Health commodity resource allocation formula to achieve universal health coverage in Tanzania: a policy perspective

Capitation formula funding is increasingly being adopted in low- and middle-income countries (LMICs) including Tanzania. The approach can address health system objectives of equity and efficiency especially when relevant risk adjustments to the capitation funding are made. This paper aimed at documenting the formula review process, identifying variables explaining current pharmaceutical expenditure by health facilities in Tanzania, and assessing their usefulness in allocating prospective health commodities funds. We followed a six phased approach. First, the desk review determined possible variables and evaluated criteria for their selection. Second, stakeholders’ meetings reviewed the current allocation formula and risk adjusters for the new proposed formula. Third, Pearson’s correlation and multiple regression analysis assessed the usefulness of the proposed variables in allocating health commodities funds. Fourth, the stakeholder consensus and data-driven weighting were used to rank the relative importance of the variables. Fifth, scenario analysis using data from national health information system was done to test the functionality and implications of the new proposed formula. Finally, validation meetings were conducted with decision makers. A total of eight risk adjusters were chosen out of the probable 18 identified. Analysis indicated a significant robust correlation (P < 0.01) between per capita health commodities expenditure and all selected variables except antenatal care, which was moderately correlated. The variables explained 77% and 74% of the variation in per capita health commodities expenditure for the model with all primary health facilities (dispensaries, health centre and district hospitals) (PHC model, n = 145) and for the model with health centres and district hospitals (HC/DH model, n = 45), respectively. Resource allocation formula review is a complex process especially in data constrained settings. Nevertheless, we demonstrated the extent to which data driven risk adjusters and basic methodologies can be applied to inform the process. Further analysis to compare funding level from the allocation formula to actual pharmaceutical expenditure would be useful. Increasingly, formula funding approaches are being adopted in Tanzania; however, risk adjustment that adequately reflect variations in health service expenditure are missing. The study demonstrates that data-driven risk adjusters can be applied to inform resource allocation, as we move away from implicit resource allocation - the predominant approach across most Low- and Middle-Income Countries. Evidence-based resource allocation can inform governments’ intended efficiency and equity objectives in meeting the health care needs of the population.