[Reliability of anthropometric measurements performed by community nutrition workers in a community-based pediatric growth-monitoring program in rural Rwanda].

BACKGROUND

In Rwanda, the community-based growth-monitoring program is implemented via volunteer community nutrition workers. These volunteers are recruited from within their communities, and receive basic training prior to providing services. Utilizing local volunteers improves access to basic nutrition services, and allows the local health jurisdictions to use qualified health care staff more efficiently. In addition to concerns raised in regards to the accountability of unpaid workers, some question the relevance of the data that is collected. We carried out a nutritional survey in the catchment area of Ruli District Hospital to evaluate the reliability of the community nutrition workers' measurements of anthropometric standards collected within the growth-monitoring framework.

METHODS

A nutritional survey was recently organized in the catchment area of the hospital in December 2006. The prevalence rates of malnutrition from the survey were compared with those from the existing community-based growth-monitoring program. Z-test was used to compare the prevalence rate of underweight from the survey with the prevalence rate determined by data collected from community nutrition workers. The concordance of children classified with moderate and severe underweight in each data set was determined by the coefficient Kappa of Cohen.

RESULTS

Our findings show that the recent survey reported an overall underweight prevalence rate of 27.2%. Community data calculated a prevalence rate of 28.8% for the same population. The difference is not statistically significant (P=0.294). Of 724 children evaluated, the survey and the community were in agreement in regards to 454 children classified in the category of good nutritional status, 143 children classified in moderate underweight and 11 children classified in the severe underweight category. The Kappa of Cohen coefficient of 0.636 indicates strong concordance between data sets.

CONCLUSION

Anthropometric measurements provided by the community are reliable. Information gathered from the community can be used for epidemiologic monitoring of malnutrition. To ensure continued reliability, health centers must provide sufficient and permanent training to community nutrition workers. In addition, continued access to essential materials used for measuring nutritional status and maintenance of these materials will be crucial to the program's ongoing success.