these responses may have been different had the respondents had the opportunity to further evaluate the study in question. In a recent study, Altwairgi and colleagues¹¹ found that 10% of randomized clinical trials involving lung cancer treatment were reported with different conclusions in their full manuscripts relative to their abstracts. We think our survey design perhaps best mimics an annual meeting environment in which participants have very limited ability to interpret studies and may rely more heavily on the factors we investigated—study design, significance of findings, and setting, all similar to information presented in an abstract—when making informed decisions. Although our response rate was only 70%, this is comparable to or better than the rates in similar survey studies that used email-based questionnaires.^12,13

Another limitation was that our survey may have forced respondents into answers they did not entirely agree with, given the limited options of the multiple-choice response format and the specific wording of the questions. Our conclusions may have been more dramatic when we were evaluating whether the study was deemed meritorious or not. However, there is no adopted standard for evaluating the extent of bias perceived by a clinician. We thought it was important to include answer options indicating a study had merit despite obvious bias in design and execution. That a study had merit can mean different things. It may change clinical practice, may require further study and reproducibility, or may not be significant enough to matter. Asking follow-up questions to evaluate this perception among the respondents could have provided validity to the term merit. Further studies in this field are needed to determine how studies are interpreted and translated into clinical practice by various clinicians.

Conclusion

Although the present study is not a quantitative analysis of the determination of bias in the orthopedic community, it is the first to evaluate orthopedic surgeons’ perceptions on the basis of key fundamentals of orthopedic research relative to COI. It is clear from our study results that introducing levels of evidence to the orthopedic milieu has had a significant impact both on the quality of research and on the foundational use of deductive reasoning when interpreting the literature. Reporting negative outcomes is perhaps the most important factor in eliminating the perception of bias among orthopedic surgeons. To what extent a perceived COI plays into medical decision-making and the ultimate treatment of patients is still relatively unknown.

As defined by the American Academy of Orthopaedic Surgeons (AAOS) in 1996, conflict of interest (COI) is the “circumstance that exists when, because of personal financial gain, an individual has the potential to be less than objective when called on to reach a judgment or interpret a result.”¹ In medical research, COIs often occur in relationships between physician-researchers and pharmaceutical, medical device, and biotechnology companies. These relationships usually take the form of research grants but can also arise when the researcher has a financial interest in the product being tested or in the company that manufactures the product.

 Although constructive collaboration between academic medicine and industry has worked to improve health care and ultimately benefit patients, potential drawbacks of such relationships include sequestration and suppression of results that may be disadvantageous to the industry sponsor,² increased likelihood of reporting positive results (pro-industry),^3-7 and biased study designs.⁸ The nature of such relationships may threaten the integrity of scientific studies, the objectivity of medical education, the quality of patient care, and the public’s trust in medicine.⁹

Financial relationships and affiliations are increasing as we seek to answer a growing number of clinical questions—with funding often being a limiting factor. At national scientific meetings, the number of presentations reporting COIs reflects this trend. Paper and poster presentations accepted for annual meetings of the Orthopaedic Trauma Association (OTA) and reporting a COI increased from 7.6% in 1993 to 12.6% in 2002 (P = .0129).²

Medical subspecialties outside of orthopedics are experiencing similar trends. Most notable is the American Psychiatric Association (APA). After the APA published a mandatory financial COI disclosure policy in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), the percentage of task force members reporting industry relationships increased by 12%.¹⁰ Analysis of the DSM-5 panels demonstrated that the panels with the largest percentage of reported COIs are those for which pharmacological treatment is the first-line intervention, including the panels for mood disorders (67%), psychotic disorders (83%) and sleep/wake disorders (100%).¹⁰ Moreover, the industry ties reported are to the pharmaceutical companies that manufacture the medications used to treat these disorders or to companies that service the pharmaceutical industry.¹⁰

The degree to which financial COIs affect the interpretation of the orthopedic literature has never been quantified. Although it is clear that COIs can confound the results and reporting of data, how the medical community uses disclosures when interpreting the literature and when formulating opinions that may or may not affect their practice patterns is largely unknown.

We conducted a study to evaluate how a hypothetical financial COI disclosure would influence the interpretation of data by orthopedic clinicians. We also wanted to determine the reliability of the data as perceived in association with different study designs, levels of evidence, research institutional settings, and reporting of positive or negative results.

Methods

We asked members of the Arthroscopy Association of North America (AANA) and the American Orthopaedic Society for Sports Medicine (AOSSM) to complete a multiple-choice situational questionnaire (Table). The questionnaire assesses the degree to which respondents use COI disclosures when interpreting the literature. It further explores the perceived clinical value of a study with a given reported COI, assuming variations in study design, research institutional setting, and significance of results. The fictional research team disclosed the project was funded by a pharmaceutical company and all team members received consulting compensation. The survey and study were reviewed and approved by our institutional review board. The survey consisted of 14 multiple-choice questions that allowed for only 1 answer selection per person and allowed survey takers to skip questions they did not wish to answer. The survey questions and associated response options appear in edited form in the Table. A link to the questionnaire (https://www.surveymonkey.com/s/MPCCLCX) was sent with a message explaining the study. The responses to the questionnaire constituted the data.

Results

We sent a request to participate in the survey to 750 physicians and received 522 responses (overall response rate, 70%). The response rate for each question equaled or exceeded 98%.

The majority of respondents (95.6%) were male. Ninety-nine percent of respondents were orthopedic surgeons. The Northeast (US) was the most common geographical practice location of respondents (32%), followed by the Midwest (19.1%) and the Southeast (16.6%). Most respondents (40%) had been in practice for more than 20 years; 67% had been in practice a minimum of 10 years. The majority (68.8%) were employed by private practice groups, either single specialty (57.8%) or multispecialty (11%).

Eighty percent of respondents strongly agreed that COI disclosure is important when interpreting study results, 62% reported always reading the disclosure slide during academy or other meeting presentations, and 41% reported always using this information when deciding how to interpret scientific data.

Seventy-five percent of respondents thought the study—an academic-center case series with significant results in favor of the pharmaceutical company funding the study—was biased (42% indicated biased with merit, 33% biased without merit). Twenty-three percent thought the study was possibly biased, but likely trustworthy given the academic institutional affiliation. When the study setting was changed to community hospital, 95% thought the study was biased (51% biased with merit, 44% biased without merit). With the same study performed at an academic center, and no statistically significant results (not in favor of the pharmaceutical company funding the study), 88% thought the study had merit (46% biased with merit, 42% unbiased with merit).

When the study design was changed to a randomized controlled trial (level I evidence) conducted at an academic center with negative results, an overwhelming 95% of respondents thought the study had merit (33% biased with merit, 62% unbiased with merit). Given the same study design at an academic center, with positive results, 78% still thought the study had merit (39% biased with merit, 39% unbiased with merit). An additional 18% thought the study was biased, but still likely trustworthy given the academic institutional affiliation. Finally, given a randomized controlled trial and positive results, but with the research setting a small community practice, 90% thought the study had merit (51% biased with merit, 39% unbiased with merit). The percentage of respondents who found the study biased and likely without merit increased from 3.7% to 9.5% when the institutional affiliation changed from academic to community.

Discussion

As governmental funding sources become increasingly limited, the role of industry sponsorship of orthopedic research has grown. Potential drawbacks and biases of such research support have been well described—most notably, increased positive result reporting, suppression of results that may be disadvantageous to the industry sponsor, and biased study designs.^2-8 However, the extent to which financial COIs affect the orthopedic medical community’s interpretation of the literature has never been quantified. To our knowledge, the present study is the first to quantify the impact of reported COI on study interpretation.

Our goal was to examine how reported financial COIs influence the interpretation of the literature by the orthopedic medical community. Moreover, we wanted to determine the perceived reliability of the data when variables (study design, institutional affiliation, positive vs negative results) were changed. The results of our survey indicate that, when a financial COI is reported, study reliability is perceived as highest when negative results were found.

Our survey noted a discrepancy between the documented importance of the hypothetical research team’s COI disclosure and the use of such disclosures when interpreting study results. Eighty percent of respondents agreed that COI disclosure is important when interpreting study results, but only 62% reported always reading disclosures, and even fewer (41%) reported always using the information when interpreting results. It is unclear exactly why this trend exists, as one would expect the percentages to be more similar. These particular survey questions were formed around using COI disclosures when interpreting study results during academic presentations at national meetings and not during the review of published literature. It is possible that positioning the COI disclosure at the beginning of a presentation has an effect, but only 3.7% of respondents indicated they seldom remembered the disclosure by the end of the presentation. The results of our survey may have varied if the questions had targeted reading and interpreting the literature.

Interestingly, the results of these survey questions tended to be more consistent with rates of reported financial COI by presenters at national orthopedic meetings. A study published in the New England Journal of Medicine found that less than 80% of orthopedic surgeons reported their disclosures at a large annual meeting (AAOS), even when the disclosure involved payments pertinent to the research they were presenting.⁵ When the payments were indirectly related to the research, the percentage of surgeons reporting disclosures was 50%, almost the same as the disclosure rate for unrelated payments.⁵

When the study was changed to a level I randomized controlled trial, more survey respondents found it to be less biased and have more merit. Although it would seem intuitive for a study with a higher level of evidence to carry more clinical value during interpretation, this may not hold true in the setting of industry-sponsored clinical trials. Several studies have documented a significant association between the reporting of positive results and industry-sponsored randomized clinical trials. In 2008, Khan and colleagues³ examined 100 orthopedic randomized clinical trials reported in 5 major orthopedic subspecialty journals over a 2-year period. The association between industry funding and favorable outcome in all original randomized clinical trials was strong and significant (P < .001). This is not surprising, given the amount of time and money required for a well-designed clinical study. Commercial products with preclinical promise are pushed to testing in a clinical trial, whereas resources would not be wasted on products lacking preclinical merit.

The most important variable affecting interpretation of study merit by survey respondents was the reporting of negative results. As more researchers are developing COIs, many studies are discovering a relationship between COIs and outcomes of research studies. Reviewing the adult total joint literature, Ezzet⁸ found an industry funding rate of 50%. Positive results were reported in 93% of cases in commercially funded studies versus 37% of cases in independently funded studies. Furthermore, no negative results were reported by investigators who were receiving royalties from the respective companies.

Studies across the medical literature have also found this association between industry sponsorship and reporting of positive results. One such study, reported by Valachis and colleagues⁷ in the Journal of Clinical Oncology, examined more than 80 economic analyses of targeted oncologic therapies and found the studies funded by pharmaceutical companies were more likely to report favorable qualitative cost estimates. In addition, when studies with a COI disclosure were examined, those reporting any financial relationship with a manufacturer (eg, author affiliation, funding) were more likely than those without such a relationship to report favorable results.

Our study had several limitations. First, as most of the survey respondents were orthopedic surgeons, extrapolating their data to the medical community at large may not be appropriate, as each specialty may view industry affiliations differently. In addition, respondents were asked to base their interpretations of a study on conclusions we predetermined—no direct visualization of the data set or statistical testing methods. It is possible that these responses may have been different had the respondents had the opportunity to further evaluate the study in question. In a recent study, Altwairgi and colleagues¹¹ found that 10% of randomized clinical trials involving lung cancer treatment were reported with different conclusions in their full manuscripts relative to their abstracts. We think our survey design perhaps best mimics an annual meeting environment in which participants have very limited ability to interpret studies and may rely more heavily on the factors we investigated—study design, significance of findings, and setting, all similar to information presented in an abstract—when making informed decisions. Although our response rate was only 70%, this is comparable to or better than the rates in similar survey studies that used email-based questionnaires.^12,13

Another limitation was that our survey may have forced respondents into answers they did not entirely agree with, given the limited options of the multiple-choice response format and the specific wording of the questions. Our conclusions may have been more dramatic when we were evaluating whether the study was deemed meritorious or not. However, there is no adopted standard for evaluating the extent of bias perceived by a clinician. We thought it was important to include answer options indicating a study had merit despite obvious bias in design and execution. That a study had merit can mean different things. It may change clinical practice, may require further study and reproducibility, or may not be significant enough to matter. Asking follow-up questions to evaluate this perception among the respondents could have provided validity to the term merit. Further studies in this field are needed to determine how studies are interpreted and translated into clinical practice by various clinicians.

Conclusion

Although the present study is not a quantitative analysis of the determination of bias in the orthopedic community, it is the first to evaluate orthopedic surgeons’ perceptions on the basis of key fundamentals of orthopedic research relative to COI. It is clear from our study results that introducing levels of evidence to the orthopedic milieu has had a significant impact both on the quality of research and on the foundational use of deductive reasoning when interpreting the literature. Reporting negative outcomes is perhaps the most important factor in eliminating the perception of bias among orthopedic surgeons. To what extent a perceived COI plays into medical decision-making and the ultimate treatment of patients is still relatively unknown.