We reviewed the first analysis of mortality among patients receiving #ConvalescentPlasma therapy from #COVID19 survivors through the US Expanded Access Program. The effect of #ConvalescentPlasma on #COVID19 mortality has not previously been well described bit.ly/3gpFGDn 
In full disclosure, the lead author is based at Hopkins; however, none of the NCRC reviewers were direct collaborators. Thus, the article had an independent external review by someone without current or past professional collaborations with the author and no Hopkins affiliation.
In this preprint on @medrxivpreprint, Joyner et al analyzed 35,322 hospitalized #COVID19 patients from 1,809 sites overseen by @MayoClinic, from April 4-July 4. They assessed if timing of #ConvalescentPlasma reduced #mortality and if higher doses of #antibodies reduced mortality.
They found fewer deaths among people who got #ConvalescentPlasma in the first 3 days (vs ≥4 days) of their #COVID19 hospitalization. However, other factors differed between those who received plasma early and late (such as time to hospitalization, BMI, race, clinical status)
The titer, or dose, of #SARSCoV2 #antibodies was only known for about 10% of people in the study. Among them, higher doses were associated with lower #COVID19 mortality (14% vs 9% at 7 days, and 30% vs 22% at 30 days).
STRENGTHS: Large sample size, so the study was well-powered to detect differences in mortality. Adjustment for several potential clinical confounders. Stratification by month (April, May & June) to account for time trends in disease presentation, course and treatment.
LIMITATIONS: There was no control group that did not receive convalescent plasma included in this study. Cutoffs defining high and low titer plasma were not determined a priori.
LIMITATIONS: There were also significant differences in mortality over time due to other interventions and between comparison groups that may have resulted in the observed effects, rather than #ConvalescentPlasma itself.
SUMMARY: This large study of #ConvalescentPlasma suggests that earlier treatment w/higher titer plasma may have modest survival benefit. However, it does not provide convincing evidence of a benefit and highlights the need for randomized controlled studies medrxiv.org/content/10.110…
