A Study On The Probability Of Health State Transition In The Elderly In China

Currently,China is facing an increasingly serious problem of population aging,with both the number of disabled people and the proportion of disabled people increasing.At the same time,in a specific era,China’s family structure has shown a trend of miniaturization,coupled with the greater pressure on young people to live,and the elderly care function of families is gradually weakened.In this context,the Chinese government and society pay high attention to the living conditions of the elderly,especially the long-term care needs faced by the elderly after their disability.Both the design of long-term care security systems in society and the calculation of long-term care insurance premiums need to be based on accurate health status transition probabilities.For this purpose,this article uses relevant theoretical knowledge such as insurance and actuarial science to study how to calculate a more accurate transition probability of health status.As the basis of research on issues such as the health of the elderly,different scholars have estimated the health transition probability of the elderly using China’s micro database containing health status information of the elderly and establishing relevant state transition probability estimation models,but the results have significant differences.After analysis,it is believed that on the one hand,the state transition probability estimation models are different,and on the other hand,the micro databases selected by scholars are different.Especially when the estimation models of different students are similar,there are significant differences in the final simulation results due to the different databases used.In this article,we have selected two micro databases,CLHLS and CHARLS,which are most commonly used in estimating the probability of health state transition among the elderly in China,for comparative analysis.Firstly,it compares the two from the perspectives of survey object,survey time,survey content,sampling method,sample size,sample distribution,etc;Secondly,the definition criteria for different health states are clarified,and the models and assumptions for calculating the probability of state transition are determined.The tracking data of health status from 2015 to 2018 in CLHLS and CHARLS databases were analyzed in detail by gender,age group,and initial health status.By comparing the differences between the two databases in calculating the intensity of health state transitions using rough transition probabilities,the advantages of CLHLS and CHARLS databases and the selection of merge items are clarified.Finally,the combined database is used to calculate the health state transition probability.Based on the merged data,a health state transition probability matrix with a time interval of three years was calculated.Then,the health state transition probability matrix was converted into a health state transition intensity matrix using Kolmogorov forward equation.After correcting and smoothing the health state transition intensity matrix according to its properties,a more accurate ground health state transition probability matrix with a time interval of one year was obtained,and the final results were analyzed.The results showed that:(1)compared with male elderly,female elderly have more survival advantages;(2)The risk of death increases not only with age,but also with worsening health conditions;(3)The health status transition of the elderly is not only influenced by age and gender,but also depends on their initial health status;(4)Although female elderly people have more survival advantages than male elderly people,at the same time,surviving male elderly people have a more significant trend towards improved health,but the likelihood of such a trend will continue to decrease with age.