Estimating the effects of time-varying exposures in observational studies using Cox models with stabilized weights adjustment.

PURPOSE

Assessing the safety and effectiveness of medical products with observational electronic medical record data is challenging when the treatment is time-varying. The objective of this paper is to develop a Cox model stratified by event times with stabilized weights (SWs) adjustment to examine the effect of time-varying treatment in observational studies.

METHODS

Time-varying SWs are calculated at unique event times and are used in a Cox model stratified by event times to estimate the effect of time-varying treatment. We applied this method in examining the effect of an antiplatelet agent, clopidogrel, on events, including bleeding, myocardial infarction, and death after a drug-eluting stent was implanted in coronary artery. Clopidogrel use may change over time on the basis of patients' behavior (e.g., non-adherence) and physicians' recommendations (e.g., end of duration of therapy). We also compared the results with those from a Cox model for counting processes adjusting for all covariates used in creating SWs.

RESULTS

We demonstrate that the (i) results from the stratified Cox model without SWs adjustment and the Cox model for counting processes without covariate adjustment are identical in analyzing the clopidogrel data; and (ii) the effects of clopidogrel on bleeding, myocardial infarction, and death are larger in the stratified Cox model with SWs adjustment compared with those from the Cox model for counting processes with covariate adjustment.

CONCLUSIONS

The Cox model stratified by event times with time-varying SWs adjustment is useful in estimating the effect of time-varying treatments in observational studies while balancing for known confounders.