ABSTRACT
To explore the relationship between the ingestion of Agouti paca (AP) and human leptospirosis in Guyana, 19 febrile men who said they had hunted and eaten A. paca were screened for malaria, using bloodsmears, and for leptospirosis, using an enzyme immuno-assay that detects Leptospira -specific IgM. Those found positive for anti-Leptospira IgM were then evaluated further, with a microscopical agglutination test based on a limited panel of serovars from three pathogenic species of Leptospira. Although six of the 18 patients who provided suitable samples for the serology were found seropositive for acute leptospirosis, only three of the 19 patients were found smear-positive for malaria. A clinical-decision model, based on medical histories, the results of physical examinations, and the use of routine urine dipsticks, and enabling prediction of the serological results, was developed. This model, which had 83% sensitivity and 100% specificity for leptospirosis, indicated that, in the absence of serology, most febrile patients reporting AP ingestion could be correctly treated if each was checked for malaria using traditional bloodsmears. The smear-positives should be treated with antimalarial drugs whereas the smear-negatives should be treated for leptospirosis if they had any of the following: a skin rash; lymphadenopathy; abnormal urine sediment (proteinuria or haematuria); and/or no previous history of malaria. In the present study, the relative risk of leptospirosis among the patients who were smear-negative for malaria and fulfilled at least one of these four criteria was 13 (P = 0.0007). In Guyana at least, leptospirosis appears to be common among men who hunt, prepare and ingest AP. Vaccines may be the best, practical form of protection among such men.