Open Access

ARTICLE

Predicting unplanned readmissions to a pediatric cardiac intensive care unit using predischarge Pediatric Early Warning Scores

by

1 Department of Pediatrics, Section of Critical Care Medicine, Monroe Carell Jr Children’s Hospital at Vanderbilt, School of Medicine, Vanderbilt University, Nashville, Tennessee, USA
2 School of Medicine, Vanderbilt University, Nashville, Tennessee, USA

* Corresponding Author: Ashley Kroeger MD, Department of Pediatrics, Section of Critical Care Medicine, Monroe Carell Jr Children’s Hospital at Vanderbilt, School of Medicine, Vanderbilt University, 2200 Children’s Way, 5121 DOT, Nashville, TN 37232-9075. Email:

Congenital Heart Disease 2018, 13(1), 98-104. https://doi.org/10.1111/chd.12525

Abstract

Objective: Unplanned readmission to the pediatric cardiac intensive care unit (CICU) is associated with significant morbidity and mortality. The Pediatric Early Warning Score (PEWS) predicts ward patients at risk for decompensation but has not been previously reported to identify at-risk patients with cardiac disease prior to ward transfer. This study aimed to determine whether PEWS prior to transfer may serve as a predictor of unplanned readmission to the CICU.
Design: All patients discharged from a tertiary children’s hospital CICU from September 2012 through August 2015 were included for analysis. PEWS assessment was performed following transfer to the cardiac ward, and starting in January 2014, PEWS scores were also assigned by bedside CICU nurse prior to transfer from the CICU. Scores exceeding a predetermined threshold prompted further stability assessment by provider team prior to transfer.
Results: Among 1320 discharges of 1082 patients during the study period, there were 130 unplanned readmissions during their hospitalization. Following implementation of pretransfer PEWS scoring, there was no significant reduction in unplanned readmission frequency (10.2% vs 9.2%, P = .39). A secondary analysis of PEWS scores revealed cardiac scoring as a strong discriminator of those likely to experience an unplanned readmission, independent of other significant clinical predictors of readmission (OR 1.78, 95% CI 1.17–2.71, P = .007). The resultant multivariate model was a good predictor of unplanned readmission (AUC 0.77, 95% CI 0.71-0.83, P < .001).
Conclusion: While implementation of a pretransfer PEWS assessment did not reduce the frequency of unplanned readmissions in this small single-center cohort, a multivariate model including pretransfer elements of an early warning scoring system, along with other patient characteristics serves as a good discriminator of patients likely to experience an unplanned readmission following CICU discharge. Further prospective investigation is needed to define objective measures of pretransfer discharge readiness to potentially reduce the likelihood of unplanned readmissions.

---

Keywords

---