Longitudinal Evaluation of a Pediatric Rapid Response System with Realist Evaluation Framework.

Background: Rapid response (RR) systems' impact on clinical outcomes is influenced by institutional social factors. This study applied the realist evaluation (context-mechanism-outcomes) framework to review significant RRs defined as REACT (Rapid Escalation After Critical Transfer) events for appraising a pediatric RR system.

Methods: REACT events included all RRs with cardiopulmonary arrest (CPA) and/or ventilation and/or hemodynamic support instituted within 24 hours after RR. A continuous quality improvement process was employed to identify, debrief, and review REACT events to recognize and act on RR mechanistic and contextual deficiencies. The aim was to decrease REACT events with mechanistic/contextual gaps categorized into crisis resource management (CRM) themes by 25% over three years while ensuring process sustainability.

Results: From 2015 to 2019, 5,581 RR events occurred, of which 67.2% transferred to ICU, and 1,392 (24.9%) were identified as REACTs. In the first two years, 100% identification and review within three months of 90% REACTs was accomplished. One hundred percent of the 17 providers ascertained that the process was safe and transparent, and 80.0% of respondents expressed their commitment from perceived benefit to patient care. Over five years, the proportion of REACTs with CRM gaps decreased from 62.3% to 26.5%, those with multiple deficiencies reduced from 72.5% to 23.2%, and CPAs outside ICUs decreased from 15 to 3 per year. Improvement actions included modifications to RR system (activation, process, and management), hospital (resources and policies), dedicated RR training, and sharing of positive feedback.

Conclusion: The realist evaluation framework facilitated holistic assessment of an RR system. Review of REACTs was feasible, sustainable, and yielded useful information to guide systemwide improvement.