-
oa Comparative analysis of clinical scores for predicting COVID-19 severity and mortality: A retrospective prognostic study
- Source: Avicenna, Volume 2025, Issue 2, Nov 2025, 11
-
- 19 April 2025
- 03 July 2025
- 20 August 2025
Abstract
Introduction: The COVID-19 pandemic has posed significant challenges to healthcare systems, underscoring the need for effective triage tools to predict disease severity and guide clinical decision-making. Several prognostic scoring systems have been developed for this purpose, yet their relative effectiveness remains unclear. This study evaluates the performance of five commonly used scoring systems – CRB65, modified PRIEST (mPRIEST), NEWS2, Modified Early Warning Score (MEWS), and Quick COVID-19 Severity Index (qCSI) – in predicting in-hospital mortality and intensive care unit (ICU) admissions among COVID-19 patients.
Methods: A retrospective prognostic accuracy study was performed involving 7,674 adult patients diagnosed with COVID-19 at Hamad General Hospital, Qatar. Data were extracted from electronic medical records, and severity scores were calculated based on patients’ initial clinical presentation. The predictive accuracy of each score for mortality and ICU admission was assessed using the area under the receiver operating characteristic curve (ROC AUC), along with sensitivity, specificity, and Youden’s index.
Results: In the study cohort, the overall mortality rate was 1.1%, with an ICU admission rate of 5.2%. Among the various scoring systems evaluated, mPRIEST demonstrated the highest AUC for predicting mortality (0.899) and ICU admission (0.888), followed by qCSI (0.870 for mortality, 0.865 for ICU admission) and NEWS2 (0.858 and 0.866, respectively). In contrast, CRB65 and MEWS showed lower predictive accuracy, with AUC values of 0.776 and 0.672 for mortality and 0.692 and 0.670 for ICU admission, respectively.
Conclusion: The findings indicate that mPRIEST and qCSI are the most accurate prognostic scores among those evaluated for predicting mortality and ICU admission, thus serving as valuable tools for COVID-19 triage in emergency settings. NEWS2 also demonstrated strong predictive performance, whereas CRB65 and MEWS were comparatively less effective. These insights help optimize early risk stratification and resource allocation in clinical practice.
