Combined elevated C-reactive protein and detectable plasma RNA has predictive value for hospitalization during an outpatient COVID-19 convalescent plasma clinical trial.

Background: Laboratory biomarkers that predict progression from mild COVID-19 to hospitalization are not well defined. Objective: To evaluate the performance of C-reactive protein (CRP) and plasma SARS-CoV-2 RNA for predicting hospitalization among mild COVID-19 outpatients enrolled in a randomized controlled trial of COVID-19 convalescent plasma (CCP). Methods: CRP and plasma SARS-CoV-2 RNA were measured in 487 outpatient trial participants screened before March 2021. Weighted analyses accounting for sampling probabilities were used to estimate positive predictive values (PPVs) and sensitivities for progression to hospitalization, with performance compared across trial arms. Results: Elevated CRP (>1 mg/dL) was present in 27% (131/487) of participants, and plasma SARS-CoV-2 RNA was detectable in 15% (74/487). In weighted analyses, PPVs for hospitalization were 15% (95% CI, 10-21) for elevated CRP alone, 25% (95% CI, 16-35) for detectable plasma RNA alone, and 37% (95% CI, 23-52) when both biomarkers were present. Among participants with elevated CRP and plasma RNA >500 copies/mL, PPV increased to 72% (95% CI, 37-100). Elevated CRP demonstrated 100% sensitivity for hospitalization in CCP-transfused participants and 54% in control plasma-transfused participants, while detectable plasma RNA demonstrated sensitivities of 60% and 50%, respectively. Among individuals meeting severe COVID-19 risk criteria based on sex, age, obesity, or comorbidities, the combination of normal CRP and undetectable plasma RNA have a high negative predictive value for hospital progression risk. Conclusions: The combination of elevated CRP and detectable plasma SARS-CoV-2 RNA improves laboratory-based risk stratification for progression to hospitalization among outpatients with mild COVID-19.