Statistical Methods In Clinical Efficacy Evaluation Cohort Study Of Traditional Chinese Medicine And Its Application In Chronic Heart Failure

Objective1.In view of the problems in the application of statistical methods,this paper puts forward the selection ideas of statistical methods in the clinical efficacy evaluation cohort study of traditional Chinese medicine,and provides methodological suggestions for related studies.2.To establish a cohort study of patients with chronic heart failure in traditional Chinese medicine hospital,so as to provide reference for the use of statistical methods in the cohort study of clinical efficacy evaluation of traditional Chinese medicineMethod1.Preliminary establishment of methodological suggestions for statistical analysis in clinical efficacy evaluation cohort study of traditional Chinese MedicineThrough searching the related literature of clinical efficacy evaluation,cohort study and statistics in HowNet,Wanfang,PubMed and other databases,consulting the works and books on the application of statistical methods,understanding the principles of statistics,understanding the existing problems of statistical methods in the current clinical efficacy evaluation of traditional Chinese medicine,summarizing and summarizing the statistical knowledge in the literature,works and books,and comparing with the clinical efficacy of traditional Chinese Medicine Combining the characteristics of evaluation and cohort study,according to the existing problems,this paper gives methodological suggestions for statistical analysis in the cohort study of clinical efficacy evaluation of traditional Chinese medicine,and makes improvement and modification under the guidance of statistics professional teachers and clinical experts of traditional Chinese medicine.2.Application of statistical methods in the cohort study of chronic heart failure patients in traditional Chinese medicine hospitalEstablish the research queue of chronic heart failure in traditional Chinese medicine hospital:relying on the medical record management system of seven Class-A hospitals of Shandong traditional Chinese medicine hospital,Weifang traditional Chinese medicine hospital,Rizhao traditional Chinese medicine hospital,Jinan traditional Chinese medicine hospital,Qingdao Haici medical group,Zibo traditional Chinese medicine hospital and Weihai traditional Chinese medicine hospital,the admission time meeting the inclusion and exclusion criteria will be from January 1,2012 to December 31,2017 Patients with failure were put into the study queue,and the basic information,treatment drugs,end events and other data of patients during hospitalization and follow-up were collected.The database was established and data was entered with epidate software.Based on the collected cohort study data,describe the basic characteristics of patients with chronic heart failure in 7 TCM hospitals during hospitalization,measure the mean,standard deviation and percentage of grade data,and show them by statistical table and chart;then analyze and end-point by logistic regression and survival analysis according to the occurrence of end-point events during hospitalization and follow-up Event related factors,and the results were analyzed and compared.Among them,propensity score method was used to balance the baseline differences between the exposure group and the control group during the follow-up period,and to compare the balanced results.So as to show the practical application of statistical methods in the cohort study of chronic heart failure.Result1.Cohort study is conducive to reflect the characteristics of clinical efficacy evaluation of traditional Chinese medicine,but there are some problems in the application of statistical methods,such as misuse,not screening single factor before multivariate analysis,not considering the impact of time on the outcome.The selection of statistical methods needs to consider research purposes,research methods,data types and other aspects.When balancing the inter group baseline,the known confounding factors can be selected as propensity score,stratified analysis and multi factor analysis:propensity score is suitable for the team research with more confounding factors and large sample size,stratified analysis and multi factor analysis are suitable for the team research with less confounding factors and not particularly large samples To deal with unknown and known confounding factors,tool variable method can be used.In statistical inference,there are many statistical methods that can be used.In the application of single factor analysis,attention should be paid to the normality and homogeneity of variance of the data;logistic regression analysis and Cox regression are applicable to the multi factor analysis with the outcome variable of two categories of data.The difference between them is that the latter considers the influence of time on the survival outcome;the generalized linear mixed model does not require the data type,And can handle missing data.2.A total of 2691 patients with chronic heart failure were included in the cohort study,and 2591 patients were followed up.Using logistic regression,we found that NYHA grade and kidney disease were the influencing factors of cardiac death during hospitalization;propensity score balanced the baseline between exposure group and control group,NYHA grade and age before balance were the influencing factors of cardiac death during follow-up,and only NYHA grade after balance was the influencing factors of cardiac death during follow-up.Conclusion1.Statistical methods such as propensity score,stratified analysis and multi factor analysis have been applied in the clinical efficacy evaluation of traditional Chinese medicine;2.The treatment of traditional Chinese medicine can reduce the incidence of cardiac death in patients with chronic heart failure and prolong the survival period.