Poster 31: Community-acquired pneumonia in veterans with spinal cord injury

Objective: To describe the characteristics of community-acquired pneumonia (CAP) in persons with spinal cord injury (SCI) and how management is related to outcomes. Design: Cross-sectional retrospective review of administrative and clinical data. Setting: Department of Veterans Affairs (VA) facilities, and for substudy, 3 VA SCI centers (October 1998-September 2000). Participants: Veterans with SCI: 260 inpatients with CAP; in the substudy, 41 inpatients and outpatients with CAP from 3 sites. Interventions: Not applicable. Main Outcome Measures: Percentage of patients with an etiologic diagnosis, mortality rate, mean length of stay (LOS), and number and types of procedures and treatments. Results: Of the 260 inpatients with SCI identified from administrative data with CAP, only 24% had an etiologic diagnosis. Etiologic diagnosis was not associated with mortality after adjusting for several factors (OR=1.38; CI, 0.45-4.20), however, it was associated with an increase in LOS (P=.024). For the substudy, almost 75% of the 41 patients were hospitalized (mean LOS=16.3d) and 3 patients died. Most received chest radiographs (85%), but up to 54% did not receive other tests standard for management of CAP during the first day of care (eg, blood cultures, CHEM 7). Of the 16 patients with sputum cultures, an organism was identified in 44% through microbiology testing. Over 90% received antibiotics within 24 hours of admission. Conclusions: Many patients do not receive the minimum recommended testing. Empiric treatment appears to have been the predominant type of management used in this population. Further research to assess the relationship between clinical characteristics and management with patient outcomes is