Reducing Failure to Rescue Events with a Synergy Driven Nursing Bundle

Abstract

Problem: A four-month chart review of inpatients seen by the Rapid Response Team within an academic medical center detected up to 20% of patients admitted to the ICU were admitted for less than 48 hours. Current processes to identify patients at risk for decompensation do not incorporate increased nursing workload seen with COVID-19. Purpose: Primary goals were 10% reduction of RRT activations with ICU transfer from a 21-bed med/surg observation floor for patients admitted <48 hours. Out of 273 inpatient calls assessed, 6 patients required RRT from the floor. Secondary aims included increased communication and improved working relationship between the floor and RRT staff with an acuity tool incorporating nursing assessment. Methods: Patients admitted <48 hours were assessed with the AACN Synergy Model Patient Acuity Tool (ADT-SMAT) every 12 hours. Scores were entered into REDCap software by the floor charge nurse. Patients scoring >11 points were "high-risk," and communicated to RRT with secure messaging to follow the patients remotely. Any further interventions or needs communicated by the floor charge were implemented to prevent a failure to rescue event. Patients transferred to the ICU with documentation of high-risk ADT-SMAT scores were tracked through the hospital's electronic health record. Results: A total of 340 patients were scored using the ADT-SMAT in the 15-week implementation; 6 required RRT activation. Four were high-risk and transferred to ICU. This is a 33% reduction in unplanned ICU transfers from the pre-implementation period. Communication between the implementation floor and RRT increased to identify these at-risk patients. Conclusion: The ADT-SMAT tool is successful at identifying high-risk patients on the floor at risk for clinical decompensation. This is a pilot study to determine if the tool is beneficial to the relationship between bedside staff and RRT, further testing on additional floors for sustainability is recommended.