基于影像学指标构建永久性乙状结肠造口旁疝发病风险预测模型

Development of a nomogram prediction model for the risk of permanent sigmoid parastomal hernia based on radiologic predictors

目的对永久性乙状结肠造口旁疝的危险因素进行分析,并基于这些风险因素,构建并验证永久性乙状结肠造口旁疝发病风险的列线图预测模型,以期为造口患者提供精准的个体化治疗方案提供依据。方法回顾分析2010年1月至2018年1月在青岛大学附属医院接受永久性乙状结肠造口术的291例患者的病历资料,构建列线图预测模型;同时,回顾性收集2010年1月至2018年1月在青岛市立医院接受永久性乙状结肠造口患者63例用于预测模型的外部验证。采用单因素分析和多因素logistic回归分析筛选永久性乙状结肠造口术后发生造口旁疝的独立危险因素,基于独立危险因素建立列线图预测模型。使用Bootstrap方法进行内部验证并在验证组进行外部验证。结果2010年1月至2018年1月,291例患者接受了永久性乙状结肠造口术,2年内61例患者发生造口旁疝,230例患者未发生造口旁疝。经多因素Logistic回归分析发现,年龄(OR 1.063,95%CI 1.014~1.114,P=0.011)、性别(OR 3.927,95%CI 1.327~11.621,P=0.013)、体重指数(OR 1.553,95%CI 1.228~1.963,P=0.001)、白蛋白(OR 0.876,95%CI 0.783~0.981,P=0.022)、造口直径(OR 4.426,95%CI 1.976~9.915,P=0.001)、皮下脂肪指数(OR 1.028,95%CI 1.006~1.050,P=0.011)、腹直肌指数(OR 0.591,95%CI 0.352~0.993,P=0.047)是永久性乙状结肠造口患者发生造口旁疝的独立危险因素。基于独立危险因素建立的列线图预测模型,训练组和验证组的受试者工作特征曲线下面积分别为0.941和0.887,Hosmer-Lemeshow检验(P>0.05)和校准曲线显示实际概率与预测概率具有良好的一致性。结论根据永久性乙状结肠造口患者发生造口旁疝的独立危险因素建立的列线图预测模型,对预测永久性乙状结肠造口旁疝的发生具有较高的临床意义。为专业外科医生进行造口手术的患者提供了个性化和疾病相关评估。

Objective To analyze risk factors for permanent sigmoid parastomal hernias(PSH),and to develop and validate a nomogram prediction model for the risk of PSH based on these risk factors,in order to provide precise individualized treatment plan for stoma patients.Methods The medical records from January 2010 to January 2018 of 291 patients who underwent permanent sigmoid colostomy surgery at Qingdao University Affiliated Hospital were retrospectively analyzed to construct a nomogram prediction model.Simultaneously,we retrospectively collected data of 63 patients who received permanent sigmoid colostomy at Qingdao Municipal Hospital during the same period for external validation of the model.Independent risk factors for parastomal hernia after permanent sigmoid colostomy were identified using univariate and multivariate logistic regression analysis.The nomogram prediction model was established based on these risk factors.Internal validation was conducted using the Bootstrap method,and external validation was performed in the validation group.Results From January 2010 to January 2018,291 patients underwent permanent sigmoid colostomy,61 patients developed parastomal hernia within 2 years and 230 patients did not.Results of multivariate logistic regression analysis showed that independent risk factors of developing parastomal hernia in patients with permanent sigmoid colostomy including age(OR 1.063,95%CI 1.014-1.114,P=0.011),sex(OR 3.927,95%CI 1.327-11.621,P=0.013),body mass index(OR 1.553,95%CI 1.228-1.963,P=0.001),serum albumin levels(OR 0.876,95%CI 0.783-0.981,P=0.022),stoma diameter(OR 4.426,95%CI 1.976-9.915,P=0.001),subcutaneous fat index(OR 1.028,95%CI 1.006-1.050,P=0.011),and rectus abdominis index(OR 0.591,95%CI 0.352-0.993,P=0.047).The nomogram prediction model based on these independent risk factors,exhibited an area under the receiver operating characteristic curve of 0.941 for the training group and 0.887 for the validation group.The Hosmer-Lemeshow test(P>0.05)and calibration curves underscored the commendable alignment between actual and predicted probabilities.Conclusion Based on the independent risk factors for the occurrence of parastomal hernia in patients with permanent sigmoid colostomy,the established nomogram prediction model has significant clinical significance in predicting the occurrence of parastomal hernia.This model provides personalized and disease-specific assessments for patients undergoing ostomy surgery by professional surgeons.