资源描述:
《社区老年冠心病患者死亡危险因素评估》由会员上传分享,免费在线阅读,更多相关内容在工程资料-天天文库。
1、社区老年冠心病患者死亡危险因素评估【摘要】目的:探究社区老年冠心病患者死亡危险因素,为相关医务人员提供参考。方法:选择社区350名老年冠心病患者作为研究对象,系统收集疾病的相关信息。首先进行单因素Logistic回归分析,筛选有意义的变量作为潜在指标,将入选的变量作为协变量进行逐步向前多因素Logistic回归分析,制作模型预测ROC曲线并进行检验,同时对模型进行Hosmer・Lemeshow(H)拟合优度检验。结果:单因素Logistic回归分析共有15项指标纳入,相关系数B在0.02到0.48之间,Exp(B)即OR值在1.02~1.62范围内。逐步向前多因素Logistic
2、回归分析共纳入8项指标,B值范围为0.96~2.25,Exp(B)范围为2.61~9.49,Logistic回归建模以后的预测符合度图形ROC曲线下面积为0.88,Hosmer-Lemeshow(H)检验为P二0.39。结论:Logistic回归显示以下8项指标为老年冠心病死亡的危险因素:配偶是否健在、周围血管病变、脑血管病变、IgG、冠心病史、糖尿病史、肾功能不全、缺乏锻炼,本次回归模型的拟合效果较好。【关键词】老年冠心病;死亡;危险因素;Logistic回归分析;社区Assessmentofriskfactorsofdeathinelderlypatientswithcoro
3、naryheartdiseaseinthecommunityAbstract[Objective]:Toexploretheriskfactorsofdeathinelderlypatientswithcoronaryheartdiseaseincommunity,andproviderefereneeforrelevantmedicalstaff.[Methods]:Weselectedcommunityof350elderlypatientswithcoronaryheartdiseaseastheobjectofstudy,systematicallycollectingt
4、hedisease-relatedinformation.First,singlefactorLogisticregressionanalysiswasperformedtoselectmeaningfulvariablesaspotentialindicators,andtheselectedvariablesascovariateswereanalyzedbystepwiseforwardmultipleLogisticregressionanalysis.TheROCcurvewaspredictedandtested,andthegoodnessoffitofthemod
5、elwastestedbyHosmer-Lemeshow(H).[Results]:Atotalof15indexeswereineludedinthesinglefactorLogisticregressionanalysis.ThecorrelationcoefficientBwasbetween0.02and0.48,andExp(B)waswithintherangeof1.02~1.62.8indexeswereineludedinthestepwisemultiplefactorLogisticregressionanalysis.TherangeofBvaluewa
6、s0.96to2.25,andtherangeofExp(B)was2.61~9.49.AftertheLogisticregressionmodeling,theareaundertheROCcurveis0.88,andtheHosmer-Lemeshow(H)testisP=0.39.[Conclusion]:Logisticregressionshowedthatthefollowing8indicatorswereriskfactorsforthedeathofelderlycoronaryheartdisease:whetherthespousewashealthy,
7、peripheralvasculardisease,cerebrovasculardisease,IgG,coronaryheartdiseasehistory,diabeteshistory,renalinsufficiency,andlackofexercise.Thefittingeffectofthisregressionmodelisfit.[Keywords]Coronaryheartdiseaseintheelderly;death;riskfactors;Logi