ABSTRACT
We address here two conceptual and two methodologic issues related to the estimation of incidence from field survey data. The conceptual issues are (1) the distinction of first from total incidence, and (2) the threshold at which disease is defined as present. The methodological issues are (1) the effect of systematic error in case detection on incidence, and (2) the effect of random error in case detection on incidence. Incidence data are presented for eight categories of mental disorder and examples are given from our work in psychiatric epidemiology.
