Use of a sepsis predictor made little difference in time to antibiotic administration for septic patients in the emergency department, based on data from more than 200 patients.
“One of the big problems with sepsis is the lack of current tools for early and accurate diagnoses,” said Daniel Burgin, MD, an internal medicine resident at Louisiana State University, Baton Rouge, in a presentation at the annual meeting of the American College of Chest Physicians.
The EPIC Sepsis Model (ESM) was designed to help facilitate earlier detection of sepsis and speed time to the start of antibiotics, but its effectiveness has not been well studied, Dr. Burgin said.
In Dr. Burgin’s facility, the ESM is mainly driven by systemic inflammatory response syndrome (SIRS) and blood pressure and is calculated every 15 minutes; the system triggers a best-practice advisory if needed, with an alert that sepsis may be suspected.
To assess the impact of ESM on time to antibiotics, Dr. Burgin and colleagues reviewed data from 226 adult patients who presented to a single emergency department between February 2019 and June 2019. All patients presented with at least two criteria for SIRS. An ESM threshold of 6 was designed to trigger a set of orders to guide providers on a treatment plan that included antibiotics.
The researchers compared times to the ordering and the administration of antibiotics for patients with ESM scores of 6 or higher vs. less than 6 within 6 hours of triage in the ED. A total of 109 patients (48.2%) received antibiotics in the ED. Of these, 71 (74.5%) had ESM less than 6 and 38 (40.6%) had ESM of 6 or higher. The times from triage to antibiotics ordered and administered was significantly less in patients with ESM of 6 or higher (90.5 minutes vs. 131.5 minutes; 136 minutes vs. 186 minutes, respectively; P = .011 for both).
A total of 188 patients were evaluated for infection, and 86 met Sepsis-2 criteria based on physician chart review. These patients were significantly more likely than those not meeting the Sepsis-2 criteria to receive antibiotics in the ED (76.7% vs. 22.8%; P <.001).
Another 21 patients met criteria for Sepsis-3 based on a physician panel. Although all 21 received antibiotics, 5 did not receive them within 6 hours of triage in the ED, Dr. Burgin said. The median times to ordering and administration of antibiotics for Sepsis-3 patients with an ESM of 6 or higher were –5 and 38.5 (interquartile range), respectively.
“We hope that the ESM would prompt providers to start the order [for antibiotics],” Dr. Burgin said in his presentation. However, the researchers found no consistent patterns, and in many cases the ESM alerts occurred after the orders had been initiated, he noted.
The study findings were limited by the use of data from a single center; the implementation of the EPIC tool is hospital specific, said Dr. Burgin. However, the results suggest that he said.
“While this research proved useful in assessing the impact of ESM on time to antibiotics, more research is needed to understand how to operationalize predictive analytics,” Dr. Burgin said of the study findings. “The goal is to find the balance between early identification of sepsis and timely antimicrobial therapy and the potential harm of overalerting treatment teams.”
The study was supported in part by Cytovale, a sepsis diagnostics company. Several coauthors disclosed financial relationships with Cytovale. Dr. Burgin reports no relevant financial relationships.
A version of this article first appeared on Medscape.com.