Each year, approximately 3000–3500 patients undergo carotid endarterectomy in the UK, and over 150,000 worldwide [1, 2]. It is thought that the selection of anaesthetic method – whether cervical plexus block, general anaesthesia or a mix of both – can impact haemodynamic parameters differently and the oxygenation and perfusion of the brain and heart. This variability may have an impact on the risk of stroke, myocardial infarction and mortality [3]. We conducted a systematic review to compare the safety and clinical outcomes of different anaesthetic techniques in patients undergoing carotid endarterectomy.

This analysis was based on data from a systematic review, the protocol of which was registered prospectively. We included randomised clinical trials that enrolled patients aged ≥ 18 y undergoing carotid endarterectomy. The studies included comparisons between any two anaesthesia techniques (general, local/regional, combined regional and general). Our primary outcomes were stroke and myocardial infarction within 30 days of surgery. We also included studies reporting data on death within 30 days; postoperative pain within 24 h; arteries shunted; postoperative haematoma; time course of arterial occlusion; patient satisfaction; surgeon satisfaction; postoperative cognitive dysfunction; duration of surgery, ICU stay and hospital stay; need for reintervention; cranial nerve injury; respiratory complications; quality of life one month after surgery; and haemodynamic parameters. Screening and data collection were conducted in duplicate by independent reviewers. We assessed risk of bias in individual studies using the Risk of Bias 2 tool and judged the certainty of evidence according to GRADE recommendations (online Supporting Information Figure S1).

We categorised the interventions into four groups for the Bayesian network meta-analyses: epidural; regional (i.e. cervical plexus block and/or local infiltration); general; and combined regional and general anaesthesia. We also performed Bayesian pairwise meta-analyses comparing regional with general anaesthesia. Overall, 24 studies encompassing 5341 patients were included in the analyses.

We did not find any significant differences between the interventions for the primary outcomes of stroke and myocardial infarction within 30 days of surgery (Table 1 and Fig. 1). However, there was significantly less use of an arterial shunt with regional compared with general anaesthesia (relative risk (95% credible interval) 0.33 (0.17–0.66), Table 1). No significant differences were found for the remaining outcomes.

Our study did not yield results to inform decision-making regarding choice of anaesthetic technique for carotid endarterectomy. No significant differences were observed for most outcomes. However, readers should not interpret these findings as evidence of equivalence as they may reflect insufficient data.

While our findings show reduced use of arterial shunt with the use of regional anaesthesia, there is a possibility that this result is due to chance, given the number of analyses performed. Applying a Bonferroni correction would render this significant difference non-significant. Our results also indicate that regional anaesthesia may reduce mortality within 30 days of surgery (Table 1 and Fig. 1). However, it is important to emphasise that this was not statistically significant, the certainty of evidence is very low, and this finding should not be used to inform clinical practice.

Other researchers have summarised data from randomised trials and observational studies [3-5]. While observational data suggest improved outcomes with regional anaesthesia [3, 5], including a reduced incidence of stroke, myocardial infarction and death, these results were not confirmed by a previous summary of randomised studies [4], and not by our updated analysis.

While regional anaesthesia may improve outcomes by blocking nociceptive pathways and increasing cerebral blood flow, it is also possible that uncovered nerve fibres or inadequate sedation might lead to tachycardia and hypertension, increasing the risk of myocardial ischaemia in this high-risk cohort of patients. This adds complexity, as different techniques may affect key outcomes like stroke and myocardial infarction differently. Future research might consider evaluating composite or more critical outcomes, such as mortality.

Although our findings do not provide strong evidence to inform clinical decision-making, they do suggest a potential difference between anaesthetic techniques in relation to important outcomes such as the need for arterial shunt and mortality. These findings are consistent with observational data and may be associated with the incidence of stroke and myocardial infarction [3, 5]. As such, our results underscore the need for further research to determine whether any specific technique or combination of techniques may improve patient outcomes.

In conclusion, the current evidence does not provide robust support for the selection of any specific anaesthetic technique for carotid endarterectomy. We did not observe significant differences for most outcomes, and the one observed difference may be attributable to chance.

英国每年约有 3000-3500 名患者接受颈动脉内膜切除术，全球超过 150,000 名患者 [1， 2]。据认为，麻醉方法的选择——无论是颈丛神经阻滞、全身麻醉还是两者的混合——都会对血液动力学参数以及大脑和心脏的氧合和灌注产生不同的影响。这种可变性可能会影响中风、心肌梗死和死亡率的风险 [3]。我们进行了一项系统评价，以比较不同麻醉技术在颈动脉内膜切除术患者中的安全性和临床结局。

该分析基于一项系统评价的数据，该评价的方案进行了前瞻性注册。我们纳入了随机临床试验，这些试验招募了 ≥ 18 岁接受颈动脉内膜切除术的患者。这些研究包括任何两种麻醉技术（全身麻醉、局部/区域麻醉、区域联合麻醉和全身麻醉）之间的比较。我们的主要结局是手术后 30 天内的卒中和心肌梗死。我们还纳入了报告 30 天内死亡数据的研究;术后 24 小时内疼痛;动脉分流;术后血肿;动脉闭塞的时间进程;患者满意度;外科医生满意度;术后认知功能障碍;手术时间、ICU 住院时间和住院时间;需要再次干预;颅神经损伤;呼吸系统并发症;手术后一个月的生活质量;和血流动力学参数。筛选和数据收集由独立评价员一式两份进行。我们使用Risk of Bias 2工具评估了单个研究的偏倚风险，并根据GRADE建议（在线支持信息图S1）判断证据质量。

我们将干预措施分为四组进行贝叶斯网络荟萃分析：硬膜外麻醉;区域性（即 颈丛神经阻滞和/或局部浸润）;常规;以及区域麻醉和全身麻醉联合治疗。我们还进行了贝叶斯成对荟萃分析，比较了区域麻醉和全身麻醉。总体而言，分析纳入了 24 项研究，涉及 5341 名患者。

我们未发现干预措施在手术后 30 天内对卒中和心肌梗死的主要结局有任何显著差异（表 1 和图 1）。然而，与全身麻醉相比，区域动脉分流的使用显着减少（相对风险 （95% 可信区间） 0.33 （0.17-0.66），表 1）。其余结局未发现显著差异。

我们的研究没有产生结果来为选择颈动脉内膜切除术的麻醉技术决策提供信息。大多数结局未观察到显著差异。但是，读者不应将这些发现解释为等效性的证据，因为它们可能反映数据不足。

虽然我们的研究结果表明，使用区域麻醉减少了动脉分流的使用，但考虑到进行的分析数量，这一结果可能是偶然的。应用 Bonferroni 校正将使此显著差异不显著。我们的结果还表明，区域麻醉可能会降低手术后 30 天内的死亡率（表 1 和图 1）。然而，需要强调的是，这并不具有统计学意义，证据质量非常低，这一发现不应用于临床实践。

其他研究人员总结了随机试验和观察性研究的数据 [3-5]。虽然观察数据表明区域麻醉的结局有所改善[3,5]，包括中风、心肌梗死和死亡的发生率降低，但这些结果并未得到先前随机研究总结的证实[4]，也没有被我们更新的分析所证实。

虽然区域麻醉可能通过阻断伤害感受途径和增加脑血流量来改善结果，但未覆盖的神经纤维或镇静不足也可能导致心动过速和高血压，从而增加这一高危患者群体患心肌缺血的风险。这增加了复杂性，因为不同的技术可能会对中风和心肌梗死等关键结局产生不同的影响。未来的研究可能会考虑评估复合或更关键的结局，例如死亡率。

尽管我们的研究结果没有提供强有力的证据来为临床决策提供信息，但它们确实表明麻醉技术在重要结局（例如动脉分流需求和死亡率）方面存在潜在差异。这些发现与观察数据一致，可能与卒中和心肌梗死的发生率有关 [3， 5]。因此，我们的结果强调了进一步研究的必要性，以确定任何特定技术或技术组合是否可以改善患者的预后。

总之，目前的证据并未为选择任何特定的颈动脉内膜切除术麻醉技术提供有力支持。我们没有观察到大多数结局的显著差异，观察到的一个差异可能是偶然的。