At Olive View-UCLA Medical Center, the St John Sepsis Agent (Cerner Corporation, Kansas City, MO) was implemented for early recognition of sepsis via alert and notification to clinicians. This alert is based on an algorithm of parameters, including vital signs and laboratory values to determine patients at risk for sepsis. The purpose of this study is to evaluate the impact of the St John Sepsis alert on physician clinical decision-making in terms of the following: initiating fluid resuscitation, starting or changing current antibiotics, and time of starting antibiotics from alert notification.
Single-center, retrospective review of adult inpatients with a documented St John Sepsis notification from May 2016 to June 2016. Data collection included the following: discharge diagnoses pertaining to infection, current antibiotic orders, time when alert was notified to the physician, fluid bolus and antibiotic administration, and culture sensitivities. Appropriateness of antibiotics were reviewed according to input from Infectious Diseases Specialists and Pharmacists.
100 patients met study inclusion criteria. 14% had subsequent initiation of antibiotics or a change in current antibiotic regimen. Among these patients, 21% received antibiotics within one hour from the time of sepsis alert notification. 12 out of 14 patients had an antibiotic change that was deemed to be unwarranted, as these patients had met SIRS criteria for non-infectious diagnoses, or the change was not necessary given the patient’s diagnosis (e.g. antibiotics were broadened unnecessarily). 16% of all study patients received fluid bolus therapy within three hours of alert notification.
The implementation of the alert does not appear to correlate with timely initiation of appropriate antibiotics or positive changes in antibiotic management. The majority of alerts (98%) were not associated with evidence of true infection, with meaningful changes in antibiotics, or with appropriate broadening of antibiotics. Although the purpose of the St John Sepsis is to aid clinicians in early recognition of patients with sepsis, this study demonstrates that in the vast majority of cases, it was not beneficial and may lead to unnecessary administration or broadening of antibiotics.
A. Jeng, None