Burnout in Orthopaedic Physical Therapists

Burnout in Orthopaedic Physical Therapists Brenda j. Wandling, MPT' Barbara S. Smith, PhD, PT* urnout can be defined as "a syndrome of physical and emotional exhaustion, involving the development of negative selfconcept, negative job attitudes, and loss of concern and feeling for clients" (18). Factors that influence burnout include time constraints, poor communication among staff members, perceived high job demands, and boredom (4,8,16). Orthopaedic physical therapists may work in environments that contain many of these factors and, thus, can be candidates for burnout. Burnout may be detrimental to a person's health. Physical symptoms include frequent headaches, gastrointestinal disturbances, sleeplessness, chronic fatigue, and illness (7,9,16). Psychologically, one may experience depersonalization, increased emotional exhaustion, a reduced sense of accomplishment, increased cynicism toward others, and depression (4,11, 16,26). These physical and psychological problems are linked to alcoholism, drug use, and family stress (7,9, l6,26). The quality of one's job performance may also be affected (9). Psychologists have studied burnout among professional leaders for more than 15 years. Initially, Freudenherger related burnout to intrinsic (personality) factors after he observed the high level of stress exhibited by staff members in alternative health care institutions (8). Pines and Maslach (19) studied the social and psychological dimensions of burnout among social welfare workers, psychiatric nurses, poverty lawyers, prison personnel, and child care Few studies exist about burnout in physical therapists. No research is available on burnout in orthopaedic therapists. The purpose of this study was to determine if burnout exists in this group and if a relationship exists behveen burnout scores and selected variables. Members of the Orthopaedic Section of the American Physical Therapy Association from the east north central region of the United States were sent a survey consisting of demographic questions and the Maslach Burnout Inventory. The return rate was 38.7%. Burnout levels were low to moderate depending on the grouping used to examine the data. All of the variables examined (eg., years as a physical therapist, number of patients seen per day) showed very low correlations with subscale scores. Further research is needed to identify those at most risk for burnout and to find effective strategies for its prevention. Key Words: burnout, occupational stress, physical therapist ' Staff Physical Therapist, United Cerebral Palsy of Greater Houston, Houston, TX Associate Professor, Department of Physical Therapy, Wichita State University, 1845 Fainnount, Wichita, KS 67260-0043 workers. As a result of these studies, they felt burnout was best described using three separate subscales: personal accomplishment, emotional exhaustion, and depersonalization. Speech-language pathologists, staffs of mental health institutions, and occupational, music, and corrective therapists have been surveyed about various aspects of burnout (1,4, 7,15,19,22). However, burnout among physical therapists has received little attention from researchers, although physical therapists are an important and increaqingly growing segment of the health care profession. The status of burnout among physical therapists was introduced in 1981 by Wolfe (26). He suggested that failure and frustration, work overload, and role conflict could result in burnout. Today's orthopaedic therapists may also experience these events. As a consequence of burnout, Wolfe (26) suggested that patients could be affected in at least three ways: decreased time devoted to each patient, inadequate treatment given to each patient, and failure to meet patient goals. Little other research has examined potential factors that could lead to burnout among physical therapists. Schuster et a1 suggested that excessive job-related stress, such as excessive demands and overwork, leads to burnout (24). Deckard and Present studied physical therapists in Missouri using the Maslach Burnout Inventory and the Anxiety-Stress Questionnaire. They found a significant relationship between role conflict and emotional exhaustion (3). Donohoe et al determined that lack of communication among staff' members, unrealistic goals for patients, and time constraints contributed to burnout among physical therapists working in rehabilitation hospitals (6). Schlenz et a1 studied occupational and physical therapists working in head injury rehabilitation. They concluded that professional development activities in the workplace may augment feelings Volume 26 Number 3 September 1997 JOSFT

RESEARCH STUDY of personal accomplishment and minimize burnout as an issue in job retention (23). The last two studies (6,23) are the only ones involving physical therapists that reported levels of burnout calculated from the Maslach Burnout Inventory (MBI), a tool for delineating jobrelated stress in human service workers (10,ll). Burnout was best described using three separate subscales: personal accomplishment, emotional exhaustion, and depersonalization. The purpose of this study was to investigate the relationship between burnout and personal and professional demographic variables (eg., gender, number of years as a physical therapist). Members of the Orthopaedic Section of the American Physical Therapy Association (APTA) were chosen because no reports are available about this group's reaction to work stress as measured on the MBI. METHOD Subjects The east north central region, one of the six regions of the United States according to the U.S. Census Bureau (25), was randomly selected. The states in this region are Illinois, Indiana, Michigan, Ohio, and Wisconsin. A proportional random sample of 1,000 members of the Orthopaedic Section of the APTA was chosen from the region's 1,625 members. Subject selection was geographically restricted to avoid regional idiosyncrasies in burnout variables (1 2). How often: 0 Never 1 A few times a year 2 Once a month or less 3 A few times a month 4 Once a week 5 A few times a week 6 Every day TABLE 1. Example ot question from the Maslach Burnout Inventory and Likert scale used to quantify the response. Instrument The Maslach Burnout Inventory (MBI) was used to measure burnout. This inventory consists of 22 items that make up three subscales: Personal Accomplishment, Emotional Exhaustion, and Depersonalization. The Personal Accomplishment sub scale consists of eight items and is designed to evaluate attitudes of personal achievement and accomplishment in the respondent's work. The Emotional Exhaustion subscale is made up of nine items in which emotions such as frustration, strain, and exhaustion are expressed. The Depersonalization subscale consists of five items measuring impersonal and uncaring attitudes toward the group of people the respondent is serving. Items on the MBI are written as statements about personal feelings or attitudes. Permission was granted by the publisher to change the generic term "recipients," used to refer to clients, to "patients." An example of a question and the scale used are found in Table 1. Test-retest reliability of the Maslach Burnout Inventory has been reported for two samples consisting of graduate students in social welfare and administrators in a health agency. The coefficients for the sub scales ranged from 0.60 to 0.82: 0.80 for Personal Accomplishment; 0.82 for Emotional Exhaustion; and 0.60 for Depersonalization. The inventory's convergent and discriminant validity are considered satisfactory (14). In addition to the information gathered from the Maslach Burnout Inventory, the following demographic information was obtained: gender, age, full- or part-time employment, number of years as a physical therapist, number of years at the respondent's present job, number of patients seen per day, number of hours per week spent working with patients, number of hours per week spent outside work that is job related, and the type of work setting. Data Analysis Subscale scores for Personal Accomplishment, Emotional Exhaustion, and Depersonalization were determined for each subject according to the instructions provided by Maslach and Jackson (13). Each sub scale score wa. calculated separately; there is no composite score. Data analysis was performed using SPSS/PC (Version 5.0, SPSS Inc., Chicago, IL). Pearson product moment correlations were calculated to analyze the relationship between burnout subscale scores and each demographic variable except gender, full- or part-time work, and work setting. Eighteen correlation coefficient.. were generated. This relatively large number of calculations increases the likelihood of statistical significance occurring simply by chance (20). Consequently, the.002 (.05/18) level was used to signify significant relationships. Results at the.05 and.o1 level should be interpreted with caution. An independent t test was used to determine differences in subscale scores by gender. Alpha level for this test was set at.05. RESULTS Profiles of Respondents and Burnout Scores The return rate of usable surveys was 38.7%. This represents 23.8% of the east north central region's Ortho- JOSFT Volume 26 Number 3 September 1997

RESEARCH STUDY years as a noun per Years at Patients Seen Hours per Week Wedc Outside Age Physical Current Job Therapist Per Day in Patient Care job All respondents Males = 148, Females = 237' 36.0 7.9 11.4 8.0 5.1 5.0 15.0 10.38 32.2 10.4 7.2 9.0 Full time = 334, Part time = 53 Full-time therapii Males = 147, Females = 184' 36.0 8.3 11.3 8.4 5.3 5.2 15.5 11.0 34.4 12.5 7.7 9.8 Part-time therapists 35.8 5.2 11.9 4.9 4.5 3.6 12.2 4.0 18.7 6.8 4.6 7.5 Males = 1, Females = 50 * Not all respondents responded to this question. TABLE 2. Profile of respondents (X 5 SD). paedic Section membership. The profile of respondents is found in Table 2. The subscale scores without regard for any personal or professional variables show low or moderate burnout (Table 3). This table also contains scores from other studies for comparison. Subscale scores ranged as follows: Personal Accomplishment, 22-48; Emotional Exhaustion, 0-52; and Depersonalization, 0-23. Results of the correlation studies are found in Table 4. Scores Regarding Personal Data No significant difference was found between males and females on any of the subscales. Therefore, all respondents were considered as one group. For convenience of describing the data, age was divided into the quartiles found in Table 5. All age groups showed low burnout on the Personal Accomplishment and Depersonalization subscales; all ages showed moderate emotional exhaustion, except the 34.1- to 41-yearald group that had low burnout on this spondents had been physical therasubscale. pists was divided into the quartiles found in Table 5. All quartiles had scores ~ ~ profenioml ~ Data ~ low dburnout scores i for ~ each subscale ~ except the Emotional Exhaustion Respondents averaged 11.37 years subscale score. Respondents who had in the profession (range = 0.3-37.0 been in practice greater than 16.1 years). The number of years the re- years were the only group who had a - Personal Accomplishment* Maslach Burnout Subscale Scores Emotional Exlustion' Depersonalization' Maslach (overall sample) 34.5 7.1 21.0 10.7 8.7 5.8 Wandling and Smith (this study) 41.8 5.O 18.7 9.3 5.4 4.4 Donohoe et al (6) 37.3 8.4 23.5 10.7 7.6 5.7 Piedmont (1 7) 39.5 5.1 24.6 7.1 5.4 3.6 Rogers and Dodson (22) 36.1 4.9 19.9 8.6 5.6 5.1 Schlenz et al (23) 39.4 4.7 25.4 9.4 6.4 3.7 * 5 3 1 = High; 32-38 = Moderate; 239 = Low burnout. ' 227 = High; 17-26 = Moderate; 116 = Low burnout. * 2 13 = High; 7-12 = Moderate; 56 = low burnout. TABLE 3. Maslach Burnout Inventory subscale scores from Maslach andlackson's overall sample (13) and from studies of physical and occupational therapists (X 5 SD). Years as a Physical Therapist Years at Current job Number of Number of Hwn Number of Hours Patients Seen per Week per Week in Patient Care Outside job Personal accomplishment,1120'.1088'.1931*.0650.0691.1380' Emotional exhaustion -.1268' -.1728' -.1808*.0404.0513 -.089b8 Depersonalization -.1379' -.1627* -.1413'.0089.0593 -.lo62 * p 5.05. ' p 5.01. * p 5.001. TABLE 4. Correlations of variables with burnout scores. Volume 26 Number 3 September 1997 JOSPT

RESEARCH - STUDY MBI Subscale Scores Gender' Male (N = 148) Female (N = 237) Full or part time Full time (N = 334) Part time (N = 53) Years as a physical therapist 15 (N = 100) 5.1-9.5 (N = 95) 9.6-1 6 (N = 97) 216.1 (N = 95) Years at current job 51.5 (N = 104) 1.C3 (N = 93) 3.1-7.25 (N = 94) 27.3 (N = 96) HOUR outside job per week 12 (N = 121) 2.1-5 (N = 127) 5.1-10 (N = 84) rlo.l (N = 55) Acrompliimenti Exhaustion' Depersonalizationr setting Hospital (N = 1 18) Home health (N = 15) Rehabilitation (N = 17) Private practice (N = 148) Nursing home (N = 25) - Numbers do not alwvs add to 3Ri because not all respondents answered each demographic question. ' 131 = High; 32-38 = Moderate; 239 = low burnout. * 227 = High; 17-26 = Moderate; 1 16 = Low burnout. L 13 = High; 7-12 = Moderate; 16 = low burnout. TABLE 5. Maslach Burnout Inventory (MI) subscale scores according to demographic data. (k 2 SD). low level of burnout on this subscale. Those who had been in practice less than 5 years had the highest Em* tional Exhaustion subscale scores. Respondents had been at their current job an average of 5.1 7 years (range = 0.1-30.0 years). The number of years at the currenlt job was divided into the quartiles found in Table 5. Levels of burnout were the same as those found for years as a physical therapist. JOSFT Volume 26 Number 3 September 1997 The mean number of hours per week spent outside work on jobrelated activities was 7.23 (range = 0-70.0 hours). Using the quartiles found in Table 5, respondents had low burnout except on the Emotional Exhaustion subscale score. Overall, moderate burnout was seen in the Emotional Exhaustion subscale despite the hours per week spent on outside activities. Those who spent the most time on jobrelated activities had the lowest burnout level across all three subscales. Most respondents were in private practice (148) or worked in a hospital (1 18). Only three other work settings had more than 10 respondents: nursing homes (25), rehabilitation facilities (rehab centers with or without beds) (17), and home health (15). With respect to setting, therapists showed low burnout on the Personal Accomplishment and Depersonalization subscales except those in rehabilitation settings, who showed moderate burnout on the Depersonalization subscale. All therapists, except those in private practice, showed moderate levels of burnout on the Emotional Exhaustion subscale. The mean score for therapists in rehabilitation settings (23.71) was the highest Emotional Exhaustion score for any grouping used in the study. Full- and part-time therapists differed significantly in the mean number of patients seen per day and the number of hours per week spent on patient care. Therefore, each group was separately divided into the quartiles found in Table 6. Moderate burnout was seen in the Emotional Exhaustion subscale for all full-time therapists despite the number of hours per week devoted to patient care or the number of patients seen per day. Only part-time therapists who worked 15.1-20 hours per week or saw more than 15 patients per day had Emotional Exhaustion scores showing moderate burnout. DISCUSSION Burnout This group of physical therapists as a whole experienced less burnout based on the Maslach Burnout Inventory than other groups of physical and occupational therapists who have been evaluated using this instrument. Emotional Exhaustion was the only subscale on which the respondents consistently showed moderate burnout. This is not surprising because, of

RESEARCH STUDY MBI Subscale Scores Accomplishmenti Exhawtion' X SD 1 SD x SD Full time Number of hours per week therapist sees patients 530 (N = 102)' 41.7 4.7 19.3 9.5 5.3 4.6 30.1-35 (N = 69) 41.4 5.2 19.2 9.0 6.3 4.9 35.140 (N = 93) 41.7 5.0 20.1 8.6 5.9 4.7 240.1 (N = 65) 42.6 4.9 18.8 9.8 5.5 4.6 Number of patients seen per day 510 (N = 100) 41.4 4.9 19.6 8.8 5.2 4.5 10.1-14 (N = 79) 41.6 5.2 20.6 9.1 6.5 4.7 14.1-18 (N = 80) 41.9 4.8 18.8 8.8 5.8 4.5 218.1 (N = 74) 42.8 4.6 18.4 10.0 5.4 4.4 Part time Number of hours per week therapist sees patients 515 (N = 17) 40.3 5.8 12.4 5.7 4.0 3.9 15.1-20 (N = 14) 42.4 4.6 17.4 12.2 3.9 4.4 20.1-24 (N = 14) 40.3 4.4 15.9 9.5 4.1 3.2 224.1 (N = 8). 42.7 7.1 12.2 4.0 3.2 1.8 Number of patients seen per day 510 (N = 16) 40.4 5.4 13.2 5.5 3.1 2.0 10.1-12 (N = 12) 42.9 3.4 12.7 7.0 2.4 2.0 12.1-15 (N = 16) 41.8 5.0 15.1 12.4 5.6 5.1 215.1 (N = 8) 39.1 8.2 20.4 6.5 4.6 3.1 *,Vunil)rrs\wl dcpcnding on nun~l)rr oi rr,pondrnt, who '~nswrred thr qurstion. 53 1 = High; 32-38 = Moderate; 239 = low burnout. * 227 = High; 17-26 = Moderate; 5 16 = Low burnout. ' 2 13 = High; 7-12 = Moderate; 16 = Low burnout. TABLE 6. Maslach Burnout Inventory (MBI) subscale scores for full- and part-time therapists (X t SD). the three burnout components, emotional exhaustion is the closest to an orthodox stress variable. Stress can occur in endless types of situations (war, natural disaster), including unending types of work situations (18). Considering emotional exhaustion as the only measure of burnout, however, is insufficient because it ignores the other two components of selfevaluation (personal accomplishment) and relation to others (depersonalization) (1 1). Personal Data Comparing the effects of gender from other studies using therapists as their subjects was not possible. Respondents in those studies (6,22,23) were overwhelmingly female (113: 1 ), while the present study had a ratio of 1.6:l females to males. When both genders are more equally represented in studies of other health professionals, subscale scores are comparable with these orthopaedic therapists (2,21). The direct association between age and subscale scores is consistent with several studies of physical and occupational therapist. (6,22,23). Maslach (11) hypothesized that age reflects more than just the length of time on the job. With increased age, people show more stability and maturity, have a more balanced perspective on life, and are less prone to the excesses of burnout. Conversely, Brown and Pranger (1 ) found an inverse association with age and burnout scores. They surmised that psychiatric occupational therapy personnel who are older may not be skilled in the most current treatment techniques, and this may contribute to the feeling of burnout. In addition, older staff may be more cautious of change. Professional Data The number of years on the current job affected burnout levels, although not a. much as in the two studies involving rehabilitation therapist. (6,23). In the latter studies, Among orthopaedic therapists, Personal Accomplishment scores increased (showing lower burnout) as years at the current job increased. therapists had been at their current job an average of 3.0 and 3.6 years, respectively. Their subscales scores showed higher levels of burnout than those in the current study. Among orthopaedic therapist., Personal Accomplishment scores increased (showing lower burnout) as years at the current job increased; Emotional Exhaustion and Depersonalization scores decreased (lower burnout). Schlenz et al also found Personal Accomplishment scores increased a. years on the job increased. Correlations with this variable and the other two subscales were extremely low and not significant (23). In a sample of physicians who worked for health maintenance organizations, however, years at the current job did not affect Personal Accomplishment scores. Emotional Exhaustion and Depersonalization scores were directly related to the number of years at the current job (2). Volume 26 Number 3 September 1997 JOSPT

RESEARCH STUDY Setting appeared to influence burnout scores, at least for respondents who work in rehabilitation settings. The Emotional Exhaustion and Depersonalization subscale scores were the highest for any grouping (eg., age, number of patient. per day, etc.). The mean age of respondents in this study was older (36 years) than those in the two studies involving rehab therapist5 (27 and 32.5 years, respectively) (6,23). However, some factor(s) unique to this setting may make those who work there more prone to certain components of burnout. Because only a few respondents in this study worked in rehabilitation settings, one must be cautious in generalizing this study's results in this regard. Limitations There are several limitations to this study: 1) no effort was made to contact nonrespondent.; 2) the sample is likely biased; and 3) only a few variables that affect burnout were considered. Nonrespondent. were not contacted due to limited funding. This may have added to the bias of the sample, but, more importantly, it is possible that those who are most burned out had no time or energy to answer the survey. It is more likely that the most burned out are no longer members of the Orthopaedic Section, APTA, or even of the profession. The first bout of burnout usually happens in the first years of one's career in mental health workers (11,19). The data from this study and others involving physical and occupational therapist. would sup port this statement (6,22,23). If people have diff~culty in dealing with burnout when they are younger and newer on the job, they may leave the profession entirely. Because of this early dropout in the burned out, these therapists will not be around 5 or 10 years later to answer questions about the emotional strain of their work. Researchers examine variables they believe are associated with their hypotheses about causes and effect. of burnout. No one study can possibly examine all of the factors that may be associated with burnout. Where researchers correlated sub scale scores with the same variables used in this study, correlation coefficients with significance levels less than.o1 ranged from.o1 to.40. The strength of these coefficient5 would be considered, at best, low (5). Moderate to high correlation coefficient. are not often found when correlating variables and therapists' subscale scores. Perhaps subscale scores represent abstract constructs that are difficult to measure, and a correlation coefficient of less than.50 should be considered very strong (5). Further Research Investigation of burnout was undertaken to provide insight into factors that may be associated with burnout and, therefore, may identify areas for future study. Two possibilities are described below. Longitudinal studies tracking new graduates could concentrate on the process of burnout. This type of study would be one way to follow new graduates as they deal with job stresses and learn to cope with the many precipitating factors of burnout. Rogers and Dodson's study (22) of occupational therapists was the only one cited which evaluated burnout as it related to years of education. Their results showed a significant inverse association with level of education and Emotional Exhaustion and Depersonalization subscale scores, but the direct association of education with Personal Accomplishment scores was not significant. Maslach (1 1) found the greatest amount of burnout in providers who have earned a 4-year college degree, but had no postgraduate training. Overall, mental health care providers with less education showed less burn- out than their more educated colleagues. In particular, those who had some college experience but not a full 4 years had lower levels of burnout (11). None of the burnout studies in the field of physical therapy has included the assistant. The inclusion of subjects with differing levels of education may help clarify the relationship of education and burnout for physical therapy personnel. CONCLUSION These findings suggest that this group of physical therapists has a relatively low level of burnout. All variables thought to be associated with burnout had weak associations. However, orthopaedic therapists who had less than 5 years of experience or who worked in rehab settings demonstrated subscales scores which may indicate these therapists are at risk for burnout. JOSFT REFERENCES Brown GT, Pranger T: Predictors of burnout for psychiatric occupational therapy personnel. Can j Occup Ther 59:258-267, 1992 Deckard GI, Meterko M, Field 0: Physician burnout: An examination of personal, professional, and organizational relationships. Med Care 32:745-754, 1994 Deckard GI, Present RM: Impact of role stress on physical therapists' emotional and physical well-being. Phys Ther 69: 713-718, 1989 DePaepe j, Croce R: The existence of burn-out among corrective therapists: A descriptive analytical investigation. Am Correct Ther J 40:32-38, 1986 Domholdt E: Physical Therapy Research, Philadelphia, PA: W.B. Saunders Company, 1993 Donohoe E, Nawawi A, Wilker L, Schindler T, jette DU: Factors associated with burnout of physical therapists in Massachusetts rehabilitation hospitals. Ph ys Ther 73:750-76 1, 1993 Fimiam MI, Lieberman Rj, Fastenau PS: Development and validation of an instrument to measure occupational stress in speech-language pathologists. I Speech Hear Res 34:439-446, 1991 Freudenberger HI: Staff burn-out. j Soc Issues 30: 159-165, 1974 JOSR Volume 26 Number 3. September 1997

RESEARCH STUDY 9. Freudenberger HI: The staff burn-out syndrome in alternative institutions. Psychother Theory Res Practice 12:73-82, 1975 10. Maslach C: Burnout: A multidimensional perspective. In: Schaufeli WB, Maslach C, Merek T (eds), Professional Burnout: Recent Developments in Theory and Research, pp 19-32. Washington, DC: Taylor and Francis, 1993 1 1. Maslach C: Burnout: The Cost of Caring, Englewood Cliffs, NJ: Prentice- Hall, Inc., 1982 12. Maslach C, Jackson SE: Maslach Bumout lnventory Manual (1st Ed), Palo Alto, CA: Consulting Psychologists Press, 1 98 1 13. Maslach C, Jackson SE: Maslach Burnout lnventory Manual (2nd Ed), Palo Alto, CA: Consulting Psychologists Press, 1986 14. Offerman LR: Maslach Burnout Inventory. In: Keyser DJ, Sweetland RC (eds), Test Critiques, pp 419-426. Kansas City, MO: Test Corporation ofamerica, 1985 15. Oppenhein L: Factors related to occupational stress or burnout among music therapists. J Music Ther 24(2):97-106, 1987 16. Paine WS: Professional burnout: Some major costs. Fam Community Health 6(4): 1-9, 1984 17. Piedmont RL: A longitudinal analysis of burnout in the health care setting: The role of personal dispositions. J Pen Assess 6 1 :457-473, 1 993 18. Pines AM: Burnout: An existential perspective. In: Schaufeli WB, Maslach C, Merek T (eds), Professional Burnout: Recent Developments in Theory and Research, pp 33-52. Washington, DC: Taylor and Francis, 7 993 19. Pines A, Maslach C: Charactetistia of staff bumout in mental health settings. HHOY, Community Psychiatry 29233-237, 1988 20. Portney LC, Watkins MP: Foundations of Clinical Research: Applications to Practice, Nonvalk, CT: Appleton & Lange, 1993 21. Raquepaw JM, Miller RS: Psychotherapist burnout: A componential analysis. Professional Psycho1 Res Practice 20: 32-36, 1989 22. Rogers JC, Dodson SC: Burnout in occupational therapists. Am J Occup Ther 42:787-792, 1988 23. Schlenz KC, Guthrie MR, Dudgeon B: Burnout in occupational therapists and physical therapists working in head injury rehabilitation. Am J Occup Ther 49:986-993, 1995 24. Schuster ND, Nelson DL, Quisling C: Burnout among physical therapists. Phys Ther 64:299-303, 1984 25. Statistical Abstract of the United States (1 12th Ed), Washington, DC: U.S. Bureau of the Census, 1992 26. Wolfe GA: Burnout of therapists. Phys Ther 6l:lO46-1050, 1981 Volume 26 Number 3 September 1997 JOSFT