Robotic-Assisted Spinal Instrumentation from C1 to S1 Experience of a UK Neurosurgical Tertiary Referral Centre

Background: Robotic-assisted spinal instrumentation is increasingly used to support implant placement, reduce radiation exposure, and improve operative workflow. However, UK NHS data describing early programme experience across elective and emergency practice remain limited. This study aimed to evaluate the feasibility, safety profile, and early learning curve of robotic-assisted spinal instrumentation in an unselected UK NHS cohort. Methods: A retrospective, single-centre, single-arm observational cohort study was conducted, including the first 50 consecutive patients treated from programme inception between June 2024 and January 2026. No exclusion criteria were applied, and each record represented one patient and one operation. Missing data were not imputed, and denominators were reported per variable. Learning curve effects were assessed using Spearman correlation and cumulative sum analysis. Wilson 95% confidence intervals were reported for complication rates. Results: The mean age was 61.9 years; 50% were male; mean body mass index was 29.0; and median Charlson Comorbidity Index was 3. Indications were degenerative disease in 74%, trauma in 24%, and deformity in 2%. Emergency admissions accounted for 18/50 cases. Robot-specific adverse events included abandonment or conversion in 2/50 cases and system malfunction in 1/50. Surgical complications occurred in 7/50 patients. On surgeon-reviewed routine post-operative imaging, no screw malpositions required revision and no durotomies or vascular injuries were recorded; this represents a clinical revision rate rather than a formal radiological measure of screw accuracy, for which blinded Gertzbein-Robbins grading was not performed. Cumulative sum analysis suggested a potential fluoroscopy change point around case 30, with the median fluoroscopy events falling from 151 to 16 thereafter. Emergency cases involved more instrumented levels, longer operative times, and more open surgery than elective cases. Conclusions: Robotic-assisted spinal instrumentation was feasible across a diverse UK NHS caseload from C1 to S2, with a low observed complication rate. The findings suggest a fluoroscopy learning curve threshold around case 30. Larger comparative studies with formal radiological accuracy assessment are required.