Extension of Kaplan-Meier Methods in Longitudinal Observational Studies with Time-Varying Exposure

• Activity Number: 405
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Stanley Xu*+ and Susan Shetterly and David Powers and Marsha Raebel and Michael Ho and Thomas Tsai and David Magid
• Companies: Kaiser Permanente Colorado and Kaiser Permanente Colorado and U.S. Census Bureau and Kaiser Permanente Colorado and Denver VA Medical Center and Denver VA Medical Center and Kaiser Permanente Colorado
• Keywords: observational study ; Kaplan Meier estimates ; stabilized weights ; time-varying exposure ; drug eluting stents ; inverse propensity score weighting

Abstract

Stabilized weights have been developed to allow estimation of treatment effects and its variance directly from regression in the presence of confounders in observational studies. Generally Kaplan-Meier method ignores confounders and does not control for time-varying exposures. Recently the association of clopidogrel use with bleeding, myocardial infarction (MI) and death events was examined in 7689 patients with follow-up up to 18 months after receiving drug eluting stents. The mean clopidogrel-covered days was 257;stops and restarts of the medication were common. To compare the Kaplan-Meier estimates with time-varying exposure while adjusting for known confounding factors, we developed Kaplan-Meier like estimates with stabilized weights and developed test statistics. It was shown that clopidogrel use was associated with increased risk of bleeding and a decreased risk of MI or death.