De-escalate now?! When can we stop antibiotics in patients with suspected sepsis?

Reviewed by S. Shaefer Spires, MD, Duke Center for Antimicrobial Stewardship and Infection Prevention, Durham, NC

This question was identified by community hospital antimicrobial stewardship programs (ASP) in the Duke Antimicrobial Stewardship Outreach Network.  Since antibiotics can be life-saving in patients with sepsis, ASPs typically focus on stewardship interventions targeting the de-escalation or optimization of the antibiotics for a particular infection.  Specifically, the question arises, what do we do with patients who present with a sepsis syndrome but end up improving without a clearly identified infection?  Can we stop their antibiotics? 

These authors performed a multicenter, patient-level, randomized controlled trial of an “opt-out” de-escalation stewardship intervention. Conducted in a combination of community hospitals and academic medical centers, the study investigators screened non-ICU patients with suspected sepsis, active antibiotic orders, and negative blood cultures at 48-96 hours. They then applied a 23-item safety checklist to exclude patients with a possible active infection. Then after passing this rigorous checklist, the patient was randomized to have the “opt-out” de-escalation intervention vs routine stewardship management. 

The intervention was performed at the physician level and was a verbal discussion from a team member describing how the patient passed the initial screening and then they were asked about whether they wanted to stop the antibiotics or “opt-out” and continue the antibiotics. If they chose to continue the antibiotics, the clinician was required to give the rationale for continuation of antibiotics and to verbalize a plan for the duration of antibiotics.  

Given the rigorous safety checklist used to exclude any possible active infections, the authors screened over 9,000 patients and only enrolled 767 (8%), leading to a highly selected patient population.  The most common reason that patients were excluded was the patient received antibiotics prior to blood cultures. The second most common reason was for a positive culture from another site, followed by an abnormal chest X-ray.  Clinicians chose to opt-out most frequently due to the treatment of a localized infection (76%), followed by the belief that stopping antibiotics was unsafe (31%).

The investigators were able to stop the antibiotics in about 1 in 5 patients in the intervention group (79% vs 85% in the control group). Notably, there was also a decrease in days of therapy among the patients in the intervention group in whom the antibiotics were continued compared to control, suggesting the ASP discussion moved clinicians to formulate a plan regarding antibiotic duration.

Although this intervention is a resource intensive strategy for a seemingly small benefit, it emphasizes the importance of frequent and direct communication of the ASP team with frontline clinicians. Successful ASP interventions are based on the currency of relationship. Also, ASPs need to understand that shortening duration or optimizing antibiotic therapy is a worthy goal even if clinicians are not amenable to stopping antibiotics. Finally, there is a subset of patients initially suspected to have sepsis where no infection is identified in whom antibiotics can be targeted for safe discontinuation.

Disclosure: The reviewer is a member of the Duke Center for Antimicrobial Stewardship and Infection Prevention which conducted this trial but was not personally involved in the conduct or publication of this study.


We use cookies to help improve your experience