2026 Southeastern Residency Conference

Purpose: Time-sensitive completion of bundled interventions is the focus of the Severe Sepsis and Septic Shock Early Management Bundle (SEP-1) Centers for Medicaid and Medicare Services (CMS) quality measure, which aims to reduce sepsis-related morbidity and mortality. To avoid missed cases of sepsis, clinicians often rely on the Systemic Inflammatory Response Syndrome (SIRS) criteria as the major identifier of sepsis, which has inadvertently raised concerns for potential overdiagnoses. This study evaluated the incidence of patients diagnosed with sepsis in the emergency department (ED) who did not meet SIRS criteria and/or lacked a documented source of infection. Recognizing such cases may highlight opportunities to reduce unnecessary use of broad-spectrum antibiotics and ultimately support more targeted antimicrobial stewardship in sepsis care.
Methods: This study design was submitted to the Ballad Health System Institutional Review Board for approval. This was a retrospective electronic chart review that included adult patients aged 18 years and older who presented to the ED of Bristol Regional Medical Center, Holston Valley Medical Center, or Johnson City Medical Center between December 2025 and March 2026 with a sepsis ICD-10 diagnosis. Using the electronic health record, we identified patients who were reported to have met the SIRS criteria. Among these qualifying candidates, patients were then assessed whether there was a documented source of infection and whether broad-spectrum antibiotics were administered. Data collected included patient age, gender, ethnicity, respiratory rate, heart rate, temperature, white blood cell count, and antibiotics received. Provider documentation was reviewed to verify whether there was a documented source of infection. All data was reviewed and recorded in a manner that ensured complete patient anonymity. Statistical analysis was conducted to compare groups meeting sepsis criteria, antibiotic use, and documented infection source.
Results: Among evaluated patients, 11% of those diagnosed with sepsis in the ED did not meet greater than two SIRS criteria at initial presentation, while 89% met SIRS criteria. Additionally, 3% of the study population lacked a documented or suspected source of infection. Clinician compliance with SIRS-based sepsis diagnosis across facilities ranged from 81% to 97%, exceeding the pre-study hypothesis of 70%. Cefepime and vancomycin were the most commonly utilized antimicrobials, with broad-spectrum antibiotic use occurring in approximately 90% of patients.
Conclusions: Majority of patients diagnosed with sepsis in the ED met SIRS criteria, with compliance rates higher than initially anticipated. However, a subset of patients did not meet diagnostic criteria or lacked an identifiable infectious source, suggesting potential opportunities for improvement with diagnostic accuracy and reducing unnecessary broad-spectrum antibiotic use. Further studies with larger sample sizes and less exclusion criteria are warranted to better assess antimicrobial appropriateness based on patient-specific risk factors and presumed sources of infection.