Introduction
The prevalence of health workforce absenteeism in Uganda is estimated to range between 37% and 48%.1–3 High absenteeism among Ugandan doctors has been associated with delays in the provision of care and adverse outcomes for mothers and neonates.4 The reduction of health workforce absenteeism from 48% to 20% has been a priority for Uganda’s health sector since 2010, tracked through annual monitoring indicators.5–8 However, there is no corresponding strategy. The absence of a guiding strategy is in part attributable to a limited understanding of the underlying causes of health workforce absenteeism9 10 and the poor implementation of absenteeism policies as articulated in Uganda’s Public Service Standing Orders.11
Among the key drivers of absenteeism in low-income countries is the phenomenon of health workforce dual practice12–14—holding two or more concurrent jobs—as a means of compensating for low salaries in the public sector. In low-income countries, dual practice arises principally from healthcare workers’ perceptions of organisational injustice,13–15 and specifically the perceived inadequacy of working conditions and unfair pay.16–20 Also contributory are scheduled or unscheduled attendance to ‘off station’ duties, such as training workshops and immunisation outreaches,18 19 social activities, including prayers, burials and weddings,21 22 and administrative duties at the decentralised government.18 However, many surveys in these settings commonly report unknown reasons.1 18 23
Typically, several low-income countries define health workforce absenteeism as the proportion of healthcare workers unavailable at a health facility at one or two audit visits conducted by external supervisors such as policy makers19 24 or researchers.1 18 20 Using the audit method, also termed the Public Expenditure Tracking Survey (PETS),20 24 interviewers ask healthcare workers found on duty to list all missing or absent colleagues and to provide reasons why they are absent. The PETS surveys report the same reasons for health workforce absence in low-income countries as those found in the literature.18 19 21–23 25
Although important for understanding the non-availability of the health workforce in low-income countries, the PETS audit method is prone to reporting and social desirability biases. For example, healthcare workers may have an incentive to provide or withhold potentially incriminating information regarding a colleague’s absence. Also, the design inherently lacks validation of findings, which could be improved through accessing workers’ schedules and conducting follow-up interviews with absentees. Consequently, PETS overestimates health workforce absenteeism, reporting rates above 40% in countries where ‘off station’ working is common, such as in Uganda, Bangladesh1 and Tanzania,18 while in countries where out-of-station working is uncommon, such as Laos People’s Democratic Republic (PDR), rates are as low as 17%.20
In contrast, absenteeism data in high-income countries are mainly obtained from healthcare organisation sickness registers,26–28 sickness benefits registers,29–31 insurance records32 and interviews with healthcare managers.33 34 Absenteeism is computed as the work-time lost due to absence,35 36 thereby measuring both non-availability and productivity loss. However, data from these sources assume that absenteeism is due to sickness or work-related stress/injury.37 Moreover, only medically certified absenteeism (typically lasting 3 days or longer) is notifiable in many healthcare organisations,27 38–40 thus failing to detect short-term, often motivated absenteeism.21 41 Yet there is growing evidence in several countries of the Organisation for Economic Co-operation and Development such as Australia, the UK, USA, Canada and the Netherlands that health workforce absenteeism is due to reasons other than sickness.42–46 Nonetheless, the empirical evidence is often limited to nursing professionals working in urban hospital settings10 21 47 and is almost exclusively from high-income countries.21 48
To contribute to the limited evidence, we undertook a qualitative approach to explore the reasons and motives of front-line and supervisory healthcare workers’ engagement in absenteeism in rural Uganda. The rural setting was of relevance because health workforce absenteeism is understood to worsen existing workforce shortages,49–51 potentially widen inequalities in accessing healthcare52–54 and negatively impact on quality and outcomes of care.55
A number of theoretical frameworks have been used to analyse absenteeism across sectors, especially the ‘withdrawal’ or motivational model,56 the work-stress or sickness model,9 57 and the rational choice economic model applied to labour market forces.58 59 The withdrawal model suggests that individual absence such as lateness, absenteeism and turnover manifests in avoidance of dissatisfying work environments and conditions.35 56 Relatedly, the rational choice model suggests that individuals absent voluntarily when they perceive their jobs as secure, such as for tenured compared with contractual workers,58 and if their skills are in high demand, such as for the highly skilled migrant health professionals.60 The two preceding theories have the motivational construct central to explaining a worker’s absence. Contrary, the work-stress and sickness models suggest that individuals absent themselves when job resources are overshadowed by job effort, leading to strain, stress and consequently injury or disability.9 38 57 The empirical findings of this study were evaluated for how well or not they aligned to these prominent individual-focused absenteeism models.