Long-stay patients in academic children's hospital

• Objective: To determine the characteristics and resource utilization associated with long-stay patients in academic children's hospitals. • Study design and methods: Retrospective cohort study using the Pediatric Health Information System database, which includes data from 34 noncompeting children's hospitals in the United States. We defined long-stay patients as those with a length of stay (LOS) 2 standard deviations above the mean. For each discharge, demographic, clinical, and resource characteristics were identified. Chi-square tests, t tests, and logistic regression models were used to test the relationships between these characteristics and LOS for all long-stay patients and separately for neonatal (age ≤ 30 days) and pediatric long-stay patients. • Results: Long-stay patients comprised 2.4% of the population but represented 27.9% of hospital days and 30.2% of adjusted charges. A total of 81.9% of long-stay patients were discharged home. Death occurred in 6.7% of neonatal and 8.4% of pediatric long-stay patients. Mean adjusted charges between age-groups did not differ. Neonatal long-stay patients had an actual LOS and adjusted charges 2 times greater than expected. For pediatric long-stay patients, these variables were 4 times greater than expected. Female sex, black race, age < 30 days, increased severity of illness, complication or infection during hospitalization, and the need for additional care upon discharge increased the likelihood of prolonged LOS. • Conclusions: Long-stay patients in academic children's hospitals are rare but account for a significant proportion of hospital days and charges. Pediatric long-stay patients are responsible for a disproportionate share of increased charges. Improved case management of prolonged hospitalizations may reduce charges and improve quality of care.