Preventing mismanagement of community-acquired pneumonia at an urban public hospital: implications for institution-specific practice guidelines

To assess institutional performance of key diagnostic and therapeutic interventions and to identify areas amenable to improvement in the management of community-acquired pneumonia (CAP). A chart-based retrospective study. Cook County Hospital, a large, urban, public teaching hospital. Adult inpatients with a hospital discharge diagnosis of CAP. None. Fifty hospital admissions were reviewed. Only 25 patients (50%) had two specimens obtained for blood culture, and sputum was sent for Gram's stain and culture for only 11 patients (22%). Approximately one third of the patients had portable anterior-posterior instead of standard posterior-anterior and lateral chest radiographs performed. Physicians in the emergency department (ED) tended to be less likely to note the presence of multilobar infiltrates or pleural effusions than the attending radiologists. The antibiotic regimens employed in the ED and on the inpatient wards were widely variable. The mean time from hospital entry until administration of the first dose of antibiotics was 5.5 h for the 18 patients for whom treatment was initiated in the ED vs 16.1 h for the 27 patients admitted through the ED for whom therapy was deferred until ward admission (p < 0.001, Student's t test). Institutional variations in the performance of basic diagnostic and therapeutic interventions for patients with CAP may be substantial. The local performance of these key processes of care should be assessed to help direct the formulation of institutional practice guidelines for the management of CAP.