Diagnostic accuracy of SMART-COP Score for predicting the need of intensive respiratory and vasopressor support in adults with community-acquired pneumonia-Study from a low resource tertiary care center

Abstract

Background & Objective: Community-acquired pneumonia (CAP) is a predominant cause of emergency department (ED) visits. Bedside scoring tools are essential for early identification of at-risk patients in need of intensive respiratory and vasopressor support (IRVS). The aim of the study was to determine the diagnostic accuracy of SMART-COP as a bedside tool for predicting the need for IRVS in adult ED patients presenting with CAP.

Methodology: A prospective diagnostic accuracy study was conducted at the Indus Hospital and Health Network (IHHN) ED, Karachi from November 2020 to May 2021. Adults (> 18 years) were enrolled. The reference standard was IRVS during hospitalization. Sensitivity, specificity, predictive values and area under the receiver operating characteristic (ROC) curve (AUC) were calculated. Multivariate logistic regression identified independent predictors for IRVS.

Results: Out of the 209 participants, 86 (41.1%) required IRVS. At a cut-off of >3, SMART-COP showed 94.2% sensitivity, 27.6% specificity with AUC 0.71. Each one-point increase in SMART-COP doubled the odds of IRVS (adjusted OR 1.80, 95% CI: 1.43-2.26). Performance was lower in COVID-19 positive patients (sensitivity 94.7%, specificity 26.0%, AUC 0.695) as compared to COVID-19 negative patients (sensitivity 100%, specificity 35.3%, AUC 0.877).

Conclusion: SMART-COP demonstrated high sensitivity and modest specificity in ED for predicting IRVS in patients with CAP. Its simplicity warrants multicenter validation and integration with clinical practice guidelines.