Death and Emergency Readmission of Infants Discharged After Interventions for Congenital Heart Disease: A National Study of 7643 Infants to Inform Service Improvement

Abstract

Background: Improvements in hospital-based care have reduced early mortality in congenital heart disease. Later adverse outcomes may be reducible by focusing on care at or after discharge. We aimed to identify risk factors for such events within 1 year of discharge after intervention in infancy and, separately, to identify subgroups that might benefit from different forms of intervention.

Methods and results: Cardiac procedures performed in infants between 2005 and 2010 in England and Wales from the UK National Congenital Heart Disease Audit were linked to intensive care records. Among 7976 infants, 333 (4.2%) died before discharge. Of 7643 infants discharged alive, 246 (3.2%) died outside the hospital or after an unplanned readmission to intensive care (risk factors were age, weight-for-age, cardiac procedure, cardiac diagnosis, congenital anomaly, preprocedural clinical deterioration, prematurity, ethnicity, and duration of initial admission; c-statistic 0.78 [0.75-0.82]). Of the 7643, 514 (6.7%) died outside the hospital or had an unplanned intensive care readmission (same risk factors but with neurodevelopmental condition and acquired cardiac diagnosis and without preprocedural deterioration; c-statistic 0.78 [0.75-0.80]). Classification and regression tree analysis were used to identify 6 subgroups stratified by the level (3-24%) and nature of risk for death outside the hospital or unplanned intensive care readmission based on neurodevelopmental condition, cardiac diagnosis, congenital anomaly, and duration of initial admission. An additional 115 patients died after planned intensive care admission (typically following elective surgery).

Conclusions: Adverse outcomes in the year after discharge are of similar magnitude to in-hospital mortality, warrant service improvements, and are not confined to diagnostic groups currently targeted with enhanced monitoring.

Keywords: congenital heart defects; health policy and outcomes research; pediatrics; risk model; risk stratification.

Figure 1

Classification and regression tree (CART) analysis. The stratification tree generated by the CART analysis and evaluated across the entire data set (see Table 5 for a breakdown by development and test set). The number and rate of adverse events and the total number of patients are given for each node. HLHS indicates hypoplastic left heart syndrome; PA, pulmonary atresia; UVH, functionally univentricular heart; VSD, ventricular septal defect.