Design issues for drug epidemiology.

摘要: Despite the difficulties involved in designing drug epidemiology studies, these studies are invaluable for investigating unexpected adverse effects of drugs. The aim this paper is to discuss various aspects study design, particularly those issues that not easily found either textbooks or review papers. We have also compared and contrasted with randomized controlled trial (RCT) wherever possible. Drug especially useful many situations where RCT suitable, even base has be defined before appropriate cohort subjects assembled. If all cases identified, then a referent sample controls may assembled by random sampling base. cannot assembled, hypothetical secondary need created. Preferably, only new-users should included, risk-ratio will different acute users chronic users. Studies usually possible when researching unintended It difficult efficacy because confounding indication. In occasional circumstances it (examples given). Discussion dangers generalisability mind provided. Additionally, similarities design between discussed detail, as well design-characteristics shared two methods.