Epidemic dengue 1 in Brazil, 1986: evaluation of a clinically based dengue surveillance system

Abstract

In the last 15 years, dengue fever has emerged as a major health problem in tropical America. Prevention and control of epidemic disease are enhanced by the rapid identification of new or increased dengue activity. Most surveillance systems, however, identify cases by clinical case reports and, therefore, lack the sensitivity needed for early detection. During the 1986 dengue 1 epidemic in Rio de Janeiro, Brazil, the authors evaluated the usefulness of a clinical case definition by comparing it with laboratory-confirmed infection status of residents in two cities. The case definition had a sensitivity of 64% and a false-positive rate of 57%. Thus, for every 100 laboratory-confirmed dengue infections, 230 cases were reported. Both infected and noninfected residents who used medical services and who lived in the city with the highest transmission were more likely to meet the case definition. Thus, factors unrelated to actual infection influenced the sensitivity. With the use of stepwise logistic regression, the authors analyzed combinations of patient symptoms and produced nine new hypothetical case definitions. However, none of the new definitions had a false-positive rate lower than 38%. This study emphasizes the need for laboratory-based dengue surveillance systems.