Selective Host Mortality in a Catalytic Model Applied to Schistosomiasis

Abstract

Analysis of a "catalytic" model used to study schistosomiasis and other diseases shows that if apparent infection with the disease contributes an increment to mortality, then a standard method for finding the force of infection and the force of defection (loss of infection) gives biased estimates. Under plausible approximations, the force of defection obtained by the standard method overestimates the true rate of loss of infection by half the increment in mortality due to infection. Evidence from the literature that infection with schistosomes does increase human mortality suggests an upward bias in published estimates of the rate of loss of infection. This suggestion is confirmed in one instance by a subsequent direct measurement showing that the true rate of loss of infection is lower than that previously estimated. Published estimates of the rates of infection and defection among snails are substantially biased by mortality due to infection. Increased awareness of the impact of differences in mortality between infected and uninfected individuals should forestall uncritical acceptance of estimates of incidence and loss of infection obtained from methods which ignore these differences.