Understanding Comparative Effectiveness Research: Would Observational Data Have Predicted Results Found in Randomized Controlled Trials of Coronary Revascularization?

Purpose

Contemporary randomized controlled trials (RCT) of coronary revascularization have not found any survival advantage from aggressive intervention, while observational data suggest otherwise. We aim to compare outcomes predicted from observational data without strict exclusion criteria with results from patients who would have been eligible for RCTs.

Using a large observational dataset, the aim of this study is to explore the following questions:

1. What proportion of the observational dataset would have been eligible for inclusion in contemporaneous RCTs?

2. How do patient populations in RCT and observational studies differ?

3. In what respect does treatment efficacy in this contemporary analysis differ among RCT-eligible and ineligible populations?

4. Would observational-based statistical models predict the results observed in RCTs in the RCT-eligible observational population and in the RCT population?

Findings

Results show that RCT-eligible patients represent a small percentage of patients in an observational dataset.  Eligible patients have significantly different characteristics than ineligible patients but no significant differential survival. Applying RCT entry criteria to an observational dataset can help to determine reasons for discrepancies between RCT and observational data results.