Improving Discrimination in Predicting Level of Care Needed for Patients Admitted with Pneumonia

Background: Various risk stratification scores are used to predict outcomes among patients with pneumonia. We have developed a novel model that predicts the risk of death or intensive care unit transfer in internal medicine. Objective: To compare the ability of two prediction models to predict clinical outcomes in patients admitted for pneumonia, using information available at the time of admission. Design: Comparison of two prediction models. Participants: 3856 pneumonia admissions to the internal medicine service of a tertiary medical center. Main Measures: We compared the ability of two scores to predict in-hospital mortality and escalation of care (e.g., to the intensive care unit) among patients admitted for pneumonia. One was the CURB-65 score, which is currently in use at our hospital. The other score was one we developed, based on the Elixhauser case mix adjustment model with additional data, such as vital signs and laboratory values. Key Results: 11.8% of patients died in-hospital and 17.7% required an escalation of care. The most common CURB-65 score was 2 (44%), the lowest CURB-65 score ordinarily requiring admission. Our risk prediction score was better than CURB-65 at predicting mortality (c-statistic 0.846 vs. 0.724) and escalation (0.757 vs. 0.633). Our score was able to discriminate among patients classified as similar-risk by the CURB-65 score: of the 1681 patients with a (medium-risk) CURB-65 score of 2, our model placed 180 (11%) into the lowest-risk quintile of patients, and 309 (18%) into the highest-risk quintile. Conclusions: Our risk stratification tool is calculable with information available in the electronic medical record of most hospitals. The new score was much better able to predict the outcomes of in-hospital mortality and escalation of care among patients admitted for pneumonia, compared to CURB-65.