A Superior Prognostic Indicator for In-Hospital Mortality of Adult COVID-19 Patients

Aims: : Despite the progressive decline in the virulence of the novel coronavirus, there has been no corresponding reduction in its associated fatality rate. Our aim was to redefine an accurate predictor of mortality risk in COVID-19 patients, enabling effective management and resource allocation. Methods: : We conducted a retrospective analysis of 2917 adult patients diagnosed with COVID-19 who were admitted to our hospital during two waves of epidemics, involving the Beta and Omicron variants. Upon admission, NT-proBNP levels were measured, and we collected demographic, clinical, and laboratory data. We introduced a new concept called the NT-proBNP ratio, which measures the NT-proBNP level relative to age-specific maximum normal values. The primary outcome was all-cause in-hospital mortality. Results: : Our analysis revealed a higher in-hospital mortality rate in 2022, as shown by the Kaplan-Meier Survival Curve. To assess the predictive value of the NT-proBNP ratio, we employed the time-dependent receiver operator characteristic (ROC) curve. Notably, the NT-proBNP ratio emerged as the strongest predictor of mortality in adult hospitalized COVID-19 patients (area under the curve, AUC = 0.826; adjusted hazard ratio [HR], 3.959; 95% confidence interval [CI], 3.001–5.221; P < 0.001). This finding consistently held true for both the 2020 and 2022 subgroups. Conclusions: : The NT-proBNP ratio demonstrates superior predictive capability compared to several established risk factors, including NT-proBNP, hsCRP, and NLR, when it comes to forecasting in-hospital mortality among adult patients with COVID-19. Clinical Trial Registration: www.chictr.org.cn ChiCTR2000031587 and www.clinicaltrials.gov NCT05615792