Dynamic Trajectory Of Patient Reported Outcomes With Heart Failure Based On Growth Mixture Model

Objective:With the improvement of people's economic level,the unhealthy life style becomes more and more prominent.With the improvement of medical level and the acceleration of social aging,the population of cardiovascular disease patients is becoming larger and larger.Chronic heart failure,as one of the final outcomes of cardiovascular disease,brings great burden to patients and their families.The patient's life quality is relatively low and the treatment cost is expensive.In this study,patients' reported clinical outcome(pro)was introduced to evaluate the quality of life of patients.The multiple level model was used to analyze the influencing factors of the repeated measurement data of pro in the longitudinal follow-up study.Furthermore,the growth mixed model(GMM)was used to identify the sub groups of potential patients with different characteristics,so as to solve the problem of data heterogeneity and reveal the change track of Pro after hospital In order to improve the quality of life of patients and provide theoretical support for relevant departments to propose group intervention measures.Methods:In the First Affiliated Hospital of Shanxi Medical University and the Cardiovascular Hospital of Shanxi Province,patients who were definitely diagnosed as heart failure from January 1,2014 to October 31,2019 were selected as the subjects of this study.The information of Pro scale was collected at the time of hospitalization,3 months after discharge,6 months after discharge and 12 months after discharge.Through the two-level model,the influencing factors of the total score of pro,physiological,psychological,social and treatment scores at the individual level and in time point level are analyzed;through the growth mixed model,the real change track of the total score of pro,physiological,psychological,social and treatment scores is explored,the potential patient groups with different quality of life are identified,and the significant variables in the influencing factor analysis are further fitted As a covariate,it was incorporated into the growth mixed model to determine the real change trajectory of different patientsResults:Two,three and four potential subclass models of linear,quadratic and undefined curve types were fitted on the total scores of pro,physiological,psychological,social and therapeutic fields respectively.From the fitting index results,we can see that the optimal unconditional GMM models of the five scores are respectively two potential subclass models of linear,two potential subclass models of undefined curve types and two potential subclass models of undefined curve types Potential subclass model,linear two potential subclass model and three potential subclass models of undefined curve type;the individual level variables that have the greatest impact on each score in the two-level model are incorporated into the unconditional GMM model as covariates,which are age,age,gender,family income and past history,respectively.The results after incorporation into the covariates show that:according to the total score of pro,patients are divided into two categories Class A:79 in the slow rising group,27.431%;209 in the fast rising group,72.569%.According to the score of physiological field,the patients were divided into two categories:197 in the fast rising group,68.403%;91 in the slow rising group,31.597%.According to the score of psychological field,the patients were divided into two groups:280 in the decrease group,97.222%;8 in the increase group,2.778%According to social domain scores,patients were divided into two groups:277 in the rising group,96.181%;11 in the declining group,3.819%.According to the score of treatment field,according to the fitting results,patients should be divided into three categories,but because the number of one category is too small,two categories of models are selected,patients are divided into two categories:48 in the decline group,16.667%;240 in the fast rise group,83.333%Conclusion:In this study,by exploring the subgroups of patients with different characteristics in patients with heart failure,we can identify the group heterogeneity,understand the change trend of individual quality of life,explore the change track rule of pro with time after discharge,and further explain the differences between individuals,which is helpful to guide patients in different stages of life course to give personalized interventions,improve the prognosis of patients,and improve health The quality of storage is of certain significance.