Selecting Treatments for Peripheral Artery Disease: Differences Between Registry and a Randomized Controlled Trial...

Area of Science:

• Vascular Surgery
• Health Services Research
• Clinical Trials

Background:

• Randomized trial participants often differ from real-world patient populations.
• Understanding comparative effectiveness in diverse patient groups is crucial for clinical decision-making.

Purpose of the Study:

• To compare patient characteristics and treatment outcomes between a randomized trial (BEST-CLI) and a large real-world registry (VQI).
• To assess the validity of using registry data with advanced statistical methods for comparative effectiveness research.

Main Methods:

• Descriptive statistics and Cox regression were used to compare 1815 BEST-CLI patients with 104,877 endovascular and 32,120 bypass patients from the VQI registry (2014-2020).
• Inverse probability weighting with propensity scores adjusted for differences in patient characteristics between the trial and registry.

Main Results:

• VQI patients were older, more frequently female, and of non-Hispanic ethnicity compared to BEST-CLI participants.
• After weighting, 5-year mortality was higher with endovascular treatment (26.3%) versus bypass (23.7%) in the VQI cohort.
• Bypass was associated with an 8% lower mortality risk (HR=0.92, P=0.005) compared to endovascular treatment.

Conclusions:

• Significant differences exist between patients in the BEST-CLI trial and the VQI registry.
• Weighted registry data analysis yielded similar treatment effect estimates, supporting its use for real-world comparative questions.