Background: Our purpose was to assess the usefulness of a strategy based on the clinical presentation in order to choose antibiotics in patients with non-severe community-acquired pneumonia (CAP).
Patients and method: During one year, all patients admitted to the emergency department with a diagnosis of non-severe (Fine risk-classes I, II and III) CAP, were randomized and assigned into the following groups: GROUP 1: the clinical presentation was not taken into account and all patients were treated with levofloxacin; GROUP 2: patients with typical presentation were treated with amoxicillin and patients with atypical presentation were treated with clarithromycin. The following aspects were evaluated during the follow-up: presence of fever after 72 h of treatment, days of hospitalization and complications.
Results: The eventual population analyzed included 125 patients: 59 (47%) were assigned to Group 1 and 66 (53%) to Group 2. Patients assigned to Group 1 had a lower rate of fever after 72 h of treatment (7% vs 27%, p = 0.005); they were hospitalized for fewer days (4.8 vs 8.1 days, p = 0.01), and had less complications: changes in antibiotic treatment (10% vs 23% p = NS), admissions in ICU (0% vs 5%, p = NS), re-admissions in hospital (0% vs 5%, p = 0.05) and mortality (0% vs 2%, p = NS).
Conclusions: A strategy of empiric treatment of non-severe CAP with levofloxacin as the antibiotic of choice is more useful than that of a treatment based on the clinical presentation.