No Product? No Program: Adapting malaria commodity supply to seasonal changes

Background:  Health workers often base commodity order quantities on recent months’ consumption data. But in locations where seasonality affects malaria drug demand and facility accessibility, this Average Monthly Consumption (AMC) method is less effective. To improve the accuracy of the AMC method in locations affected by seasonality, we developed the Look-ahead Seasonality Indices (LSI).

Methods: The LSI multiplies AMC by indices compensating for seasonality, adding one variable to the standard resupply formula. The USAID | DELIVER PROJECT analyzed two years of consumption data from health facilities in Zambia, Zimbabwe, and Burkina Faso to develop LSIs for each country. We modeled the LSI approach, comparing it to four other resupply models (including AMC) and evaluating it as an inventory replenishment mechanism and forecasting method.

Results: LSI outperformed AMC in all countries. The LSI in Zambia closely tracked the most sophisticated model, and was 40 percent more accurate in forecasting than AMC.  For inventory replenishment, costs due to lost consumption were 31 percent lower for LSI in Zambia and 50 percent lower in Zimbabwe than AMC.  In Burkina Faso, weaknesses in national consumption data availability affected the ability of the modeling to predict performance.  Additional data would strengthen the model.

Conclusions: Compared to AMC, the LSI provides an enhanced but simple method of calculating resupply for seasonal commodities. Compared to more sophisticated alternatives, the LSI could be implemented easily and potentially applied in countries using AMC. A pilot assessment of the LSI will test its practicality and results in a controlled field-setting.