基于SMOTE的髋部骨折病人术后便秘预测模型的构建

打开文本图片集

Abstract  Objective：To construct constipation risk prediction model for hip fracture patients after surgery based on synthetic minority over⁃sampling technique（SMOTE）.Methods：A total of 130 hip fracture patients after surgery admitted to our hospital from January 2020 to December 2021 were selected as research subjects.According to whether patients had constipation within 30 days after surgery or not，they were divided into constipation group（41 cases） and non⁃constipation group（89 cases）.The data of the two groups of patients were collected.The independent risk factors of constipation after surgery were screened by univariate analysis and binary Logistic regression analysis，and a Logistic regression model（P1） was constructed.The data set was improved based on SMOTE，and a Logistic prediction model（P2） based on SMOTE was constructed.Prediction efficiency of the model were tested.Results：Age（X1），glycosylated hemoglobin level（X2），whether combined use opioid analgesics or not（X3），time in bed after operation（X4） and first feeding time after operation（X5） were independent influence factors of constipation for hip fracture patients after surgery（P<0.05），P2=1／［1+e-（-35.244+0.163X +1.469X +1.562X +0.313X +0.340X ）］，the coefficient of determination of P2 was 0.904，and the area under receiver operator characteristic was 0.936，both higher than P1 （coefficient of determination was 0.878，and the area under receiver operator characteristic was 0.752）.Conclusions：The constipation risk prediction model for hip fracture patients after surgery based on SMOTE has higher prediction efficiency on constipation risk for hip fracture patients after surgery.

Keywords  hip fracture； constipation； risk prediction model； nursing

摘要  目的：基于人工少数类过采样法（SMOTE）构建髋部骨折病人术后便秘预测模型。（剩余9395字）