Treatment duration for Escherichia coli bloodstream infection and outcomes: retrospective single-centre study

To investigate the impact of treatment duration on mortality and on relapse in patients with Escherichia coli bloodstream infection (BSI).

Retrospective single-centre study of patients diagnosed with E. coli BSI at our centre over a 4-year period.

age <18 years, clinical data not available, polymicrobial BSI, failure to receive in vitro active therapy, and death while receiving antibiotic therapy. Exposure variable was treatment duration dichotomized into short (≤10 days) and long (>10 days) therapy. Primary end point was all-cause mortality within 90 days after index BSI. Secondary end point was relapse, defined as repeat isolation of E. coli from blood cultures within 90 days after index BSI, in patients with documented clinical cure and completion of therapy for the initial episode.

Of the 856 analysed patients: 426 received short and 430 received long therapy. All-cause mortality at day 90 occurred in 47 patients; on multivariate analysis, short therapy was not associated with a higher risk of mortality, also after adjusting the model for the propensity score of receiving short therapy. Relapse occurred in 42 patients. Independent risk factors for relapse using death as competing risk were immunosuppression (subhazard ratio 4.67, p < 0.001), and end-stage liver disease (subhazard ratio 2.58, p 0.013). The propensity-weighted estimation of the average treatment effect for relapse reduction with long therapy (>10 days) was -1.6% (p 0.26) in the total population, and -7.1% (p 0.18) in immunocompromised patients.

We could not identify shorter treatment duration as a risk factor for mortality and for relapse in patients with E. coli BSI.