Abstract
Background Missing signs of sepsis can result in delayed diagnosis, treatment and complications. We have introduced an ED-based Children At High Risk(CAHR-AT) tool to improve recognition, team critical thinking, and patient outcomes. Delayed recognition and poor nurse-provider communication were identified as common challenges in timely treatment of children at high-risk for infection-related decompensation.
Objectives This stand-alone urban children’s hospital ED aims to improve team-based care, situational awareness, and patient outcomes through team huddles and associated interventions (figure 1).
CAHR key driver diagram CAHR key driver diagram
Methods Development of the CAHR-AT utilized vital signs data of >1 × 10^6 patients to derive standards. Logistic regression and ‘machine-learning(AI)’ identified factors showing the highest association with gold-standard sepsis cases and applied weights to each factor for optimum sensitivity. A nursing assessment form was added to the dyad assessment process and visual redesign of the tool interface went into effect using a stoplight approach with red, yellow, and green lights indicating patient acuity and resources needed (figure 2).
CAHR project timeline and stoplight activation algorithm (Training video link attached) CAHR project timeline and stoplight activation algorithm (Training video link attached)
Results It has been over 239 days (934 alerts) since the last unanswered alert by the provider/nurse dyad (figure 3). The average percent of CAHR patients with a completed initial huddle increased from 9.3% to 45.3% (figure 4). Higher CAHR-AT scores were associated with higher severity-of-index (SOI) and acute kidney injury (AKI) within 48 hrs of arrival (figures 5 and 6). Preliminary data show CAHR-AT patients with a score ≥8 who received the bundle (IV-fluid bolus and IV-antibiotics) significantly shorter length of stays (figure 7).
Number of CAHR alerts between dyad (MD/RN) unanswered alerts Number of CAHR alerts between dyad (MD/RN) unanswered alerts
Compliance of initial pit stop (Huddle) completion (p-chart) Compliance of initial pit stop (Huddle) completion (p-chart)
Association of CAHR-AT score and severity of illness Association of CAHR-AT score and severity of illness
Association of CAHR-AT score with incidence of acute kidney injury (AKI) among admitted patients within 48 hrs of arrival Association of CAHR-AT score with incidence of acute kidney injury (AKI) among admitted patients within 48 hrs of arrival
Differences in length of stay (LOS in days) between CAHR+ patients who received the bundle (IVF and IV antibiotics) vs. those who did not Differences in length of stay (LOS in days) between CAHR+ patients who received the bundle (IVF and IV antibiotics) vs. those who did not
Conclusions CAHR-AT predicts physiologic decompensation and AKI. Its processes promote team-based critical thinking and improve patient outcomes. Next steps include prescriptive order sets for both red/yellow stoplight activations and spread to inpatient units.
