Background: Wellbeing of healthcare workers (HCWs) is important for the effectivernfunctioning of health systems; it impacts the health service quality and patient outcome. rnHCWs are at an increased risk of stress and burnout both the nature of their work and theirrncontext of their work, but this has been little investigated in low income countries. The aimrnof this study was to: (1) explore the conceptualisations of wellbeing, stress and burnout, (2)rnassess the level of burnout and, professional satisfaction and (3) evaluate the short-termrnoutcome of burnout and associated factors in the context of mental health service provisionrnamong HCWs providing healthcare in a rural districts in Southern Ethiopia. rnThe study has followed a mixed methods approach divided into three overlapping studies. rnStudy one (conceptualisation of burnout) rnMethods: was a qualitative study conducted in the Silti zone. A total of 52 frontline primaryrnHCWs participated in either the in-depth interviews (n=18) or Focus Group Discussionsrn(FGDs) (4 groups, total n=34). Data were analysed using thematic analysis. rnResults: Most participants conceptualised wellbeing as absence of stress rather than as arnpositive state. Many threats to wellbeing were identified. For facility-based workers, the mainrnthreats to wellbeing were inadequate supplies leading to fears of acquiring infection andrnconcerns about performance evaluation. For community -based health workers, the mainrnthreat was role ambiguity. Workload and economic inefficiency were concerns for bothrngroups. Burnout and its symptoms were recognised and projected by most as a problem ofrnother healthcare workers. Derogatory and stigmatising terms, such as, “chronicsâ€, were usedrnto refer to those who had served for many years and who appeared to have become drained ofrnall compassion. Most participants viewed burnout as inevitable if they continued to work inrntheir current workplace without career progression. Structural and environmental aspects ofrnwork emerged as potential targets to improve wellbeing, combined with tackling stigmatisingrnattitudes towards mental health problems. An unmet need for intervention for healthcarernworkers who develop burnout or emotional difficulties was identified. rnStudy two (level of burnout) rnMethods: was a sequential, mixed quantitative and qualitative methods study of HCWsrnworking in all 66 rural primary healthcare facilities (n=145) of the Sodo district, SouthrnEthiopia, where a new integrated mental health service was being implemented. First arnquantitative cross-sectional survey was conducted. The Maslach Burnout Inventory (MBI)rnwas the primary measure of burnout, comprising three domains: emotional exhaustion,rncynicism and reduced feeling of personal accomplishment. Cynicism was excluded fromrnmost analysis because of poor internal consistency of the domain. A set of instrumentsrnmeasuring professional satisfaction and psychosocial stress were also included. Allrninstruments were self-administered. To assist understandings of the quantitative findings, fourrnFGDs were conducted with 17 community and 16 facility-based HCWs. rnResults: A total of 145 HCWs who had worked for a median of 5.0 (25rnthrn and 75rn percentilesrnor interquartile range (IQR) 0.4, 12) years participated in the study. Most participants werernfemale (62.1%; n=90) and working in healthcare facilities (65.5%; n=86); a third wererncommunity-based health extension workers. The median score (IQR) on the MBI was 3 (0, 8)rnfor emotional exhaustion, and 34 (24, 40) for reduced personal accomplishment. Highrndepression symptom scores (adjusted Odds Ratio (aOR) 1.19, 95% CI 1.05, 1.34), poor socialrnsupport (aOR 1.45, 95% CI 1.12, 1.86), and experiencing two or more stressful life eventsrn(aOR 1.41, 95% CI 1.12, 1.95) were associated with higher emotional exhaustion. Onlyrnabout half (50.8%) reported a high level of job satisfaction. FGD participants spoke of highrnlevels of burnout and job-related stress, which the participants believed was under-reported inrnthe quantitative study due to the overwhelming expectations to be strong and exemplary inrnthe eyes of the community. rnStudy three (short-term outcome of burnout) rnMethods: was a longitudinal follow up of the study two participants (145 primary HCWsrnworking in the Sodo district). Burnout was assessed at baseline (when the new service wasrnbeing introduced), as part of the initial cross-sectional survey, and after six months. rnMultivariate logistic regression and generalised estimating equations (GEE) were used tornassess the association between burnout and relevant work-related and psychosocial factors.rnResults were considered statistically significant whenever p-value was less than 0.05. rnthrnResults: A total of 136 (93.8%) of PHC workers were re-interviewed after six months. Therernwas a non-significant reduction in the burnout level between the two time points. In GEErnregression models, high depression symptom scores (adjusted mean difference (aMD) 0.56,rn95% CI 0.29, 0.83 p