Evaluation of the skin test for hookworm as an epidemiological tool.

The usefulness of a skin test with larval Necator americanus antigen for assessment of hookworm prevalence was evaluated in an endemic area of Costa Rica. In comparison with standard coprologic techniques employed to survey the population, the skin test detected 83% of infections, showing a fairly satisfactory sensitivity. The overall specificity of the test was 50%, i.e., random. No correlation was found between skin reactivity and hookworm burden. The sensitivity of the test increased moderately with age, but its specificity decreased significantly at the same time. False positive reactions were more numerous among persons formerly infected with hookworm who had been negative for as long as 5 years. There was an indication of cross reactivity with intestinal nematodes other than hookworm. The test was used to detect hookworm infected persons in the community for selective treatment, in comparison with mass treatment of all the people in another village. The selective administration of an anthelminthic drug to only skin test positive persons did not achieve the same drop in community hookworm prevalence as did the indiscriminate treatment of the whole population.