EDUCATIONAL ADVANCE

Factors Associated With Burnout During Emergency Medicine Residency James Kimo Takayesu, MD, MSc, Edward A. Ramoska, MD, MPH, Ted R. Clark, MD, MPP, Bhakti Hansoti, MBchB, MPH, Joseph Dougherty, DO, Will Freeman, MD, Kevin R. Weaver, DO, Yuchiao Chang, PhD, and Eric Gross, MD, MMM

Abstract Objectives: While the prevalence of burnout in practicing emergency physicians has been studied, little is known of the prevalence and risk factors in emergency medicine (EM) residents. The aim of this study was to assess the prevalence of burnout among EM residents and the individual-level factors associated with burnout. Methods: Eight EM residency programs were surveyed using the Maslach Burnout Inventory (MBI). Demographic data and data on job satisfaction and tolerance of uncertainty in clinical decision-making were collected using validated instruments. Results: Of 289 eligible residents, 218 completed the MBI (response rate = 75%). A total of 142 residents (65%) met the criteria for burnout. Complete data sets of the other instruments were obtained from 193 (response rate = 67%), and this group comprised our study population. Subjects having a significant other or spouse had a higher prevalence of burnout compared to single residents (60% vs. 40%, p = 0.002). Poor global job satisfaction (p < 0.0001), lack of administrative autonomy (p = 0.021), and lack of clinical autonomy (p = 0.031) correlated with burnout, as did intolerance of uncertainty (p = 0.015). Conclusions: Burnout is highly prevalent in EM residents. Interventions should be targeted at 1) improving resident autonomy in the emergency department where possible, 2) supervision and instruction on medical decision-making that may affect or teach individuals to cope with risk tolerance, and 3) social supports to reduce work–home conflicts during training. ACADEMIC EMERGENCY MEDICINE 2014;21:1031–1035 © 2014 by the Society for Academic Emergency Medicine

T

he term burnout was coined by Herbert Freudenberger in 1974.1 It has been described as a prolonged response to chronic emotional and interpersonal stressors on the job defined by three dimensions: emotional exhaustion, depersonalization, and a sense of reduced personal accomplishment.2

Attending physicians,3 residents in training,4 and even medical students5 have been found to suffer from burnout. The rate of burnout in residents of various specialties varies from 18% to 80%.4 Rates of burnout in attending emergency physicians (EPs) have been reported at 32%6

From the Department of Emergency Medicine, Massachusetts General Hospital (JKT, YC), Boston, MA; the Department of Emergency Medicine, Drexel University College of Medicine (EAR), Philadelphia, PA; the Division of Emergency Medicine, Southern Illinois University (TRC), Carbondale, IL; the Department of Emergency Medicine, Johns Hopkins University (BH), Baltimore, MD; the Department of Emergency Medicine, Ohio Valley Medical Center (JD), Wheeling, WV; the Department of Emergency Medicine, Louisiana State University Heath Baton Rouge Emergency Medicine Residency Program (WF), Baton Rouge, LA; the Department of Emergency Medicine, Lehigh Valley Health Network (KRW), Bethlehem, PA; and the Department of Emergency Medicine, Hennepin County Medical Center (EG), Minneapolis, MN. Dr. Gross is currently with the Department of Emergency Medicine, University of California at Davis, Davis, CA. Received February 4, 2014; revision received April 21, 2014; accepted April 28, 2014. Presented at the Society for Academic Emergency Medicine Mid-Atlantic Regional Meeting, Philadelphia, PA, February 2014; the Council of Emergency Medicine Residency Directors Academic Assembly, New Orleans, LA, March 2014; and the Society for Academic Emergency Medicine Annual Meeting, Dallas, TX, May 2014. The authors have no relevant financial information or potential conflicts of interest to disclose. Supervising Editor: Lalena Yarris, MD. Address for correspondence and reprints: Edward A. Ramoska, MD, MPH; e-mail: [email protected].

© 2014 by the Society for Academic Emergency Medicine doi: 10.1111/acem.12464

ISSN 1069-6563 PII ISSN 1069-6563583

1031 1031

1032

Takayesu et al. • FACTORS ASSOCIATED WITH BURNOUT DURING EMERGENCY MEDICINE RESIDENCY

to 60%,7 depending on the definition of burnout used. Burnout may lead to decreased productivity and decreased job satisfaction.6,8,9 Other negative effects include depression,10 risk of medical errors11 and adverse effects on patient safety,12,13 physical symptoms and illness,14 increased alcohol and drug use,14,15 and risk of suicidal ideation.16 Burnout in residents has been shown to be related to multiple problems, including poor health, mood disturbances, depression, alcohol abuse, and anxiety.4 Moreover, Kuhn et al.6 have shown that poor tolerance for uncertainty because of a concern for bad outcomes is correlated with burnout in practicing EPs. The prevalence of burnout in emergency medicine (EM) residents, and the individual characteristics associated with it, has not been well described. Before an effective plan to address burnout among EM residents can be developed and implemented, the scope of the problem must be clearly defined. The objective of this study was to assess the prevalence of burnout among EM residents in a geographically diverse cohort and to determine whether there are individual-level factors, such as demographics, levels of job satisfaction, and tolerance of uncertainty in clinical decision-making, associated with increased levels of burnout. METHODS Study Design This was a descriptive, multicenter, cross-sectional study of EM residents from eight training programs accredited by the Accreditation Council for Graduate Medical Education (ACGME) and/or American Osteopathic Association (AOA). The institutional review board at each institution approved the study, and written consent was obtained from each participant. Study Setting and Population All EM residents from the eight programs were invited to take part in the study. There were no exclusion criteria. The programs were located in urban, suburban, and rural settings throughout the Northeast, Midwest, and Southern United States. The participating programs were Brigham and Women’s/Massachusetts General Hospital, Drexel University College of Medicine/Hahnemann University Hospital, Hennepin County Medical Center, Lehigh Valley Hospital, Louisiana State University, Ohio Valley Medical Center, Southern Illinois University, and the University of Chicago. These eight residency programs represent the authors’ home institutions from the Medical Education Research Certificate program. Study Protocol In the fall of 2011, consenting residents took the Maslach Burnout Inventory (MBI) survey, the Physicians’ Reaction to Uncertainty (PRU) Scale, and an EM job satisfaction survey and filled out a demographic questionnaire. Data were gathered during the residents’ weekly conferences. Participants self-generated identification codes according to a prescribed algorithm to assure anonymity, while allowing for matching of the various surveys and questionnaires.

Four data collection instruments were employed. Demographic information was collected for medical degree, postgraduate year (PGY) of training, age range, sex, marital status, and number of children. The MBI survey is a validated 22-item questionnaire considered the criterion standard tool for measuring burnout.17–19 This survey measures three dimensions of burnout: emotional exhaustion, depersonalization, and reduced personal accomplishment. The MBI has a scoring rubric that categorizes each respondent into low, moderate, or high burnout levels for each scale. The revised PRU scale is a validated 15-item questionnaire that provides short, conceptually clear, psychometrically sound measures of physicians’ emotional reactions to uncertainty.20 The PRU addresses four areas: anxiety caused by uncertainty, concern about bad outcomes, reluctance to disclose uncertainty to patients, and reluctance to disclose mistakes to physicians; the scale ranges from 15 to 90 with higher numbers correlating with greater anxiety and concern regarding uncertainty. Job satisfaction was measured with a validated two-part questionnaire, which consisted of the Emergency Physician Job Satisfaction (EPJS) instrument, and the Global Job Satisfaction (GJS) instrument.21,22 The EPJS survey consists of 82 questions and addresses administrative autonomy, clinical autonomy, availability of resources, work relationships, lifestyle, and job challenges; the GJS consists of 12 questions. Both instruments use a seven-point Likert scale, which ranges from strongly disagree to strongly agree and is scored from –3 to +3. Data Analysis The raw data collection instruments were sent to a central location where they were coded and the results were entered into an Excel spreadsheet. Burnout was dichotomized into yes/no, with burnout defined as meeting the MBI definitions of high emotional exhaustion, high depersonalization, or low personal accomplishment.6 Continuous variables were summarized using mean  SD, and categorical variables were summarized using frequency and percentage. Correlations between burnout and demographic data were examined using chi-square tests with statistical significance defined as a two-sided p < 0.05. Job satisfaction and intolerance of uncertainty data were analyzed using two-tailed t-tests, because the data were roughly normally distributed. We did not adjust the error rate for multiple comparisons. With the seven demographic factors we examined, the familywise Type I error rate was 0.3. Because the number of eligible subjects was fixed, we did not carry out a formal a priori sample size calculation. As a post hoc calculation for the factors with nonsignificant differences, the maximum observed difference between groups was less than 8%. Data analysis was performed using SAS version 9.3. RESULTS There were 289 eligible residents in the eight participating residency programs. A total of 218 residents (response rate = 75%) completed the MBI and demographic data forms; of these, 142 residents (65%) met the criteria for burnout. Thirty-three percent of residents

ACADEMIC EMERGENCY MEDICINE • September 2014, Vol. 21, No. 9 • www.aemj.org

Table 1 Maslach Burnout Inventory and Burnout Scoring Variable Total Emotional exhaustion High Moderate Low Depersonalization No response High Moderate Low Personal accomplishment No response Low Moderate High Burnout scoring High Low

Number

Table 2 Demographic Data Percentage Characteristic

218 72 82 64

1033

33 38 29

1 128 51 38

0.5 59 23 17

3 129 56 30

1 59 26 14

142 76

65 35

had high degrees of emotional exhaustion, 59% had high degrees of depersonalization, and 59% had a low sense of personal accomplishment (Table 1). The demographic data are shown in Table 2. Many of the demographic characteristics did not correlate with burnout, including age group, PGY level, sex, having children, having advanced degree training, residency site, and 3-year versus 4-year program. There was no difference between ACGME-accredited programs and AOA-accredited programs. Residents responding as being married or having a significant other did have a significantly higher prevalence of burnout compared to single residents (60% vs. 40%, p = 0.002). Male residents in a relationship (married or significant other) had a higher prevalence of burnout (n = 71, 80%) when compared to female residents in a relationship (n = 43, 65%) and single female or male residents not in a relationship (n = 46, 57%; and n = 58, 53% respectively). Complete data sets of the other instruments were obtained from 193 residents (response rate = 67%); this group comprised our study population. These results are shown in Table 3. There were significant correlations between burnout and poor resident job satisfaction scores in administrative autonomy, clinical autonomy, and global job satisfaction. There was no significant correlation of burnout to resources in the clinical environment, work relationships, wellness, and challenges associated with clinical work. Intolerance of uncertainty also was correlated significantly with burnout. DISCUSSION We found a high rate of burnout among EM residents that was not related to level of training or other demographic factors. Using the more conservative approach of dichotomizing burnout into yes/no rather than rating it as low/moderate/high, our study found the prevalence of burnout to be 65%. This method is in agreement with

Respondents

Total PGY 1 2 3 4 Age range (yr) 25–29 30–34 35–39 40+ Sex Female Male Relationship status Single Married/ significant other No response/ unknown Children No Yes No response/ unknown Degree MD DO No response Residency format 3-year 4-year

%

218

Total Residents

%

pvalue*

289 0.97

70 67 63 18

32 31 29 8

92 88 88 21

32 30 30 7

115 74 21 8

53 34 10 4

164 95 22 8

57 33 8 3

89 129

41 59

122 167

42 58

104 114

48 52

122 132

42 46

0

0

35

12

154 34 30

71 16 14

219 37 33

76 13 11

144 43 31

66 20 14

230 59 0

80 20 0

0.72

0.75

Factors associated with burnout during emergency medicine residency.

While the prevalence of burnout in practicing emergency physicians has been studied, little is known of the prevalence and risk factors in emergency m...
222KB Sizes 0 Downloads 6 Views