Prospective individual patient data meta-analysis: Evaluating convalescent plasma for COVID-19

Area of Science:

• Clinical Trials Methodology
• Infectious Disease Epidemiology
• Biostatistics

Background:

• The COVID-19 pandemic necessitated rapid clinical trial initiation to evaluate treatments.
• Recruitment challenges in individual trials due to pandemic's shifting nature hinder timely results.
• Decentralized trial designs often lack mechanisms for data pooling.

Purpose of the Study:

• To propose an innovative framework for pooling patient-level data from multiple, non-networked randomized clinical trials (RCTs).
• To present the statistical analysis plan for a prospective individual patient data (IPD) meta-analysis (MA) of convalescent plasma (CP) for COVID-19.
• To demonstrate a generalizable approach for accelerating therapeutic discoveries across various diseases and settings.

Main Methods:

• Prospective individual patient data (IPD) meta-analysis (MA) of ongoing COVID-19 randomized clinical trials (RCTs).
• Adaptive Bayesian statistical approach for continuous monitoring of pooled data.
• Evaluation of safety, efficacy, and harm using accumulating posterior probabilities.

Main Results:

• The proposed framework enables continuous data pooling from disparate, ongoing RCTs.
• Adaptive Bayesian monitoring allows for real-time assessment of treatment effects.
• This approach addresses recruitment challenges and accelerates evidence generation.

Conclusions:

• The presented IPD meta-analysis framework offers a rapid and efficient method for evaluating therapeutic interventions.
• This approach is particularly valuable for time-sensitive research, such as during pandemics.
• The methodology is adaptable for pooling data from RCTs in diverse therapeutic areas and disease contexts.