Implementation of an Automated Antibiotic Time-out at a Comprehensive Cancer Center

Abstract

Background

Antimicrobial stewardship programs can optimize antimicrobial use and have been federally mandated in all hospitals. However, best stewardship practices in immunocompromised patients with cancer are not well established.

Methods

An antimicrobial time out, in the form of an email, was sent to physicians caring for hospitalized patients reaching 5 days of therapy for targeted antimicrobials (daptomycin, linezolid, tigecycline, vancomycin, imipenem/cilastatin, meropenem) in a comprehensive cancer center. Physicians were to discontinue the antimicrobial if unnecessary or document a rationale for continuation. This is a quasi-experimental, interrupted time series analysis assessing antimicrobial use during the following times: period 1 (before time-out: January 2007-June 2010) and period 2 (after time-out: July 2010-March/2015). The primary antimicrobial consumption metric was mean duration of therapy. Days of therapy per 1000 patient-days were also assessed.

Results

Implementation of the time-out was associated with a significant decrease in mean duration of therapy for the following antimicrobials; daptomycin: −0.89 days (95% confidence interval [CI], −1.38 to −.41); linezolid: −0.89 days (95% CI, −1.27 to −.52); meropenem: −0.97 days (95% CI, −1.39 to −.56); tigecycline: −1.41 days (95% CI, −2.19 to −.63); P < .001 for each comparison. Days of therapy/1000 patient-days decreased significantly for meropenem (−43.49; 95% CI, −58.61 to −28.37; P < .001), tigecycline (−35.47; 95% CI, −44.94 to −26.00; P < .001), and daptomycin (−9.47; 95% CI, −15.25 to −3.68; P = .002).

Discussion

A passive day 5 time-out was associated with reduction in targeted antibiotic use in a cancer center and could potentially be successfully adopted to several settings and electronic health records.