Predicting Hospital Readmissions Among NICU Graduates

Background: Correlation between hospital readmission & prematurity has been noted. However, little is known regarding predictive factors for readmission among NICU patients. Methods: Retrospective chart analysis of neonates discharged 2007-2008 from a Level 4 NICU. Patients who died or were transferred were excluded. Infants readmitted 6 months after discharge were identified. Data was recorded on demographics, antenatal, & neonatal issues at discharge & readmission. &Chi² & t-tests were used for analysis. Results: Total readmissions n=97 (13%). This is an interim analysis of 75 patients, 25 readmissions & 2 controls per readmission matched by gestational age & month of birth. Days to readmission were 16±3 (mean±SEM). Subjects had similar birthweights (2400g±200v2400±130). The commonest readmission reasons were apnea & respiratory illness. Issues associated with readmission (p<0.05) were reflux medication use during hospitalization (55%v22%), congenital GI anomalies (100%v27%); longer: to reach full feeds (21d±6v9±3), length of stay (53d±13v27±5), time from nippling all feeds to discharge (15d±4v6±1); higher post-conceptual age at discharge (42wk±2v38±0.3), more outpatient appointments (4±0.4v2±0.3), transferred from another hospital (50%v23%), those requiring surgery (56%v24%) or with a genetic disorder (100%v27%). The following were not associated with readmission: severe intraventricular hemorrhage, chronic lung disease, discharge on oxygen, caffeine, or a monitor. Conclusions: Neonates who take longer to reach full feeds & stay longer in the NICU have a higher risk of readmission. Transfer from another hospital, reflux medication use, surgery, and congenital anomalies increases this risk. These factors may define a high-risk group in need of special observation to help prevent avoidable readmission.