Work xx (20xx) x–xx 1

DOI:10.3233/WOR-203203
IOS Press

1 Medical student burnout: Impact of the gap

2 year in burnout prevention

f
roo
3 Stephanie A. Guanga , Adam E. M. Eltoraia , Wesley M. Duranda and Alan H. Danielsb,∗
4
a Warren Alpert Medical School of Brown University, Providence, RI, USA
b Department of Orthopaedic Surgery, Warren Alpert Medical School of Brown University, Providence, RI, USA
5
8
6

rP
7 Received 17 October 2018
Accepted 24 July 2019

tho
9 Abstract.
10 OBJECTIVE: Burnout is a common response to stress and is pervasive among medical students. An increasing proportion
11 of students are taking “gap years” following undergraduate education and before matriculation to medical school. This study
12
Au
evaluates rates of and risk factors for burnout, with a particular focus on students who took gap years.
13 METHODS: Burnout was measured utilizing the abbreviated Maslach Burnout Inventory. The primary independent variable
14 was whether a student took one or more gap years prior to medical school matriculation. Other variables included age, gender,
15 year in medical school, choice in specialty, and status in a combined baccalaureate-M.D. program. Bivariate and multivariate
16 regression was performed to elucidate relationships between student-level variables and burnout.
17 RESULTS: A total of 31.4% of respondents were found to be experiencing high levels of burnout. In multivariate analysis,
d

18 gap years were independently associated with lower levels of burnout (p = 0.041). Further, burnout decreased in a stepwise
19 fashion with students who took 0 (p = 0.350), 1 (p = 0.192), and 2+ (p = 0.048) gap years.
cte

20 CONCLUSIONS: Students taking gap years exhibited significantly lower levels of burnout than those who did not. Efforts
21 should be made in pre-medical and medical school curricula to better support students in their paths to medical school.

22 Keywords: Burnout, professional, education, medical, undergraduate, career choice, students, medical psychology, personal
23 satisfaction, life change events
rre

24 1. Introduction at many institutions and appears to increase with 30

co

time spent in medical school [2–5]. Although a small 31

25 Burnout syndrome, characterized by high levels degree of stress is beneficial for learning, various 32

26 of emotional exhaustion, depersonalization, and feel- forms of burnout are associated both psychological 33

27 ings of ineptitude, has been shown to affect up to 56% and physiological morbidity [6, 7]. 34
Un

28 of medical students, far exceeding the prevalence in Clinically, burnout is defined as “a psychologi- 35

29 similar age cohorts of the general population [1]. cal syndrome emerging as a prolonged response to 36

Poor mental health among students has been studied chronic interpersonal stressors on the job,” divided 37

into three dimensions: overwhelming emotional 38

exhaustion, feelings of cynicism and detachment 39

∗ Addressfor correspondence: Alan H. Daniels, MD, Depart- from the job, and lack of effectiveness and accom- 40
ment of Orthopaedic Surgery, Warren Alpert Medical School
of Brown University, 100 Butler Drive, Providence, RI 02906,
plishment [8]. Several studies have linked burnout 41

USA. Tel.: +1 401 330 1420; Fax: +1 401 330 1495; E-mail: in medical students with adverse consequences, 42

Alan [email protected]. including: depression, suicidal ideation, thoughts of 43

1051-9815/20/$35.00 © 2020 – IOS Press and the authors. All rights reserved
 2 S.A. Guang et al. / Medical student burnout

44 dropping out of medical school, and other serious countries, for personal maturation and more diversity 96

45 mental health concerns [9]. Burnout has also been in lived experiences [16–18]. 97

46 shown to influence specialty choice, particularly with The association between gap years and medical 98

47 respect to lifestyle and controllability of specialty student burnout has not been previously studied. This 99

48 fields as opposed to primary care [10]. Additionally, study sought to elucidate the relationship between 100

49 burnout affects performance in health care, causing student characteristics and burnout, with particular 101

50 deterioration of compassion, empathy, and altruism focus on gap years and participation in a combined 102

51 for patients. These effects may manifest in poorer baccalaureate-M.D. program (B.A./M.D.). 103

f
52 quality of care, decreased patient satisfaction, and

roo
53 increased rates of medical error [7, 11].
54 The various causes of poor mental health and 2. Methods 104

55 burnout among medical students have been previ-

56 ously studied. In addition to academic pressure and 2.1. Participants and design 105

57 stressors relating to patient care, students may also

rP
58 be subject to personal and social challenges; negative An electronic cross-sectional survey was adminis- 106

59 personal life events, in particular, have been shown to tered to Year 3 and Year 4 students at a single medical 107

60 have a strong relationship with burnout and resiliency school in the U.S at the end of the academic year. Fol- 108

61 [12]. lowing institutional review board approval, a secure 109

tho
62 The average matriculating medical student age survey was created and distributed using RED Cap 110

63 is 24 years old, and approximately 10% of stu- electronic data capture tools during 2016 and 2017. 111

64 dents are age 27 or older in the United States [13].

65 Nationally, students aged 20–23 at matriculation are 2.2. Survey content 112

66 decreasing every year compared to increased num-

Au
67 bers in older age groups, and 63.4% of students Demographic information was collected at the 113

68 report taking at least one gap year [14]. Accord- beginning of the survey. Respondents provided 114

69 ingly, a greater proportion of students are pursuing their age, year in medical school, gender iden- 115

70 other interests prior to matriculation, i.e., “taking tity, B.A./M.D. program or traditional premedical 116

71 gap years”. These “gap years” may offer students route (TPM) status, gap years between college 117

the opportunity to gain diverse experiences before and medical school, and intended specialty. Spe-
d

72 118

73 committing to a career in medicine, and students cialty choice was categorized as either those with 119
cte

74 themselves feel this is advantageous for their learning a high level of controllability (anesthesiology, 120

75 [15]. Many students pursue international fellow- dermatology, emergency medicine, neurology, oph- 121

76 ships, obtain graduate degrees, teach or engage in thalmology, otolaryngology, pathology, psychiatry, 122

77 community service, or join the workforce, all of radiology) or low controllability (family practice, 123

78 which lend experiences that helps relate to a larger internal medicine, Ob/Gyn, orthopedic surgery, pedi- 124
rre

79 population of patients. The AAMC currently lists atrics, surgery [general], urology), as categorized in 125

80 238 “career changer” post-baccalaureate programs, previous research according to the number of work- 126

81 that are geared towards applicants who have non- ing hours, on-call responsibilities, and patient load 127

82 science backgrounds and decided to pursue medicine [10]. 128

co

83 after completing college, as opposed to pre-medical Students then completed an abbreviated Maslach 129

84 students who focus their undergraduate studies on Burnout Inventory (aMBI) – Human Services, a val- 130

85 pre-requisite science classes. idated tool for measuring burnout. The condensed 131

86 Other graduate disciplines similarly value post- version consists of a 9-item inventory that evaluates 132
Un

87 graduate experience; some law schools have “Junior burnout with three questions, each for assessing 3 133

88 Deferral Programs” which accepts college students subscales of Emotional Exhaustion (EE), Deperson- 134

89 under the condition that they spend at least 2 years alization (DP), and Personal Accomplishment (PA). 135

90 working, studying or pursuing other fellowships prior For the purposes of the present study, references 136

91 to law school matriculation. Most full-time M.B.A. to clients and personal were changed to “patients” 137

92 programs also prefer candidates with at least one to and other relevant wording. The questions on the 138

93 two years of working experiences. Other studies have MBI portion of the survey are rated on a Likert 139

94 demonstrated the value of a gap year between high scale of how often the student feels a certain way 140

95 school and college, a common practice in many other to his/her work (1 = Never, 2 = A few times per year, 141
 S.A. Guang et al. / Medical student burnout 3

142 3 = Once a month or less, 4 = A few times per month, respondents were Year 3 students and 61% (31/51) 168

143 5 = Once a week, 6 = A few times per week, 7 = Every were Year 4 students. 69% (35/51) identified as 169

144 day). According to convention, high scores for deper- female and 31% (16/51) identified as male. 59% 170

145 sonalization or emotional exhaustion were used for (21/51) of respondents took one or more gap years 171

146 quantifying professional burnout. Clinically, high between their undergraduate education and medi- 172

147 levels of burnout out are often identified as MBI- cal school, with an average interval of 3.36 years 173

148 EE ≥14 or MBI-DP ≥ 12, which is proportionately (range = 13, median = 26, STD = 2.96). 57% (23/51) 174

149 scaled down for the abbreviated MBI [13]. Given lack of respondents were admitted through the traditional 175

f
150 of standardization of cut-off scores, however, argu- premedical route, and 43% (28/51) were students 176

roo
151 ments have been made against defining burnout as enrolled through the B.A./M.D. program. 177

152 a dichotomous variable for research purposes [22]. On bivariate analysis, students who took gap 178

153 Accordingly, burnout was analyzed as a continuous years before medical school exhibited lower lev- 179

154 variable in this study. els of burnout as compared to students who did 180

not (16.8 vs. 22.4, p = 0.002). Additionally, students 181

rP
155 2.3. Statistical analysis in the B.A./M.D. program exhibited higher levels 182

of burnout than those from traditional pathways 183

156 Linear regression was performed to compare lev- (21.7 vs. 17.0, p = 0.01) (Table 1). No significant 184

157 els of burnout. A multivariate linear regression association was observed between burnout and age 185

tho
158 was performed to control for potential confound- (p = 0.392), gender (p = 0.691), intended specialty 186

159 ing covariates. Independent variables that both did (p = 0.391), or medical school year (p = 0.108). 187

160 and did not demonstrate significance in the bivari- On multivariate analysis, students who took gap 188

161 ate analyses were included in the multiple regression years between college and medical school exhibited 189

162 analysis. Analyses were performed in StataSE ver- significantly less burnout than those who did not 190
Au
163 sion 15 for Mac (Stata Corp, College Station, Texas). (–4.93, p = 0.041). While students in the B.A./M.D. 191

program exhibited trending higher levels of burnout 192

as compared to students who were admitted tradition- 193

164 3. Results ally, this effect was not statistically significant after 194

adjusting for other variables (+1.94, p = 0.39). 195

Of the 246 students in Year 3 and Year 4 enrolled Frequency distributions show that B.A./M.D.’s
d

165 196

166 at the Warren Alpert Medical School (AMS), 20.7% took fewer gap years (average = 0.39 years) than 197
cte

167 (n = 51)completed the survey. The average age of their standard-admit counterparts (average 3.2 years) 198

respondents was 27 years old. 39% (20/51) of (Fig. 1) Chi-squared test of the binary of no gap 199

Table 1
Bivariate analysis of variables affecting burnout. Averages for MBI subscales of emotional exhaustion (EE), depersonalization (DP), and
rre

personal accomplishment (PA) are shown across variables. Burnout is the sum of emotional exhaustion and depersonalization
n MBI-EE MBI-DP MBI-PA Burnout p-value
Age 0.392
24–26 28 (55%) 12.0 8.71 18.1 20.7
27–37 23 (45%) 10.4 6.78 18.0 17.2
co

Year in medical school 0.108

Year 3 20 (39%) 11.9 8.55 18.2 20.4
Year 4 31 (61%) 10.8 7.31 17.9 18.1
Gender 0.691
Un

Female 35 (69%) 11.4 7.97 18.6 19.4

Male 16 (31%) 11.0 7.56 16.8 18.6
Time off 0.002
No 21 (41%) 13.1 9.29 17.9 22.4
Yes 30 (59%) 9.97 6.83 18.1 16.8
Specialty 0.391
Controllable 33 (65%) 11.5 8.18 18.3 19.7
Uncontrollable 18 (35%) 10.8 7.20 17.4 18.0
Program 0.011
Traditional premedical route 28 (55%) 10.1 6.93 18.4 17.0
B.A./M.D. program 23 (45%) 12.7 8.96 17.6 21.7
 4 S.A. Guang et al. / Medical student burnout

f
roo
rP
tho
Fig. 1. Frequency distributions of gap years between traditional medical students and B.A./M.D. students.

Table 2
Multivariate Analyses of Burnout by Student Characteristics
Au
(Reference level listed first) Coefficient 95% Confidence interval
Lower Upper
Age 0.231 –0.520 0.982
Year (Third, Fourth) 2.47 –1.29 6.22
Gender (Female, Male) 1.59 –5.13 8.31
d

Time off (No, Yes) –4.93 –9.65 –0.202

Specialty (Uncontrollable, Controllable) –1.90 –8.65 4.86
Program (TPM, B.A./M.D.) 1.94 –2.58 6.46
cte

Table 3 4. Discussion 210

Burnout within levels of gap years before medical school
B 95% Confidence Interval This study evaluated the relationship between stu- 211
rre

Lower Upper dent characteristics and burnout. The journey from 212
Gap years medical student to physician is an intense and long 213
0 –5.85 –18.3 6.63
1 –8.57 –21.6 4.45 process [20]. The growing trend for medical stu- 214

2+ –12.0 –24.4 0.44 dents to take gap years before medical school is 215
co

evident in the population of respondents, who were 216

an average of 27 years old in their clinical years. 217

200 years versus gap years yields a p-value of 0.0028. We found that students who took gap years exhibited 218

201 B.A./M.D. students were 18 times less likely to take less burnout than those who did not; a dose-response 219
Un

202 time off before medical school. effect was observed between the number of gap 220

203 Moreover, when the number of gap years taken is years and medical student burnout. While students 221

204 grouped into levels of 0, 1, vs 2+ years, there appears in the B.A./M.D. program exhibited higher levels 222

205 to be a dose response effect where increased num- of burnout in bivariate analyses, this effect was not 223

206 ber of gap years is more significantly related to lower significant after adjusting for other confounding vari- 224

207 burnout (Table 3). A multivariate analysis was per- ables. B.A./M.D. students were less likely to take gap 225

208 formed and shows a significant decrease in burnout years than students admitted through the traditional 226

209 is seen with 2 or more years before medical school route. This is anecdotally consistent, as students in 227

(adjusted R-squared = 0.215). the program tend not to take time off because they are 228
 S.A. Guang et al. / Medical student burnout 5

229 not required complete the full medical school appli- survey. In-person paper surveys may be effective as 281

230 cation process. In the United States, eligibility for a suggestion for future improvement, as medical stu- 282

231 most medical school programs requires that students dents receive a high volume of emails and may be 283

232 take the Medical College Admission Test (MCAT), easier to access in a student spaces on their cam- 284

233 fulfill prerequisites in the sciences and humanities, puses. As this is a cross-sectional study, it is difficult 285

234 and apply and interview at the medical schools them- to distinguish causal relationships from associations. 286

235 selves. Students who require extra time after college Self-reporting of data may also have introduced bias. 287

236 to complete these steps or further strengthen their Burned out students may be more inclined to com- 288

f
237 application often elect to take a gap year. Other indi- plete a survey on a personally relevant topic or, 289

roo
238 viduals pursue careers in other fields before deciding alternatively, may be either too apathetic or stressed 290

239 to complete post-baccalaureate programs to apply to to complete the survey. Future multi-institutional 291

240 medical school. research is needed to further support and extend these 292

241 Though a direct path from undergraduate to med- results. Moreover, longitudinal data following medi- 293

242 ical school has historically been preferred, there is cal students into residency could provide more insight 294

rP
243 a growing trending to pursue gap years and have into burnout than possible with a cross-sectional 295

244 different life experiences. study. Finally, students deciding to pursue gap years 296

245 There are a number of potential mechanisms to in education may be less prone to burnout, rather than 297

246 account for our observation of lower burnout with the gap year itself truly preventing burnout. Further 298

tho
247 increasing number of gap years. Students taking gap study of the type of experiences undertaken during 299

248 years often engage in work experiences, research, gap years may better parse out these findings. 300

249 international or cross-cultural exchanges, military

250 service, and completion of other graduate degrees;
251 these activities may confer a wider range of expe- 5. Conclusion 301
Au
252 rience, resulting in greater resilience among these
253 students. Other literature from the U.K., where the Students taking gap years prior to medical school 302

254 practice of taking a gap year abroad before college, exhibited significantly lower levels of burnout than 303

255 demonstrates increased civic engagement along with those who did not. Students in the combined 304

256 psychosocial and intellectual development because baccalaureate-M.D. program exhibited higher levels 305

students are presented with challenges in unique, of burnout on bivariate analyses, but not after adjust-
d

257 306

258 new environments. High school students were also ing for other characteristics; this may be attributed to 307
cte

259 found to have greater “self-authorship” which is a the fact that combined baccalaureate-M.D. students 308

260 way of approaching difficult situations in a way were less likely to take gap years than students admit- 309

261 that recognizes the contextual nature of knowledge ted through other routes. Efforts should be made in 310

262 and balances it with the students’ understanding of medical school and pre-medical school curricula to 311

263 self [21]. Analogously, medical students with non- better support medical students in the timing and 312
rre

264 academic gap year experiences outside of medical nature of their path to becoming physicians. Gap 313

265 school might better be able to contextualize and man- years should no longer be viewed as undesirable com- 314

266 age the stresses encountered than students who have pared to direct entry into medical school. Taking time 315

267 only been in higher education their whole lives [22]. off to explore other opportunities and have life expe- 316
co

268 Our findings suggest that direct matriculation into riences should be encouraged, as it can play a role in 317

269 medical school has a negative effect on student reducing burnout. 318

270 well-being. Prior research on burnout among med-

271 ical students indicates that students’ personal lives,
Un

272 learning environment, and financial concerns are Conflict of interest 319

273 associated with burnout [23, 24]; it is possible that

274 gap years may interact with these other factors. As None to report. 320

275 medical school is becoming less didactic with more

276 clinical integration in the pre-clinical years, physi-
277 cian burnout may have begun to affect the medical Ethical approval 321

278 school experience as well.

279 A notable limitation of this study is the low Exemption from Federal Regulation 45 CFR 46 322

280 response-rate of medical students to the administered as meeting the criteria in 45 CFR 46.101(b) (2) was 323
 6 S.A. Guang et al. / Medical student burnout

324 obtained from April 18, 2016 from LifeSpan IRB program. Annals of Internal Medicine. 2002;136(5):358-67. 366

325 (FWA-Rhode Island Hospital (RIH) 00001230). [12] Dyrbye LN, Thomas MR, Huntington JL, Lawson KL, 367
Novotny PJ, Sloan JA, Shanafelt TD. Personal life events 368
and medical student burnout: a multicenter study. Academic 369
Medicine. 2006;81(4):374-84. 370

326 Funding [13] AAMC. Age of Applicants to U.S. Medical Schools at 371
Anticipated Matriculation by Sex and Race/Ethnicity, 2014- 372
2015 through 2017-2018. [updated 2017 November 30; 373
327 None to report. accessed 2018 June 7]. Available from https://www.aamc. 374
org/download/321468/data/factstablea6.pdf 375

f
[14] AAMC Matriculating Student Questionnaire All Schools 376

roo
328 References Summary Report & Individual School Report. AAMC; 2018 377
[cited 2019Apr4]. Available from: https://www.aamc.org/ 378
download/494044/data/msq2018report.pdf 379
329 [1] Dyrbye LN, Massie FS, Eacker A, Harper W, Power D, [15] Paterson-Brown L, Paterson-Brown F, Simon E, Loudon J, 380
330 Durning SJ, Thomas MR, Moutier C, Satele D, Sloan J, Henderson-Howat S, Robertson J, Paterson-Brown S. The 381
331 Shanafelt TD. Relationship between burnout and profes- Influence of a ‘Gap Year’on Medical Students. Journal of 382

rP
332 sional conduct and attitudes among US medical students. Education and Training Studies. 2015;3(4):49-56. 383
333 Jama. 2010;304(11):1173-80. [16] Martin AJ. Should students have a gap year? Motiva- 384
334 [2] Clark DC, Zeldow PB. Vicissitudes of depressed mood dur- tion and performance factors relevant to time out after 385
335 ing four years of medical school. Jama. 1988;260(17):2521- completing school. Journal of Educational Psychology. 386
336 8. 2010;102(3):561. 387
337 [3] Dyrbye LN, Thomas MR, Shanafelt TD. Medical student [17] Johan N, Sadler-Smith E, Tribe J. Informal and inciden- 388

tho
338 distress: causes, consequences, and proposed solutions. In tal learning in the liminal space of extended independent 389
339 Mayo Clinic Proceedings 2005 Dec 1 (Vol. 80, No. 12, pp. (gap year) travel. Academy of Management Learning & 390
340 1613-1622). Elsevier. Education. 2018 Dec 20(ja). 391
341 [4] Thomas NK. Resident burnout. Jama. 2004;292(23):2880- [18] Clerkin A. A three-wave longitudinal assessment of socioe- 392
342 9. motional development in a year-long school-based ‘gap 393
343 [5] Dunn LB, Iglewicz A, Moutier C. A conceptual model of year’. British Journal of Educational Psychology. 2019 Feb 394
Au
344 medical student well-being: promoting resilience and pre- 3. 395
345 venting burnout. Academic Psychiatry. 2008;32(1):44-53. [19] Dyrbye LN, West CP, Shanafelt TD. Defining burnout 396
346 [6] Guthrie EA, Black D, Shaw CM, Hamilton J, Creed FH, as a dichotomous variable. Journal of General Internal 397
347 Tomenson B. Embarking upon a medical career: psycho- Medicine. 2009;24(3):440. 398
348 logical morbidity in first year medical students. Medical [20] Viswanath O, Viswanath N, Viswanath K. Navigating the 399
349 Education. 1995;29(5):337-41. progressive intensity of lifelong learning from medical 400
350 [7] Guthrie E, Black D, Bagalkote H, Shaw C, Campbell M, student to attending physician. Work (Reading, Mass.). 401
d

351 Creed F. Psychological stress and burnout in medical stu- 2018;61(4):509. 402
352 dents: a five-year prospective longitudinal study. Journal of [21] O’shea J. Delaying the academy: A gap year education. 403
cte

353 the Royal Society of Medicine. 1998;91(5):237-43. Teaching in Higher Education. 2011;16(5):565-77. 404
354 [8] Novack DH, Epstein RM, Paulsen RH. Toward creat- [22] Magolda MB. Making their own way: Narratives for trans- 405
355 ing physician-healers: fostering medical students’ self- forming higher education to promote self-development. 406
356 awareness, personal growth, and well-being. Academic Stylus Publishing, LLC.; 2004. 407
357 Medicine. 1999;74:516-20. [23] Dyrbye LN, Thomas MR, Harper W, Massie Jr FS, Power 408
358 [9] Dyrbye LN, Thomas MR, Massie FS, Power DV, Eacker A, DV, Eacker A, Szydlo DW, Novotny PJ, Sloan JA, Shanafelt 409
rre

359 Harper W, Durning S, Moutier C, Szydlo DW, Novotny PJ, TD. The learning environment and medical student burnout: 410
360 Sloan JA. Burnout and suicidal ideation among US medical a multicentre study. Medical Education. 2009;43(3):274-82. 411
361 students. Annals of Internal Medicine. 2008;149(5):334-41. [24] Dahlin ME, Runeson B. Burnout and psychiatric morbidity 412
362 [10] Enoch L, Chibnall JT, Schindler DL, Slavin SJ. Association among medical students entering clinical training: a three 413
363 of medical student burnout with residency specialty choice. year prospective questionnaire and interview-based study. 414
co

364 Medical Education. 2013;47(2):173-81. BMC Medical Education. 2007;7(1):6. 415

365 [11] Shanafelt TD, Bradley KA, Wipf JE, Back AL. Burnout and
self-reported patient care in an internal medicine residency
Un