Abstract
Background
Convalescent plasma (CP) was demonstrated promising benefit for clinical practice involved in efficacy and safety in previous corona virus pandemics, however, the efficacy of CP from COVID-19 sufferers are still controversial and unascertainable based on current randomized controlled trials (RCTs). The urgent needs for affirmative replies on the efficacy and safety of CP for COVID-19 patients must be developed as soon as possible.
Objective
To corroborate the efficacy and safety of CP based on high-quality double-blinded, parallel-arm placebo-control randomized clinical trials and provide evidence-based support for clinical application of CP against COVID-19.
Methods
Such medical electronic databases as Embase, PubMed, and Web of Science were retrieved from inception to March 12, 2022. This meta-analysis synthesizes such dichotomous outcomes as the incidences of 28-day mortality, hospitalization rate, invasive mechanical ventilation, adverse events (AEs)and serious AEs using intention-to-treat (ITT) analysis. Statistical analysis, using Review Manager (RevMan) 5.4.1 software, Mantel-Haenszel (M-H) statistical method and random effects (RE) analysis model, risk ratios (RRs) plus their 95% confidence intervals (CIs) as effect measures, were performed. Two reviewers independently searched, screened, included the eligible clinical trials, extracted data of concern from the mand assessed the risks of bias (ROB) of the included articles with the Cochrane ROB tool 1.0 and Rev Man 5.4.1 software. The effect measures of RRs plus their 95% CIs in this meta-analysis will be computed as dichotomous outcomes of interest. Statistical heterogeneities, subgroup analysis and sensitivity analysis will be fulfilled to explore the heterogeneities and their causes. We evaluate the quality of evidence and put forward strength of recommendations for clinical practice based on the GRADE approach. This prospective meta-analysis protocol has been registered on PROSPERO.
Results
697 references were preliminarily identified from the databases of concern and manual retrieves, and 9 eligible double-blinded, parallel-arm, placebo-control randomized clinical trialswith 1898 subjects in the intervention group and 1696 participants in the control group were ultimately included in the meta-analysis. 7, 4, 3, 3 and 3 eligible trials are adjudged as low ROB for mortality, the rate of hospitalization, the incidence of invasive mechanical ventilation, AEs and serious AEs, respectively; all the rest of included trials are defined as high risk corresponding to the respective outcome. The meta-analysis on the hospitalization rate was abandoned because of high heterogeneity (I2 = 92%) among the inclusion trials. The RRs, 95%CIs and P-values were 0.78 [0.62, 0.97], P = 0.03 on mortality; 0.84 [0.50, 1.42], P = 0.51 on invasive mechanical ventilation; 1.01 [0.78, 1.32],P = 0.92 on AEs; 0.96 [0.73, 1.28], P = 0.80 on serious AEs, respectively, with low or medium levels of heterogeneity; which indicate that CP infusion in COVID-19 patients can efficaciously reduce mortality by 22%, and exhibit excellent safety and not decrease the incidence of invasive mechanical ventilation. Sensitivity analysis on mortality with the combining effect measure (RR 0.83 [0.66, 1.06], I2 0%, Z-value 1.46, P = 0.14) after deleting the study by O’Donnell showed that there is not different between the intervention group and control group, hinting that the deleted study may be more efficacious for reducing mortality. Subgroup analysis on mortality based on age showed that CP therapy in COVID-19 patients aged ≤ 60 years old may more efficaciously reduce mortality by 36%. Sensitivity analyses and subgroup analyses on the other outcomes present robust pooling outcomes. The registration code on PROSPERO is CRD42022324324.
Conclusions
Administration of CP to COVID-19 patients, especially to COVID-19 patients aged ≤ 60 years old, may efficaciously reduce mortality with excellent safety, but does not reduce the incidence of invasive mechanical ventilation.