Fig. 2

Illustration of benchmark multiplier adjustments to passive malaria surveillance data using a health care utilization index. Each dot represents the average health care utilization index (orange) or resulting multiplier (teal) for one of the 195 Fokontany in Ifanadiana over the study period. In this example, average per capita health care utilization index is normalized from 0.1 to 1, where the maximum is equivalent to 2 visits per year (excluding malaria). Both variables are plotted as a function of distance between each Fokontany and its nearest health center. The solid lines are smoothed conditional means (LOESS method) and the grey areas are the 95% confidence intervals. Fokontany with smaller health care utilization indices have larger multipliers, resulting in greater adjustments after the benchmark multiplier method was applied