Surgical Outcomes for Degenerative Cervical Myelopathy: A British Spinal Registry and Hospital Episode Statistics Study Exploring Patient Characteristics and Outcomes

Sources of Funding: This work is funded by the NIHR SPCR Post-doctoral fellowship (C111) and the Open Innovation Platform (OIP/23-24/012). LW has also been supported by the NIHR Exeter BRC and the Bateman Professorial Fellowship. Conflicts of Interest: No conflicts of interest Purpose Surgical registry and administrative datasets differ in their coverage, completeness and case ascertainment. We compared degenerative cervical myelopathy (DCM) case ascertainment, patient characteristics, outcomes and risk factors for early postoperative complications across two large UK datasets. Methods We analysed Hospital Episode Statistics (HES) administrative data and unlinked British Spinal Registry (BSR) registry data (April 2017–March 2024), comparing DCM diagnosis, demographics and outcomes. Mixed-effects logistic regression explored associations with early postoperative complications (adjusted odds ratios (AOR)). Results In HES, 44,623 cases underwent cervical decompression surgery of which 36,441 cases underwent cervical decompression surgery for likely degenerative causes, of which 24,883 (68%) were coded as DCM. The BSR included 27,111 records (74% HES coverage), with 12,085 DCM cases (45% of all BSR entries; 49% of HES-coded DCM cases). BSR patients were more likely to be older and male, while higher proportions from the most deprived quintiles were recorded in HES. Intraoperative complications were reported by 610/12,085 (5%) BSR respondents; 30-day readmissions occurred in 1,361/24,883 (5%) HES cases. Both datasets identified increased complication risk with posterior surgical approach (HES: AOR 1.26, 95% CI 1.06 to 1.50; BSR: AOR 1.78 95% CI 1.12 to 2.83) and most deprived quintile (HES: AOR 1.3, 95% CI 1.05 to 1.52; BSR: AOR 2.73, 95% CI 1.15 to 6.47). Conclusion The BSR captured 74% of HES cervical surgeries, but case ascertainment differed between sources. Future studies should link datasets and adjust for selection bias to improve validity.