What is the Right Indicator to Evaluate Impact of Interventions? Examples from a Rice Fortification Trial in Cambodia

Abstract

Objectives: Large-scale programs often use proxy indicators for evaluation purposes but little is known on whether use of these indicators might lead to erroneous conclusions regarding impact of programs. In the FORISCA-UltraRice+NutriRice trial in Cambodia, fortified rice was introduced through the WFP schoolmeal program (WFP-SMP). A wide range of impact indicators collected during the study was explored to test the usefulness of 2 proxy indicators (anemia and stunting) for program evaluation. Methods: In a placebo-controlled, double blinded trial, 20 schools (n=9500) were randomly allocated to no SMP (control, 4 schools), normal schoolmeal (placebo, 4 schools) or a schoolmeal with fortified rice (3 different types, 4 schools each). Data on absenteeism and morbidity was available for all schoolchildren, whereas biochemical data (hemoglobin, iron, zinc and vitamin A status), anthropometry, parasite infection and cognitive performance was available for ~25% of children. Results: At baseline, anemia prevalence was significantly lower in non-SMP (9.5% vs 16.6%, P<0.05) and iron status significantly higher in non-SMP schools than in intervention schools (P<0.005). Stunting prevalence was high (41.5%) and not different among the schools. After 6 months of consumption, fortified rice had no overall impact on anemia prevalence or hemoglobin concentrations, but improved micronutrient status and cognitive performance in some intervention groups. Conclusions: Anemia prevalence is a valid indicator to select schools for SMP, but stunting prevalence had little value in this study. Changes in anemia prevalence did not correspond to changes in functional outcomes, showing that program evaluation requires better proxy indicators. Funded by USDA/FAS, WFP-DSM consortium and IRD.