Multimorbidity Pattern Mining And Prevention And Control Mode Of Chronic Diseases In The Elderly

Objective:With the rapid development of economy and the improvement of medical and health services,people's life expectancy is gradually extended.Multimorbidity is a common phenomenon in elderly patients with chronic diseases.Multimorbidity has become an important hidden danger that threatens human survival and health.Compared with a single chronic disease,multimorbidity leads to a decline in the quality of life of patients,an increase in the risk of adverse drug events,an increase in the risk of death,and an increase in the consumption of medical resources,which brings many challenges to the prevention and management of chronic diseases.However,the current management of chronic diseases only focuses on single disease management,which is inconsistent with the common phenomenon of multimorbidity in elderly people with chronic diseases.It is of great significance to carry out the research on multimorbidity pattern mining to understand the current situation of multimorbidity and improve the comprehensive prevention and control system of chronic diseases.Nowadays,in the context of medical big data,how to mine multimorbidity patterns and extract valuable information to assist multimorbidity management has become an urgent problem.Therefore,based on data mining methods and techniques,this study has carried out an in-depth discussion on the theories and methods of multimorbidity pattern mining and utilization,which could provid data and theoretical support for the management and prevention and control of multimorbidities.Methods:(1)Through the investigation of relevant domestic and foreign literature,the relevant theoretical basis of the research in this paper is clarified,and the related concepts of multimorbidity are sorted out and defined.Discuss the types of influencing factors of multimorbidities and their mechanism.Clarify the process of data-driven multimorbidity management decision-making.Based on the DIKW system of information management,a multimorbidity pattern mining and utilization model is constructed.(2)Based on the basic idea of the complex network theory,a multimorbidity pattern mining method integrating influencing factors is proposed.Use chi-square test and Logistic regression model to analyze the influencing factors of multimorbidities,extract disease-related information and disease-influencing factor-related information to establish a multimorbidity network that integrates influencing factors,and on this basis,introduce overlapping community discovery algorithms to identify community structures and discover common diseases mode.(3)Combining the results of multimorbidity pattern mining,based on the basic ideas of chronic disease management theory and collaborative management theory,construct a multimorbidity prevention and control mechanism in the elderly,and put forward specific suggestions to provide a theoretical basis for government departments to formulate public health policies.Results:(1)The occurrence and development of multimorbidities are affected by biological factors,psychological factors,social environmental factors,and life behavior factors.A petal model of multimorbidity influencing factors is constructed,which is composed of biological factors,psychological factors,social environmental factors and life behavior factors.And the interaction between biological factors,psychological factors,social environmental factors and life behavior factors are systematically explained.(2)The mining and utilization model of multimorbidity pattern based on the DIKW system is constructed.The mining and utilization model of multimorbidity pattern is composed of multimorbidity data(data layer),association information in multimorbidity(information layer),multimorbidity pattern(knowledge layer)and multimorbidity Management(wisdom layer).The transformation between basic elements is realized through data-driven.(3)This paper selects CHARLS 2015 project of China Health and pension tracking survey as the data source,and uses the multimorbidity pattern mining method proposed in this study to identify the multimorbidity patterns of the elderly patients with chronic diseases in China.The results of influencing factor analysis showed that age,pain,sleep time,basic activities of life disorder,obesity,depression and life satisfaction were the influencing factors of multimorbidity in elderly chronic patients;The identified multimorbidity patterns included 5 communities,including{hypertension,dyslipidemia,diabetes mellitus,heart disease,24?BMI<28,BMI?28},{chronic lung disease,asthma,70-79 years old,BMI<18.5,general satisfaction with life,sleep time <4 hour},etc.(4)Through the analysis of the current situation of multimorbidty management,we found that there are some problems such as imperfect system and asymmetric information.Combined with the results of multimorbidity pattern mining,we proposed the basic principles of whole person management,classified guidance,prevention first and comprehensive coordination,constructed the multimorbidity prevention and control mode of chronic diseases,and analyzed the responsibilities of the government,medical institutions,communities and patients in multimorbidity management,and put forward the main measures for the prevention and control of multimorbidities.Conclusions:(1)The mining and utilization model of multimorbidity pattern includes data collection and preprocessing,multimorbidity association relationship extraction,multimorbidity pattern mining and multimorbidity management decision-making.Its core task is to use multimorbidity pattern mining results to drive multimorbidity management decision-making.(2)The proposed multimorbidity pattern mining method integrating influencing factors can not only discover the association relationship between diseases,but also the potential association between multimorbidity patterns and influencing factors.(3)Using the proposed multimorbidity pattern mining method integrating influencing factors,identify the multimorbidity pattern of Chinese elderly people with chronic diseases.(4)Based on the results of multimorbidity pattern mining,it can provide services for multimorbidity management decision-making and optimize multimorbidity prevention and control strategies.