Are convalescent plasma and intravenous immunoglobulin effective in critically ill COVID-19 patients? – An Update

Objective: To test whether selection bias has affected the reported effect estimate for convalescence plasma (Risk Difference, 0.03; 95%CI: -0.12 to 0.18; p = 0.72) to reduce all-cause mortality at up to day 28. Method: Trial adjusted, simulated comparator trial (SCT) based I2 test for single randomised controlled trials, using revised I2 threshold limits. Result: A baseline covariate imbalance, either due to chance or due to selection bias was established, being equivalent to selection bias level = B% 50-60 (Bias category = 3). This level suggests that 50-69% of (95 – 131 of the total 190) subjects were non-randomly allocated in favour of one treatment group above the other. Conclusion: The effect estimate reported by the tested trial (RD 0.03; 95%CI: -0.12 to 0.18; p = 0.72) does not suggest any significant differences in all-cause mortality at up to day 28 for convalescence plasma compared to human immunoglobulin. The selection bias level, estimated by use of the trial adjusted, SCT based I2 test for single RCTs (B% 50-60) appears to be high enough to have affected the reported effect estimate. In line with new findings, this report updates previous report to the same RCT: EBSM Volume 1, Issue 1, 2025 Pages 51-53, DOI: 10.5281/zenodo.17569898.