SECTION I: SYMPOSIUM I: C. T. Brighton/ABJS Workshop on Orthopaedic Education
Residency training in orthopaedic surgery is known to be challenging and stressful. In a previous survey of residents and faculty in orthopaedic surgery, residents showed high levels of burnout, emotional exhaustion, and depersonalization. (Supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article then click on the Article Plus link on the right.) Over 30% of residents showed psychological decompensation compared with only 8% of faculty.5 A number of parameters correlated with the incidence and degree of burnout, most notably the number of hours worked per week.
In recent years the number of hours worked by residents has been under increased scrutiny. In 2003 the Accreditation Council on Graduate Medical Education (ACGME) implemented guidelines limiting the number of hours worked on average per week, the number of hours worked continuously on a single shift, and the number of days worked per week without a day free of clinical responsibilities.3 This policy was implemented on July 1, 2003.
We wondered whether the change in work hours would achieve some of the goals of the ACGME by reducing measures of burnout.
MATERIALS AND METHODS
The first survey was completed in 2002, the year before ACGME duty hour standards went into effect. The study group for this first survey consisted of 21 orthopaedic surgery residents from a university-based training program and 25 full-time orthopaedic surgery faculty members from two institutions. The second survey was administered in 2005, two years after the duty hour standards had gone into effect. Twenty faculty and 34 residents from the same two institutions completed the second survey. The study was approved by the institutional review boards of both institutions.
The survey instrument was divided into six sections. Section one contained background information, such as age, gender, year in training, race, marital status, number of children, and family background information. Section two consisted of the Maslach Burnout Inventory, a validated instrument consisting of 22 questions that evaluate various aspects of emotional functioning affected by work stress. The three subscales of the Maslach Burn-out Inventory assess emotional exhaustion, depersonalization, and personal achievement.4 The first two correlate with burnout while the third is inversely related to burnout.
Section three comprised the General Health Questionnaire-12, a widely used validated grading scale, used to assess psychiatric morbidity. The questionnaire is a reliable indicator of anxiety, depression, somatic symptoms, and social dysfunction.6 Scores of four or higher are consistent with decompensation.2
'Section four consisted of 23 questions that assessed subjective ratings of sources of worklife stress. Section five consisted of 18 questions focusing on stress coping methods. Section six consisted of the 14-item revised dyadic adjustment scale, which is another validated instrument that assesses marital adjustment on a 69-point scale. Scores of less than 46 are considered indicative of a distressed relationship.1
Descriptive statistics, including pair-wise correlations, on the graded questionnaires were computed. Bivariate relationships were tested with a simple correlation coefficient. Spearman correlations and the Pearson correlation coefficient were used to determine clinical significance. A p value < 0.05 was considered to be significant, p values between 0.05 and 0.1 were considered indicative of a trend, and p values < 0.15 were considered suggestive of a possible trend.
A detailed discussion of the results of the first survey have been previously published.5 There were no significant differences in age, gender, or marital status among the groups between the pilot study and the current study, although there was a trend towards a higher percentage of married residents in the current study (p = 0.1, Table 1). A brief summary of the findings of the first survey follows. The residents rated in the upper third of the burnout scale for emotional exhaustion and depersonalization and the middle third for personal achievement, while the faculty scored in the lower third for emotional exhaustion, the middle third for depersonalization, and in the upper third for personal achievement. On the general health questionnaire, seven residents (33%) and two faculty members (8%) had a score ≥ 4, indicating psychiatric morbidity (p < 0.05).
Three residents and six faculty members scored in the distressed range, which was not a significant difference. Factors that significantly correlated positively with resident burnout included anxiety about competence, increased work hours, work/home life conflict, and stressed relationships with nurses, faculty and senior residents. Factors correlating positively with the presence of faculty burnout also included anxiety about competence, as well as alcohol use, worry about a future with orthopaedics, and stress in relationships with other faculty members. Factors that were inversely proportional to resident burnout consisted of potentially protective factors including having a father who was a physician, time alone with spouse, parenthood and satisfaction in talking with colleagues. Factors inversely proportional to faculty burnout included perceived support from faculty, increase in work hours, quality of marriage, and work-life balance, as well as hours that the spouse worked outside the home (Tables 2 and 3). A decrease in resident work hours, however, seemed to have a positive effect. It is interesting to note that an increase in reported work hours among faculty was inversely related to burnout and one can only speculate as to the reason for this. Faculty members certainly generally have greater control over their workload and it is possible that those faculty members that are busier and choose to work longer hours may have a greater sense of personal achievement and be less prone to burnout. It remains to be seen whether this finding will be substantiated by a larger, more broadly based study, but this is one of many questions that will be explored further in the follow-up study.
There was a decline (p < 0.0001) in the reported number of hours worked by residents between the first and second survey (70.4 versus 88.3). A decline in hours was seen at every year level with the greatest difference at Year 4 and the least at Year 3 (Table 4).
Results of the second survey showed improvements in several aspects of resident functioning, but not in faculty functioning. Compared to preduty hour standards, scores for emotional exhaustion were lower (p = 0.056) among the current group of residents (22.3 versus 27.5), while the score of the current faculty group remained virtually identical to those in the original survey (16.4 versus 16.2). The depersonalization score was also somewhat lower (p = 0.14) among residents in the more recent survey (12.5 versus 15.2) and this indicated a possible trend. A sample size calculation indicates that an increase in the study group size to 120 would be necessary to achieve statistical significance for emotional exhaustion and to 172 subjects for depersonalization. The score for personal accomplishment was higher (p < 0.01) among the current group of residents (40.4 versus 34.8), while again the scores among the faculty were essentially identical to the original survey (42.2 versus 42.8). The scores on the General Health Questionnaire indicated five of 34 residents (15%) scored ≥ 4 compared with seven of 21 (33%) in the first survey (p < 0.2). The incidence of scores ≥ 4 in the faculty group stayed essentially the same (15% vs 8%). There were no differences observed in the revised dyadic adjustment scale (assessment issues) of relationship between the first and second survey.
The ACGME duty hour standards include an 80-hour week limit averaged over four weeks, at least one day out of seven free of clinical and educational activities, continuous work hour limit of 24 hours plus an additional 6 hours of administrative time, and an on-call limit of every third night. The first survey was administered a year before the implementation of these standards when there was no formal tracking of resident duty hours. The results of the first survey indicated high levels of stress and burnout among residents, substantially more so than among the faculty. The duty hour standards had a number of potential benefits, including improving patient care and minimizing medical errors, minimizing sleep deprivation, and potentially improving the quality of life and educational experience of residents. Our results indicate it is likely the duty hour standards have had some positive impact on decreasing the burnout observed among orthopaedic residents. A positive impact on the score of all three subcategories of the Maslach Burnout Inventory was observed. The improvement in the score for personal accomplishment was high, the decline in emotional exhaustion showed a very strong trend toward association, and the decline in the depersonalization score also showed a possible trend towards association.
We note several limitations of these surveys. The sample size is small and there were only two training programs involved in the study. The same group of residents did not complete the survey before and after the implementation of the duty hours. It is possible other changes in the training programs could have had an impact on the scores between the two time intervals. A more robust study would have had a larger sample size, higher number of programs surveyed, and preferably the same group of residents and faculty completing the survey before and after the implementation of the standards. Finally, it was not possible to determine what percentage of faculty and residents completed both surveys because they were completed voluntarily and anonymously.
Despite these limitations, it is striking that changes were observed with such a small sample size. It does seem likely implementing the duty hour standards has been associated with a positive effect on the incidence of burnout among orthopaedic surgery residents. A larger scale survey has been recently completed involving 50 training programs and over 1000 responses from orthopaedic surgery residents, faculty, and their significant others to examine how coping methods are involved with orthopaedic surgery training.
1. Busby DM, Christensen C, Crane DR, Larson JH. A revision of the dyadic adjustment scale for use with distressed and nondistressed couples: construct hierarchy and multidimensional scales. J Marital Fam Ther
2. Goldberg D, Williams P. A User's Guide to the General Health Questionnaire
. Windsor, UK: NFER-Nelson Publishing; 1988.
3. Graduate Medical Education Directory 2004-2005. Chicago. IL: American Medical Association
. 2004;18- 21:196-201.
4. Maslach C, Jackson SE, Leiter MP. Maslach Burnout Inventory Manual
. 3rd ed. Palo Alto, CA: Consulting Psychologists Press; 1996.
5. Sargent MC, Sotile W, Sotile M, Rubash H, Barrack RL. Stress and coping among orthopaedic surgery residents and faculty. J Bone Joint Surg Am
© 2006 Lippincott Williams & Wilkins, Inc.
6. Weir E. Jung at heart: assessing one's suitability for medical training. CMAJ