Process evaluation of the Senegal-Community Nutrition Project: an adequacy assessment of a large scale urban project
Although essential for understanding the reasons for success or failure of large scale nutritional interventions, process evaluation results are rarely reported. Our objective was to assess whether the process output objectives of the Community Nutrition Project (CNP) in Senegal, West Africa, were adequately met. Methods
An adequacy assessment study based on monitoring data for individuals collected during the CNP was used to assess ‘fidelity’, ‘extent’ and ‘reach’ of participants recruitment and of the services provided. The CNP provided underweight or nutritionally at risk 6- to 35-month-old children in poor districts with monthly growth monitoring and promotion and a weekly food supplementation for 6 month periods, provided that mothers attended weekly nutrition education sessions. An exhaustive sample of the participating children (n = 4084) in Diourbel was used for evaluation over the first 2 years. Results
At recruitment, only 66% of children were underweight (vs. 90% expected) varying with the CNP center and cohort, and the child's sex and age. Attendance at growth monitoring reached expected levels (93%vs. 90%) whereas numbers of food supplements distributed and education sessions attended were lower than expected (45%vs. 90% and 62%vs. 80%, respectively). At the end of follow-up, 61% of underweight children recovered vs. 80% expected. Conclusions
Because of CNP design for underweight diagnosis and bias in the targeting process, respect for selection criteria was low and consequently under coverage and leakage occurred. Besides a globally satisfactory process, wide discrepancies were observed between CNP centres concerning the utilization and effectiveness of services. This formative evaluation helped diagnose weaknesses; ongoing feedback enabled the CNP to improve targeting and supply of supplements. It also informed a larger impact evaluation. Some generalizable lessons for similar programmes have been highlighted.