Comparison of individual and pooled diagnostic examination strategies during the national mapping of soil-transmitted helminths and Schistosoma mansoni in Ethiopia

BACKGROUND: Laboratory-based studies have highlighted that pooling stool and urine samples can reduce costs and diagnostic burden without a negative impact on the ability to estimate the intensity of soil-transmitted helminth (STH, Ascaris lumbricoides, Trichuris trichiura and hookworms) and schistosome infections (Schistosoma mansoni and S. haematobium). In this study, we compare individual and pooled stool examination strategies in a programmatic setting. METHODS: Stool samples were collected from 2,650 children in 53 primary schools in Amhara Regional State, Ethiopia, during the national mapping of STHs and schistosome infections. Eggs of STHs and S. mansoni were quantified in both individual and pooled samples (pools were made from 10 individual samples) using a single Kato-Katz smear. PRINCIPAL FINDINGS: A pooled diagnostic examination strategy provided comparable estimates of infection intensity with higher fecal egg count (expressed in eggs per gram of stool (EPG)) than those based on individual strategy (Ascaris: 45.1 EPG vs. 93.9, p = 0.03; Trichuris: 1.8 EPG vs. 2.1 EPG, p = 0.95; hookworms: 17.5 EPG vs. 28.5 EPG, p = 0.18; S. mansoni: 1.6 EPG vs. 3.4 EPG, p = 0.02), but had lower sensitivity (Ascaris: 90.0% vs. 55.0%; Trichuris: 91.7% vs. 16.7%; hookworms: 92.6% vs. 61.8%; S. mansoni: 100% vs. 51.7%, p < 0.001). A pooled approach resulted in a ~70% reduction in time required for sample testing, but reduced total operational costs by only ~11%. CONCLUSIONS/SIGNIFICANCE: A pooled approach holds promise for the rapid assessment of intensity of helminth infections in a programmatic setting, but it is not major cost-saving strategy. Further investigation is required to determine when and how pooling can be utilized. Such work should also include validation of statistical methods to estimate prevalence based on pooling samples. Finally, the comparison of operational costs across different scenarios of national program management will help determine whether pooling is indeed worthwhile considering.