ObjectiveTo evaluate the relationship between job satisfaction, burnout syndrome (BS) and depressive symptoms (DS) based on the job demand–control framework model on a nationally representative sample of physicians working in the Peruvian Health System.SettingWe carried out a secondary data analysis of the National Survey of Satisfaction of Users in Health 2016 in Peru.Primary and secondary outcome measuresOur study assessed the development of the predictive model and had two parts: (1) to evaluate the association among the variables based on the job demand–control framework, and (2) to assess the proposed model acceptability using the structural equation modelling approach to estimate goodness-of-fit indices (GOFIs).ParticipantsWe excluded physicians older than 65 years, who did not report income levels or who had missing data related to the workplace. Thus, we analysed 2100 participants.ResultsThe prevalence of DS was 3.3%. Physicians’ work-related illnesses had more probability to result in DS (prevalence ratio=2.23). DS was moderately related to BS dimensions (r>0.50); nevertheless, the relationships between DS and the three job satisfaction scales were weak (r<0.30). The first predictive model based on the variables, DS, BS and job satisfaction, had low GOFIs (comparative fit index (CFI)=0.883; root mean square error of approximation (RMSEA)=0.125). In a second evaluation, we used models with correlated errors obtaining optimal GOFIs (CFI=0.974; RMSEA=0.060).ConclusionsOur study identified a stable model to explain the relationship between job satisfaction, BS and DS among physicians. The results are consistent with the job demand–control framework. They could be applied to decision-making in occupational contexts in Latin American low/middle-income countries.

We evaluated the relationship between job satisfaction, burnout syndrome, and depressive symptoms among Peruvian physicians using the job demand–control framework. We carried out a secondary data analysis of the National Survey of Satisfaction of Users in Health 2016 in Peru. The first part was based on the association among the variables based on the job demand–control framework. The second part focused on estimating the acceptability of the proposed model based on the study variables through the structural equation modeling technique. A total of 2,100 physicians were included, and the prevalence of depressive symptoms was 3.3%. Physicians who had a work-related illness were more than twice as likely to have depressive symptoms (PR=2.23) compared with those who did not. The first predictive model based on the variables, depressive symptoms, burnout syndrome, and job satisfaction had a low goodness-of-fit index. Therefore, for a second evaluation, models with correlated errors were considered, and optimal goodness-of-fit indices were found (CFI=0.974; RMSEA=0.060). Our study identified a stable model to explain the relationship between job satisfaction, burnout, and depressive symptoms among physicians. The results are consistent with the job demand–control framework and can be applied to decision making in occupational contexts in Peru.