DataFigure3.

Background
Monitoring the nociception/antinociception balance for analgesic guidance during general anesthesia may improve the quality of anesthesia. The University Hospital of Lille (France) has developed an expert software system for automatic remifentanil administration based on the continuous monitoring of the Analgesia Nociception Index (MDoloris, France). We assessed the clinical efficacy and safety of the ANI-REMI-LOOP “expert-system software” during burn surgery in a monocentric randomized controlled trial.
Methods
The trial was approved by the French Ethics Committee, and all patients gave written informed consent. From 2018 to 2022, 52 adults were randomized into two groups: manual remifentanil infusion (standard practice) or automatic remifentanil infusion (expert-system software) during BIS-guided propofol anesthesia at the burn center of the hospital. In the standard practice group, remifentanil administration was based on Minto’s model and guided by the analgesia nociception index. In both groups, propofol was administered based on Schnider’s model and guided by the BiSpectral Index (Covidien). The primary endpoint was the cumulative remifentanil dose administered during anesthesia and secondary endpoints were related to the clinical safety of automatic remifentanil administration with the incidence and duration of hypotension, bradycardia, hypertension or tachycardia related to nociception. After anesthesia, the endpoints were pain and analgesic requirements during 2 hours. A p value < 0.05 was considered statistically significant. Data are presented as median [1st to 3rd quartile].
Results
The cumulative remifentanil dose was significantly lower in the automatic group 0.125 µg.kg-1.min-1 [0.106 to 0.149] vs. 0.152 µg.kg-1.min-1 [0.137 to 0.237], p = 0.004), and the cumulative proportion of time with hemodynamic impairment or reactivity was significantly lower in the expert-system automatic group 4.2% [2.5 to 5.7] vs. 19.4% [6.9 to 59.9], p = 0.010). There were no safety issues, and pain and analgesic requirements were similar in both groups after surgery.
Conclusions
Automatic remifentanil administration demonstrated good clinical performances during propofol anesthesia for burn surgery. It is likely that these results can be extrapolated to any surgical setting under general anesthesia, but this needs to be tested with further randomized clinical trials.

背景
全身麻醉期间，监测伤害感受（nociception）与抗伤害感受（antinociception）的平衡以指导镇痛管理，可有效提升麻醉质量。法国里尔大学附属医院基于镇痛伤害感受指数（Analgesia Nociception Index）的持续监测，开发了一款瑞芬太尼（remifentanil）自动输注专家软件系统。本研究通过单中心随机对照试验，评估了ANI-REMI-LOOP“专家系统软件”在烧伤手术中的临床疗效与安全性。
方法
本试验已获法国伦理委员会批准，所有患者均签署书面知情同意书。2018年至2022年，52名成年患者被随机分为两组，在医院烧伤中心接受脑电双频指数（Bispectral Index, Covidien）引导的丙泊酚（propofol）全身麻醉：一组采用手动瑞芬太尼输注（标准操作组），另一组采用自动瑞芬太尼输注（专家系统软件组）。标准操作组中，瑞芬太尼输注基于Minto模型，并以镇痛伤害感受指数为指导；两组丙泊酚输注均基于Schnider模型。本研究的主要终点为麻醉期间瑞芬太尼的累积输注剂量；次要终点与自动瑞芬太尼输注的临床安全性相关，包括与伤害感受相关的低血压、心动过缓、高血压或心动过速的发生率及持续时间。术后2小时内的疼痛程度与镇痛药物需求量为麻醉后终点指标。本研究以p值<0.05为具有统计学显著性，数据以中位数[第1至第3四分位数间距]形式呈现。
结果
自动输注组的瑞芬太尼累积输注剂量显著低于手动组：0.125 μg·kg⁻¹·min⁻¹[0.106~0.149] vs 0.152 μg·kg⁻¹·min⁻¹[0.137~0.237]，p=0.004；专家系统自动输注组的血流动力学异常或应激反应的累积时间占比亦显著降低：4.2%[2.5~5.7] vs 19.4%[6.9~59.9]，p=0.010。两组均未出现安全性相关不良事件，且术后疼痛程度与镇痛药物需求量无显著差异。
结论
烧伤手术中丙泊酚全身麻醉下的自动瑞芬太尼输注展现出良好的临床性能。该结果或可推广至所有全身麻醉下的外科手术场景，但仍需进一步的随机对照试验予以验证。