Background

Peroral endoscopic myotomy (POEM) is nowadays a standard method for treatment of achalasia; nevertheless, it remains an invasive intervention with corresponding risk of adverse events (AEs). The classification and grading of AEs are still a matter of discussion. The aim of our retrospective study was to assess the occurrence of all “undesirable” events and “true” adverse events in patients undergoing POEM and to compare the outcomes when either Clavien–Dindo classification (CDC) or American Society of Gastrointestinal Endoscopy (ASGE) lexicon classification applied.

Methods

This was a retrospective analysis of prospectively managed database of all patients who had undergone POEM between December 2012 and August 2018. We assessed the pre-, peri-, and early-postoperative (up to patient’s discharge) undesirable events (including those not fulfilling criteria for AEs) and “true” AEs according the definition in either of the classifications.

Results

A total of 231 patients have successfully undergone 244 POEM procedures (13 × re-POEM). Twenty-nine procedures (11.9%) passed uneventfully, while in 215 procedures (88.1%), a total of 440 undesirable events occurred. The CDC identified 27 AEs (17 minor, 10 major) occurring in 23/244 (9.4%) procedures. The ASGE lexicon identified identical 27 AEs (21 mild or moderate, 6 severe or fatal) resulting in the severity distribution of AEs being the only difference between the two classifications.

Only the absence of previous treatment was found to be a risk factor [p = 0.047, OR with 95% CI: 4.55 (1.02; 20.25)] in the combined logistic regression model.

Conclusion

Undesirable events are common in patients undergoing POEM but the incidence of true AEs is low according to both classifications. Severe adverse events are infrequent irrespective of the classification applied. CDC may be more appropriate than ASGE lexicon for classifying POEM-related AEs given a surgical nature of this procedure.

Graphical abstract

中文翻译：

---

经口内窥镜肌切开术 (POEM) 的围手术期安全性概况——根据两种不同分类对不良事件的回顾性分析

背景

经口内窥镜肌切开术 (POEM) 是当今治疗贲门失弛缓症的标准方法；尽管如此，它仍然是一种侵入性干预，具有相应的不良事件 (AE) 风险。AE 的分类和分级仍在讨论中。我们的回顾性研究的目的是评估接受 POEM 的患者中所有“不良”事件和“真实”不良事件的发生，并比较 Clavien-Dindo 分类 (CDC) 或美国胃肠内窥镜检查协会 (ASGE) 时的结果应用词典分类。

方法

这是对 2012 年 12 月至 2018 年 8 月期间接受 POEM 的所有患者的前瞻性管理数据库的回顾性分析。我们评估了术前、围手术期和术后早期（直至患者出院）的不良事件（包括那些不符合标准的事件） AEs）和“真正的”AEs 根据任一分类中的定义。

结果

共有 231 名患者成功进行了 244 次 POEM 手术（13 × re-POEM）。29 次手术 (11.9%) 顺利通过，而在 215 次手术 (88.1%) 中，总共发生了 440 次不良事件。CDC 在 23/244 (9.4%) 的程序中确定了 27 次 AE（17 次轻微，10 次严重）。ASGE 词典确定了 27 种相同的 AE（21 种轻度或中度，6 种严重或致命），导致 AE 的严重程度分布是两种分类之间的唯一区别。

在组合逻辑回归模型中，仅发现先前未接受治疗是一个风险因素 [ p  = 0.047，OR with 95% CI：4.55 (1.02; 20.25)]。

结论

不良事件在接受 POEM 的患者中很常见，但根据这两种分类，真正的 AE 发生率都很低。无论应用何种分类，严重不良事件都很少见。鉴于此程序的手术性质，CDC 可能比 ASGE 词典更适合对 POEM 相关的 AE 进行分类。

图形概要