Korean J Fam Med. 2013;34:36-42 http://dx.doi.org/10.4082/kjfm.2013.34.1.

36

Original
Association of Intern and Resident Burnout Article
with Self-Reported Medical Errors

Eui-Kyu Kang, Ho-Seob Lihm, Eun-Hee Kong*

Department of Family Medicine, Kosin University College of Medicine, Busan, Korea

Background: Burnout is a common problem for interns and residents. It may be related to medical error, but little is known
about this relationship. The purpose of this study was to determine the association between burnout and perceived
medical errors among interns and residents.
Methods: The study group consisted of interns and residents working in a university hospital in Busan. Data were provided
by 86 (58.5%) of 147 interns and residents. They completed a questionnaire including self-assessment of medical errors,
a linear analog self-assessment of overall quality of life (QOL), fatigue, the Epworth Sleepiness Scale (ESS) score, the
Maslach Burnout Inventory, and a validated depression screening tool.
Results: According to univariate logistic regression analyses, there was an association between perceived medical errors
and fatigue (odds ratio [OR], 1.37 per unit increase; 95% confidence interval [CI], 1.12 to 1.69; P < 0.003) and ESS scores
(OR, 1.13 per unit increase; 95% CI, 1.03 to 1.23; P < 0.009). Perceived medical errors were also associated with burnout
(ORs per 1-unit change; emotional exhaustion OR, 1.07; 95% CI, 1.02 to 1.13; P < 0.005; depersonalization OR, 1.11; 95%
CI, 1.02 to 1.21; P < 0.013), a negative depression screen (OR, 0.29; 95% CI, 0.11 to 0.76; P < 0.013), and overall QOL (OR,
0.80; 95% CI, 0.70 to 0.98; P < 0.033). In multivariate logistic regression analyses, an association was identified between
perceived medical errors and emotional exhaustion (OR, 1.06; 95% CI, 1.00 to 1.11; P < 0.046) when adjusted for ESS, and
depersonalization (OR, 1.01; 95% CI, 1.01 to 1.19; P < 0.04) when adjusted for fatigue.
Conclusion: Higher levels of burnout among interns and residents were associated with perceived medical errors.

Keywords: Burnout; Resident; Medical Errors

INTRODUCTION and residents must pass intense training courses to acquire the
required skills. A resident receives training in an educational
hospital or medical institute in order to acquire a license as a
Becoming a physician requires a high level of medical
medical specialist. According to data collected by the Korean
knowledge and skill in order to deal with issues of life and death,
Medical Association, there were 14,633 residents as of December
Received: July 29, 2011, Accepted: January 4, 2013 31, 2008, accounting for 18.6% of the 78,518 total physicians
*Corresponding Author: Eun-Hee Kong working in the medical field. Most residents work in educational
Tel: +82-51-990-6476, Fax: +82-51-990-3045 hospitals, where they often play important roles as primary
E-mail: [email protected] physicians. Like students of medical schools or graduate schools
Korean Journal of Family Medicine of medicine, they are also in training to become qualified medical
Copyright © 2013 The Korean Academy of Family Medicine specialists. However, residents work in educational hospitals
This is an open-access article distributed under the terms of
and have a lower economic status and social position relative to
the Creative Commons Attribution Non-Commercial License
(http://creativecommons.org/licenses/by-nc/3.0) which permits the larger community of physicians. Residents often commit a
unrestricted noncommercial use, distribution, and reproduction in variety of medical errors because of heavy job stress and fatigue.
any medium, provided the original work is properly cited.

36 | Vol. 34, No. 1 Jan 2013 Korean J Fam Med

 Eui-Kyu Kang, et al: Association of Intern and Resident Burnout with Self-Reported Medical Errors

In addition, hospitals often fail to provide management to reduce explained in the questionnaire. Fatigue was estimated using linear
job stress, and residents often feel that they cannot state their analog self-assessment (LASA) questions. Respondents measured
needs due to their position as trainees. Hospitals often consider their fatigue from 1 (not fatigued) to 10 (always fatigued); a higher
them as low-cost operating labor and consider the improvement score corresponded to greater fatigue. Sleepiness was assessed with
of systems or facilities associated with medical training as a the Epworth Sleepiness Scale. This tool measures daytime sleepiness
hindrance to hospital operations, and therefore little effort is according to eight items for which respondents were instructed to
made to improve conditions for residents and their training rate on a Likert scale from 0 (not sleepy at all) to 3 (very sleepy).
program. A total score higher than 10 points was considered indicative of
Medical errors and the safety of patients are important daytime sleepiness.7,8) Quality of life (QOL) was assessed with
issues for both patients and physicians. In the US in 1999, it LASA questions by measuring the overall quality of life from 1
was reported that about 100,000 patients died from previously (very bad) to 10 (very good).9-11) Burnout was defined as a feeling
preventable side effects and this has significant implications of continued and repeated emotional pressure. Maslach et al.12)
for both physicians and patients.1) Studies regarding medical suggested that the sub-areas of burnout consist of emotional
errors report that resident fatigue and sleepiness are major exhaustion, depersonalization, and reduced professional efficacy
1-4)
factors in medical errors. Another study reports that resident or personal accomplishment. Resident burnout was measured
burnout from stress is a major cause of self-reported medical with a burnout assessment tool that had been translated and
5) 6)
errors. Continued and repeated stress results in burnout by used in a previous study of the factors affecting social worker
inducing physical, emotional, and mental exhaustion. It is easy exhaustion.13,14) A total of 22 questions (9 questions for emotional
to understand why fatigue may cause medical errors. However, exhaustion, 8 questions for personal accomplishment, and 5
because fatigue or sleepiness may affect burnout, there is a need questions for depersonalization) assessed burnout, and each was
for studies solely on burnout apart from fatigue and sleepiness. measured on a 7-point Likert scale from “1: very weak, scarcely
This study intends to help improve both patient safety and felt” to “7: very strong, very frequently felt.” The Maslach burnout
resident training conditions by understanding the relationship inventory (MBI) scale is a tool produced by Maslach et al.12)
between resident burnout and subsequent medical errors. in 1981 for measuring burnout and has proven both reliable
and valid in estimating the degree of burnout experienced by
subjects. In addition, because the fact that this tool has been
METHODS used in a number of previous studies across many occupations,
it may provide a basis for interpreting results of this case.15-18)
1. Subjects and Study Period In the screening test for depression, two questions provided by
In order to investigate the relationship between resident Spitzer et al.19) were used.20) These two questions were “1) during
burnout and self-reported medical errors, a survey of 147 the past month, have you often been bothered by feeling down,
residents in a university hospital in Busan was performed depressed, or hopeless?” and “2) during the past month, have
from July to August, 2010. Residents completed a self-answer you often been bothered by little interest or pleasure in doing
questionnaire. Among the 147 distributed questionnaires, 61 things?” If the subject answered “yes” to either of these questions,
were not returned. Therefore, a total of 86 questionnaires were the screening was considered positive for depression. In previous
collected, yielding a response rate of 58.5%. studies, this screening test was proved to be reasonable compared
to other screening tests for depression.21, 22)
2. Methods
Self-reported medical errors were assessed by the question, 3. Analysis
“have you committed a medical error in the last three months?” The t-test and the chi-square test were used to identify the
Medical errors were defined as subjective errors recognized by relationship between self-reported medical errors of residents
residents rather than events that harmed patients, and this was and fatigue, daytime sleepiness, quality of life, burnout, and

Korean J Fam Med Vol. 34, No. 1 Jan 2013 | 37

Eui-Kyu Kang, et al: Association of Intern and Resident Burnout with Self-Reported Medical Errors

depression. In order to determine whether burnout is associated of 35 subjects who reported no medical errors were positive in
with self-reported medical errors, univariate and multivariate the screening test for depression.
logistic regression analyses were performed and P < 0.05 was
considered statistically significant. All data were analyzed using
SPSS ver. 12.0 (SPSS Inc., Chicago, IL, USA).
Table 1. Participant demographics (n = 86)

Variable Value
RESULTS Age (y) ≤30 67 (77.9)

>30 19 (22.1)
1. General Features of Subjects Sex Men 64 (74.4)
Data were provided by 86 of 147 interns and residents and Women 22 (25.6)
the characteristics of participants are shown in Table 1. Course Intern 19 (22.1)

Resident 1 20 (23.3)
2. Comparison of Self-Reported Medical Errors Resident 2 23 (26.7)
with Respect to Survey Parameters
Resident 3 18 (20.9)
In this study, 51 (59%) subjects reported medical errors.
Resident 4 6 (7.0)
Summary measures to identify general associations between self-
Marital status Single 61 (70.9)
reported medical errors and degree of fatigue, sleepiness, QOL,
Married 25 (29.1)
burnout, and symptoms of depression are shown in Table 2.
No. of child 0 72 (83.8)
Interns and residents who reported medical errors experienced
1 7 (8.1)
greater fatigue and sleepiness and had significantly lower overall
2 7 (8.1)
QOL and higher levels of burnout as evidenced by increased
depersonalization and emotional exhaustion. In the screening test Working hour per day 15.6 ± 4.8

for depression, 42 (82.4%) of 51 subjects who reported medical Sleeping hour per day 5.0 ± 1.1

errors were positive in the depression screening and 20 (57.1%) Values are presented as number (%) or mean ± SD.

Table 2. Comparison of residents reporting no perceived errors vs. reporting perceived errors

Medical errors
Variable Metric (scale) Difference (95% CI) P-value*
No (n = 35) Yes (n = 51)

Fatigue LASA fatigue (0–10) 5.9 ± 2.5 7.5 ± 2.0 −1.6 (−2.6 to −0.6) 0.001

Sleepiness ESS (0–24) 12.0 ± 5.5 15.2 ± 5.1 −3.2 (−5.5 to −1.0) 0.006

QOL LASA overall QOL (0–10) 4.9 ± 2.5 3.8 ± 2.0 1.1 (0.1 to 2.0) 0.029

Burnout MBI-EE (9–63) 33.9 ± 9.6 40.3 ± 9.6 −6.5 (−10.7 to −2.3) 0.003

MBI-PA (8–56) 34.4 ± 7.1 32.4 ± 7.8 2.0 (−1.3 to 5.3) 0.236

MBI-DP (5–35) 15.9 ± 5.6 19.1 ± 5.5 −3.2 (−5.6 to −0.8) 0.010

Depression Any positive 2-item depression screen 20 (57.1) 42 (82.4) −25.2% (−44.7% to −5.8%) 0.015

Values are presented as mean ± SD or number (%).

CI: confidence interval, LASA: linear analog self-assessment, ESS: Epworth Sleepiness Scale, QOL: quality of life, MBI-EE: Maslach burnout
inventory-emotional exhaustion, MBI-PA: Maslach burnout inventory-personal accomplishment, MBI-DP: Maslach burnout inventory-
depersonalization.
*Obtained by t-test or chi-square test.

38 | Vol. 34, No. 1 Jan 2013 Korean J Fam Med

 Eui-Kyu Kang, et al: Association of Intern and Resident Burnout with Self-Reported Medical Errors

Table 3. Unadjusted association of fatigue, sleepiness, QOL, burnout, and symptoms of depression with self-reported medical errors (n = 86)

Variable Metric OR (95% CI) P-value*

Fatigue LASA fatigue (0–10) 1.37 (1.12–1.69) 0.003

Sleepiness ESS (0–24) 1.13 (1.03–1.23) 0.009

QOL LASA overall QOL (0–10) 0.80 (0.70–0.98) 0.033

Burnout MBI-EE (9–63) 1.07 (1.02–1.13) 0.005

MBI-PA (8–56) 0.97 (0.91–1.02) 0.235

MBI-DP (5–35) 1.11 (1.02–1.21) 0.013

Depression Any positive 2-item depression screen 0.29 (0.11–0.76) 0.013

QOL: quality of life, OR: odds ratio, CI: confidence interval, LASA: linear analog self-assessment, ESS: Epworth Sleepiness Scale, MBI:
Maslach burnout inventory, EE: emotional exhaustion, PA: personal accomplishment, DP: depersonalization.
*Univariate logistic regression analysis.

Table 4. Adjusted association of fatigue, sleepiness, QOL, burnout, and depression symptoms with self-reported medical errors (variable
adjusted for sleepiness)

Variable Metric OR (95% CI) P-value*

Quality of life LASA overall QOL 0.85 (0.69–1.05) 0.13

Burnout MBI-EE 1.06 (1.00–1.11) 0.046

MBI-PA 0.96 (0.91–1.03) 0.255

MBI-DP 1.08 (0.99–1.18) 0.085

Depression Any positive 2-item depression screen 0.40 (0.14–1.14) 0.086

QOL: quality of life, OR: odds ratio, CI: confidence interval, LASA: linear analog self-assessment, MBI: Maslach burnout inventory, EE:
emotional exhaustion, PA: personal accomplishment, DP: depersonalization.
*Multivariate logistic regression analysis adjusted for sleepiness.

Table 5. Adjusted association of fatigue, sleepiness, QOL, burnout, and depression symptoms with self-reported medical errors (variable
adjusted for fatigue)

Variable Metric OR (95% CI) P-value*

Quality of life LASA overall QOL 0.86 (0.70–1.07) 0.185

Burnout MBI-EE 1.05 (0.99–1.11) 0.104

MBI-PA 0.94 (0.88–1.01) 0.089

MBI-DP 1.01 (1.01–1.19) 0.04

Depression Any positive 2-item depression screen 0.43 (0.15–1.24) 0.118

QOL: quality of life; OR: odds ratio, CI: confidence interval, LASA: linear analog self-assessment, MBI: Maslach burnout inventory, EE:
emotional exhaustion, PA: personal accomplishment, DP: depersonalization.
*Multivariate logistic regression analysis adjusted for fatigue.

Korean J Fam Med Vol. 34, No. 1 Jan 2013 | 39

Eui-Kyu Kang, et al: Association of Intern and Resident Burnout with Self-Reported Medical Errors

3. Univariate Analysis for Factors Related to study conducted by West et al.21) that found higher degrees of
Medical Errors emotional exhaustion and depersonalization and lower feelings
In the univariate logistic regression analysis of medical errors, of personal accomplishment were associated with a higher
it was found that the odds ratio (OR) of fatigue was 1.37 (95% frequency of medical errors.27) In a study of internal medicine
confidence interval [CI], 1.12 to 1.69) and the OR of daytime physicians at the Mayo, West et al.21) found that burnout was
sleepiness was 1.13 (95% CI, 1.03 to 1.23). The OR of quality of related to self-reported medical errors, independent of fatigue.
life was 0.80 (95% CI, 0.70 to 0.98) and in the analysis of burnout, In particular, when controlling for daytime sleepiness, severe
the OR of emotional exhaustion was 1.07 (95% CI, 1.02 to 1.13), emotional exhaustion and depersonalization were associated with
the OR of depersonalization was 1.11 (95% CI, 1.02 to 1.21), higher rates of medical errors. When controlling for fatigue, severe
and the OR of personal accomplishment was 0.97 (95% CI, 0.91 emotional exhaustion and depersonalization and a lower sense
to 1.02). However, the results for personal accomplishment were of personal accomplishment were also associated with medical
not found to be statistically significant. For those subjects who errors. In this study, the results of univariate analysis suggest that
were found to be negative for depression, the OR was 0.29 (95% emotional exhaustion and depersonalization are associated with
CI, 0.11 to 0.76) (Table 3). medical errors. In the multivariate logistic regression analysis
controlling for daytime sleepiness, only emotional exhaustion was
4. Multivariate Analysis for Medical Error- significantly related with medical errors and depersonalization
Related Factors Adjusted for Sleepiness and was associated with daytime sleepiness. Depersonalization is
Fatigue defined as a state in which patients are regarded as cases rather
After adjusting for sleepiness and fatigue, a multivariate than as people and physicians may come to develop a negative or
logistic regression analysis was performed. When controlling for sarcastic attitude toward patients or their care-givers and become
daytime sleepiness, the OR of emotional exhaustion was 1.06 uncooperative.12) Because residents go without sleep for long
(95% CI, 1.00 to 1.11) and when controlling for fatigue, the OR periods during their training, they often feel sleepy in the daytime
of depersonalization was 1.01 (95% CI, 1.01 to 1.19) (Tables 4, in spite of sufficient sleep at night. As a result, it is thought that
5). residents often lose motivation for their work and come to have
passive attitudes toward their patients. When controlling for
fatigue, a higher degree of depersonalization was associated
DISCUSSION with significantly more medical errors. It is thought that if
someone is tired and his/her energy is depleted, mental and
Although stress experienced by residents in hospitals physical resources are also reduced, finally leading to emotional
is unavoidable, the degree of stress may be greater than that exhaustion. Comparing this study with the study by West et al.21)
experienced by those in other professions.23) Some amount of there were some statistically significant differences with respect
stress may be helpful for the management of patients, but when to the sub-areas of burnout, but these differences were due to
sustained unbearable stress is experienced, this may affect physical the longer duration of their study period (from 2003 to 2008,
24,25)
and mental health and result in burnout. Burnout, a symptom with three month intervals) and larger number of subjects (430
caused by continued stress, is a result of emotional pressure residents of internal medicine). Maslach et al.12) defined burnout
developed by long-term repeated exposure to continuous as over 27 points for emotional exhaustion or over 10 points for
26) 12)
occupational stress. According to Maslach et al., burnout can depersonalization, and emphasized that the three areas of burnout
be divided into emotional exhaustion, depersonalization, and should be considered separately because they represent different
personal accomplishment, and can be measured by the MBI scale. aspects of this condition.13) Therefore, this study suggests that
In this study, we found that subjects who experienced high burnout is associated with self-reported medical errors regardless
degrees of emotional exhaustion and depersonalization reported of fatigue and daytime sleepiness.
more medical errors. This result was partially consistent with a The study by West et al.21) indicated that burnout, as well as

40 | Vol. 34, No. 1 Jan 2013 Korean J Fam Med

 Eui-Kyu Kang, et al: Association of Intern and Resident Burnout with Self-Reported Medical Errors

fatigue, must be addressed in the discussion of the improvement REFERENCES

of medical training courses in the US.27) Similarly, it may be
beneficial when planning medical training courses in Korea, 1. Kohn LT, Corrigan J, Donaldson MS; Institute of Medicine
to recognize a need to develope a strategy to alleviate burnout (US); Committee on Quality of Health Care in America.
because burnout negatively affects the quality of medical service. To err is human: building a safer health system. Washington
17)
Furthermore, like the study by Gopal et al. reporting that when (DC): National Academy Press; 1999.
resident working hours were reduced, the degree of burnout 2. Landrigan CP, Rothschild JM, Cronin JW, Kaushal R,
decreased, more studies that affect burnout are needed. Burdick E, Katz JT, et al. Effect of reducing interns’ work
There are several limitations to this study. First, although hours on serious medical errors in intensive care units. N Engl
the purpose of this study was to assess the relationship between J Med 2004;351:1838-48.
burnout and medical errors, it was difficult to identify a causal 3. Barger LK, Ayas NT, Cade BE, Cronin JW, Rosner B, Speizer
relationship because medical errors were reported for the FE, et al. Impact of extended-duration shifts on medical
previous three months and burnout was assessed at the time errors, adverse events, and attentional failures. PLoS Med
of the survey. Because of this issue, results are best interpreted 2006;3:e487.
as associations rather than as definitive evidence of causation. 4. Lockley SW, Barger LK, Ayas NT, Rothschild JM, Czeisler
Second, because medical errors were evaluated via self-report, CA, Landrigan CP, et al. Effects of health care provider work
it did not directly reflect clinically documented medical errors hours and sleep deprivation on safety and performance. Jt
and could be arbitrary. Third, severity of medical errors was not Comm J Qual Patient Saf 2007;33(11 Suppl):7-18.
considered. It was impossible to judge precisely if self-reported 5. Fahrenkopf AM, Sectish TC, Barger LK, Sharek PJ, Lewin
medical errors resulted in actual harm to patients. However, it is D, Chiang VW, et al. Rates of medication errors among
necessary to recognize that self-reported medical errors may be depressed and burnt out residents: prospective cohort study.
more likely to represent preventable medical errors. The fourth BMJ 2008;336:488-91.
is the fact that the response rate was only 58.5%, and an analysis 6. Maslach C. Burned-out. Hum Behav 1976;5:16-22.
of characteristics of the non-respondents was not carried out. 7. Johns MW. A new method for measuring daytime sleepiness:
Fifth, because this study was performed at a university hospital the Epworth sleepiness scale. Sleep 1991;14:540-5.
in Busan, results may not be applicable to all medical residents. 8. Johns MW. Reliability and factor analysis of the Epworth
In the future, larger sample sizes may be necessary to better Sleepiness Scale. Sleep 1992;15:376-81.
generalize this association. However, considering the fact that 9. Gudex C, Dolan P, Kind P, Williams A. Health state valuations
Korean studies of medical errors of residents are lacking, this from the general public using the visual analogue scale. Qual
study is an important step in identifying the relationship between Life Res 1996;5:521-31.
resident burnout and self-reported medical errors in spite of its 10. Shanafelt TD, Novotny P, Johnson ME, Zhao X, Steensma
limitations. DP, Lacy MQ, et al. The well-being and personal wellness
promotion strategies of medical oncologists in the North
Central Cancer Treatment Group. Oncology 2005;68:23-32.
CONFLICT OF INTEREST 11. Rummans TA, Clark MM, Sloan JA, Frost MH, Bostwick
JM, Atherton PJ, et al. Impacting quality of life for patients
No potential conflict of interest relevant to this article was with advanced cancer with a structured multidisciplinary
reported. intervention: a randomized controlled trial. J Clin Oncol
2006;24:635-42.
12. Maslach C, Jackson SE, Leiter MP. Maslach burnout
inventory manual. 3rd ed. Palo Alto: Consulting Psychologists
Press; 1996.

Korean J Fam Med Vol. 34, No. 1 Jan 2013 | 41

Eui-Kyu Kang, et al: Association of Intern and Resident Burnout with Self-Reported Medical Errors

13. Maslach C, Jackson SE. Maslach burnout inventory. Palo 21. West CP, Huschka MM, Novotny PJ, Sloan JA, Kolars JC,
Alto: Consulting Psychologists Press; 1981. Habermann TM, et al. Association of perceived medical
14. Choi HY. A study on affecting factors of social workers errors with resident distress and empathy: a prospective
burnout [master’s thesis]. Seoul: Yonsei University; 1994. longitudinal study. JAMA 2006;296:1071-8.
15. Shanafelt TD, Bradley KA, Wipf JE, Back AL. Burnout and 22. Williams JW Jr, Noel PH, Cordes JA, Ramirez G, Pignone M.
self-reported patient care in an internal medicine residency Is this patient clinically depressed? JAMA 2002;287:1160-70.
program. Ann Intern Med 2002;136:358-67. 23. Claus KE. The nature of stress. In: Claus KE, Bailey JT,
16. Thomas NK. Resident burnout. JAMA 2004;292:2880-9. editors. Living with stress and promoting well-being: a
17. Gopal R, Glasheen JJ, Miyoshi TJ, Prochazka AV. Burnout handbook for nurses. St. Louis: Mosby; 1980. p. 9-15.
and internal medicine resident work-hour restrictions. Arch 24. Freudenberger HJ. Burn-out: the organizational menace.
Intern Med 2005;165:2595-600. Train Dev J 1977;31:26-7.
18. Rosen IM, Gimotty PA, Shea JA, Bellini LM. Evolution 25. Pines AM, Kanner AD. Nurses’ burnout: lack of positive
of sleep quantity, sleep deprivation, mood disturbances, conditions and presence of negative conditions as two
empathy, and burnout among interns. Acad Med 2006;81:82- independent sources of stress. J Psychosoc Nurs Ment Health
5. Serv 1982;20:30-5.
19. Spitzer RL, Williams JB, Kroenke K, Linzer M, deGruy FV 26. Melamed S, Shirom A, Toker S, Berliner S, Shapira I. Burnout
3rd, Hahn SR, et al. Utility of a new procedure for diagnosing and risk of cardiovascular disease: evidence, possible causal
mental disorders in primary care: the PRIME-MD 1000 paths, and promising research directions. Psychol Bull 2006;
study. JAMA 1994;272:1749-56. 132:327-53.
20. Whooley MA, Avins AL, Miranda J, Browner WS. Case- 27. West CP, Tan AD, Habermann TM, Sloan JA, Shanafelt TD.
finding instruments for depression. Two questions are as Association of resident fatigue and distress with perceived
good as many. J Gen Intern Med 1997;12:439-45. medical errors. JAMA 2009;302:1294-300.

42 | Vol. 34, No. 1 Jan 2013 Korean J Fam Med