Sepsis screening… does addressing one problem lead to another?

Abstract

In the wake of widely publicized events where children died from sepsis misidentified as another condition, numerous states have enacted laws requiring hospitals to implement evidence-based processes to quickly recognize sepsis and intervene where appropriate (https://www.cdc.gov/hai/pdfs/sepsis/VS-Sepsis-Policy-FINAL.pdf). These include Gabby's law in Illinois and Rory's Regulations in New York. Although these initiatives have sparked lively conversation about the proper role for legislation in clinical care, particularly when prompted by single cases, the mandates have demonstrated success (JAMA 2018;320:358-67). The problem of appropriately recognizing sepsis is real, particularly in emergency departments not accustomed to caring for children. However, one argument frequently advanced is that such mandates may lead to unnecessary and inappropriate interventions in children who do not need them. The study by Baker et al in this volume of The Journal sought to answer that question. This single center study implemented a sepsis screening algorithm in the electronic health record (EHR) (Figure). As is common when implementing electronic alerts, the screen was first silent to providers to ensure that it functioned appropriately in the EHR. Subsequently the alert was activated for use in real time. This provided a unique opportunity to compare 2 time epochs, where the screening alert identified patients in an identical fashion, but during 1 period was functioning silently in the background and then in a later period was visible to the clinical team. In this study the investigators focused on false positive screens. Sepsis screens are well known to have a high false positive rate, as a number of conditions may produce the constellation of clinical findings for which sepsis should be in the differential diagnosis. Their intent is to prompt a careful patient evaluation, not a prescribed intervention, to enhance clinical acumen, not substitute for clinical judgment. In comparing the 2 periods they found that there was no difference in receipt of antibiotics or fluid boluses for patients who initially screened positive but did not have sepsis. Clinicians did not react to a positive screen by providing unnecessary interventions. This is reassuring news, and refutes the contention of skeptics that sepsis screens lead to unwarranted and costly care. However, it does not completely eliminate the concern. This study reflects practice in a quaternary care pediatric hospital with broad experience in both pediatric sepsis and in look-alike presentations. A parallel study in non-pediatric settings will help refine these findings and could further enhance pediatric sepsis care. Article page 193 ▸ Effect of a Sepsis Screening Algorithm on Care of Children with False-Positive Sepsis AlertsThe Journal of PediatricsVol. 231PreviewTo determine if implementation of an automated sepsis screening algorithm with low positive predictive value led to inappropriate resource utilization in emergency department (ED) patients as evidenced by an increased proportion of children with false-positive sepsis screens receiving intravenous (IV) antibiotics. Full-Text PDF