Commentary

Caution is required when using health facility-based data to evaluate the health impact of malaria control efforts in Africa

Alexander K Rowe¹ , S Patrick Kachur¹ , Steven S Yoon¹ , Matthew Lynch² , Laurence Slutsker¹ and Richard W Steketee³

¹  Malaria Branch, Division of Parasitic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, USA

²  Global Program on Malaria, Center for Communication Programs, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland, USA

³  Program for Appropriate Technology in Health (PATH), Ferney-Voltaire, France

author email corresponding author email

Malaria Journal 2009, 8:209doi:10.1186/1475-2875-8-209

Published:	3 September 2009

Abstract

The global health community is interested in the health impact of the billions of dollars invested to fight malaria in Africa. A recent publication used trends in malaria cases and deaths based on health facility records to evaluate the impact of malaria control efforts in Rwanda and Ethiopia. Although the authors demonstrate the use of facility-based data to estimate the impact of malaria control efforts, they also illustrate several pitfalls of such analyses that should be avoided, minimized, or actively acknowledged. A critique of this analysis is presented because many country programmes and donors are interested in evaluating programmatic impact with facility-based data. Key concerns related to: 1) clarifying the objective of the analysis; 2) data validity; 3) data representativeness; 4) the exploration of trends in factors that could influence malaria rates and thus confound the relationship between intervention scale-up and the observed changes in malaria outcomes; 5) the analytic approaches, including small numbers of patient outcomes, selective reporting of results, and choice of statistical and modeling methods; and 6) internal inconsistency on the strength and interpretation of the data. In conclusion, evaluations of malaria burden reduction using facility-based data could be very helpful, but those data should be collected, analysed, and interpreted with care, transparency, and a full recognition of their limitations.