Efficiency in Adult Spinal Deformity Surgery: A Multicenter Comparison of Resource Use

Abstract

Introduction Decision makers and payers are placing increasing scrutiny on the costs and outcomes associated with complex spine surgery. To justify higher cost practice patterns, there must be evidence of additional patient benefit; however, little is currently known regarding the variation in resources used and associated outcomes in complex spine surgeries. Our study examines multicenter variability in patient-level surgical resource use, including implants, biologics, and length of stay, alongside health-related quality-of-life (HRQoL) improvements following adult spinal deformity (ASD) surgery. Material and Methods This is a retrospective analysis of a multicenter prospective database of consecutive patients with ASD. HRQoL outcomes were calculated from the ODI, SRS-22, and SF-36 domain scores. Changes in HRQoL were estimated as the difference between baseline and 2-year values. Patient-level surgical resources included blood use, bone morphogenetic protein (BMP), volume (ccs), length of stay (LOS), and implants. Patients were classified by sagittal modifier and analyzed across centers. The three sagittal modifier types are defined as follows: mild (SVA < 4 cm, PT < 20 degrees, PI–LL within 10 degrees), moderate (at least 1 of SVA 4–9.5 cm, PT 20–30 degrees; PI–LL 10–20 degrees, and no severe modifiers); and severe (at least 1 of SVA > 9.5 cm, PT > 30 degrees, PI–LL > 20 degrees). Statistical analysis was performed using analysis of variance (ANOVA) and multivariable regression methods comparing HRQoL outcomes and resource use across center, considering the overall patient cohort and each modifier group. Results Baseline and 2-year HRQoL data were available for 226 surgical patients with ASD, with an average age of 56 (range, 18–84 years) years who were predominantly female ( N = 189, 84%). Significant differences were found in the average 2-year change in HRQoL across centers, however, this difference becomes insignificant after controlling for patients within the same major modifier groups ( p > 0.05). Significant differences were found across centers in average resources used per surgery ( p < 0.05), with only LOS not reaching significance ( p > 0.05). Significant differences were found in average BMP and screw use across all modifier groups. After accounting for clinical, demographic, and regional characteristics at the patient level, variation among centers persisted in both screw use and total BMP volume with no corresponding statistical differences in HRQoL outcomes. Conclusion The use of additional surgical resources does not appear to impact 2-year HRQoL outcomes following surgery for ASD. To improve efficiency in ASD surgery, standardization of physician practice patterns and resource use could help curb costs without negatively impacting patient HRQoL.