The POISE study: Quantitative findings from a mixed methods study investigating burnout in family physicians in low- and middle-income countries

Problem

The World Health Organisation (WHO) has identified primary care as essential for the delivery of Universal Health Coverage (UHC) and a healthy workforce is crucial for this. While burnout is a significant issue for health workers globally, there is a dearth of data from low- and middle-income countries (LMICs) on burnout in family physicians. Therefore, the aim of this study was to estimate the prevalence of burnout among family physicians in LMICs and identify factors associated with burnout. Burnout is defined as exhaustion and disengagement from work and has significant consequences for physician health, patient safety and staffing levels. The aetiology of burnout is an imbalance in the demands and resources of a job, therefore, risks may be increased for healthcare workers in lower resource settings, where significant demands are combined with significant resource shortages.

Approach

An online survey containing the Oldenburg Burnout Inventory (OLBI), as well as demographic and workplace questions, was disseminated to family physicians in Pakistan, South Africa, Malaysia, Sri Lanka, and Zambia. Burnout was defined as a high mean item score of ≥ 2.25 on the exhaustion subscale and ≥ 2.1 on the disengagement subscale. Prevalence percentages were calculated, and tests of association were used to explore associations of 47 variables with burnout. Statistical significance was set at p ≤ .05. Hierarchical Poisson regression was performed to identify risk and protective factors.

Findings

The sample size was 451, of which 49.9% (n=225) were women, 45.2% (n=204) were in the 30-39 age-group, and 31.5% (n=142) in the 40-49 age-group. Overall burnout prevalence was 77.4%. Country breakdown as follows: Zambia 84.4%; South Africa 81.7%; Pakistan 75.9%; Sri Lanka 75.5%; and Malaysia 74.8%. Tests of association found 20 out of 47 variables had significant relationships with burnout with some variation between countries. The Poisson regression model found significant risk factors for burnout were working over 50 hours per week (IRR, 1.06; 95% CI, 1.01 – 1.11; p = .03) and experiencing conflict at work (IRR, 1.12; 95% CI, 1.00 – 1.27; p = .05). Significant protective factors were manager support (IRR, 0.82; 95% CI, 0.74 – 0.91; p = <.001), and a lack of conflict between work and home responsibilities (IRR, 0.82; 95% CI, 0.74 – 0.92; p = .001).

Consequences

This high prevalence of burnout has implications for physician health and performance and the associated high turnover and migration has implications for staffing. The number of variables with a statistically significant relationship with burnout underlines the complexity of burnout syndrome. While these findings are essential to inform the development of interventions, additional data is needed from each country to confirm these results and further work required to develop interventions to prevent and mitigate burnout.