ABSTRACT
This chapter reviews some of the problems encountered in these National Morbidity Surveys, with particular reference to the analysis of data provided by the longitudinal file of the Second Survey and suggests that some possibilities for solving them. The first problem in making any inferences concerning psychiatric morbidity from records provided by the Surveys is that they do not provide records of illness per se, but of consultations at which a psychiatric problem has been recognized. Many individuals suffer from various forms of psychological distress without ever seeking help from a general practitioner. A serious flaw in the design of the Second National Morbidity Survey is its reliance on records of episodes of disorder in the absence of any clear operational definition of what constitutes an episode. It was simply left to the individual doctor to decide which consultations belonged to which episodes.
