RESUMO
OBJECTIVE: To assess the efficacy of an antibiotic protocol to avoid empirical use of third-generation cephalosporins in community-acquired pneumonia (CAP). DESIGN AND SETTING: Retrospective case review of patients with CAP one year after implementing the protocol. Comparison was made with patients with CAP treated at a metropolitan tertiary referral hospital (where use of third-generation cephalosporins was common). PARTICIPANTS: 86 patients (district hospital with an antibiotic protocol) and 72 patients (metropolitan tertiary referral hospital), January - June 1999. OUTCOME MEASURES: Rate of staff adherence to the protocol; patient characteristics associated with poor protocol adherence; demographic and prognostic features of both groups at presentation; duration of intravenous therapy, time to defervescence, length of stay; inpatient mortality rates; and drug cost savings per patient treated according to the protocol. RESULTS: Overall protocol adherence rate was 60%. Patients with penicillin allergy were significantly less likely to receive treatment according to the protocol (P<0.001). At the district hospital, patients were generally older and taking more regular medications. Patients at each hospital had similar prognostic factors and demographic features at presentation. Inhospital mortality (P=0.92; 95% CI, -0.08 to 0.07), duration of fever (P=0.57) and length of stay (P=0.78) were not significantly different between patients treated empirically with penicillin and those treated empirically with third-generation cephalosporins. Treating a patient according to the protocol saved an average of $77.44 in drug costs. CONCLUSION: One year after implementation, our protocol for treating CAP is proving efficacious, although levels of adherence could improve.