Online Program

Selecting Optimal Observational Methods for Comparative Effectiveness Research

*Douglas Landsittel, University of Pittsburgh 
Joyce Chang, University of Pittsburgh 
Elan Cohen, University of Pittsburgh 
Andrew Topp, University of Pittsburgh 
Ester Saghafi, University of Pittsburgh 
Sally Morton, University of Pittsburgh 

Keywords: observational, comparative effectiveness research, systematic review, simulations, decision tool

Patient-centered comparative effectiveness research (PC-CER) seeks to evaluate which treatment works best when. Observational data represent an important and increasingly available tool for CER and PCOR. Observational data have many strengths, such as generalizability and real-world applications, but are also subject to many limitations. There are many methods to address those limitations, but there are no clear answers to the most fundamental questions, such as "Which method should I trust more if instrumental variables and propensity scores lead to different results for my data?" This presentation summarizes findings from a Patient-Centered Outcomes Research Institute (PCORI)--funded study that aims to 1) conduct a systematic review, 2) conduct simulations to address gaps in the literature, and 3) develop a decision tool for recommending optimal methods for a given data set. This presentation will summarize results of the systematic review and preliminary results of simulations and propose a structure for a decision tool. The final impact will be providing clinicians, and ultimately patients, with the best possible information about which treatment works best when.