腹腔镜脾切除联合贲门周围血管离断术中转开腹风险预测模型的构建

A risk prediction model of conversion to open surgery during laparoscopic splenectomy and esophagogastric devascularization

目的建立腹腔镜下脾切除联合贲门周围血管离断手术(LSED)中转开腹的风险预测模型。方法回顾性分析2011年2月至2020年11月空军军医大学第二附属医院普外科收治的358例肝硬化门静脉高压患者的临床资料,所有患者均首先尝试LSED。采用单变量Logistic和最小化绝对收缩和选择算子(LASSO)回归分析中转开腹的独立危险因素,并用R语言构建中转开腹列线图预测模型。比较腹腔镜组和中转开腹组患者术中及术后情况。结果358例患者中31例中转开腹,中转开腹率为8.7%。单因素分析中,终末期肝病模型(MELD)评分高、BMI≥24 kg/m2、有上腹部手术史、食管胃底静脉曲张红色征、PLT计数低及凝血酶原时间延长均是发生中转开腹的危险因素(均P<0.05)。LASSO回归最终确定5个因素:MELD评分、BMI、PLT、上腹部手术史及食管胃底静脉曲张红色征。成功构建列线图预测模型,模型的ROC曲线下面积为0.831。相比于腹腔镜组,中转开腹组手术时间延长、出血量增加、术后腹腔引流时间和住院时间均延长、围手术期并发症的发生率高,差异均有统计学意义(t=-2.167,P=0.031;Z=-4.350,P<0.01;Z=-3.102,P=0.002;Z=-3.454,P=0.001;χ^(2)=8.773,P=0.003)。结论采用LASSO回归筛选出5个对中转开腹影响最大的指标,即MELD评分、BMI、PLT计数、食管胃底静脉曲张红色征及既往上腹部手术史,据此建立的列线图预测模型具有良好的预测能力。中转开腹的患者有更差的短期预后。

Objective To establish a risk prediction model of conversion to open surgery during laparoscopic splenectomy and esophagogastric devascularization(LSED)and evaluate the impact of this conversion on patients'short-term prognosis.Methods A total of 358 cirrhotic portal hypertension patients admitted to the Department of General Surgery,Second Affiliated Hospital,Air Force Military Medical University from Feb 2011 to Nov 2020 were retrospectively analyzed.All patients underwent attempted LSED.Univariate and least absolute shrinkage and selection operator(LASSO)Logistic regression were used to analyze the independent risk factors for conversion to laparotomy,and the R language was used to build a nomogram prediction model for conversion to laparotomy.The intraoperative and postoperative conditions of the two groups were compared.Results A total of 358 patients were included in this study,of which 31(8.7%).patients were converted to open surgery.In univariate analysis,high MELD score,BMI≥24 kg/m2,history of upper abdominal surgery,red sign of the varicose,low platelet count and prolonged PT are risk factors for conversion.LASSO regression finally identified 5 factors:MELD,BMI,PLT,history of surgery,and red sign.In the nomogram prediction model the area under ROC curve was 0.831.The conversion led to longer operation time;increased blood loss;prolonged postoperative abdominal drainage,longer hospital stay,and increased perioperative complications(t=-2.167,P=0.031;Z=-4.350,P<0.01;Z=-3.102,P=0.002;Z=-3.454,P=0.001;χ^(2)=8.773,P=0.003).Conclusions LASSO regression selected five indicators with greatest impact on intraoperative conversion:MELD,BMI,PLT,red sign,and previous history of abdominal surgery.The nomogram prediction model established has good prediction ability.Patients converted to open surgery had worse short-term outcomes.