OBJECTIVESpinal fusion is used in the treatment of pediatric neuromuscular scoliosis (NMS) to improve spine alignment and delay disease progression. However, patients with NMS are often medically complex and require a higher level of care than those with other types of scoliosis, leading to higher treatment costs. The purpose of this study was to 1) characterize the cost of pediatric NMS fusion in the US and 2) determine hospital characteristics associated with changes in overall cost.METHODSPatients were identified from the National Inpatient Sample (2012 to the first 3 quarters of 2015). Inclusion criteria selected for patients with NMS, spinal fusion of at least 4 vertebral levels, and elective hospitalization. Patients with no cost information were excluded. Sociodemographics, treating hospital characteristics, disease etiology/severity, comorbidities, length of stay, and hospital costs were collected. Univariable analysis and multivariable gamma log-link regression were used to determine hospital characteristics associated with changes in cost.RESULTSA total of 1780 weighted patients met inclusion criteria. The median cost was $68,815. Following multivariable regression, both small (+$11,580, p < 0.001) and medium (+$6329, p < 0.001) hospitals had higher costs than large hospitals. Rural hospitals had higher costs than urban teaching hospitals (+$32,438, p < 0.001). Nonprofit hospitals were more expensive than both government (–$4518, p = 0.030) and investor-owned (–$10,240, p = 0.001) hospitals. There was significant variability by US census division; compared with the South Atlantic, all other divisions except for the Middle Atlantic had significantly higher costs, most notably the West North Central (+$15,203, p < 0.001) and the Pacific (+$22,235, p < 0.001). Hospital fusion volume was not associated with total cost.CONCLUSIONSA number of hospital factors were associated with changes in fusion cost. Larger hospitals may be able to achieve decreased costs due to economies of scale. Regional differences could reflect uncontrolled-for variability in underlying patient populations or systems-level and policy differences. Overall, this analysis identified multiple systemic patterns that could be targets of further cost-related interventions.
