糖尿病足患者溃疡愈合风险的列线图预测模型

Nomogram prediction model for ulcer healing risk in diabetic foot patients

目的探讨预测糖尿病足溃疡(DFU)患者溃疡愈合风险的列线图模型。方法收集2021年1―12月于新疆医科大学第一附属医院就诊的200例DFU患者的临床资料,依据DFU是否愈合将其分为愈合组(n=57)和未愈合组(n=143)。收集患者的年龄、性别、体重指数(BMI)、糖化血红蛋白、知-信-行(KAP)模式量表评分以及患者血液中微小RNA(miRNA)-15b-5p的相对含量,采用Logistic回归模型对临床特征数据进行单因素及多因素分析,向后法筛选出相关预测因子并建立预测模型,使用增强Bootstrap验证法对模型内部进行验证,使用一致性指数(C-index)、列线图、校准图、受试者工作特征(ROC)曲线及校准曲线、决策曲线分析(DCA)进行评估。结果通过二元Logistic回归分析及向后法筛选变量,从8个总变量中确定了3个预测因子,即年龄、miRNA-15b-5p相对含量、K评分,其构建的预测模型呈现出中等预测能力,ROC曲线下面积(AUC)为0.923。结论年龄、K评分及miRNA-15b-5p相对含量均为DFU患者溃疡未愈合相关的预测因子,基于这些预测因子建立的预测模型具有优秀的综合性能,可从多个层面预测DFU患者溃疡未愈合的发生风险,为DFU患者溃疡愈合提供了个体化治疗依据。

Objective To explore the nomogram prediction model for ulcer healing risk in diabetes foot ulcer(DFU)patients.Method Clinical data of 200 patients with DFU treated in the First Affiliated Hospital of Xinjiang Medical University from January to December 2021 were collected.According to whether DFU healed or not,they were divided into healing group(n=57)and non-healing group(n=143).Age,gender,body mass index(BMI),glycosylated hemoglobin,knowledge-attitude-practice(KAP)model scale score and the relative content of microRNA(miRNA)-15b-5p in the two groups were collected.Logistic regression was used to carry out univariate and multi-factor analysis of clinical feature data,the relevant predictors were screened backward,the prediction model was established.The enhanced Bootstrap validation method was used to verify the model internally.Evaluation was performed using C-index,nomogram,calibration diagram,receiver operator characteristic(ROC)curve and calibration curve,and decision curve analysis(DCA).Result By using binary Logistic regression analysis and backward method to screen variables,three predictors were identified from the eight total variables,namely,age,miRNA-15b-5p relative content and K score.The predictive model constructed by this method showed moderate predictive ability,and the area under curve(AUC)of ROC was 0.923.Conclusion Age,K score and the relative content of miRNA-15b-5p are all predictors of non-healing of ulcers in DFU patients.The prediction model established based on these predictors had excellent comprehensive performance,and could predict the risk of non-healing of ulcers in DFU patients from multiple levels,providing individualized treatment basis for ulcer healing in DFU patients.