硬膜外減痛分娩效果不佳──相關因素之回溯性研究

Ineffectiveness of Epidural Analgesia for Labor Pain in a Northern Taiwan Medical Center a Retrospective Analysis of Risk Factors

前言：我們的研究希望了解產婦施行硬膜外止痛的盛行率及硬膜外止痛的失敗率，並分析與硬膜外止痛效果不佳相關的因素。過程與方法：這是一個回溯性研究。我們蒐集了台北市新光吳火獅紀念醫院從2005年1月到2006年12月所有做硬膜外止痛的產婦的資料。資料隨機的被分為兩組，訓練組（training group）與確認組（validating group）。我們定義硬膜外止痛效果不佳為疼痛分數大於3，在給予硬膜外止痛藥物三十分鐘後。使用卡方檢定跟t檢定分析訓練組所有的變項，找出跟硬膜外止痛效果不佳的因素。有差異的變項再放入邏輯式回歸分析，試著建立預測模式。確認組的資料則用來確定此模型的準確度。結果：共有5809位產婦，其中1015位接受了硬膜外減痛分娩，盛行率為17.47%。在訓練組中止痛效果不佳的比例為26%，止痛效果不佳者有較短時間的第一產程（310.7比264.43分鐘），給藥三十分鐘後子宮頸擴張程度較大（3.25比2.9公分），以宮頸擴張速度較快（1.52比0.67公分／小時），較少使用產箝或真空吸引接生（11.25%比21.69%），對減痛分娩較不滿意（21.77%比51.49%表示非常滿意）。硬膜外止痛使用的藥物種類也有顯著差異，在止痛效果不佳者較多產婦使用Lidocaine及Bupivacaine一次給藥，較多使用Bupivacaine持續給藥。邏輯式回歸分析建立的預測模型，選擇出有意義的變項，此預測模型的操作特性曲線下的面積（AUC, area under ROU curve）為0.6712。當可能性的切點為0.5時，確認組的準確度為0.6873。結論：結果顯示跟硬膜外減痛分娩較果不佳相關的因素包括子宮頸擴張速度較快，使用Lidocaine或Bupivacaine一次給藥，使用Bupivacaine持續給藥。未來的研究可以加入更多的因子來分析。

Objectives: Although epidural analgesia is the most effective and least depressant treatment for labor pain, inadequate pain relief still occurs. The cause of failed epidural analgesia in labor pain is uncertain. In this study we analyzed the failure rate of epidural analgesia in labor pain and evaluated the factors associated with inadequate pain relief. Methods: A retrospective chart review was performed in parturients who underwent epidural analgesia for labor pain in a northern Taiwan medical center. Ineffectiveness of epidural analgesia of labor pain was defined as pain NRS (numeric rating score) > 3 at 30 minutes after epidural drug administration. All data were randomized to two groups: training group and validating group. The data of the training group were analyzed to establish a logistic regression model for predictive ineffectiveness of epidural painless labor. Then validating group was used to confirm the accuracy of this model. Results: From 2005 to 2006, 1015 of 5809 parturients under parturition received the epidural painless labor management. The prevalence was 17.47%. The failure rate of the training group was 26%. The failure group had shorter phase I duration, more cervical dilatation in 30 minutes, faster progression of cervical dilatation, less instrumentation delivery, and less satisfaction about epidural painless labor. When compared with other drugs, the usage of Ropivacaine in loading analgesia showed less incidence of analgesia failure than Lidocaine and Bupivacaine, and also showed less failure rate than Bupivacaine in continuous infusion. The predictive model of ineffective epidural painless labor was established. The area under receiver operating characteristic curve (AUC) was calculated as 0.6712. When the cut point of probability was 0.5, the accuracy of validating group was 0.6873. Conclusion: This is the first study about the determinants of ineffective epidural analgesia for labor pain in Taiwan. Our results reveal that faster speed of cervical dilatation, loading with Lidocaine or Bupivacaine, and infusing with Bupivacaine are key factors associated with epidural anesthesia in labor.