罗哌卡因规律性间断硬膜外注射分娩镇痛中右美托咪定剂量预测模型的建立

Establishment of Dose Prediction Model of Dexmedetomidine in Labor Analgesia with Regular Intermittent Epidural Injection of Ropivacaine

目的建立罗哌卡因规律性间断硬膜外注射分娩镇痛中右美托咪定剂量的预测模型。方法选取广东省中山市小榄人民医院2020年12月至2022年5月收治的硬膜外自控镇痛(PCEA)分娩的初产妇120例作为研究对象。根据单盲随机原则,将其分为D组(人工神经网络模型组)、A组(单变量线性回归方程模型组)和M组(多元线性回归方程模型组),每组40例。比较分析3组产妇的一般资料、疼痛视觉模拟评分法(VAS)、体温、新生儿情况、镇痛药物使用情况、镇痛效果、分娩方式以及产妇满意度。采用Kaplan-Meier方法绘制生存曲线。结果T1时D组疼痛VAS评分以及镇痛起效时间均显著低于A组和M组(P<0.05),镇痛维持时间显著更高(P<0.05)。在T2时,D组的PCEA按压次数和疼痛VAS评分与M组比较差异无统计学意义(P>0.05),但显著低于A组(P<0.05)。右美托咪定用量M组显著低于A组和D组(P<0.05)。生存曲线结果显示,3组总产程时间差异有统计学意义(P<0.05);A组和D组与M组总产程时间比较,差异无统计学意义(P>0.05);D组和A组总产程时间比较差异有统计学意义(P<0.05)。D组满意度为优的产妇比例最高,且无满意度为差者,但3组产妇满意度比较差异无统计学意义(P>0.05)。结论多元线性回归方程模型以及人工神经网络模型预测计算右美托咪定剂量应用于罗哌卡因PCEA无痛分娩中有效性、满意度优于主要参考标准为疼痛VAS评分的单变量线性回归方程,并且人工神经网络模型表现佳,这表明右美托咪定剂量预测模型具有一定的临床应用价值。

Objective To establish a model for predicting the dose of dexmedetomidine in labor analgesia with regular intermittent epidural injection of ropivacaine.Methods 120 primiparas who were delivered by PCEA in our hospital from December 2020 to May 2022 were selected as study subjects.According to the principle of single-blind randomization,they were divided into D group(artificial neural network model group),A group(univariate linear regression equation model group),and M group(multivariate linear regression equation model group),with 40 cases in each group.The general data,pain VAS score,body temperature,newborn condition,use of analgesic drugs,analgesic effect,delivery mode,and maternal satisfaction of the three groups were compared and analyzed.Kaplan-Meier method was used to draw the survival curve.Results The visual analogue scale(VAS)of pain at T1 and the onset time of analgesia in D group were significantly lower than those in A and M groups(P<0.05),and the duration of analgesia was notably higher than those in A and M groups(P<0.05).The PCEA pressing times and pain VAS score in D group at T2,compared with M group,showed no significant difference(P>0.05),but were significantly lower than those in A group(P<0.05).The dosage of dexmedetomidine in M group was significantly lower than that in A group and D group(P<0.05).The results of the survival curve demonstrated notable differences in the total labor process time among the three groups(P<0.05);there was no notable difference between A group,D group,and M group(P>0.05);however,there was a notable difference in the total labor process time between D group and A group(P<0.05).The proportion of women with excellent satisfaction in D group was the highest,and those without poor satisfaction were none,but there was no notable difference among the three groups(P>0.05).Conclusion The effectiveness and satisfaction of dexmedetomidinedose calculated by multiple linear regression equation model and artificial neural network model in labor pains with regularintermittent epidural injection of ropivacaine were better than those of the univariate linear regression equation with VAS scoreas the main reference index,and the artificial neural network model was the best.This indicates that the dose prediction model ofdexmedetomidine has certain clinical application value.