CUR-65 score for community-acquired pneumonia predicted mortality better than CURB-65 score in low-mortality-rate settings

Abstract

Background: It is not clear whether low blood pressure criterion could be removed from CURB-65 score to orchestrate an improvement in identifying patients with community-acquired pneumonia (CAP) in low-mortality-rate settings. Methods: A retrospective cohort study of 1230 CAP patients was performed to simplify the CURB-65 scoring system by excluding low blood pressure variable. The simplification was validated in a prospective two centre cohort of 1409 adults with CAP. Results: The hospital mortalities were 1.3% and 3.8% in the retrospective and prospective cohorts, respectively. The mortality rates in the two cohorts increased directly with the increasing scores, showing significant increased odds ratios for mortality. The pattern of sensitivity, specificity, PPV, and Youden9s index of a CUR-65 score of ≥ 2 for prediction of mortality was better than that of a CURB-65 score of ≥ 3 in the retrospective cohort. Higher values of corresponding indices were confirmed in the validation cohort. The higher accuracy of CUR-65 score for predicting mortality was illustrated by the area under the receiver operating characteristic curve of 0.937, compared with 0.915 for CURB-65 score in the retrospective cohort (p = 0.0073). The validation cohort confirmed a similar paradigm (0.953 vs 0.907, p = 0.0002). Conclusions: CURB-65 score could be simplified by removing low blood pressure to orchestrate an improvement in predicting mortality in CAP patients who have a low risk of death. A CUR-65 score of ≥ 2 might be a more valuable cut-off value for severe CAP.