Efficient and robust propensity?score?based methods for population inference using epidemiologic cohorts

Most epidemiologic cohorts are composed of volunteers who do not represent the general population. To improve population inference from cohorts, propensity-score (PS)-based matching methods, such as PS-based kernel weighting (KW) method, utilise probability survey samples external references to develop PSs for membership in cohort versus survey. We identify a strong exchangeability assumption (SEA) that underlies existing methods whose failure invalidates inferences, even if propensity model is correctly specified. Herein, we framework estimation and relax SEA weak (WEA) methods. recover efficiency, propose scaled KW (KW.S) method by scaling weights estimation. prove consistency KW.S estimators means/prevalences under WEA provide consistent finite variance estimators. In simulations, had smallest mean squared error (MSE). Our data example showed estimates requiring large bias, whereas proposed MSE.