Volume-outcome relationship in halo vest utilization for C2 fractures

BACKGROUND CONTEXT: The prevalence of C2 fractures has increased in recent years. The treatment of these fractures include halo-vest immobilization (HVI), rigid cervical collar, or spinal fusion. There is controversy regarding the management of these fractures with different institutions having their own protocols based on individualized experience. The volume-outcome relationship of HVI use for C2 fractures has not been studied. Evaluation of such relationships are important as they suggest that patients may benefit from referral to and treatment at high-volume institutions. PURPOSE: To evaluate the volume-outcome relationship in HVI use for C2 fractures in New York State. STUDY DESIGN: Retrospective analysis of a statewide database. PATIENT SAMPLE: We queried the New York Statewide Planning and Research Cooperative System database for the International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis code 805.02 (closed fracture of second cervical vertebra) and procedure code 029.4 (insertion or replacement of skull tongs or halo traction device) to identify all patients who received HVI for a fracture of the second cervical vertebra between the years 2001 and 2014. Those who had isolated C2 fractures were selected. OUTCOME MEASURES: Outcomes of interest included resource utilization characteristics (hospitalization charges and length of stay), perioperative complications, comorbidities, 30-day mortality, any readmission, and any future cervical fusion surgery. METHODS: The 2001 to 2014 Statewide Planning and Research Cooperative System database was used to identify patients with C2 fractures who received HVI. Our key independent variable was institution volume modeled as high- (>25 halos/year), medium-, (10-25 halos/year), or low-volume (<10 halos/year) based on the total number of HVI procedures reported by hospitals during the study period. We compared outcomes with respect to hospital volume. We also compared patients by age groups: <40, 40 to 60, 60 to 80, and >80. Multivariate logistic regressions were performed for the binary variables any complication and any readmission while controlling for covariates hospital volume, age, sex, race, insurance status, and Elixhauser comorbidity mean. Statistical significance was set at a value of p<.05 for all analyses. RESULTS: In all, 625 patients with C2 fractures managed with HVI were included. Most patients were male (53%) and Caucasian (76%) with a mean age of 57. Patients at high-volume hospitals were younger (52 vs. 59 and 60 for medium- and low-volume, respectively; p<.01) and had fewer future readmissions (40% vs. 54% and 84% for medium- and low-volume, respectively; p<.01). On multivariable analysis, those with private insurance and worker's compensation had lower likelihood of future readmission compared to Medicaid patients. Patients >80 had higher rates of major in-hospital complications (52% vs. 40%, 18%, and 19% for groups 60-79, 40-59, and <40, respectively; p<.01), mortality (14% vs. 5%, 1%, and 1% for groups 60-79, 40-59, and <40, respectively; p<.01), and readmissions after the initial HVI (62% vs. 50%, 54%, and 37% for groups 60-79, 40-59, and <40, respectively; p<.01). The annual rate of HVI use for C2 fractures decreased significantly from 2001 to 2014 (0.32 to 0.06 HVI procedures per 100,000 people; p<.01) with the rate of decline being less pronounced in high-volume institutions (70% decrease vs. 85% and 90% for medium- and low-volume, respectively). CONCLUSIONS: Halo vest utilization for C2 fractures in New York State has been declining over the past decade, with the decline being less pronounced in high-volume hospitals. Our hospital volume analysis suggests that HVI use in high-volume institutions is associated with a lower rate of future readmissions. This finding suggests that patients with C2 fractures may benefit from treatment at high-volume institutions. Further research to help improve referral of appropriate patients and increase access to such institutions is warranted. (C) 2020 Elsevier Inc. All rights reserved.