[Minimizing losses and maximizing efficiency in the detection of acute severe malnutrition].

OBJECTIVE One of the many challenges faced by epidemiologists is to adequately plan and optimize subject selection procedures in terms of effectiveness and efficiency. In the context of a case-control study involving severe acute malnutrition, a two-step subject selection procedure is used. The aim of the article is to establish an appropriate cut-off point for the screening phase and to achieve a common ground for standards, efficiency in detecting severe malnutrition and the two-step procedure. METHODS The study includes 154 children under the age of 2 from two different hospitals. To determine the ideal cut-off point of weight-for-age (WFA), the following estimators are of interest: the proportion of false negatives (PFN), false positives (PFP) and the percentage of total gain by time (ptg). Weight-for-height (WFH) (cut-off point at -2 SDs) is used as reference for establishing severe acute malnutrition. RESULTS The magnitude of false negatives declines steadily until the 3rd WFA percentile (P3) and reaches zero close to P9. At this point, the PFP is around 0.4. The ptg decreases sharply up to P4, declining smoothly towards P10 thereafter (54.5%). CONCLUSIONS The WFA P10 can be recommended for the screening phase. At this cut-off point, there is still efficiency whereas losses of true cases of severe acute malnutrition are minimized.