Variability in spinal surgery outcomes among children's hospitals in the United States.

BACKGROUND: We performed a retrospective cohort study of 7637 spinal fusion surgical cases from 2004 to 2006 at 38 children's hospitals participating in the Pediatric Health Information System database to evaluate the variability of in-hospital outcomes by patient factors and between facilities in children who underwent spinal surgery.
METHODS: Outcomes were stratified by whether children did or did not have neurological impairment. Multilevel multivariate logistic regression models were used to determine patient and hospital factors associated with in-hospital infections, surgical complications, and length of stay (LOS)≥10 days.
RESULTS: Neurologically impaired (NI) children (N=2117 out of 7637) represented 28% of the cases. The interhospital interquartile range of LOS for NI children was 6 to 8 days (median 7 d) and for non-neurologically impaired (NNI) children was 5 to 6 days (median 5 d). Children with NI had roughly 6 times higher rates of in-hospital infection and 3 times higher complication rates: major interhospital variation was seen for both of these outcomes. Hospital rates of infection ranged from 0% to 27% (median 10%) for NI and from 0% to 14% (median 2%) for NNI children. Complication rates ranged from 0% to 89% (median 33%) for NI and from 3% to 68% (median 9%) for NNI children. The following factors were associated with a LOS≥10 days: in-hospital infection (P<0.0001), surgical complication (P<0.0001), and anterior/posterior versus posterior-only surgery (P<0.0001). Hospital case volume was not associated with infection, surgical complication, or LOS≥10 days.
CONCLUSIONS: Substantial variation exists in reported outcomes for children undergoing spinal surgery in children's hospitals within the United States. Further study is needed to characterize hospital-level factors related to surgical outcome to direct future quality improvement.