 Longitudinal attrition

Loss of participants or outcome observations across follow-up. If remaining under observation depends on exposure and determinants of the outcome, analyzing only retained participants can induce selection bias.