Objective: To compare the predictive accuracy for 30-day mortality of the CURB65 score adopted by the British Thoracic Society and the simpler CRB65 score to APACHE (acute physiology and chronic health evaluation) II in patients with methicillin-resistant Staphylococcus aureus (MRSA) pneumonia.
Design: A retrospective, single-center, observational cohort study.
Setting: Barnes-Jewish Hospital, a 1,200-bed urban teaching hospital.
Patients: Adult patients requiring hospitalization identified to have MRSA pneumonia.
Interventions: Retrospective data collection from automated hospital, microbiology, and pharmacy databases.
Measurements and main results: Two hundred eighteen patients with MRSA pneumonia were identified over a 3-year period. Forty-four patients (20.2%) died during hospitalization. All three prediction rules had high negative predictive values but relatively low positive predictive values at most cut-off points examined. APACHE II had the greatest area under the receiver operating characteristic curve (0.805; 95% confidence interval [CI], 0.743 to 0.866) compared to CURB65 (0.634; 95% CI, 0.541 to 0.727) and CRB65 (0.643; 95% CI, 0.546 to 0.739) [p < 0.05 for both comparisons]. Similar results were obtained when the subgroups of community-acquired MRSA pneumonia and health-care–associated MRSA pneumonia were examined separately.
Conclusions: APACHE II outperformed CURB65 and CRB65 for initial prognostic assessment in MRSA pneumonia.