Aim: Leptospirosis is the most frequent zoonotic disease worldwide and it requires an efficient epidemiologic surveillance. This study evaluated the operative definitions of the Costa Rican Protocol for the prevention and control of leptospirosis.Methods: this work is a two-stage (descriptive, analytic) cross-sectional study using a clinical-predictive model by means of logistic regression. Data were obtained from the National Reference Center for Virology and Leptospirosis, Costa Rican

more »

... e for Research and Education in Nutrition and Health (January 2001 to June 2003).Results: Five hundred and sixty-eight records were initially found. One hundred and fifty-four were eliminated for not having the correct information for the classification of the leptospirosis protocol. The other 414 records were analyzed and none of them complied with the criteria for a confirmed case. For this reason, a broader definition taken from the literature was used (confirmed case = serology of 1:800). Consequently, 52 cases were confirmed and 368 were classified as suspicious. Only cefalalgia (OR=0.5; CI 95% 0.2 -1.1) and male gender (OR=3,01; CI 95% 1.2-8.1) showed a significant association with the diagnosis of leptospirosis. When clinical and epidemiologic variables were grouped, the combinations of cephalalgia+myalgia+epidemiologic history(OR=3,8; CI 95% 1.1-14.9) and systemic symptoms+epidemiologic history (OR=0.01; CI 95% 1.2-18.9) showed significant association with the diagnosis, although with a high correlation between them (Kappa > 0.8). Conclusion: with the existing data and by means of the methodology used for the analysis it was not possible to validate the definitions established by the protocol or to generate operative definitions that could be applied on a national scale. It was also impossible to establish a definition of a probable case.