P3.J194: Journey to Reduce Unplanned Pediatric Readmissions: A Seven-Year Multifaceted Approach

Purpose/Objectives: Unplanned Readmission Rates (URR) are a critical focus in delivering high-value care, given their impact on patient outcomes and hospital quality metrics. Prior to implementing any intervention, our hospital's 30-day URR stood at 10.7%, significantly exceeding the national average of 6.5%. This disparity underscored the need for action. In 2017, our hospital joined the Children's Hospitals' Solutions for Patient Safety (SPS), a multicenter collaborative network, and introduced a comprehensive 5-element readmission prevention bundle. To reaffirm our institutional commitment to reducing unplanned readmissions, we implemented a validated web-based fault tree tool and introduced a live dashboard to monitor trends in real time. The primary aim of this study was to decrease 30-day unplanned pediatric readmissions by 5% within 12 months.

Design/Methods: This project was conducted at a 500-bed, tertiary care, non-profit freestanding children’s hospital with ~18,400 annual discharges. From 2017-2023, readmission data was reviewed by a multidisciplinary committee, comprised of physicians, nurses, case managers, social workers, pharmacists, and administrative personnel. Planned readmissions were excluded.

Key interventions included the SPS bundle, fault tree tool to determine readmission preventability, and a statistical dashboard to report data for all-cause and unplanned 30-day readmissions (Figure 1). The main outcome measure was 30-day URR, while process measures included physician awareness of URR and identification of preventable readmissions (Figure 1). Total discharges and case-mix index (CMI) were also tracked. Average length of stay (ALOS) was closely monitored as a balancing measure to ensure a reduction in readmissions did not result in an unintended increase in ALOS.

Data analysis involved tracking 30-day readmissions and calculating percentages by dividing the number of readmissions by the total cases. Excel's average and standard deviation functions were used, and an SPC chart was created, applying standard statistical rules for adjusting the mean. This study was reviewed and considered exempt by our hospital’s IRB committee.

Results: From 2017-2023, there was a decline in 30-day URR (10.7% to 9.5%) (Figure 2). While the SPS bundle did not result in any immediate changes, the mean was lowered twice in this range (Figure 1). After implementation of the fault-tree tool and the live dashboard, no clinically significant trends have yet been identified (Figure 1).

Conclusion/Discussion: Since implementation of the SPS Readmission Bundle (2017), we showed a decline in unplanned readmissions. Our initial timeline extended past 12 months, as it took time to fine-tune interventions within our institution. Although we have not reduced URR by 5%, the implementation of a live dashboard will allow our committee to better quantify and qualify URR in real-time. Coupled with the fault-tree tool, these tools will shed light on other targets for intervention in future studies.