• G W Waterer, G W Somes, and R G Wunderink.
• Department of Medicine, University of Western Australia, Royal Perth Hospital, GPO Box X2213, Perth 6847, Western Australia.
• Arch Intern Med. 2001 Aug 13; 161 (15): 1837-42.

BackgroundAlthough monotherapy for pneumococcal pneumonia is standard, a survival benefit of combination beta-lactam and macrolide therapy has been suggested.HypothesisInitial empirical therapy with a combination of effective antibiotic agents would have a better outcome than a single effective antibiotic agent in patients with bacteremic pneumococcal pneumonia.MethodsA review of adult bacteremic pneumococcal pneumonia within the Methodist Healthcare System, Memphis, Tenn, between January 1, 1996, and July 31, 2000. Empirical therapy was defined as all antibiotic agents received in the first 24 hours after presentation. On the basis of culture results, empirical therapy was classified as single effective therapy (SET), dual effective therapy (DET), or more than DET (MET). Acute Physiology and Chronic Health Evaluation II (APACHE II)-based predicted mortality, and Pneumonia Severity Index scores were calculated.ResultsOf the 225 patients identified, 99 were classified as receiving SET, 102 as receiving DET, and 24 as receiving MET. Compared with the other groups, patients who received MET had statistically significantly more severe pneumonia as measured by the Pneumonia Severity Index score (P =.04) and predicted mortality (P =.03). Mortality within the SET group was significantly higher than within the DET group (P =.02, odds ratio, 3.0 [95% confidence intervals, 1.2-7.6]), even when the DET and MET groups (P =.04) were combined. In a logistic regression model including antibiotic therapy and clinical risk factors for mortality, SET remained an independent predictor of mortality with a predicted mortality-adjusted odds ratio for death of 6.4 (95% confidence intervals, 1.9-21.7). All deaths occurred in patients with a Pneumonia Severity Index score higher than 90, and the predicted mortality-adjusted odds ratio for death with SET in this subgroup was 5.5 (95% confidence intervals, 1.7-17.5).ConclusionsWe found that SET is associated with a significantly greater risk of death than DET. Therefore, monotherapy may be suboptimal for patients with severe bacteremic pneumococcal pneumonia who have Pneumonia Severity Index scores higher than 90.

26 keywords...

hide…