Contag SP, Golub JS, Teknos TN, Nussenbaum B, Stack BC, Arnold DJ, Johns MM. Professional Burnout Among Microvascular and Reconstructive Free-Flap Head and Neck Surgeons in the United States. Arch Otolaryngol Head Neck Surg. 2010;136(10):950-956. doi:10.1001/archoto.2010.154
To determine the prevalence of professional burnout among microvascular free-flap (MVFF) head and neck surgeons and to identify modifiable risk factors with the intent to reduce MVFF surgeon burnout.
A cross-sectional, observational study.
A questionnaire mailed to MVFF surgeons in the United States.
A total of 60 MVFF surgeons.
Main Outcomes Measures
Professional burnout was quantified using the Maslach Burnout Inventory– Human Services Study questionnaire, which defines burnout as the triad of high emotional exhaustion (EE), high depersonalization (DP), and low personal accomplishment. Additional data included demographic information and subjective assessment of professional stressors, satisfaction, self-efficacy, and support systems using Likert score scales. Potential risk factors for burnout were determined via significant association (P < .05) by Fisher exact tests and analyses of variance.
Of the 141 mailed surveys, 72 were returned, for a response rate of 51%, and 60 of the respondents were practicing MVFF surgeons. Two percent of the responding MVFF surgeons experienced high burnout (n = 1); 73%, moderate burnout (n = 44); and 25%, low burnout (n = 15). Compared with other otolaryngology academic faculty and department chairs, MVFF surgeons had similar or lower levels of burnout. On average, MVFF surgeons had low to moderate EE and DP scores. High EE was associated with excess workload, inadequate administration time, work invading family life, inability to care for personal health, poor perception of control over professional life, and frequency of irritable behavior toward loved ones (P < .001). On average, MVFF surgeons experienced high personal accomplishment.
Most MVFF surgeons experience moderate professional burnout secondary to moderate EE and DP. This may be a problem of proper balance between professional obligations and personal life goals. Most MVFF surgeons, nonetheless, experience a high level of personal accomplishment in their profession.
Burnout is a syndrome defined by the work-related triad of high emotional exhaustion (EE), high depersonalization (DP), and a low sense of personal accomplishment (PA). Emotional exhaustion is the feeling of being emotionally overextended and exhausted by one's work. Depersonalization is the adoption of a callous or dehumanized perception of others. Low personal accomplishment is the feeling of dissatisfaction with one's job-related achievements.1,2 Aside from the psychological unpleasantness associated with burnout, the disorder can lead to professional and personal dysfunction, which among physicians can negatively affect patient care.3,4
The burnout phenomenon has been quantified in many professions using the widely employed Maslach Burnout Inventory (MBI)–Human Services Study (HSS) survey.1,2,5- 8 Studies have examined the problem of physician burnout and found it pervasive in various fields of medicine. The value behind identifying physician burnout and its correlated risk factors lies in the potential to modify those risk factors and prevent the development of burnout as well as its consequences on patient care.8- 10
Our group has recently undertaken a series of studies to analyze the problem of burnout in academic otolaryngology–head and neck surgery.11- 13 Higher burnout was observed among residents and department chairs compared with more moderate levels of burnout among faculty. In addition, higher burnout was seen among associate professors than full professors. In the analysis of academic faculty, the highest predictors of burnout were dissatisfaction with the balance between personal and professional life, sense of low self-efficacy, inadequate research time, and inadequate administration time. Several of these risk factors could potentially be modified to decrease burnout, such as protecting time for research and administrative activities.11 Finally, a study of academic faculty members in the United Kingdom showed levels of burnout qualitatively similar to those found in their American counterparts.11
Our preliminary data suggested that microvascular free-flap (MVFF) surgeons might experience higher burnout than all other subspecialists in otolaryngology.11 A more careful study of this highly specialized field was thus warranted. We report herein results of the first nationwide investigation to our knowledge of burnout in academic MVFF head and neck surgeons.1We surveyed numerous possible covariates for burnout, including demographics, professional stressors, self-efficacy, and support systems. Using statistical correlation analyses, we identified risk factors for burnout. These new data may thus allow risk factor modifications, with the ultimate goal of decreasing the burnout rate in this important subspecialty.
The investigational design was an observational, cross-sectional, questionnaire-based study of MVFF surgeons in the United States. Questionnaires were mailed to 141 otolaryngologists who had completed a fellowship in microvascular and reconstructive surgery in the United States from 2002 to 2007 and to other known otolaryngologists practicing free-flap surgery. Each participant received a questionnaire, a postage-paid return envelope, a cover letter explaining the study's general purpose, and an instruction sheet. The mailing was repeated to nonresponders after 1 month to increase the response rate. Participation was voluntary. Confidentiality was strictly maintained: participants were instructed not to mark any personally identifiable information on the questionnaire or return envelope. The Emory University institutional review board approved the protocol and questionnaire.
The survey content was modified from a previous version used to study burnout in academic otolaryngology faculty.11 The survey consisted of 68 total questions divided into the 6 categories described herein. The questions were numbered 1 through 28, with 6 questions containing multiple subquestions. A space for general comments was included at the end.
Fourteen questions collected demographic and work schedule information, including age, sex, years in practice, type of practice, average number of hours worked per week, and average number of free-flap procedures performed per year. Questions addressing work schedule were phrased to elicit typical values for an average week.
Participants rated the degree to which 13 stressors affected them on a 5-point Likert scale from 1 (not at all) to 5 (extreme amount). Examples of stressors included billing audits, inadequate administration time, lack of sleep, excessive workload, malpractice suits, lack of time for personal growth and development, and work invading relationship with family.
Eight questions concerned overall career satisfaction. Three questions assessed the likelihood of continuing to perform free-flap surgery, satisfaction with the decision to perform free-flap surgery, and satisfaction with the balance between professional and personal life on a Likert scale from 1 (very satisfied) to 5 (very dissatisfied). Four “agree/disagree” statements addressed satisfaction with free-flap surgery. One question asked participants to list their top 3 reasons for continuing to perform free-flap surgery.
Self-efficacy was evaluated using a previously developed metric6 modified to reflect issues specific to otolaryngology. Participants rated their sense of control over their professional life, their sense of personal effectiveness and time management skills, and their sense of certainty (ie, positivity and assurance) regarding the attainment of personal success. Seven responses were scored on 5-point Likert scale from low to high self-efficacy. Individual response scores were then added to calculate a total self-efficacy score. The allowable range was thus 7 to 35. Finally, scores were stratified into low (7-15), moderate (16-24), and high (25-35) levels.
Participants quantified the amount of support derived from personal relationships, which was characterized by the frequency of negative behavior between loved ones secondary to work-related stress. Three questions addressed relationship stability with spouses or significant others and/or family members. A second set of 3 questions addressed spousal support among married participants. Responses to each question were scored on a 5-point Likert scale.
The 3 components of burnout (EE, DP, and PA), were assessed using the MBI-HSS questionnaire.1 The instrument consisted of 22 randomly arranged questions, each written in the form of a subjective statement. Participants were asked to rate how often they felt or experienced each statement on a scale from 0 (never) to 6 (every day). Nine statements addressed EE (eg, “I feel like I’m at the end of my rope”); 5 statements addressed DP (eg, “I feel like I treat some of my patients as if they were impersonal objects”); and 8 statements addressed PA (eg, “I have accomplished many worthwhile things in my job”).
Total scores were calculated for EE (possible range, 0-54), DP (possible range, 0-30), and PA (possible range, 0-48) by summing the individual response scores. Scores were then stratified into high, moderate, and low levels according to the scoring guidelines determined in the original MBI questionnaire.1 The score stratification scheme is summarized in Table 1.
Statistical analysis was performed using SAS software, version 9.1 (SAS Institute Inc, Cary, North Carolina). Descriptive statistics were used to characterize the study population. Demographic and work schedule data were described using means and standard deviations (SDs). The remaining variables including professional stressors, career satisfaction, self-efficacy, and support systems were characterized according to their Likert score frequencies and modal scores.
Analysis of association between surveyed variables and burnout was performed by post hoc multiple hypotheses testing. The original authors of the MBI-HSS previously determined that the statistical reliability of individual subscales is higher than the reliability of the composite burnout score.1 Therefore, each burnout subscale (EE, DP, and PA) was analyzed individually. For each burnout subscale, MVFF surgeons were grouped for analysis according to their score stratification of low, moderate, or high. Associations between stratified burnout subscale groups and continuous variables were analyzed by analysis of variance for comparison of means. Associations between stratified burnout subscale groups and categorical variables, including those measured using Likert scores, were analyzed using Fisher exact tests. To correct for the inflation of type 1 error in our multiple hypotheses testing, a higher significance level was imposed on our analysis. Bonferroni correction was applied to the Fisher exact P values in the following manner: professional stressors, P = .05/13; satisfaction determinants, P = .05/7; determinants of support, P = .05/6; and determinants of self-efficacy, P = .05/7.
Of 141 distributed questionnaires, 72 were returned in the mail, for a participation rate of 51%. Of the 72 returned questionnaires, 60 were entirely completed by practicing MVFF surgeons and used for subsequent analysis. Demographic and work schedule information is summarized in Table 2. Of the 60 participants, 88% were men (n = 53) and 12% were women (n = 7). Eighty-seven percent of participants worked in academic practice (n = 52); 3% in private practice (n = 2), and 10% in a practice associated with academic practice (n = 6). Average time in practice was 9 years, with 91% of participants having performed 1 to 15 years of free-flap surgery (n = 55), and 9% having performed more than 16 years of free-flap surgery (n = 5). The average number of free-flap procedures performed per surgeon per year was 39 (range, 1-140). Respondents worked 65 h/wk on average.
Participants rated the degree to which 13 common professional stressors affected their professional life on a scale of 1 (no effect) to 5 (extreme effect). Inadequate research time was the highest-ranked stressor, with a modal score of 4, indicating a large stress effect. Thirty-nine percent of participants reported that this problem contributed a large or extreme degree of stress to professional life (n = 26). Forty-two percent of participants reported that it caused a slight or moderate degree of stress (n = 25). Only 13% of participants reported that it was not a cause of stress at all (n = 8).
Three stressors were ranked equally with modal scores of 3 each, indicating a moderate degree of stress: (1) lack of time for personal growth and development, rated to have a large to extreme stress effect by 39% of participants (n = 23) and a slight to moderate effect by 55% (n = 33); (2) inability to socialize with friends outside of medicine, rated to have a moderate to large effect by 48% (n = 29) and slight to no effect by 47% (n = 28); and (3) low collections and billing, rated to have moderate to large effect by 46% (n = 28) and slight to no effect by 43% (n = 26).
Seven equally ranked stressors each had a modal score of 2, indicating a slight stress effect: (1) work invading relationship with family was rated to have a slight to moderate stress effect by 66% of participants (n = 40); (2) lack of sleep and/or irregular sleep pattern, rated slight to moderate effect by 70% (n = 42); (3) inability to care for personal health, rated slight to moderate effect by 65% (n = 39); (4) excessive workload, rated large to extreme effect by 38% (n =23) and slight to moderate by 57% (n = 34); (5) lack of independence, rated to have slight or no effect by 85% (n = 51); (6) inadequate administration time, rated slight or no effect by 63% (n = 25), large effect by 22% (n = 13); and (7) inadequate monetary compensation, rated slight to moderate by 55% (n = 37).
Two stressors each had a modal score of 1, indicating no stress effect at all on professional life: billing audits, rated to have slight or no effect by 83% (n = 50), and being a defendant in a malpractice case, rated to have slight to no effect by 95% (n = 57).
Overall, US MVFF surgeons were very satisfied with their chosen work. Fifty-five percent of surgeons reported being very satisfied with their choice to perform MVFF surgery (n = 33). Thirty-seven percent reported being only somewhat satisfied (n = 22), and the remaining 8% felt neutral about their choice (n = 5). In congruence with overall high satisfaction about career choice, 75% of surgeons agreed or strongly agreed they were meeting their career expectations (n = 45). Ninety-five percent of MVFF surgeons agreed or strongly agreed that their work provided a personal sense of meaning (n = 57). This was despite a more mixed opinion on whether stress altered their physician-patient relationship: 49% agreed or strongly agreed that stress altered their relationship with patients (n = 29); 33% disagreed or strongly disagreed (n = 20); and 18% felt neutral on the issue (n = 11).
Satisfaction with the balance between personal and professional life was more varied. Thirty-five percent of surgeons reported feeling somewhat dissatisfied (n = 21), and 10% reported feeling very dissatisfied (n = 6). Conversely, 20% felt very satisfied (n = 12), and 25% felt at least somewhat satisfied (n = 15). Ten percent of surgeons felt neutral (n = 6). Despite these data, most MVFF surgeons (72%) reported that they were not likely to stop performing free-flap surgery within the next 1 to 2 years (n = 43). Furthermore, only 25% had seriously considered early retirement (n = 15), while 52% had not entertained the possibility (n = 31).
Participants were asked to list their top 3 reasons for continuing to perform free-flap surgery. The most common responses, in decreasing frequency, were enjoyment of the surgical challenge and satisfaction in performing the free-flap technique; performance of free-flap surgery was required to meet the needs and demands of the surgical practice; better patient outcomes; and financial incentives and demands.
Eleven questions addressed self-efficacy and control. On average, participants reported a moderate level of perceived self-efficacy, with a mean (SD) total score of 20 (4) (range, 14-30). Forty-three percent of participants reported having moderate control over their professional life (n = 26); 35% felt that they had a large amount of control (n = 21); and 20% felt only a slight amount of control (n = 12). Most participants (62%) believed that achieving good control of their professional life was determined by a combination of their own efforts and factors beyond their control (n = 37). Twenty-seven percent believed that achieving good control was determined mostly by their own effort (n = 16), whereas 10% believed it to be mostly or entirely due to other factors (n = 6). Finally, most participants (75%) felt they were very or moderately effective in coping with time management in their professional life (n = 45). The remaining 25% felt slightly or not at all effective at this skill (n = 15).
Self-efficacy was further characterized by examining the amount of certainty (positivity and assurance) felt in regard to personal and professional life outcomes. Seventy-five percent of surgeons felt large or moderate certainty about their lives (n = 45). Twenty-two percent believed that improvement in their professional lives was very or extremely likely in the next several years (n = 13); 60% believed improvement was only slightly or moderately likely (n = 36); and 17% thought improvement was not at all likely (n = 10). The trend in opinion was similar when participants were asked to subjectively consider the probability of personal life improvement in the next few years: 28% responded that it was very likely to improve (n = 17); 60% that it was slightly or moderately likely to improve (n = 36); and 10% that it was not at all likely to improve (n = 6). Seventy three percent of surgeons thought their professional role interfered to a large or moderate degree with the development of other life goals (n = 44).
The questionnaire assessed the amount of support derived from personal relationships, which was measured as the frequency of negative behaviors between loved ones secondary to work-related stress. Seventy-seven percent of MVFF surgeons felt that they were only sometimes or once in a while withdrawn or quiet with loved ones when preoccupied with work (n = 46). Most MVFF surgeons felt that they were only sometimes or once in a while irritable with loved ones when preoccupied with work (79%; n = 47) or with the amount of time spent at work (67%; n = 40). The remaining 20% reported feeling irritable or withdrawn frequently with loved ones when preoccupied with work (n = 12).
Married MVFF surgeons reported how often their spouses demonstrated job-related supportive behavior. Most felt well supported. Seventy-five percent reported that their spouse was always or frequently willing to listen to work-related problems (n = 39). The remaining 25% reported that their spouse was willing to listen only sometimes or once in a while (n = 13). Most surgeons felt that their spouse was usually understanding when they had to work extra hours: 77% reported always or frequently (n = 40); and 21% reported sometimes or once in a while (n = 11). When asked how often participants felt encouraged by the spouse to take advantage of professional opportunities, the response was more varied: 56% said always or frequently (n = 29), while 44% said sometimes or once in a while (n = 23).
Most MVFF surgeons (73%) experienced a moderate level of overall burnout (n = 44); 25% experienced low overall burnout (n = 15); and 2% experienced high overall burnout (n = 1) (Table 3). The average level of EE was low to moderate, with a mean (SD) total score of 20 (11) (range, 2-46). Forty-eight percent of participants had low EE (n = 29); 25% had moderate EE (n = 15); and 27% had high EE (n = 16). The average level of DP was low to moderate, with a mean (SD) total score of 7 (6) (range, 0-28). Fifty-five percent had low DP (n = 33); 23% had moderate DP (n = 14); and 22% had high DP (n = 13). The average level of PA was high, with a mean (SD) total score of 40 (7) (range, 18-48). Sixty-two percent had high PA(n = 37); 28% had moderate PA (n = 17); and 10% had low PA (n = 6) (Table 4 and Table 5).
Post hoc analysis of the data was performed to determine possible risk factors associated with EE, DP, and PA. Participants were stratified into low, moderate, and high levels of each burnout subscale. Demographic predictors were analyzed with analysis of variance multivariate analysis for association with burnout subscales. There were no significant associations between burnout subscale scores and demographic predictors, including age, sex, years in practice, years performing free-flap surgery, hours worked per week, and number of free-flap procedures performed per year.
Fisher exact tests were performed to determine correlations between burnout subscales and professional stressors, perceived career satisfaction, self-efficacy, and personal support. Only EE revealed significant associations with several predictors after P values were subjected to Bonferroni correction to control for multiple hypothesis testing. Table 6 summarizes the positive associations between various factors and EE. Professional stressors significantly associated with EE (P ≤ .001) included: work invading family life, excessive workload, inadequate administration time, and inability to care for personal health. High EE was inversely associated with high satisfaction with career choice (P ≤ .001) and was positively associated with the consideration of early retirement (P ≤ .001). Furthermore, high EE was associated with poor perception of control over professional life (P = .001). Finally, support systems had significant association with burnout. Increased frequency of irritable behavior toward loved ones about time spent at work correlated positively with EE (P < .001).
Our results demonstrate that most MVFF surgeons in the United States experience moderate professional burnout due to moderate to low levels of EE and DP. Participants had a high sense of PA, which was likely protective against burnout. This trend was similar to that seen in other otolaryngology academic faculty members.11 The existing EE and DP in microvascular surgery may be attributed to an imbalance of professional and personal life, which is a pervasive problem among American surgeons in general.14 High PA among MVFF surgeons is likely driven by high satisfaction and enjoyment experienced by performing complex and highly specialized MVFF operations.
Surprisingly, MVFF surgeons did not have a significantly higher level of burnout than academic otolaryngologists, residents, faculty members, and chairs (Table 5). Microvascular free-flap surgeons reported low to moderate levels of EE and DP, similar to that reported by academic faculty members and chairs.11,13 Interestingly, MVFF surgeons, like academic faculty, reported a higher sense of PA than department chairs. Microvascular free-flap surgeons worked 65 h/wk on average, which was more hours than otolaryngology academic faculty as a whole (mean, 60 h/wk), but fewer hours than otolaryngology academic chairs (mean, 68 h/wk) and otolaryngology residents (mean, 71 h/wk).11- 13 Hours worked did not have an association with higher burnout. While this finding is consistent with previous studies of American surgeons in general, it is contrary to what was observed in otolaryngology academic faculty and residents.11,12,14 There was a sizeable range in the number of free-flap procedures performed per MVFF surgeon per year (range, 1-140; mean, 39), with no correlation to burnout. Similarly, a study of 582 American surgeons by Campbell et al15 found no association between case load and burnout. Instead, burnout was associated with factors such as the imbalance between work and family life and lack of autonomy and decision involvement at work.
Interestingly, sex, age, and years in practice had no association with burnout. This observation is contrary to several other studies on burnout within the medical profession that often correlate younger age, fewer years in practice, and female sex with higher burnout.6,15- 18 Peisah et al18 characterized burnout among 158 physicians across multiple specialties and determined that age and years in practice had a strong negative association with burnout. Older physicians, by age and years in practice, experienced less burnout than their younger colleagues who had fewer years in practice. Physicians with more years in practice had learned how to cope with the demands of the profession and protect against burnout through protective mechanisms including acquired confidence, maturity, and degree of control over professional life. It was also noted that older physicians were more likely to be in private practice than younger physicians. The authors noted that the “burnout protective” effect could have been derived from either years of experience or working in a private practice vs academic setting. In our study, most MVFF surgeons, regardless of age, were in academic practice. Therefore, the pressures of academic life may be the common denominator for burnout across the spectrum of age or years in practice.
Overall, perceived control over professional life had a negative association with EE, consistent with previous studies.17,19 In addition, perceived excess workload had a strong positive association with EE. In this largely academic subspecialty, most MVFF believed inadequate research time to be a large source of professional stress, ranking it the number 1 stressor. This problem is not specific to MVFF surgeons; it was also rated the highest professional stressor by otolaryngology academic faculty.11 This stressor was not significantly associated with burnout in MVFF surgeons. Inadequate administration time, however, was positively associated with EE.
Our present analysis of perceived professional stressors that correlate with burnout provides fertile ground for a more in-depth characterization of their relationship to burnout. Future work may aim to determine exactly how these aspects of academic life contribute to the development of professional burnout among MVFF surgeons. For example, studies may look beyond number of free-flap procedures performed yearly and quantify the proportion of time devoted to research, teaching, administrative, and clinical responsibilities, and then analyze how the distribution influences burnout. After all, the expectation for greater clinical productivity and academic output in addition to administrative obligations may together increase the risk of burnout. Shanafelt et al8 recently demonstrated that the extent to which faculty physicians are able to focus on the aspect of work that is most meaningful to them, whether patient care, research, teaching, or administration, has a strong inverse relationship to the risk of burnout. In this regard, departmental support of faculty is key. Restructuring responsibilities to protect time for preferred duties (whether it be administrative or research) may decrease EE among MVFF surgeons as well as all otolaryngology academic faculty.
The quality of personal support systems significantly influenced professional burnout. Decreased reliance on support systems, as evidenced by a high frequency of withdrawn or irritable behavior toward loved ones, had a strong association with higher EE. Most MVFF surgeons reported good personal support systems. Most felt well supported and encouraged by their loved ones. Spouses were rated as being always or frequently understanding, even when surgeons had to work extra hours, as well as encouraging of surgeons pursuing professional opportunities. Therefore, individual emphasis on cultivating and maintaining personal support systems and pursuing personal goals outside of work may help protect against professional burnout. This is of particular importance because the “psychology of postponement” has been described previously as the widespread, compulsive tendency among physicians to consistently delay attending to relationships and other personal pursuits to fulfill work responsibilities.20,21
Overall, MVFF surgeons were very satisfied with their chosen work and felt that they were meeting career expectations. Remarkably, 95% strongly agreed that their work provided a sense of personal meaning (n = 57), in spite of a 73% majority feeling that work interfered with developing other life goals (n = 44). The willingness to sacrifice personal life for work may be explained by the most common self-reported reason for doing the job: the intrinsic satisfaction and enjoyment derived from performing a highly complex, precise, and specialized surgical skill.
The limitations of this study include its cross-sectional design and the potential for recall bias in the self-reported measures. Furthermore, the 51% response rate may reflect sampling bias in that those surgeons most burned out might have been less likely to complete and return the survey secondary to a lack of interest or time. Therefore, there exists the possibility that the extent of burnout was underreported in our study population. This sampling bias was difficult to eradicate from the study design because participation was anonymous and voluntary. Also, the 51% response rate may have decreased the power of our study to detect correlations between burnout and the possible risk factors of interest.
In conclusion, MVFF surgeons have levels of burnout comparable to those of their academic faculty peers. The general perception that free-flap surgery is plagued by high professional burnout may be invalid; indeed only 2% of MVFF surgeons fell into this category (n = 1). Departmental efforts to reduce burnout should focus on allowing increased or protected time for preferred work-related activities. Whenever possible, measures to reduce excess workload on individual physicians should be instituted. Individual efforts as well as departmental encouragement of balancing professional and personal life goals might help reduce EE. Finally, professional support, mentorship, and encouragement of those involved in or interested in MVFF surgery may help promote growth of the field and protect against burnout.
Correspondence: Michael M. Johns III, MD, Emory Voice Center, Department of Otolaryngology–Head and Neck Surgery, 550 Peachtree St NE, Ninth Floor, Ste 4400, Atlanta, GA 30308 (firstname.lastname@example.org).
Submitted for Publication: March 26, 2010; final revision received June 1, 2010; accepted June 24, 2010.
Published Online: August 16, 2010. doi:10.1001/archoto.2010.154
Author Contributions: Dr Johns had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Contag, Golub, Teknos, Nussenbaum, Stack, Arnold, and Johns. Acquisition of data: Contag, Golub, Teknos, and Johns. Analysis and interpretation of data: Contag, Golub, Teknos, Nussenbaum, and Johns. Drafting of the manuscript: Contag, Golub, Teknos, and Johns. Critical revision of the manuscript for important intellectual content: Contag, Golub, Nussenbaum, Stack, Arnold, and Johns. Statistical analysis: Contag. Administrative, technical, and material support: Johns. Study supervision: Golub, Teknos, Nussenbaum, Stack, and Johns.
Financial Disclosure: None reported.
Additional Contributions: Oswaldo Henriquez, MD, collated and disseminated the survey, and Stephen A. Contag, MD, assisted with statistical analyses.