R. M. Dorman1,2, H. Naseem1, K. D. Bass1,2, D. H. Rothstein1,2 1Women And Children’s Hospital Of Buffalo,Department Of Surgery,Buffalo, NY, USA 2State University Of New York At Buffalo,Department Of Surgery,Buffalo, NY, USA
Introduction: Hospital readmission after discharge for trauma care in adults confers significant morbidity, mortality, and resource utilization. Less is known about hospital readmission in pediatric patients after an index admission for trauma care. In this study, we examine pediatric intensive care unit (PICU) admission as a risk factor for hospital readmission after trauma care.
Methods: This is a retrospective cohort study of patients aged 1-19 years discharged with a trauma diagnosis from hospitals in the Pediatric Health Information System database between March, 2010, and February, 2015. Patients with inadequate clinical information, those transferred to the PICU after initial ward admission, and those who died after admission were excluded. Demographic patient variables included age, gender, payer status, and race/ethnicity. Clinical variables included length of stay, presence of mechanical ventilation, APR-DRG severity of illness (SOI), and disposition upon discharge. The main outcome variable was hospital readmission within 30 days of discharge. Odds ratios (OR) were calculated in both univariate and multivariate analyses with corresponding 95% confidence intervals (C.I.).
Results: During the study period, 87,401 patients were admitted with a trauma diagnosis. Of these, 14,770 (16.9%) were admitted directly to the PICU. The overall population was 65.3% male and 62.4% white, and had an average age of 9.26 (SD 4.83) years. The most common payers were private (43.9%) and Medicaid (43.4%). Nearly half of the patients had a low SOI (49.3%). Most were discharged without home health services (96.1%). Hospital readmissions within 30 days occurred in 3.4% of patients. On univariate analysis, patients directly admitted to the PICU had more than twice the risk of 30-day hospital readmission compared to those never admitted to the PICU, 6.4% vs 2.8% (OR 2.36, C.I. 2.18-2.56). On multivariate analysis, controlling for demographic and clinical variables (excluding SOI), the OR for hospital readmission in patients initially admitted to the PICU was 1.45 (C.I. 1.32-1.59) compared to those never admitted to the PICU. When including SOI, the OR dropped to 1.11 (C.I. 1.002-1.228).
Conclusion: Direct admission to the PICU during an index hospitalization for trauma care is an independent risk factor for hospital readmission within 30 days of discharge. The majority of this effect is likely related to the patient’s severity of illness. Further risk stratification may help appropriately focus resources on high risk patients in order to improve clinical outcomes and reduce unnecessary hospital readmissions.