妇科癌症化疗致周围神经病变患者跌倒风险的Nomogram模型构建与验证

Construction and verification of a Nomogram model for fall risk in patients with chemotherapy-induced peripheral neuropathy for gynecological cancer

目的探讨妇科癌症化疗致周围神经病变患者跌倒的影响因素,基于独立影响因素构建并验证Nomogram模型。方法回顾性选取2021年2月至2023年4月在中国人民解放军陆军军医大学第一附属医院接受妇科癌症化疗且出现周围神经病变的患者216例,使用随机数字表法按7∶3比例分为建模集和验证集,根据建模集中患者有无发生跌倒分为跌倒组和未跌倒组,通过单因素分析及多因素Logistic回归分析得出独立影响因素,并基于回归分析法构建预测模型,使用R语言软件绘制对应的Nomogram图,采用受试者工作特征(ROC)曲线、校准曲线检验预测效能,临床决策曲线评价模型预测收益。结果纳入216例患者中建模集151例,验证集65例;建模集患者中跌倒组53例,年龄(62.89±8.58)岁,未跌倒组98例,年龄(59.48±8.12)岁。化疗周期、下肢神经损害评分、Barthel指数、居家跌倒风险筛查工具评分、规律锻炼习惯、糖尿病、抑郁症均是妇科癌症化疗致周围神经病变患者跌倒的独立影响因素(OR值为0.312~3.727,均P<0.05),构建的Nomogram模型曲线下面积(AUC)为0.873,模型的区分能力良好。校准曲线平均绝对误差(MAE)为0.048,校准曲线贴合理想曲线,表示模型具有较好的校准性能。验证集AUC为0.873,校准曲线MAE为0.073,提示模型具有较好的外部预测效能。结论妇科癌症化疗致周围神经病变患者发生跌倒的Nomogram模型预测效能良好,可为临床预防患者跌倒提供依据。

ObjectiveTo investigate the influencing factors of falls in patients with chemotherapy-induced peripheral neuropathy for gynecological cancer,and establish a Nomogram model based on independent influencing factors.MethodsA total of 216 patients with peripheral neuropathy who received chemotherapy for gynecological cancer in The First Affiliated Hospital of Army Military Medical University of the People′s Liberation Army from February 2021 to April 2023 were retrospectively selected,and were divided into the modeling set and the validation set according to the ratio of 7∶3 using random number table method.According to the occurrence of falls in the modeling set,the patients were divided into the fall group and the non-fall group.Independent influencing factors were obtained through univariate analysis and multivariate Logistic regression analysis,and a prediction model was constructed based on the regression analysis.A Nomogram map was drawn using R language software.Receiver operating characteristic curve(ROC)and calibration curve were used to test the prediction efficiency,and the predictive benefits were evaluated by clinical decision curve.ResultsAmong the 216 patients,151 were included in the modeling set and 65 in the validation set.Among the patients in the modeling set,53 cases were in the fall group,aged(62.89±8.58)years old,and 98 cases were in the non-fall group,aged(59.48±8.12)years old.Chemotherapy cycle,Neuropathy Impairment Score-Lower Limbs,Barthel index,Home Falls and Accidents Screening Tool,regular exercise habits,diabetes and depression were all independent influencing factors for falls in patients with chemotherapy-induced peripheral neuropathy of gynecologic cancer(OR values were 0.312-3.727,all P<0.05).The area under the curve(AUC)of the established Nomogram model was 0.873,with good discriminating ability.The mean absolute error(MAE)of calibration curve was 0.048,which fitted the ideal curve,indicating that the model has good calibration performance.The AUC of the validation set was 0.873 and the MAE of calibration curve was 0.073,indicating that the model has good external prediction efficiency.ConclusionsThis Nomogram model can predict falls in patients with chemotherapy-induced peripheral neuropathy of gynecological cancer and provide evidence for clinical prevention of falls.