The first aim of this study was to determine the degree of burnout experienced by healthcare staff that provides service and care to pregnant, labouring, and postnatal women in Malawi. Across MBI subscales, numerous participants experienced burnout in the moderate and high levels in all three subscales (EE: 72%; DP: 43%; PA: 74%). When comparing those in the high category alone, the prevalence among our participants was higher than what has been reported in previous studies [33
]. When compared to one of the Malawian studies, slightly more of our sample experienced EE (33% vs. 31%), an even larger percentage for DP (23% vs. only 5%). The largest percentage difference between the two study groups was for PA (50% of our sample vs. 27% of McAuliffe and colleagues' sample experienced PA) [21
]. The finding could suggest that our participants felt less competent, and thus unsuccessful, than the mid-level providers in McAuliffee and colleagues' sample. In the other Malawian study, Peltzer reported that 90% of the nurses complained about burnout symptoms of which 53% were so seriously affected that their work performance suffered [20
]. Unfortunately the MBI was not used in that study making it difficult to compare findings directly.
Mean scores (EE: 23.1; DP: 6.2; and PA: 37.8) were similar to those obtained for American medical workers, which included both nurses and physicians (22.2, 5.2, and 36.5). Similarly, Hayter reported mean scores for HIV/AIDS specialist community nurses; however a lower EE score was reported (19.2) [10
]. In general, our data suggest that maternal health staff might be at a higher risk of burnout than their colleagues working in other areas. However, the classification of the group by the Maslach cut-off points must be interpreted with caution because these points tend to vary by country based on social and cultural reasons [37
]. Therefore, these cut-off points have yet to be established for Malawian healthcare workers, which warrant further research.
The second aim of the paper was to determine whether or not there were any characteristics that were significantly associated with the level of burnout that participants experienced. Of the characteristics analyzed, five appeared to be associated with burnout: age, number of years married, number of children, number of years in current position and total number of years in the profession. However, in the stepwise regressions, number of children was the only significant predictor for DP and PA.
It can be inferred that domestic duties, such as child rearing and being accountable to one's spouse, may introduce additional and sometimes intensive responsibilities. The preoccupation with providing and caring for one's child and or completing domestic duties, with possibly low spousal support, may create distractions that thwart job performance. This illustrates family-to-work conflicts where the demands from the family/home and work domains are mutually incompatible such that fulfilling the parental role makes it difficult to perform the work role satisfactorily [38
]. This may manifest in the expression of cynicism or feelings of professional inefficacy [42
]. Our finding is consistent with a previous study that showed that as the number of children respondents had increased so did evidence of increasing burnout for the dimension of PA [34
]. Demir and colleagues reported that problems in childcare were significantly associated with EE and PA, adversely influencing Turkish nurses participating in their study [43
]. They also cited another study where having problems related to family and childcare remained stressors for public health nurses. Like these studies, our findings suggest that the hospitals should be concerned with family-work conflict as a source of burnout and as a potential liability in terms of quality of care. Future studies may explore other characteristics of family life that influence burnout, such as social support and quality of parent-child relationships.
Past work supports the assertions that gender, being a nurse, and years of education are associated with burnout [10
], so it was unexpected that none of these variables were significantly associated with emotional exhaustion in our study. This may be due in part to EE being strongly correlated to excessive workload, intensity of contacts with patients, patient-to-healthcare worker ratio, and shift-working [44
]; all of which were not collected or analyzed in this study. Another possibility is that rates of moderate and high EE were quite high in the current study (72%) which may have limited the variance in this sample, and therefore, the ability to find variables that distinguished workers with EE and those without.
Limitations inherent to the study merit discussion. First, the sample size was small and the population narrowly defined. However, given the fact there is a small number of permanent staff employed primarily in maternal health within the department of obstetrics and gynaecology in question, it was unavoidable. The focus on maternal health staff should not discount the possibility of obtaining similar findings in other departments. Nevertheless, a more diverse population, cross cutting various departments and medical settings in Malawi may increase the generalizability of the findings. In addition, a number of interviews or focus groups with participants and/or key informants might have helped add depth to the findings.
Second, the assessment of overtime hours and its association to burnout was not performed. They could explain some of the varying levels of burnout [5
]. Intuitively, working overtime in the healthcare setting potentially increases the caseload size, time spent in direct patient contact, and encounters of unmanageable complications. In our setting, administrative data suggested 25% of the healthcare workers work an extra 17 hours per week, which translate to 75 hours month and only one night of sleep per month (Personal Communication: Meguid, T. Fwd: clearance changes URGENT [email]. Message to: Viva C. Thorsen. 2010 Mar 6, 6:26 am [cited 2010 Mar 6]). This may lead to greater emotional exhaustion depersonalization and personal accomplishment.
The current study was exploratory in nature and as a first step, the MBI was used. However, the low Cronbach's alpha values (.67 for EE, .42 for DP, and .60 for PA) suggest that the MBI may not be as culturally appropriate as assumed. Future studies will have to resolve this limitation by, for example, adding more items to the respective scales and performing item and factorial analyses. Moreover, the use of additional instruments, such as organizational commitment questionnaire, job satisfaction survey, workload scale, or interpersonal conflict at work scale to supplement the MBI will enrich future analyses and provide more alternative explanations for what is and is not found [33
Last, the data were cross-sectional and collected from self reports which do not allow for causal conclusions. Individuals with high negative affectivity may perceive their work context more negatively, which would artificially strengthen the associations between burnout symptoms and work environment.