Bayesian stopping guidelines for heart valve premarket approval studies.

Area of Science:

• Biomedical Engineering
• Clinical Trial Monitoring
• Statistical Analysis in Medicine

Background:

• Data Monitoring Committees (DMCs) oversee premarket approval (PMA) studies for new medical devices, like heart valve prostheses.
• DMC decisions on study continuation are often subjective, relying on expert opinion regarding safety and efficacy.
• An objective method is needed to evaluate complication rates and inform DMC decisions.

Purpose of the Study:

• To develop an objective aid for DMCs to evaluate complication rates during PMA studies.
• To provide a method for predicting the probability of a heart valve prosthesis study meeting objective performance criteria (OPC).
• To assist DMCs in making informed decisions about study continuation based on safety and efficacy data.

Main Methods:

• Utilized a Bayesian approach to calculate the probability of passing predefined objective performance criteria (OPC) for heart valve complications.
• Applied the method to data from an ongoing PMA study at interim review points.
• Calculated probabilities based on accumulated patient-years and observed event counts.

Main Results:

• Developed a predictive method for DMCs to estimate the probability of passing the OPC test for each specific complication.
• Created a graphical aid to visualize these probabilities at 100 patient-year intervals.
• Demonstrated an objective assessment of the likelihood of the valve prosthesis ultimately meeting study endpoints.

Conclusions:

• The proposed Bayesian method provides an objective assessment to supplement the subjective deliberations of DMCs.
• This tool aids DMCs in evaluating the probability of a heart valve prosthesis study successfully meeting its objective performance criteria.
• Enhances the decision-making process for stopping or continuing PMA studies based on statistical evidence.