Hospital admission, duration of stay and mortality in community-acquired pneumonia in an acute care hospital.: Correlation between a pneumonia prognosis index and conventional clinical criteria for assessing severity

INTRODUCTION. The objective of this study was to evaluate the management of community acquired pneumonia (CAP) according to conventional clinical criteria applied in the emergency room as compared to a pneumonia prognosis index (PPI) (Fine et al. NEJM 1997). We also analyzed which factors were associated with the need for inpatient treatment in PPI risk category III patients. METHODS. We prospectively enrolled all adults with CAP seen in the emergency room during 1999. The data required to calculate the PPI were collected at admission. Mortality and length of stay were recorded at discharge. RESULTS. A total of 447 patients with CAP were collected, 55.7% in the high-risk classes IV and V. Twenty-seven patients died (6.1%) and 97% of these were within the high-risk classes. There were 362 hospitalizations; 302 (83%) were classes III, IV and V. The readmission rate increased with increases in the risk class, with a range of 4% for class I to 18% for class IV. Eighty-five patients (19%) were treated on an outpatient basis. Risk class III included 80 patients; 63 (79%) were hospitalized, with a length of stay of 7.89 days. The factor most highly associated with hospitalization in this group was abnormal findings on physical examination or on laboratory testing and radiographic studies. (OR: 7.62 [1.5-35.2]). CONCLUSION. In our cohort, the PPI was effective for identifying low-risk patients with CAP who could be treated as outpatients. In risk class III patients, the severity of the disease was the strongest predictor of hospitalization, rather than the presence of comorbid conditions.