Cause versus prediction: the history of river-bathing as a risk factor and a predictor of infection by Schistosoma mansoni

Abstract

The analysis of the relationship between those parameters used to measure risk and those used to measure efficiency of diagnostic or screening tests is used, in this paper, to demonstrate the limitations of the use of risk factors as predictors of outcomes distribution in a population. The issue was illustrated by the study of the association, in children, of reported bathing in rivers by asking children if they bathed in rivers and the occurrence of Schistosoma mansoni infection. This question in the above context is important because it is part of the clinical anamnesis taught in several Brazilian medical schools. However, while the history of bathing in rivers proved to be a high risk factor it was a bad discriminator between infected and non-infected children. This finding is demonstrated to be mainly related to the low frequency of people with a negative history of bathing in rivers. But, in general, it was shown that the use of isolated or grouped risk factors in the prediction of outcomes is possible and desirable. The field of knowledge discussed in this paper is not yet well explored in its conceptual and practical aspects. However, it represents an important point of convergence between the epidemiology of the causes and the epidemiology of the interventions.