Cost-Effectiveness of Rural Incentive Packages for Graduating Medical Students in Lao PDR

Background The dearth of health workers in rural settings in Lao People’s Democratic Republic (PDR) and other developing countries limits healthcare access and outcomes. In evaluating non-wage financial incentive packages as a potential policy option to attract health workers to rural settings, understanding the expected costs and effects of the various programs ex antecan assist policy-makers in selecting the optimal incentive package.   Methods We use discrete choice experiments (DCEs), costing analyses and recent empirical results linking health worker density and health outcomes to estimate the future location decisions of physicians and determine the costeffectiveness of 15 voluntary incentives packages for new physicians in Lao PDR. Our data sources include a DCE survey completed by medical students (n = 329) in May 2011 and secondary cost, economic and health data. Mixed logit regressions provide the basis for estimating how each incentive package influences rural versus urban location choice over time. We estimate the expected rural density of physicians and the cost-effectiveness of 15 separate incentive packages from a societal perspective. In order to generate the cost-effectiveness ratios we relied on the rural uptake probabilities inferred from the DCEs, the costing data and prior World Health Organization (WHO) estimates that relate health outcomes to health worker density.   Results Relative to no program, the optimal voluntary incentive package would increase rural physician density by 15% by 2016 and 65% by 2041. After incorporating anticipated health effects, seven (three) of the 15 incentive packages have anticipated average cost-effectiveness ratio less than the WHO threshold (three times gross domestic product [GDP] per capita) over a 5-year (30 year) period. The optimal package’s incremental cost-effectiveness ratio is $1454/QALY (quality-adjusted life year) over 5 years and $2380/QALY over 30 years. Capital intensive components, such as housing or facility improvement, are not efficient.   Conclusion Conditional on using voluntary incentives, Lao PDR should emphasize non-capital intensive options such as advanced career promotion, transport subsidies and housing allowances to improve physician distribution and rural health outcomes in a cost-effective manner. Other countries considering voluntary incentive programs can implement health worker/trainee DCEs and costing surveys to determine which incentive bundles improve rural uptake most efficiently but should be aware of methodological caveats.