Review of costing methods in national HIV strategic planning documents — An Equitable Financing report

Stagnating HIV funding means efficient resource allocation is increasingly vital for an effective HIV response. Strategic planning documents (SPDs), including investment cases, national strategic plans (NSPs), and their respective resource-needs estimates, are central to this aim. This study reviewed SPD costing methods to identify common methodological strengths, weaknesses and gaps, with a view to improving national planning and budgeting and promoting transparency. Latest costed SPDs from nine countries were reviewed using purposive sampling, including six costed NSPs and four investment cases. Cost estimates were extracted and evaluated for transparency and validity using the Global Health Cost Consortium checklist. SPD narratives, and their respective costings and primary data sources, were compared to evaluate alignment in the scope of interventions and for appropriate adaptation to settings and programmatic contexts. Across all SPDs, a total of 426 cost estimates were extracted, categorized and compared with SPD narratives and source data. Costs were adapted from existing unit cost estimates in the literature (n = 101, 24%), estimated using ingredients-based costing (IBC) (n = 279, 65%) or unstated (n = 46, 11%). When literature sources were publicly accessible (n = 50, 50%), two-thirds of the corresponding estimates were found, of which half matched in activity scope. When examining peer-reviewed literature sources, primary cost literature was found to be robust for prevention interventions and testing and treatment, but sparse for the remaining four service areas. Of the 279 IBC estimates, 38% presented itemized inputs. Within the NSP subsample of SPD, a tenth (11%) included some interventions that were not included in their costings. Among the 256 costed interventions in NSPs (Figure 1), 63% of estimates defined specific activities, of which 91% matched with NSP intervention details and 72% had sufficient detail to assess. Within the NSP subsample of SPDs, a 44% information gap was found between NSPs and their respective costings. A tenth (11%) of interventions specified in NSPs were not included in their costings. Among the 256 costed interventions (Figure 1), 8% did not match in scope with NSP intervention details and 25% of cost estimates had insufficient details to assess scope alignment. This review identifies methodological strengths and weaknesses in SPD costings and identifies opportunities to improve future costings of national HIV planning documents. Although analysts make very good use of existing data sources, lack of empirical cost data and adaptation to specific settings often make IBC necessary to generate appropriate setting- and time-specific cost estimates. As formal guidance on how to do this does not exist, the approaches taken differ. Overall, there are no formalized templates or standards for reporting the full set of adaptations and assumptions for unit costs, rendering the estimates hard to interpret, and missing the opportunity to build on these unit costs to adapt them for future iterations and other settings. This review recommends that bespoke guidance on generating cost estimates for SPDs be developed, including a reporting checklist to institutionalize transparent reporting of methods. This reporting checklist and the resulting cost estimates could be submitted to a live public database that can be scrutinized, adapted and used for costings elsewhere