Establishment Of Early Warning Model Of Heart Failure And Evaluation Of Association Rules Of TCM Syndromes

Objective:1.Based on the literature retrieval method,determine the high risk factors of heart failure,construct a heart failure Nomogram early warning model,and evaluate its application value.2.We use the method of reviewing the medical records of heart failure to analyze the association rules between symptoms of TCM syndromes in heart failure.Methods:1.The establishment of the Nomogram model:By searching for risk factors or risk factors of heart failure,summarizing the documents that meet the criteria,and summarizing the risk factors of heart failure according to general information,basic diseases,laboratory tests,and special examination classification methods.Then use SPSS22.0 statistical software to analyze the indicators by univariate analysis,screen the obtained indicators for binary logistic regression analysis,use the ROC curve to determine the threshold of early warning indicators,establish a Nomogram model,and apply DCA curve evaluation.2.Association rules of TCM syndromes and symptoms:Check the medical records from 2016 to the present,statistically analyze the TCM syndromes with high frequency of heart failure,conduct frequency statistics on the symptoms of each syndrome,and screen the top 10 TCM symptoms with high frequency of occurrence.Then use the Apriori algorithm in SPSS Modeler 18.0 software to analyze the high-frequency symptoms,and draw a network diagram between the symptoms to explore the potential correlation between the high-frequency symptoms of heart failure and the syndrome.Results:1.Binary logistic regression analysis results:NT-pro BNP(OR=1.002,95%CI:1.001-1.004)、 BNP(0R=1.009,95%CI : 1.004-1.013)、 RDW(0R=21.893,95%CI :96.008-4.992)、AGE(0R=1.135,95%CI:0.998-1.290)、QTd(0R=1.224,95%CI:1.349-1.110)、CA-125(0R=1.099,95%CI:1.044-1.158),These factors are independent risk factors for heart failure(P<0.10).We constructed an early warning model of heart failure based on the results of logistic regression analysis:Logit P=NT-pro BNP+3.5BNP+ 1018RDW+82QTd+43CA-125;The ROC curve analysis results are as follows:NT-pro BNP、BNP、RDW、AGE、QTd and CA-125 have significant predictive implications for heart failure,and the combined prediction of each factor is significantly better than the single-index independent prediction;We determined the critical point of each forecast based on the most approximate index:NT-pro BNP(834.245ng/L)、BNP(277.225pg/m L)、RDW(14.020%)、QTd(57.725ms)、AGE(62.56 years old)、CA-125(25.305U/ml);The C-index of the constructed Nomogram scoring model is 0.997,and initially established a heart failure warning model.2.TCM syndrome symptoms association rules display:There were 33 cases of heart and kidney Yang deficiency,which constituted 66 symptoms and tongue veins.The main symptoms were fatigue,palpitations,chest tightness,edema of both lower extremities,oliguria,dark purple tongue.Among them,there was a close relationship between palpitation and edema of both lower limbs.There were 24 cases of qi and Yin deficiency,which constituted 77 symptoms and tongue veins.The main symptoms were bosom frowsty,lack of power,pulse is fine,breath is short,tongue is dark red.Among them,palpitations,shortness of breath,chest tightness and fatigue are closely related.There were 21 cases of Yang deficiency and water flooding syndrome,which constituted 55 symptoms and tongue veins.Chest tightness,edema of both lower limbs,shortness of breath,palpitation and wheezing were the main symptoms,and the association between the edema of both lower limbs and chest tightness was the strongest.There were 20 cases of qi deficiency and blood stasis syndrome,which formed 51 symptoms and tongue veins.The main symptoms were asthenia,tongue purple and dark,chest tightness,wheezing and shortness of breath.Among them,the relationship between pant,asthenia and tongue purple and dark is close,and the correlation between asthenia,shortness of breath and chest tightness is strong.There were 18 cases of heart-yang deficiency syndrome,which constituted 58 symptoms and tongue veins.The heart palpitations,shortness of breath,chest tightness,dark purple tongue and edema of both lower extremities were the main symptoms.The relationship between shortness of breath,palpitations and chest tightness was relatively close.The relationship between white,shortness of breath and edema of both lower extremities is relatively close.There were 17 cases of blood stasis syndrome,which constituted 59 symptoms and tongue veins.The tongue was purple and dark,the lower extremity edema,fine pulse,shortness of breath and pulse Shen were the main symptoms.Shortness of breath,edema of both lower extremities,dark purple tongue,fine veins,and deep pulse are closely related and closely related.Conclusion:1.We initially established a heartbeat Nomogram early warning model to predict the risk probability of heart failure;2.According to the confidence and support of the association rules,we form a combination of symptoms with strong correlation between symptoms of heart failure syndrome,and construct a correlation model to provide reference for clinical syndrome diagnosis.