Economic evaluations are used by decision makers when deciding to reimburse or not a given health technology. Although most are based on randomized controlled trials, interest by decision makers for economic evaluations based on observational studies is now rising. Unfortunately, economic evaluations based on observational studies are prone to numerous biases and limitations.
My research program aims to address some of these biases in hopes of improving the validity of future economic evaluations conducted in Quebec and abroad. Specifically, my program aims to address two potential limits of these types of studies; 1) problems due to confounding bias and 2) problems due to the age of data included within economic evaluations.
Confounding bias is frequent problem within observational studies. Although there exists many methods that can be used to address this bias, performance of these methods have scarcely been assessed within observational economic evaluations. In my research program, I will examine the performance of various confounding adjustment methods. Performance of these methods will be examined in the context of an economic evaluation based on health administrative claims data. Analyses will be repeated, each time using alternative methods, in order to identify which is best to account for this bias.
There exists certain techniques that can be used to update historical data and therefore include them within economic evaluations. However, validity of these techniques remains uncertain. In my research program, I also aim to examine the strengths and limits of these techniques in hopes of identifying under which conditions the results they procure remain valid.
Results that I will obtain will help improve the validity of economic evaluations based on observational studies. Ultimately, this will help to improve the medical resources allocations offered to patients.