腹腔镜前列腺癌根治术后发生动力性肠梗阻风险因素列线图预测模型的构建

Construction of a nomogram prediction model for risk factors of intestinal obstruction after radical laparoscopic prostate cancer surgery

目的 探究腹腔镜前列腺癌根治术后发生动力性肠梗阻的风险因素,构建列线图预测模型并进行评估。方法 选择2022年1月-2023年6月青岛市胶州中心医院接受腹腔镜前列腺癌根治术治疗且在术后发生动力性肠梗阻的46例患者作为病例组,选择同期前列腺癌根治术后未发生动力性肠梗阻的92例患者作为对照组。记录患者的一般资料、术前及术后24 h内的白蛋白、C反应蛋白(CRP)、肿瘤坏死因子-α(TNF-α)和白介素-6(IL-6)水平,采用焦虑自评量表(SAS)评估患者术前焦虑程度,记录评分。对前列腺癌患者术后发生动力性肠梗阻的危险因素进行单因素分析,将具有统计学意义的单因素变量纳入二元Logistic多因素回归分析,采用R软件构建列线图预测模型。结果 与对照组比较,病例组患者合并高脂血症、心脏疾病、慢性阻塞性肺疾病,阿片类镇痛药物使用的比例、手术时间、术前焦虑评分、血清TNF-α和IL-6水平均升高,术后白蛋白水平降低,差异有统计学意义(P<0.05)。二元多因素Logistic回归分析结果表明,患有心脏疾病(X_(1))、阿片类镇痛药物的使用(X_(2))、较高的术前焦虑评分(X_(3))、较低的术后血清白蛋白水平(X_(4))及较高的术后血清TNF-α水平(X_(5))、术后血清IL-6水平(X_(6))是前列腺癌患者术后发生动力性肠梗阻的独立危险因素(P<0.05),构建的Logistics回归模型为Log(P)=-14.019+1.336X_(1)+1.622X_(2)+0.129X_(3)-0.145X_(4)+0.050X_(5)+0.072X_(6)。校准曲线显示列线图模型的一致性较好,临床决策曲线评价结果显示该模型具有较好的应用价值。ROC曲线分析结果表明模型预测的曲线下面积为0.856(0.812~0.926),灵敏度和特异度分别为75.2%和76.1%。结论 患有心脏疾病、阿片类镇痛药物的使用、术前焦虑评分、术后白蛋白、TNF-α及IL-6水平是前列腺癌患者术后发生动力性肠梗阻的独立危险因素,构建的列线图模型具有较好的准确性和临床应用价值。

Objective To explore the risk factors of dynamic ileus after laparoscopic radical resection of prostate cancer,and toconstruct a nomogram prediction model and to evaluate it.Methods 46 patients who underwent laparoscopic radical prostatectomy in thehospital from January 2022 to June 2023 and developed dynamic intestinal obstruction after surgery were selected as case group,and 92 patients who did not experience dynamic intestinal obstruction after radical prostatectomy during the same period were selected as control group.Record the patient's general information,preoperative and postoperative levels of albumin,C-reactive protein(CRP),tumor necrosis factor alpha(TNF alpha) and interleukin-6(IL-6) with in 24 hours.The researchusedSelf Rating Anxiety Scale(SAS) to evaluate the patient′s preoperative anxiety level and to record the score.A univariate analysis was conducted on the risk factors for postoperative dynamic intestinal obstruction in prostate cancer patients.Statistically significant univariate variables were included in the binary Logistic multiple regression analysis,and a column chart prediction model was constructed using R software.Results Compared with control group,the proportion of the patients in the case group with hyperlipidemia,heart disease,chronic obstructive pulmonary disease,and the use of opioid analgesics,surgical time,preoperative anxiety score,average increase in serum TNF-α and IL-6 levels,and decrease in postoperative albumin levels were statistically significant(P<0.05).The results of binary multivariate Logistic regression analysis showed that heart disease(X_1),use of opioid analgesics(X_2),higher preoperative anxiety score(X_3),lower postoperative serum albumin level(X_4) and higher postoperative serum TNF-α level(X_5) and IL-6 level(X_6) were independent risk factors for postoperative dynamic intestinal obstruction in prostate cancer patients(P<0.05).The constructed Logistic regression model was Log(P)=-14.019+1.336X_1+1.622X_2+0.129X_3-0.145X_4+0.050X_5+0.072X_6.The calibration curve showed good consistency of the column chart model,and the clinical decision curve evaluation results showed that the model had good application value.The ROC curve analysis results indicated that the area under the curve predicted by the model was 0.856(0.812~0.926),with sensitivity and specificity of 75.2% and 76.1%,respectively.Conclusion Heart disease,opioid analgesics,preoperative anxiety score,postoperative albumin,TNF-α and IL-6 levels are independent risk factors for dynamic ilobstruction in patients with prostate cancer after surgery.The established nomograph model has good accuracy and clinical application value.