Interpreting the term selection bias in medical research.

摘要: The issue of selection bias is often raised in the critical appraisal medical studies, but it poorly defined and misunderstood. This paper, describes three common patterns use term their effects on study results. ways which used are related to 1) representative subjects, 2) subjects exposures, 3) at outcome. Avoidance first these issues, enhances ability generalize a study's other two uses relate internal validity studies. exposures without randomization observational studies can distort results because confounding variables. outcome case-control cross-sectional findings if into distorted according exposure status. Readers should understand different implications improve evaluation Writers discussants be aware differences provide clarifying details when they term.