The pervasive use of prevalent cohort studies on disease duration,increasingly calls for appropriate methodologies to account for the biases thatinvariably accompany samples formed by such data. It is well-known, forexample, that subjects with shorter lifetime are less likely to be present insuch studies. Moreover, certain covariate values could be preferentiallyselected into the sample, being linked to the long-term survivors. The existingmethodology for estimation of the propensity score using data collected onprevalent cases requires the correct conditional survival/hazard function giventhe treatment and covariates. This requirement can be alleviated if the diseaseunder study has stationary incidence, the so-called stationarity assumption. Wepropose a nonparametric adjustment technique based on a weighted estimatingequation for estimating the propensity score which does not require modelingthe conditional survival/hazard function when the stationarity assumptionholds. Large sample properties of the estimator is established and its smallsample behavior is studied via simulation.
展开▼
