Mortality rates in matched cohort, pseudo-randomised and randomised trials of convalescent plasma given to COVID-19 patients
Abstract
Introduction
For more than 80 years convalescent or immune sera has been used in severe life threatening infections. Since March of this year a rapidly increasing number of publications have reported series of Convalescent plasma (CP) investigations in severely ill COVID-19 patients.
Objective
a brief CP scoping review focusing on early mortality
Methods
We searched available data bases. Three randomised trials, two pseudo-randomised observations and twelve matched cohort studies were identified. Random-effects meta-analyses were performed on extracted data
Results
A total of 2,378 CP treated and 5188 “controls” in 17 studies. Individually only two studies were significant for reduction of deaths to 30 days, but all showed a similar percentage reduction. When pooled, meta-analysis was undertaken. It showed that the overall reduction of death was significant for all series RR 0.710 (p=0.00001), all matched cohort series RR = 0.610 (p-value = 0.001) and the two pseudo-randomised series RR 0.747 (p=0.005) but not the three technically inadequate randomised trials, RR 0.825 (p=0.397). In two of these randomised trials, there was faster clearance of Viral DNA at 72 hours after CP than placebo
Conclusion
It is hoped the significance of this less than perfect data will increase interest in completing the delayed randomised trials as the results suggest they could be better than currently licenced drugs. Given increasing published evidence of increased risk of both diagnosis and death from COVID-19 in patients with severe Vitamin-D deficiency, future studies should also study influence of Vitamin-D status of donor and recipient on outcome.
<table-wrap id="utbl1" orientation="portrait" position="float"> <graphic xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="20234757v1_utbl1" position="float" orientation="portrait"/> </table-wrap>Related articles
Related articles are currently not available for this article.