Burnout and Associated Factors Among Medical Students in a Public University in Uganda: A Cross-Sectional Study

Introduction

Burnout is, according to the 11 revision of the International Classification of Diseases (ICD-11),

“A syndrome conceptualized as resulting from chronic workplace stress that has not been successfully managed. It’s characterised by three dimensions: feelings of energy depletion or exhaustion; increased mental distance from one’s job or feelings of negativism or cynicism related to one’s job; and reduced personal efficacy.“¹

Burnout is an increasingly recognised problem among medical professionals.² Previous studies in medical students found 14.8–79.9% of them experience burnout during their medical training.^3–7 Its prevalence has been found to increase with medical school academic progression^8,9 but studies have suggested that by equipping medical students with evidence-based stress management approaches to their education, the incidence of burnout is reduced irrespective of their academic progression.³²

In working-class healthcare professions, burnout has been linked majorly to job- or work-related chronic stress. In medical students, it is most likely related to personal factors such as mental health disorder, physical activity, personal financial satisfaction, high parental expectations, alcohol consumption, and smoking status or individual learning environment.^1,10 A 2017 systematic review suggested that, on average, one medical student out of two suffered from burnout as defined by the ICD-11. Medical students from the Middle East and Oceania countries were found to have a higher prevalence of burnout than those from countries in other continents, with no African data available at the time of the review.¹¹ Higher levels of stress coupled with poor coping strategies may be the major factors responsible for the development of burnout among medical students,³ due to the cyclic relation between stress and emotional exhaustion.¹² Medical students, when compared to age-matched members of the general public, have been shown to have a poorer psychological quality of life.¹³ A 2015 systematic review showed that medical students displayed a variety of coping mechanisms for burnout such as personal engagement, extracurricular activities including mainly musical and physical activities, student autonomy, mentorship programs, evaluation systems, career counseling and life coaching.³

The prevalence of burnout among medical students at Mbarara University of Science and Technology (MUST) was unknown. We sought to identify the degree of burnout and identify any causal factors such as learning environment, personal life factors and outside school life factors. This cross-sectional study aimed at investigating the prevalence and factors associated with burnout among medical students in their pre-clinical and clinical years of study at MUST.

Materials and Methods

Study Design

An online descriptive cross-sectional study was conducted between 6^th September 2020 until 20^th September 2020, using a Google form.

Study Setting

The study was carried out at Mbarara University of Science and Technology, commonly called Mbarara University. It is a public university accredited by the National Council for Higher Education, offering both undergraduate and postgraduate programs. MUST was established in 1989 and has two campuses – Kihumuro campus and Mbarara campus. The Faculty of Medicine is located at the Mbarara campus approximately 266 kilometers southwest of Kampala. The MBChB (Bachelor of Medicine and Bachelor of Surgery) program at Mbarara University takes 5 years. There were approximately 423 medical students enrolled for the 2019/2020 academic year in the Faculty of Medicine at MUST.

Study Sample

Undergraduate students from first to final years at MUST, pursing a medical program of Bachelor of Medicine and Bachelor of Surgery, were surveyed using an enrollment criteria including students being in the official WhatsApp groups of the respective class years and voluntarily consenting to take part in the study (see Figure 1). Representatives of each class were contacted by the principal investigator and informed about the study procedures, ethical issues and data collection. The participants were enrolled using a consecutive sampling technique. Students were recruited into the study through class representatives. The survey tool was then disseminated through class official WhatsApp groups. The inclusion criteria were medical students at MUST enrolled for the academic year 2019/2020. Other health profession students were not eligible for enrollment in this study.

Figure 1 Flow chart for enrollment of participants.

The required sample size was calculated using the Kish–Leslie formula.

where:

N = Sample size

Z = Standard deviation at 95% Confidence Interval (1.960)

p = Expected prevalence of depression among medical students

d = Precision of the estimate (5%)

The prevalence of burnout was taken to be an average of 0.5 since there had not been any previous study carried out at MUST. This point prevalence, of 50% was also found in a recent study assessing burnout among medical students in Ethiopia.¹⁶

We used the modified Kish–Leslie equation³³ to calculate the available sample size; where K was the estimated overall population of the study population.

To cater for non-correspondents assuming a response rate of 90%, a total of 10% of the sample size was added. This led to the required sample size of 229 students.

Measurements

We developed a questionnaire that would collect participants’ characteristics (demography and socioeconomic factors) and their MBI-SS scores. We also asked about factors that might be associated with burnout among medical students. Participants’ characteristics included sex, age, year of study, religious/spiritual affiliation, source of tuition funding, marital status, family income, agreement with the effectiveness of wellness initiatives at their institution, and family residence.

Burnout was assessed using the Maslach Burnout Inventory – Student Survey tool which was adapted from^17,18 and had been validated in Brazil.¹⁹ The MBS-SS is a self-administered tool consisting of 15 questions. It is subdivided into three subscales: emotional exhaustion (five items), cynicism (four items), and professional efficacy (six items). All items are assessed by frequency using aL ikert scale ranging from never (0) to every day (6).

The tool has an internal consistency of each latent factor greater than 0.7 with values of 0.77, 0.72, and 0.81 for emotional exhaustion, professional efficacy, and cynicism, respectively.²⁰ A test–retest reliability determined at about 4 weeks later, a yielded reliability coefficients of 89, 84 and 67 respectively for exhaustion, cynicism and academic efficacy.²¹ The sensitivity, specificity, positive and negative predictive values of the of the 15-item MBI-SS were 91.9% (95% CI = 82.5–96.5%), 93.2% (95% CI = 87.5–96.4%), 86.4% (95% CI = 76.1–92.7%) and 96.1% (95% CI = 91.2–98.3%) respectively. The positive and negative likelihood ratios were 13.48 (95% CI = 7.15–25.44) and 0.09 (95% CI = 0.04–0.20) respectively.²⁰

There are two validated calculation methods used to compare burnout scores among participants ie, the summation (SUM) and average (AVE) method.¹⁷ Burnout is determined by the total score of the MBI-SS based on the participant’s responses by using both the tridimensional and bidimensional diagnostic criteria for burnout syndrome as listed in the ICD-11. On a two-dimensional model, it is identified by high scores on both emotional exhaustion and cynicism subscales. In the three-dimensional model, low-level burnout was determined based on the following scores: emotional exhaustion (0 - 9), cynicism (0 - 1), and professional efficacy <27. Moderate level burnout was determined based on the following scores: emotional exhaustion (10 - 14), cynicism (2 - 6), and professional efficacy (23 - 27). At a high level, burnout was identified based on the following scores: emotional exhaustion >14, cynicism >6, and professional efficacy <23.¹⁸ High scores on emotional exhaustion and cynicism with low scores on professional efficacy are indicative of burnout.²⁰

Factors associated with burnout among medical students were adapted from Haile et al,^16,18,22 and modified to suit the institution’s low-income setting. Most responses were dichotomised.

The questionnaire was piloted by 23 (10% of the sample size) Gulu University medical students. A Cronbach alpha of 0.8099 was obtained, demonstrating the high reliability of the tool in assessing the study outcomes.

Data Analysis

Data analyses were performed using StataCorp. 2019. Stata: Release 16. Statistical Software. College Station, TX: StataCorp LLC. Medical students’ characteristics were summarized with proportions. The prevalence of burnout was determined from the total score of the MBI-SS tool using the tridimensional criteria based on the participants’ responses to Likert scale questions related to emotional exhaustion, professional efficacy, and cynicism. Proportions were used to estimate the prevalence of burnout in the student sample. To quantify possible associations between the independent variable and dependent variables, we calculated crude odds ratios. Backward regression analysis was used to assess the predictive relationship between the associated factors obtained from the medical students and their respective burnout prevalence scores.

Results

Participant Characteristics

One hundred and fifty (35.5%) responses were received. We removed one incomplete response and four duplicate responses, leaving behind 145 responses that were analysed (34.3% response rate). The participants included 102 (70.3%) male and 43 (29.7%), female medical students. The majority (49%, n=71) of the students were under government sponsorship and over 90% (n=134) had a religious affiliation. The majority of the students were aged less than 24 years, and the rest of the sociodemographic characteristics were comparable across the 5 years of studies as shown in Table 1.

Table 1 Participants’ Characteristics

Prevalence of Burnout Among Medical Students

Table 2 presents the prevalence of burnout among medical students. More than half (54.5%, n= 79) of the medical students fitted the definition of burnout using the tridimensional dimensional criteria. Of the 145 students, 93.1% (n= 135) had high emotional exhaustion scores, 62.1% (n= 90) had low professional efficacy scores and 97.2% (n= 141) had high cynicism scores.

Table 2 Prevalence of Burnout Among Medical Students

Factors Associated with Burnout Among Medical Students

In the unadjusted logistic regression analysis, several factors were found to be independently associated with burnout. These including factors intrinsic to the student such as being optimistic about their future career (OR=2.70, 95% CI 1.11 to 6.55; p=0.028), feeling fulfilled and happy at work (OR=2.77, 95% CI 1.31 to 5.86; p=0.008), willingly choosing to pursue MBChB (OR=9.44, 95% CI 2.04 to 43.58; p=0.004), feeling satisfied with medicine as their career choice (OR=5.07, 95% CI 2.09 to 12.31; p=0.000), feeling motivated with their studies (OR=2.43, 95% CI 1.08 to 5.45; p=0.031, and having ever developed any thoughts of dropping out of school (OR=0.17, 95% CI 0.06 to 0.48; p=0.001). Key factors, outside of the students locus included an exhaustive day-to-day routine of studies (OR=0.43, 95% CI 0.19 to 0.96; p=0.040), the challenge of commuting long distances (OR=0.30, 95% CI 0.13 to 0.68; p=0.004), insufficient leisure time (OR=2.2, 95% CI 1.12 to 4.30; p=0.021), and considering one’s self important to their family member (OR=6.08, 95% CI 1.27 to 29.22; p=0.024). More details in Table 3.

Table 3 Distribution of MUST Medical Students According to Variables Related to Participant Characteristics, Personal Aspects, Learning Environment, Outside School Environment and Prevalence of Burnout

In multiple logistic backward regression only one factor, choosing MBChB willingly (OR =7.20, 95% CI 1.40 to 36.93; p=0.018) was found to be independently associated with burnout. Table 4 contains more details.

Table 4 Outcome of Logistic Regression Analysis for Related Variables and Burnout in Medical Students

Discussion

Our cross-sectional study looked at the prevalence of burnout among medical students at MUST, as well as possible associations. More than half of the students described burnout. Willingly choosing a career in medicine was the only identified associated factor. The high prevalence of burnout in our study is consistent with a similar study in Egypt which reported a burnout prevalence of 79.9%.⁷ One possible reason could be that medical students were unaware of the issue of burnout. A 2010 systematic review, suggested that there existed an association between a lower level of tolerance for uncertainty or ambiguity and self-reported burnout among medical students.²³ Studies conducted among medical students in Ethiopia and Spain found low levels of burnout, 34% and 14.8%, respectively.^6,16 Such discrepancies in the burnout prevalence in various studies may be attributed to the different burnout measurement tools used. Additionally, students in different countries are exposed to different socio-economic, political, health, and conflict-related stress factors.³

Burnout was higher among males (70.9%) compared to females (29.1%), a consistent finding with a systematic review that found male medical students to have a higher burnout prevalence.¹⁵ This could be due to the many male medical student participants as indicated in most studies. However, whether these many male survey responders are a true representative of their various institutions' male medical student population, remains a question to answer to come up with a conclusive finding. It has been proposed, by some, that female medical students are at a high risk of burnout than their male counterparts.¹⁴ Female students are more likely to suffer from stress factors than male medical students,⁴ with more demands placed on them outside of the university.⁸

Third and fourth-year medical students had the highest level of burnout in this study. This correlates with prior studies that showed a higher degree of burnout in senior medical students.^16,17 The transition from preclinical years to clinical years of training appears to be a risk factor for academic stress in medical school.^5,24 The low burnout prevalence of fifth years may be due to them having learned some coping strategies such as life coaching, reflective practice, self-care, musical and physical activities.

One interesting finding in our study was that willingly choosing to undertake MBChB was found to be significantly associated with an increased risk of burnout. This is contrary to a Cameroonian study that reported that students who regretted choosing their medical studies had higher burnout scores than their peers.²⁵ Medical schools in countries like America have established wellness programs that could be adapted to address the “bio-psycho-social-spiritual burnout” epidemic among medical students.^23,26 In Australasia, several wellness initiatives have been trialed. These include the integration of a well-being curriculum alongside the traditional medical curriculum such as the SAFE-DRS program at the University of Auckland³⁰ and the Health Enhancement Program at Monash University Medical School.³¹ A U.S based study evaluating the success of a student-led program teaching stress management and self-care skills to first-year medical students, found it to be helpful (based on longevity, student participation, and student feedback).²⁷ Additionally, the GMC – UK has provided guidance to medical schools on the creation of peer support networks, group sessions on stress management, and mentoring schemes.²⁹

Medical schools promote competition rather than cooperation. Training-related stress factors including annual examinations, an inability to cope with academic demands, and pressure with an obligation to succeed all compound the problem. Additionally, a sense of helplessness coupled with sleep deprivation, financial challenges, and the emotional burden of witnessed suffering stress medical students.^5,24 Other studies have found associations between burnout in medical students with not seeing coursework as a source of pleasure, feeling uncomfortable with course activities, not acquiring skills necessary to become a doctor, poor social support, being less than satisfied with the education system, self-perception of health, optimism, and motivation in studies.^16,18,22 Our study did not find any such associations.

Burnout is not only harmful to the medical students, their relatives, medical schools, and peers but to also the patients they care for. There is thus a heightened need to address burnout among medical students to ensure that delivery of compassionate patient care is kept paramount.^4,5,28

Limitations

Our study had several limitations. First, this was a cross-sectional study. It could not allow us to study the variation of burnout among medical students throughout their entire 5-year study period in medical school at MUST. Second, the small response rate might have led to a type 2 error. Census involving these students would reveal more reliable findings. Recall bias may also have occurred amongst students since, by the time the study was conducted, students were off-campus due to the cessation of educational activities due to the COVID-19 pandemic. This, in itself, may have also contributed.

Implications

Our study findings should be generalizable to many other low- and middle-income countries. We used a standardised burnout measurement tool to increase the reliability and validity of our study findings. This has been the first study to be conducted among medical students in Uganda and it has revealed high levels of emotional exhaustion, cynicism, and low levels of professional efficacy. There needs to be an investment in the provision of mental health support to medical students through interventions to raise awareness, understanding, and prevention of the development of psychological distress, burnout, and mental health disorders like depression and anxiety among medical students. The efficacy of this study could be measured by conducting a multi-centre, prospective longitudinal study.

Some student life activities may be present at MUST but these activities are not united under a comprehensive wellness objective. We hope that this study will help drive the establishment of wellbeing interventions at MUST.