Background: The optimum timing of carotid intervention for symptomatic carotid stenosis remains unclear. The objective was to investigate outcomes of very urgent (< 48 h from neurological event) in comparison to urgent ($ 48 h from neurological event) carotid intervention for symptomatic carotid disease.Methods: A systematic literature review was carried out of randomised control trials (RCTs) and observational studies reporting peri-procedural outcomes of carotid intervention in relation to the length of time since the neurological event (PROSPERO registration number: CRD 42017075766). Ipsilateral stroke and death were defined as the primary outcome endpoints. Transient ischaemic attack (TIA) and myocardial infarction (MI) were secondary outcome parameters. Comparative outcomes were calculated and reported as dichotomous outcome measures using the odds ratio (OR) and associated 95% confidence interval (CI) for very urgent (< 48 h since neurological event) versus urgent ($ 48 h) intervention. The combined overall effect size was calculated using a random effects model.Results: Twelve observational studies and one RCT representing 5751 interventions, 5385 carotid endarterectomies (CEAs) and 366 carotid artery stenting (CAS) procedures, were included in quantitative synthesis. Very urgent carotid intervention was associated with increased risk of stroke within 30 days of treatment compared with urgent carotid intervention (OR 2.19, 95% CI 1.46e3.26, p < .001). No significant difference was found in mortality (OR 1.55, 95% CI 0.81e2.96, p = .19), TIA (OR 1.33, 95% CI 0.55e3.19, p = .52) or MI (OR 1.33, 95% CI 0.41e4.33, p = .64).Conclusions: Very urgent carotid intervention was found to be associated with increased risk of stroke.

INTRODUCTIONThe retrojugular approach for carotid endarterectomy (CEA) has been reported to have the advantages of shorter operative time and ease of dissection, especially in high carotid lesions. Controversial opinion exists with regard to its safety and benefits over the conventional antejugular approach. METHODS A systematic review of electronic information sources was conducted to identify studies comparing outcomes of CEA performed with the retrojugular and antejugular approach. Synthesis of summary statistics was undertaken and fixed or random effects models were applied to combine outcome data. FINDINGS A total of 6 studies reporting on a total of 740 CEAs (retrojugular approach: 333 patients; antejugular approach: 407 patients) entered our meta-analysis models. The retrojugular approach was found to be associated with a higher incidence of laryngeal nerve damage (odds ratio [OR]: 3.21, 95% confidence interval [CI]: 1.46-7.07). No significant differences in the incidence of hypoglossal or accessory nerve damage were identified between the retrojugular and antejugular approach groups (OR: 1.09 and 11.51, 95% CI: 0.31-3.80 and 0.59-225.43). Cranial nerve damage persisting during the follow-up period was similar between the groups (OR: 2.96, 95% CI: 0.79-11.13). Perioperative stroke and mortality rates did not differ in patients treated with the retrojugular or antejugular approach (OR: 1.26 and 1.28, 95% CI: 0.31-5.21 and 0.25-6.50). CONCLUSIONS Currently, there is no conclusive evidence to favour one approach over the other. Proof from a well designed randomised trial would help determine the role and benefits of the retrojugular approach in CEA.