This is an original book, different from other texts in epidemiology.It is a short course focused mainly on the validity of epidem- iological investigations.The first chapter on cohort design is followed by chapters on systematic and random errors.The suc- ceeding chapters address various designs of case-control studies.The book ends with a short discussion ofchoice of study design and interpretation of results.The discussion on misclassification and its impact of the estimation of relative risk is excellent.There is a useful table giving esti- mated relative risks for different exposure frequencies, true relative risk levels, sens- itivities, and specificities.There is another table on power calculation.Efficiency, ie, the precision in relation to the cost of a study, is well discussed throughout the text.The author classifies case-control studies into three groups.In "type A" case-control studies, information on exposure is collected prior to follow up.It is analysed when the follow up is completed, and only for the cases and a random sample of controls.In "type B" case-control studies, information on exposure is collected only after the cases have occurred.In "type C" case-control studies, controls selected at random from the population are not available but are, for example, replaced by hospital controls.The chapters on case- control studies are quite extensive.Unfortunately, in the following chapter, dis- cussion ofrandomised control trials and other experiments is only given two pages.The chapter on interpretation of results is also very short.This book does not deal directly with the analysis of epidemiological data and is non-mathematical. Strong points are that there are plenty of case studies, exercises with answers, and a list of 706 classical references on study design, validity, and efficiency in epidemiology.There is also a five page glossary.
No takes yet. Share an insight, caveat, or question.
FRCP Anthony R Green (1992) studied this question.