An RDW-based score predicts mid-term mortality in patients hospitalized with community acquired pneumonia (CAP): a derivation and validation study

Background: red-blood-cell distribution width (RDW) has been associated with mortality in several clinical conditions, including CAP. Aim: derivation and validation of an RDW-based score to predict mid-term mortality in patients admitted for CAP in a general ward.: Methods: this retrospective multicentric observational study consisted of a derivation and a validation cohort of patients admitted for CAP in Internal Medicine wards in 2016 and 2017. 87 clinical and laboratory candidate variables were collected. Main outcome was all cause of death, median follow up was 18 months; survival was estimated with the Kaplan–Meier method. Cox multivariate analysis was used to develop the score. In the validation cohort the predictive performance of the score was assessed by discrimination and calibration analysis. Results: in derivation cohort (N=405, median age: 80 years, 56% males, median CURB65=2), 18-months mortality was 44% whereas in the validation one (N=217, median age 79 years, 54% males, median CURB65=2), was 34%. Cox multivariate analysis identified as independent predictors for death: RDW (HR 1.18; 95%CI 1.10-1.25, p<0.001), systolic BP < 90 mmHg (HR 3.38; 1.76-6.48, P<0.001), BUN > 21 mg/dl (HR 1.58; 1.13-2.20, p<0.005), temperature (HR 0.84; 0.73-0.97, p=0.02), confusion (HR 2.84; 2.09-3.86, p<0.001). The predictive accuracy of the score showed good discrimination (c index 0.75 (95% CI 0.70-0.78) and 0.81 (0.77-0.84) in the derivation and validation cohort respectively) and calibration. Conclusions: we derived and validated a simple clinical score including RDW to predict mid-term mortality after admission for CAP in a general ward.