摘要
目的探究老年糖尿病周围神经病变(DPN)患者发生低血糖感知受损的影响因素,并构建列线图预测模型。方法便利抽样法选取2021年3月至2023年12月苏州市立医院收治的老年DPN患者。采用一般资料调查表、低血糖感知受损(Gold评分法)、中文版低血糖恐惧调查-忧虑量表(CHFSⅡ-WS)进行调查。通过多因素Logistic回归分析低血糖感知受损的影响因素;R软件构建列线图预测模型;ROC曲线评估模型预测价值。结果共发放188份问卷,有效问卷182份,有效回收率96.81%。182例患者中发生低血糖感知受损59例(32.42%),低血糖感知受损组患者DPN中重度、有严重低血糖史、未接受低血糖预防教育的占比高于非低血糖感知受损组,CHFSⅡ-WS评分也高于非低血糖感知受损组,差异有统计学意义(P<0.05)。多因素Logistic回归分析显示,DPN中重度(OR=7.457,95%CI=2.123~26.194)、严重低血糖史(OR=7.597,95%CI=1.495~38.608)、未接受低血糖预防教育(OR=6.299,95%CI=1.741~22.795)、CHFSⅡ-WS评分(OR=2.189,95%CI=1.685~2.844)是老年DPN患者发生低血糖感知受损的危险因素(均P<0.05)。基于DPN严重程度、严重低血糖史、低血糖预防教育、CHFSⅡ-WS评分共四项指标构建的列线图模型Hosmer-Lemeshow检验卡方值4.254,P值0.750,拟合度、校准度较好;且列线图模型预测老年DPN患者发生低血糖感知受损的AUC为0.905(95%CI=0.862~0.949),敏感度83.1%,特异度82.1%。结论DPN中重度、严重低血糖史、未接受低血糖预防教育、低血糖恐惧会增加老年DPN患者低血糖感知受损的发生风险,基于四项指标构建的列线图模型对低血糖感知受损的发生具有良好预测效能。
Objective To explore the influencing factors of impaired awareness of hypoglycemia in elderly patients with diabetic peripheral neuropathy(DPN),and to construct a prediction model of nomogram.Methods The elderly DPN patients admitted to Suzhou Municipal Hospital from March 2021 to December 2023 were selected by convenient sampling method.General data questionnaire,impaired perception of hypoglycemia(Gold scoring method)and Chinese version Hypoglycemia Fear SurveyⅡ-Worry Scale(CHFSⅡ-WS)were used for investigation.Multivariate Logistic regression was performed to analyze the influencing factors of impaired hypoglycemia awareness.The R software was used to construct a nomogram prediction model,and the ROC curve was used to evaluate the predictive value of the model.Results A total of 188 questionnaires were distributed,182 were valid,with an effective recovery rate of 96.81%.Among the 182 patients,59 cases(32.42%)had impaired awareness of hypoglycemia.The proportion of patients in the impaired awareness of hypoglycemia group with moderate to severe DPN,history of severe hypoglycemia and no hypoglycemia prevention education was higher than that in the non-impaired awareness of hypoglycemia group,and the CHFSⅡ-WS score was also higher than that in the non-impaired awareness of hypoglycemia group(P<0.05).Multivariate Logistic regression analysis showed that DPN being moderately severe(OR=7.457,95%CI=2.123-26.194),having a history of severe hypoglycemia(OR=7.597,95%CI=1.495-38.608),not receiving hypoglycemia prevention education(OR=6.299,95%CI=1.741-22.795)and CHFSⅡ-WS score(OR=2.189,95%CI=1.685-2.844)were risk factors for impaired awareness of hypoglycemia in elderly DPN patients(all P<0.05).The Hosmer-Lemeshow test chi-square value of the nomogram model constructed based on four indicators:DPN severity,history of severe hypoglycemia,hypoglycemia prevention education,and CHFSⅡ-WS score,was 4.254 with a P-value of 0.750,indicating good fit and calibration.Moreover,the nematic model predicted that the AUC of impaired awareness of hypoglycemia in elderly DPN patients was 0.905(95%CI=0.862-0.949),sensitivity was 83.1%,specificity 82.1%.Conclusions Moderate to severe DPN,history of severe hypoglycemia,lack of hypoglycemia prevention education,and fear of hypoglycemia increase the risk of impaired awareness of hypoglycemia in elderly DPN patients.The nomogram model built based on the four indicators has a good predictive effect on impaired awareness of hypoglycemia.
作者
吴丽娟
王雪
潘智怡
WU Lijuan;WANG Xue;PAN Zhiyi(Geriatrics Department,Suzhou Municipal Hospital,Suzhou 215000,China)
出处
《医药高职教育与现代护理》
2025年第3期248-254,共7页
Medical Higher Vocational Education and Modern Nursing
关键词
糖尿病周围神经病变
低血糖感知受损
影响因素
列线图
预测模型
Diabetic peripheral neuropathy
Impaired awareness of hypoglycemia
Influencing factors
Nomogram
Prediction model
