Effectiveness of and obstacles to antibiotic streamlining to amoxicillin monotherapy in bacteremic pneumococcal pneumonia

Abstract Background Antibiotic streamlining is pivotal to reduce the emergence of resistant bacteria. However, whether streamlining is frequently performed and safe in difficult situations, such as bacteremic pneumococcal pneumonia (BPP), has still to be assessed. Methods All adult patients admitted to Dijon Hospital (France) from 2005 to 2013 who had BPP without complications, and were alive on the third day were enrolled. Clinical, biological, radiological, microbiological and therapeutic data were recorded. A first analysis was conducted to assess factors associated with being on amoxicillin on the third day. A second analysis, adjusting for a propensity score, was performed to determine whether 30-day mortality was associated with streamlining to amoxicillin monotherapy. Results Of the 196 patients hospitalized for BPP, 161 were still alive on the third day and were included in the study. Treatment was streamlined to amoxicillin in 60 patients (37%). Factors associated with not streamlining were severe pneumonia (OR 3.11, 95%CI [1.23–7.87]) and a first-line antibiotic combination (OR 3.08, 95%CI [1.34–7.09]). By contrast, starting with amoxicillin monotherapy correlated inversely with the risk of subsequent treatment with antibiotics other than amoxicillin (OR 0.06, 95%CI [0.01–0.30]). The Cox model adjusted for the propensity-score analysis showed that streamlining to amoxicillin during BPP was not significantly associated with a higher risk of 30-day mortality (HR 0.38, 95%CI [0.08–1.87]). Conclusions Streamlining to amoxicillin is insufficiently implemented during BPP. This strategy is safe and potentially associated with ecological and economic benefits; therefore, it should be further encouraged, particularly when antibiotic combinations are started for severe pneumonia.