Predictors and impact of time to clinical stability in community-acquired pneumococcal pneumonia.

BACKGROUND: A clinical stability (CS) evaluation is thought to be important in community-acquired pneumonia (CAP) treatment, but evidence concerning the time to CS (TCS) remains lacking. METHODS: Among consecutive patients hospitalized with pneumococcal pneumonia, relationships between TCS and other clinical outcomes were examined, and predictors and a predictive TCS score were derived from patient characteristics on admission. RESULTS: A total of 144 patients were enrolled, including 46% and 27% with moderate and severe pneumonia, respectively, defined by the pneumonia severity index (PSI). The median TCS was 2 days, and was significantly correlated with the length of hospital stay (r = 0.595); a longer TCS was significantly associated with the more presence of poor clinical outcomes and ICU stays (adjusted odds ratios: 1.359 and 1.366, respectively). A multivariate Cox proportional hazard model revealed an absence of bilateral pneumonia (hazard rate (HR): 2.107) or bacteremia (HR: 2.520), and mild or moderate pneumonia (HR: 2.798 and 2.515, respectively, versus severe) as predictors of CS. A predictive score had moderate discriminating power for the prolonged TCS (area under the curve: 0.76), and provided similar predictive values for poor clinical outcomes and ICU stays. A score of 3 or more points indicated the prolonged TCS, with a sensitivity and specificity of 73.3% and 70.9%, respectively. CONCLUSIONS: Because TCS has a significant relationship with other clinical outcomes of pneumococcal CAP, the prediction of TCS might lead to the prevention of complications or an earlier transition to oral therapy. Future studies are warranted to validate these results.