How to analyses long-term treatment effects using Real Word Data?"
Read world data (RWD) are "Data relating to patient health status and/or delivery of health care routinely collected from a variety of sources", such as electronic health records or claims and billing activities. Such data can be used to perform cohort studies that compare the effectiveness of medical products. Some RWD sources allow for cohort studies with a long follow up time but assessing long-term treatment effects poses some challenges. First, the treatment assignment is not random and potential confounding should be accounted for. Furthermore, patients may discontinue or switch treatment during the follow up period. We illustrate these difficulties in a study with a time-to-event outcome where patients are right-censored if they discontinue the index treatment. We also consider an "intention-to-treat" analysis – where patients are followed after treatment discontinuation. We also discuss alternatives, such as considering discontinuation as a competing risks and inverse probability of censoring weighting and compare and contrast the interpretation of these analysis strategies.
Epidemiology department of Merck KGaA, Germany
Group Leader : Statistical Methods for Precision Medicine
Domain 3 - U900 - CBIO - Bioinformatics, Biostatistics Epidemiology and Computational Systems