Exploring Study Design Foibles in Randomized Controlled Trials on Convalescent Plasma in Hospitalized COVID-19...

Area of Science:

• Clinical Trials
• Biostatistics
• Infectious Diseases

Background:

• Sample size estimation is crucial for the success of randomized controlled trials (RCTs).
• Inadequate sample size estimates may have contributed to varied efficacy results in COVID-19 convalescent plasma (CCP) trials.
• Assessing sample size calculations is vital for understanding the reliability of treatment effect evaluations.

Purpose of the Study:

• To evaluate sample size estimations in RCTs assessing CCP treatment for hospitalized COVID-19 patients.
• To identify factors contributing to discrepancies in sample size attainment and their impact on trial outcomes.
• To analyze the relationship between expected effect size and study outcomes in CCP trials.

Main Methods:

• Systematic literature search of RCTs evaluating CCP in hospitalized COVID-19 patients (Jan 2020-Mar 2024).
• Data extraction included expected effect size, statistical power, significance level, and measured efficacy.
• Analysis focused on sample size adequacy, recruitment challenges, and outcome reporting.

Main Results:

• 32 RCTs were identified; power and significance levels were consistent, but expected effect sizes varied significantly.
• Approximately one-third of RCTs failed to reach their planned sample size, primarily due to recruitment difficulties during pandemic peaks.
• RCTs favoring CCP treatment showed a lower median absolute difference in expected effect size compared to unfavorable trials (20.0% vs. 33.9%, P=0.04).

Conclusions:

• Many RCTs for CCP treatment overestimated efficacy, leading to underestimated sample sizes and reduced statistical power.
• Lower-than-planned participant recruitment further compromised the ability of these trials to detect treatment effects.
• These sample size and recruitment issues likely negatively influenced the primary outcomes of numerous CCP RCTs.