Background and Objectives: The COVID-19 pandemic has necessitated the development of reliable prognostic tools to predict patient outcomes and guide clinical decisions. This study evaluates the predictive utility of several clinical scores—PAINT, ISARIC4C, CHIS, COVID-GRAM, SOFA, and CURB-65—for in-hospital mortality among COVID-19 patients, comparing their effectiveness at admission and seven days post-symptom onset. Methods: In this retrospective cohort study conducted at the Clinical Emergency Hospital Pius Brînzeu in Timișoara, adult patients hospitalized with confirmed SARS-CoV-2 infection were included. The study was approved by the Local Ethics Committee, adhering to GDPR and other regulatory standards. Prognostic scores were calculated using patient data at admission and Day 7. Statistical analyses included ROC curves, Kaplan–Meier survival analysis, and multivariate Cox regression. Results: The study comprised 269 patients, with a notable distinction in outcomes between survivors and non-survivors. Non-survivors were older (mean age 62.12 years) and exhibited higher comorbidity rates, such as diabetes (55.56% vs. 31.06%) and cardiovascular diseases (48.15% vs. 29.81%). Prognostic scores were significantly higher among non-survivors at both time points, with PAINT and ISARIC4C showing particularly strong predictive performances. The AUROC for PAINT increased from 0.759 at admission to 0.811 by Day 7, while ISARIC4C demonstrated an AUROC of 0.776 at admission and 0.798 by Day 7. Multivariate Cox regression indicated that a PAINT score above 8.10 by Day 7 was associated with a hazard ratio (HR) of 4.9 (95% CI: 3.12–7.72) for mortality. Conclusions: The study confirms the strong predictive value of the PAINT, ISARIC4C, CHIS, COVID-GRAM, SOFA, and CURB-65 scores in determining mortality risk among hospitalized COVID-19 patients. These scores can significantly aid clinicians in early-risk stratification and resource prioritization, potentially enhancing patient management and outcomes in acute care settings.
Read full abstract