300 - Pediatric sepsis screening and clinical decision support: successful deployment of pediatric tools for use across a large statewide health wide system

Poster Number: 300
Publication Number: 300.347

Jessica Kanis, IU Health, Indianapolis, IN, United States; Emily Webber, Indiana University School of Medicine, Indianapolis, IN, United States; Jason Schaffer, Indiana University School of Medicine, Indianapolis, IN, United States

Background: Pediatric sepsis is a significant cause of morbidity and mortality, requiring prompt antibiotic and fluid resuscitation. Delayed recognition is common due to the non-specific signs and symptoms of pediatric sepsis. Electronic health records can provide continuous background screening for earlier detection with additional clinical decision support (CDS). Several pediatric centers have shown improved treatment with embedded sepsis screening algorithms; however, limited data exists regarding implementation in GEDs where the majority (~90%) of pediatric patients present for care nationwide.

Objective: To effectively implement a pediatric sepsis screening tool with embedded CDS in a quaternary care PED with subsequent expansion to 15 additional GEDs.

Design/Methods: A system-wide initiative developed and implemented adult and pediatric specific sepsis screening algorithms that triggered secondary screening forms and embedded decision support and treatment plans. A cloud based, age-specific algorithm monitoring vital signs and indicators of end organ function runs in the background. Patients alerted as SIRS/sepsis then had a secondary screen further altering the algorithm. Following pilot implementation, adult and pediatric sepsis screening tools were implemented simultaneously at (15) additional GEDs throughout the health system. The number of positive alerts, secondary screens, and response to alerts were monitored, in addition to, powerplan usage and time to intervention. Sensitivity and specificity of the tool were also determined by region.

Results: A total of 2845 alerts fired; majority (90%) SIRS. Secondary sepsis screens were performed in 84.2% of alerts and the median time from alert to screen varied from 3-62 min (median 22 min). Screens were performed 88% of the alerts fired at the children’s hospital with a median time to screen of 27 min. Positive screens were found in 8.1% of all alerts. Powerplan usage increased from 34.3% to 60.8% and antibiotics for severe sepsis were given within an hour in 46.2% of patients. Specificity ranged from 94-96% with sensitivity for 49% (PED) compared to 95% at all other sites.Conclusion(s): Implementation of a pediatric sepsis screening tool in GEDs was feasible and performed similarly among sites. The majority of pediatric alerts are SIRS necessitating a secondary screen to further identify sepsis and given the large volume of pediatric patients in the PED affects the sensitivity of the tool. Similar screening rates and time to screen were performed in the PED when comparted to GEDs. Further studies will evaluate time to antibiotics and mortality outcomes.