skip to main content

Advanced Approaches to Controlling Confounding in Pharmacoepidemiologic Studies

Schneeweiss, Sebastian ; Suissa, Samy Strom, Brian L ; Hennessy, Sean ; Kimmel, Stephen E

Pharmacoepidemiology, 2019, p.1078-1107

Chichester, UK: John Wiley & Sons, Ltd

Texto completo disponível

Citações Citado por
  • Título:
    Advanced Approaches to Controlling Confounding in Pharmacoepidemiologic Studies
  • Autor: Schneeweiss, Sebastian ; Suissa, Samy
  • Strom, Brian L ; Hennessy, Sean ; Kimmel, Stephen E
  • Assuntos: bias ; case–cohort design ; case–crossover design ; confounding ; disease risk scores ; instrumental variable analysis ; matching ; nested case–control design ; new user design ; propensity scores
  • É parte de: Pharmacoepidemiology, 2019, p.1078-1107
  • Descrição: Systematic errors in pharmacoepidemiologic studies, particularly confounding, often constitute serious concerns to inference on the causal relationship between drug use and intended and unintended health outcomes. Such concerns can be addressed by choosing study designs that are robust towards investigator errors and restrict the study population to comparable treatment groups, limiting confounding. The new user cohort design, including incident and prevalent new users, is a robust starting point including various sampling strategies within cohorts, the case–cohort and nested case–control designs. Self‐controlled designs help pinpoint triggers of acute events and are often of use when studying vaccines or in the absence of appropriate comparison medications. Covariate balancing tools like the propensity score and disease risk score and its refinements are analytic strategies that have proven to be particularly useful in the analysis of secondary healthcare databases. Exposure variation among providers, regions, or time periods can be exploited for improved confounder control with instrumental variable analyses if several assumptions are fulfilled. Sensitivity analyses will help investigators understand the robustness of findings regarding residual confounding and structural assumptions.
  • Editor: Chichester, UK: John Wiley & Sons, Ltd
  • Idioma: Inglês

Buscando em bases de dados remotas. Favor aguardar.