Predicting Pediatric Deterioration Earlier:
CONCERN 3.0 Brings Early Warning to the Frontlines
The CONCERN 3.0 study (Optimization of Pediatric CONCERN EWS to non-AMC settings) will result in a well-performing predictive Early Warning Score (EWS) for community-based pediatric hospitals and inform future spread to other types of lower-resourced, non-academic settings. Our original CONCERN EWS was evaluated in a randomized clinical trial at 4 adult hospitals and successfully reduced risk of in-patient mortality by 35.6% and sepsis by 7.5%. The success of CONCERN EWS is attributed to the very early identification of subtle signals of
deterioration based on expert nursing surveillance and the resulting action of early ICU transfers.
Early identification and escalation of care is particularly critical at community-based pediatric hospitals that may lack specialist expertise and require a transfer to a tertiary academic medical center. An early transfer is typically the best chance for survival, however, due to the resources required the decision to transfer a patient very early in their deterioration requires confidence in the subtle early clinical indicators of deterioration – an EWS can provide valuable role informing this decision. A well-designed EWS may also help to ensure decisions are equitable.
An EWS that was validated at a high-resourced tertiary care setting needs to be validated separately when applying it to non-academic, community-based, lower-resourced environments due to differences in documentation patterns and models of care, such as constrained resources (eg., limited access to specialists), lower staffing, and inadequate infrastructure (eg., insufficient ICU beds and advanced monitoring systems). CONCERN EWS maintains high predictive performance even with the use of minimal model features; therefore, we believe CONCERN EWS has high potential in modeling the documentation patterns and models of care in low-resourced environments for early prediction of deterioration.
Specific Aims
Aim 1: Optimize the pediatric model of CONCERN EWS for non-academic, community-based inpatient pediatric settings.
Aim 2: Develop, implement in silent mode, and evaluate the optimized pediatric version of CONCERN 3.0 EWS at 2 non-academic, community-based inpatient pediatric hospitals that are part of Children’s Hospital Colorado Health System for significantly improved lead time in predicting increased risk of deterioration and other predictive performance measures compared to the PEWS.
CONCERN 3.0 Team Members
Columbia University
- Sarah Rossetti, PhD, RN
- Rachel Lee, PhD, RN
- Elena Insley, MD
- Brandon Lau
- Temmi Daramola
University of Pennsylvania
University of Colorado