Leverage multiple real-world data sources in single-arm medical device clinical studies

Area of Science:

• Clinical Trials Methodology
• Real-World Data Analytics
• Biostatistics

Background:

• Increasing utilization of real-world data (RWD) in medical product development.
• Need for robust methods to generate real-world evidence (RWE) for decision-making.
• Challenges in integrating diverse RWD sources into clinical study designs.

Purpose of the Study:

• To propose a study design and data analysis approach for prospective clinical studies using multiple real-world data sources.
• To enhance the estimation of parameters of interest by borrowing information from RWD.
• To provide a framework for leveraging RWD in medical product evaluation.

Main Methods:

• Prospective, single-arm clinical study design.
• Integration of patient-level covariate and outcome data from multiple RWD sources.
• Application of the propensity score-integrated composite likelihood method for parameter estimation.
• Weighted averaging of estimates from each RWD source for a final robust estimate.

Main Results:

• The proposed approach effectively integrates RWD into prospective studies.
• Simulation studies demonstrate the performance of the propensity score-integrated composite likelihood method.
• The method allows for borrowing information from external RWD sources to strengthen study findings.

Conclusions:

• The proposed approach offers a statistically sound method for incorporating RWD into clinical studies.
• This methodology can improve the efficiency and informativeness of medical product development and evaluation.
• The study provides a practical framework and hypothetical example for implementing the approach.