Epidemiological Methods and Inferences in Studies of Noninfectious Diseases: PHR Review

DURING THE PAST FEW YEARS, thne epideemiological approaclh has been ap- plied increasingly to the studly of noninifectious dliseases, such as cancer and lieart disease.Vari- otis imetlhods of stuldy have been used, and vari-oIIs inferences hiave been derived from the observationis.In several inistances, suclh as the relationship betweeni lung cancer and cigarette smoking, the inferences have provoked consid- erable discussioni.This in turn hlas led to conisideration of certain selected aspects of the conceptual framiiework of tlhese inferences (1-5).However, there still exists a need for a miiore genieral ieview of the methodology anid of the coinsiderations that may influence the deii- vatioln of inifer enices froml the observations.This paper iiiteuids to provide such a reviewv, althoughl it does not pretend to cover all as- pects of the subject.This review may give perspective oIn some of the issues involved.It may stimulate furtlher discussion anid investi- (,atio1i so that the nmethodological problems confronting us can be gradually resolved.