Introduction: The management of sepsis includes the prompt administration of intravenous antibiotics.There is concern that sepsis treatment protocols may be inaccurate in identifying true sepsis andexposing patients to potentially harmful antibiotics, sometimes unnecessarily. This study was designed toinvestigate those concerns by focusing on in-hospital Clostridium difficile infection (CDI), which is a knowncomplication of exposure to antibiotics.

Methods: Our emergency department (ED) recently implemented a protocol to help combat sepsis andincrease compliance with the 2017 Sepsis CMS Core Measures (SEP-1) guidelines. In this single-center,retrospective cohort analysis we queried the electronic health record to gather data on nosocomial CDIand antibiotics prescribed over a five-year period to analyze the effect of the introduction of a sepsisprotocol order set. The primary goal of this study was to measure the hospital-wide CDI rate for threeyears prior to implementation of the sepsis bundle, and then compare this to the hospital-wide CDI rate twoyears post-implementation. As a secondary outcome, we compared the number of antibiotics prescribed inthe ED 12 months prior to administration of the sepsis protocol vs 12 months post-initiation.

Results: Over the course of five years, the hospital averaged 9.4 nosocomial CDIs per 10,000 patienthours. Prior to implementation of the sepsis bundle, the average CDI rate was 11.6 (±1.11, 95%) and afterimplementation the average rate dropped to 6.2 (±1.27, 95%, p<0.01). The mean number of antibioticsordered per patient visit was 0.33 (±0.015, 95%) prior to bundle activation, and, following sepsis bundleactivation, the rate was 0.38 (±0.019, 95%, p<0.01). This accounted for 38% of all ED patient visitsreceiving antibiotics, a 5% increase after the sepsis bundle was introduced.

Conclusion: In this study, we found that CDI infections declined after implementation of a sepsisbundle. There was, however an increase in the number of patients being exposed to antibiotics afterthis hospital policy change. There are more risks than just CDI with antibiotic exposure, and thesewere not measured in this study. Subsequent studies should focus on the ongoing effects of timed,protocolized care and the associated risks.

Introduction: Sepsis order sets improve compliance with the established guidelines, but clinicians must be careful to initiate these protocols on appropriate patients. Many conditions can mimic sepsis as defined by SEP-1 (two or more SIRS* criteria and a suspected infection) such as trauma, COPD, etc. SEP-1 criteria alone can lead to initiating a sepsis protocol without true infection based solely on vital signs.

Objective: To assess the incidence of patients who had a sepsis order set, but an infection was not discovered during their hospital course.

Methods: This study is a single-center retrospective chart review of all “SIRS positive” patients >21 years old who presented to a busy community ED who had the sepsis order set initiated from the emergency department in 2017. A total of 1577 encounters met inclusion criteria. The discharge diagnoses were reviewed to identify unique diagnoses. Similar diagnoses (e.g. RLQ abdominal pain and abdominal pain) were grouped together into the more generalized diagnosis. Several of the unique discharge diagnoses (161) were vague and required individual chart review by two people.

Results: Two hundred fifty-one unique discharge diagnoses were identified and then categorized as infectious or not. Conditions which may be inflammatory versus infectious (e.g. diverticulitis), but are classically treated with antibiotics were counted as infectious. One hundred sixty-one charts were reviewed by two physicians, of which, 130 (81%) were identified as having an infectious condition (K = 0.87). The most common sepsis mimic was abdominal pain, followed by COPD, and cough. A third (33.6%) did not have an infection identified.

Conclusion: SEP-1 criteria for diagnosis and treating sepsis are not specific, with one-third false positives. Identification criteria with higher specificity is needed, and may reduce healthcare expense.

*SIRS (Systemic Inflammatory Response Syndrome) is defined as temperature > 38C° or < 36C°, heart rate > 90 beats per minute, respiratory rate > 20 or PaCO2 < 32 mmHg, and WBC > 12k or < 4k/mm3.