imEqhOSGE VjRpBu

Sepsis: the case for automatic antibiotic de-escalation

Sepsis is a life-threatening condition caused by an extreme immune response to infection which has often invaded the blood. In severe cases, this can lead to mortality in up to 50% of patients. Luckily, the treatment for this condition is straightforward: antibiotics. But identifying sepsis early enough for effective treatment can be complex. Some of the same hallmarks of sepsis can also be caused by various other illnesses including low blood pressure, lactic acidosis, and kidney or liver failure. This often leads to overprescription of antibiotics which can have serious consequences on patients (GI upset, C. Diff, medication side effects) and the community at large (increased antimicrobial resistance). But the risks of missing a sepsis diagnosis are profound, often leading to the empiric use of antibiotics in patients suspected of sepsis anyways. So, how do we balance antibiotic stewardship with ensuring a serious case of sepsis doesn’t go untreated?

This is exactly the question explored in a recent article in Clinical Infectious Diseases. UNMC ID’s Dr. Erica Stohs (pictured left) recently outlined this article as a SHEA journal club article, which we repost below. Read on to learn about a new strategy for antibiotic de-escalation in potential sepsis patients.


Reviewed by Erica Stohs, MD, MPH, University of Nebraska Medical Center

This multicentered randomized controlled trial evaluated an opt-out protocol to decrease unnecessary antibiotic use prescribed for non-ICU hospitalized patients with suspected sepsis.

Stewarding antibiotics has been challenging in the era of SEP-1 (sepsis management bundle regulation required by CMS), which did not incorporate the balancing measure of antibiotic overuse and its consequences. This RCT studied an opt-out intervention in which clinicians had to actively engage to continue antibiotics for carefully selected patients in whom broad-spectrum agents were initiated due to suspected sepsis. Suspected sepsis was defined as having no positive blood cultures at 48-96 hours and active orders for broad-spectrum antibiotics.

The study was directed at non-critically ill adults on broad-spectrum agents with negative blood cultures in ten hospitals between September 2018 to May 2020. Through detailed expert panel review and rigorous protocol development, investigators developed a 23-item safety checklist to determine if patients were eligible for the de-escalation opt-out protocol. For example, patients with ongoing signs and symptoms of infection (fever, leukocytosis, pneumonia, or related complications), concerning or incomplete microbiologic data, antibiotic pre-treatment, and immunocompromised status were excluded. If clinicians opted out (i.e., continued antibiotics), they discussed their rationale and future plan to de-escalate. In essence, this involved thorough audit and feedback. The primary outcome was post-enrollment days of therapy (DOT) up to 30 days. Secondary outcomes: 30-day safety events, including C diff infection, readmission, ICU admission, death, and relapse of suspected sepsis.

Of the nearly 10 thousand patients screened, only 767 (8%) were enrolled. Intervened patients had 32% lower odds of antibiotic continuation and were exposed to fewer days of extended-spectrum antibiotics (36% vs 44%). For patients in whom antibiotics were continued, DOT was similar. No safety issues were noted. Interestingly, despite a quite conservative safety checklist limiting eligibility, safety concerns were commonly cited by clinicians as reason to continue antibiotics. Addressing diagnostic uncertainty remains a challenge to expansion of an opt-out approach to antibiotic de-escalation in suspected sepsis.

Reference: Moehring RW, et al. Clinical Infectious Diseases 2023;76(3):433–442. DOI: 10.1093/cid/ciac787

Leave a comment

Your email address will not be published. Required fields are marked *

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

This site uses Akismet to reduce spam. Learn how your comment data is processed.