Performances of Prognostic Scoring Systems in Patients With Healthcare-Associated Pneumonia

Background. There are limited data on the performance of the pneumonia severity index (PSI) and CURB-65 (confusion, urea, respiratory rate, blood pressure, age >= 65) score, which were originally developed for community-acquired pneumonia (CAP), for patients with healthcare-associated pneumonia (HCAP). Methods. The performances of PSI and CURB-65 were retrospectively evaluated in patients with HCAP compared to patients with CAP using prospectively collected data between January 2008 and December 2010. Results. In total, 938 patients hospitalized with pneumonia were eligible for this study, consisting of 519 (55%) with CAP and 419 (45%) with HCAP. The PSI and CURB-65 scores had similar trends of increasing mortality with worsening risk class in both the HCAP and CAP groups. In the HCAP group, however, the low-risk patients identified using CURB-65 had a higher aggregate 30-day mortality compared with the low-risk patients identified using PSI. Although the performances of PSI and CURB-65 in the HCAP group showed similar trends to those observed in the CAP group, the estimated areas under the receiver operating characteristic curve for PSI (0.679, 95% confidence interval [CI], .619.739) and CURB-65 (0.599, 95% CI, .522.675) in the HCAP group were significantly lower than those in the CAP group (0.835, 95% CI, .768-.759 for PSI and .686.832 for CURB-65). Conclusions. The performances of PSI and CURB-65 for predicting 30-day mortality in patients with HCAP were comparable to those in patients with CAP. However, the discriminatory powers of PSI and CURB-65 for 30-day mortality were significantly lower in the HACP group than those in the CAP group.