High-Dimensional Propensity Score in Residual Confounding Control in Pharmacoepidemiologic Studies

The uses of retrospective health care claims datasets are frequently criticized for lacking complete information on potential confounders. Ultimately, the treatment effects estimated utilizing such data sources may be subject to residual confounding. Digital electronic administrative records routinely collect a large volume of health-related information; many of whom are usually not considered in conventional pharmacoepidemiological studies. A high-dimensional propensity score (hdPS) algorithm was proposed in 2009 that utilizes such information as surrogates or proxies for mismeasured and unobserved confounders to reduce residual confounding bias. Since then, many machine learning and double robust (e.g., TMLE) extensions of this algorithm have been proposed to properly exploit the wealth of high-dimensional proxy information. This workshop will (i) demonstrate steps, and implementation guidelines of hdPS utilizing an open data source as an example (with reproducible R codes), (iii) explain the rationale for using the double robust extensions of hdPS, and (iv) discuss advantages, controversies, and hdPS reporting guidelines while writing a manuscript.

More Resources

R/Medicine Event site: https://events.linuxfoundation.org/r-...
Main site: https://www.r-consortium.org/
Blog: https://www.r-consortium.org/news/blog
Join: https://www.r-consortium.org/about/join
Twitter:   / rconsortium  
LinkedIn:   / r-consortium  
Mastodon: https://fosstodon.org/@RConsortium