Prediction Of Aged Disability And Dementia And Study On Financial Sustainability Of Long-term Care Insurance

With the increasing trend of aging population,our government needs to establish and improve a social long-term care insurance system.This paper forecasts the changes in the income and expenditure of the long-term care insurance fund,taking the long-term care insurance benefits and financing mechanism implemented in Chengdu on July 1,2022 as an example,to put forward effective policy suggestions and data reference for the development and improvement of China’s long-term care insurance system.At the same time,the prediction model of disability and dementia in the elderly constructed by the paper can provide more effective theoretical framework and policy oriented support for the formulation of long-term care insurance system policies.Specifically,in this paper,based on CHARLS database,the Barthel Index Rating Scale and the MMSE scale were used as indicators to measure dysfunctional dementia.The disabled and dementia population is divided into six groups: mild disability,moderate disability,severe disability,mild dementia,moderate dementia,and severe dementia.Then,four machine learning algorithms,including multiple linear regression,support vector regression,random forest and XGBoost,were used to build prediction models for elderly disability and dementia.Finally,the model with the best prediction accuracy and stability was used to analyze the factors affecting elderly disability and dementia,and the combination of factors with the greatest influence on elderly disability and dementia was obtained.In the supply and demand forecast part of the long-term care fund,this paper firstly calculates the population of the disable elderly at all ages from 2020 to 2070 by referring to the 2019 World Population Outlook issued by the United Nations.Secondly,based on the Guiding Principles for Graded Nursing in General Hospitals(Trial)and the China Statistical Yearbook over the years,and fully considering the factors such as nursing prices and wage growth,this paper calculates the expenditure of different levels of disability in urban and rural areas from 2020 to 2070.Then,based on Chengdu Long term Care Insurance Implementation Rules issued by Chengdu Medical Insurance Bureau in 2022,and taking full account of inflation,the annual balance of China’s long-term care insurance fund from 2020 to 2007 under this system was calculated.Finally,this paper will conduct sensitivity tests on the subsidy standards and payment standards of the long-term care insurance system to further explore a financial sustainable long-term care insurance system in line with China’s national conditions.The main contributions and innovations of this paper are as follows: Firstly,this paper introduces the XGboost algorithm in machine learning to establish a prediction model for the state of disability and dementia of the elderly in China,and obtains the most important combination of factors affecting the state of disability and dementia of the elderly,which to some extent simplifies the prediction of the financial sustainability of China’s long-term care insurance fund in the future.Secondly,based on Chengdu Long-term Care Insurance Implementation Rules issued by the Chengdu Medical Insurance Bureau in 2022,this paper forecasts the income and expenditure of the long-term care insurance fund and the annual balance,ensuring the representativeness and reliability of the prediction results.