Predicting prognosis in patients with Chagas disease: why are the results of various studies so different?

The clinical course of Chagas disease is highly variable, and the identification of patients at risk of dying is a critical issue. In the last decades multiple factors, such as NYHA functional class III/IV, cardiomegaly, low ejection fraction, and non-sustained ventricular tachycardia, have been reported as important markers of a poor outcome. Although prognostic models (usually obtained by techniques of multivariate regression analysis) are important tools to provide estimates of patient outcomes, they can yield problematic results if methodological guidelines and mathematical assumptions are not strictly followed. In this letter we address some of our concerns regarding prediction of death in patients with Chagas disease.