• Chien-Chung Lin, Cheng-Hung Lee, Chiung-Zuei Chen, Yuan-Chin Chu, Tsung-Jen Hung, Han-Yu Chang, and Tzuen-Ren Hsiue.
• Department of Internal Medicine, College of Medicine, National Cheng Kung University, Tainan, Taiwan.
• J Formos Med Assoc. 2005 Mar 1; 104 (3): 164-7.

Background And PurposeThe value of the Pneumonia Severity Index (PSI) in predicting the mortality of patients with community-acquired pneumonia has not been reported in Taiwan. This study investigated the value of this scoring system in estimating mortality of inpatients with community-acquired pneumonia.MethodsThis was a prospective observational study of 118 inpatients and a retrospective chart review of 115 inpatients with radiographically-confirmed community-acquired pneumonia treated at a tertiary referral medical center in southern Taiwan. Patients were stratified into 5 risk classes according to PSI score. Data on demographic characteristics, comorbidities, baseline clinical and laboratory features, in-hospital mortality and length of hospital stay were analyzed.ResultsThe mortality rates according to risk classification were 0% for class I and II, 2.5% for class III, 8.2% for class IV, and 31.2% for class V. A significant correlation was found between these risk classes and medical outcome (p < 0.001). The length of hospital stay was significantly associated with risk class, and ranged from 6.3 days for class I patients to 18 days for class V (p < 0.001).ConclusionThe PSI provided a useful prediction of medical outcome in patients with community-acquired pneumonia. To decrease unnecessary admission, further prospective studies are needed to determine whether outpatient therapy is appropriate for class I or class II patients with community-acquired pneumonia.

14 keywords...

hide…