Adaptive propensity score procedure improves matching in prospective observational trials

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Epidemiology

Background:

• Randomized controlled trials (RCTs) are the gold standard but not always feasible.
• Alternative trial designs are needed when randomization is not possible.
• This study addresses the need for adaptive designs in case-control trials.

Purpose of the Study:

• To propose a prospective and adaptive matched case-control trial design.
• To incorporate an interim analysis for estimating the matching rate.
• To enable sample size recalculation based on observed data.

Main Methods:

• A prospective and adaptive matched case-control design is proposed.
• An interim analysis with a resampling step estimates the matching rate.
• Sample size recalculation is performed based on the observed mean resampling matching rate.
• The approach was evaluated using simulation and real data.

Main Results:

• The proposed design achieved at least a 10% higher matching rate compared to a naive approach.
• This leads to a better estimation of the true matching rate.
• An interim analysis fraction of [Formula: see text] to [Formula: see text] of control patients is recommended.

Conclusions:

• The proposed resampling step improves the final matching rate estimate in prospective matched case-control trials.
• This leads to increased statistical power through sensible sample size recalculation.
• The adaptive design offers a valuable alternative to traditional methods when randomization is not feasible.