Peri-operative Outcomes after Carotid Endarterectomy in Selected Operated Patients with Pre-operative Modified Rankin Scale Score 3: Analysis of the UK National Vascular Registry.

Abstract

Randomised controlled trials supporting carotid endarterectomy (CEA) for symptomatic carotid stenosis provide little direct evidence for patients with disabling stroke. This study evaluated whether selected operated patients with a pre-operative modified Rankin scale (mRS) score of 3 had a higher peri-operative risk than operated patients with mRS scores of 0 - 2. This was a retrospective observational cohort study of prospectively collected national registry data. The UK National Vascular Registry was analysed for adults undergoing CEA for stroke between 2021 and 2023 with 50 - 99% ipsilateral carotid stenosis. Primary outcomes were thirty day stroke, registry recorded thirty day death, and thirty day stroke or death. Multivariable logistic regression adjusted for age, sex, smoking, diabetes, hypertension, ischaemic heart disease, congestive cardiac failure, atrial fibrillation, chronic renal failure, cancer, and chronic lung disease. Pre-specified sensitivity analyses compared mRS 3 separately with mRS 0 - 1 and mRS 2, added operative and medical therapy covariates, excluded thrombolysis, assessed death follow up completion, and provided sex disaggregated outcomes. Of 6 632 patients, 946 (14.3%) had mRS 3. Follow up stroke data were complete in 6 578 of 6 632 (99.2%), whereas the dedicated thirty day death field was complete in 4 301 of 6 632 (64.9%). Thirty day stroke occurred in 23 of 946 patients (2.4%) with mRS 3 and 106 of 5 686 (1.9%) with mRS 0 - 2 (p = .24); observed thirty day death in one of 946 (0.1%) and 18 of 5 686 (0.3%) (p = .51); and thirty day stroke or death in 24 of 946 (2.5%) and 117 of 5 686 (2.1%) (p = .34). After adjustment, mRS 3 was not associated with thirty day stroke (odds ratio [OR] 1.32, 95% confidence interval [CI] 0.83 - 2.08; p = .24), observed thirty day death (OR 0.36, 95% CI 0.05 - 2.72; p = .32), or thirty day stroke or death (OR 1.26, 95% CI 0.80 - 1.97; p = .31). Respiratory complications, post-operative confusion, any medical complication, and hospital length of stay were higher in the mRS 3 group. Sensitivity analyses were materially unchanged, and no convincing interaction by sex was identified. Among selected patients who underwent surgery, pre-operative mRS 3 was not associated with a statistically significant increase in registry recorded thirty day stroke or thirty day stroke or death compared with operated patients with mRS 0 - 2, although medical morbidity and resource use were higher. Because there was no medically treated comparator and mortality follow up was incomplete, these findings inform peri-operative risk in selected surgical candidates only and should not be interpreted as evidence that CEA benefits all patients with mRS 3.