Analysis of the Influencing Factors on the Preferences of the Elderly for the Combination of Medical Care and Pension in Long-Term Care Facilities Based on the Andersen Model

In May 2016, the total population of Xiamen reached 2.2055 million, of which the elderly population over 60 years old accounted for 14%. In July–October 2016, the project team surveyed the elderly aged 60 and over in Xiamen City (residents of more than 6 months), using a multi-stage stratified sampling method to select 44 communities in 16 sub-districts in 6 districts of Xiamen City. After receiving identical training, investigators conducted a questionnaire survey on the elderly in the selected communities through face-to-face interviews. Questions mainly included the elderly’s sociodemographic variables, family conditions, health status, and CMCP service preferences. The inclusion criteria of the subjects in this study were as follows: (1) home-based elderly; (2) age ≥ 60 years; (3) able to communicate effectively; (4) gave informed consent. Exclusion criteria: (1) inability to communicate effectively; (2) living in LTC institutions.

The quality control strategy adopted in this survey was as follows: (1) During the research design phase, expert consultation and discussion ensured the scientific legitimacy and feasibility of the research plan and questionnaire. (2) Pre-surveys were conducted to improve the questionnaire, to develop standard operating procedures, and to develop a consistent training protocol for investigators. (3) During on-site investigations, quality control personnel were set up to conduct on-site audits and quality control, and to adopt double-blind protocols for later data entry and verification.

Based on Andersen’s model of healthy behavior [19], this study explores a theoretical framework for the analysis of factors that influence CMCP service preferences among the elderly (Figure 1). This model is widely used in studies of healthcare demand. The factors of influence were subdivided into factors that predispose elderly individuals to choose CMCP, factors that enable them to successfully make use of CMCP, and factors of personal needs that influence demand for CMCP. On the basis of the Andersen behavior model, this paper constructed the adjusted Andersen theoretical model of influencing factors on CMCP needs according to the research purpose and variable information availability: (1) predisposing factors affecting model variables, including type of residence, gender, age, and level of education; (2) enabling variables, including living style, marriage, number of children, income, primary caregivers, and type of medical insurance; (3) personal needs variables, including self-assessment of health, activities of daily living (ADL), and feelings of loneliness.

Descriptive statistics were presented for distribution characteristics of the surveyed subjects and the distribution of CMCP preferences. The chi-square test was then used to analyze differences in the distribution of different CMCP preferences of the elderly. Taking the variables with chi-square test p < 0.1 as the independent variables, a binary logistic regression model was constructed to analyze the factors that influence the elderly’s preferences for CMCP services. First, the predisposing variables were taken as control variables and put into model 1 as the benchmark model. Then, based on the benchmark model, enabling variables and personal needs variables were added successively to form model 2 and model 3, to explain the variation of the variance of the dependent variables by the addition of other types of factors to the benchmark model. Finally, model 4 was constructed with all variables as independent variables to determine the overall explanatory power of the model. In addition, 4 binary logistic regression models (Model I–IV) were constructed by including all three types of independent variables (predisposing, enabling, and personal needs) into the regression model and eliminating the independent variables one by one, so as to compare the relative influence of the three kinds of factors. The specific models are as follows:

Model I: Logit (Yi) = predisposing variables + enabling variables + need variables

Model II: Logit (Yi) = enabling variables + need variables

Model III: Logit (Yi) = predisposing variables + need variables

Model IV: Logit (Yi) = predisposing variables + enabling variables

All statistical analyses were conducted using SAS 9.3. The odds ratio and 95% confidence interval indicated the effect of each predictor and whether it met statistical significance. A value of p < 0.05 was considered statistically significant.

A total of 3160 questionnaires were completed of which 3119 were valid, giving an effective rate of 98.7%. Among these, males accounted for 51.0%. The mean age was 70.27 ± 7.81 years old. A total of 35.8% of the elderly live in rural areas, 40.1% cannot read, 30.2% are unmarried, 1.0% have no children, 16.6% have daily expenses that exceed their income, 29.8% have no caregivers, 5.7% are incapacitated, and 39.2% reported often feeling lonely.

Among the 3119 elderly respondents, 402 (12.89%) indicated that they wanted to be admitted to a CMCP nursing facility, 82.01% preferred home-based care, and 5.1% preferred community care. The analysis of the distribution of elderly people’s preference for different life care options shows that differences in preference depend on factors such as residence, age, education level, occupation, number of existing children, marital status, income and expenditure, primary caregivers, and activities of daily life (Table 1).

Table 2 shows the binary logistic regression analysis of factors influencing CMCP service preferences. In the predisposing factors, differences in residence, age, and education level were found to be statistically significant. It was found that the elderly living in rural areas have a higher preference for CMCP services than those in urban areas, (odds ratio (OR) = 2.1, 2.07, 2.3, 2.16) and that the OR values are greater than those in the town. As compared to the 60–69 years age range, the elderly in the 70–79 age group have a higher demand for CMCP services (OR = 2.21, 1.65, 1.74), but the difference between these two groups is not statistically significant. With regard to literacy, the CMCP service preference of those with only an elementary school education is lower (OR value is less than 1,) as compared to those above the high school level (OR value is greater than 1.)

Marriage, income and expenditure, and primary caregivers are among the enabling factors that proved to be statistically significant in Models 2 and 4. The OR values of the CMCP service needs of unmarried elderly are 1.61 and 1.64, respectively. The more an individual’s income exceeded their expenditures, the greater the OR value of their preference for CMCP service. For the elderly who are mainly cared for by their spouses, the OR value of preference for CMCP service is less than 1, while the elderly who are cared for by others (such as son-in-law or nephew) are more willing to accept CMCP service (OR values of 2.05 and 1.77, respectively.) Activities of daily life was a statistically significant factor in Models 3 and 4. As compared to the elderly who were completely independent, the worse the degree of capacity for self-care, the greater was the OR value of preference for CMCP.

Table 3 shows the comparison of prediction probabilities and goodness of fit of each model. Compared with Model I, the changes of -2LL (-2 Log Likelihood) in Models II and III are the largest, and the changes of Cox and Snell R2 and Nagelkerke R2 in Models III are similar. These indicate that the enabling factors contribute the most to Model I among the three types of influencing factors; the second-largest contributors were the predisposing factors.

The results of this study show that, in general, the elderly have higher needs for life care and medical care than others (Table 4). The elderly with higher age and education level, non-remarried, with better economic status, and with poorer health status have a demand for a wider variety of CMCP services. Compared to those in urban areas, the elderly in rural areas have greater needs, mainly related to personal care, medical care, and psychological counseling. In addition, the elderly with moderate to severe disabilities tend to choose medical care services such as medical care, health examination, and health education.

In this survey, the majority of the elderly still chose traditional home-based care (82.01%), consistent with the results of previous studies [20,21,22]. One can infer that, under the influence of traditional Chinese ideology and culture (such as filial piety) and long historical practice, family care will remain the most important way of providing for the aged for a considerable time. The government should, therefore, focus on strengthening policy support for family care. Moreover, socialized pension modes such as CMCP emerged in the process of industrialization in western developed countries and has a relatively short history of development in China. As a result, it is a comparatively new pension mode unfamiliar to the elderly [23,24].

However, it is worth noting that in this survey, the proportion of elderly people choosing CMCP institutions is much higher than the proportion of elderly people willing to choose CMCP institutions, as surveyed by our project team in 2013 (2.86%) [25,26,27]. A cross-sectional survey was performed by our project team across 173 communities in Xiamen, China, in 2013. A total of 261,043 individuals were aged ≥ 60 years in Xiamen at the survey time and about 5.5% of overall elderly populations were covered in that survey. In total, 14,292 valid questionnaires were obtained. The results of the two surveys reflect the fact that, with the vigorous advocacy and policy promotion of CMCP service by the Chinese government, the elderly have gradually realized the advantages of institutional care. It can solve a range of elder care needs caused by rapid aging of the population and by family miniaturization, taking into account both the “medical” and “care” aspects of the elderly’s most serious concerns [28]. Therefore, the government should formulate and improve the relevant CMCP policies as soon as possible, establish a coordination department for medical care, improve access and service standards, explore innovative service models, improve the service quality of pension institutions, and enhance the public’s awareness and acceptance of the CMCP model.

Most of the existing research on CMCP preference lacks the support of theoretical models and does not classify impact factors. Individual studies have only classified the factors simply, and did not conduct comparative studies on different categories of factors [29,30,31]. Based on the Andersen model, this paper constructed a framework for theoretical analysis of the elderly’s preferences for CMCP services. Through field investigation of the elderly in Xiamen community in China, this study measured the relative influence of factors of predisposition, enabling, and personal need on the elderly’s elder care choices. As a class, enabling factors have the greatest influence, followed by predisposing factors, indicating that personal family resources are the main factor affecting CMCP service choices.

This study found that older people with higher age and education are more willing to accept CMCP services. With the increase in age, the elderly’s sense of loneliness becomes stronger, their sense of security gradually weakens, and their ability to take care of themselves gradually declines [32,33]. With increase in education, the elderly become more aware of and more accepting of CMCP institutional care, whereas the less educated tend to have more conservative ideas. In contrast to previous research [34,35], this study found higher acceptance of the CMCP endowment institutions in rural populations. The main reason for this may be that the phenomenon of rural empty-nest in the process of rapid urbanization is serious, and there is a lack of existing old-age resources in rural areas. This makes it difficult to satisfy the life care and medical care needs of most of the elderly in those areas.

The results of this study show that elderly people who are poor and are mainly cared for by spouses have a lower willingness to choose institutional care services, while those elderly who are better off or who are mainly cared for by others (other than spouses) are more willing to accept CMCP institutional care services. The reason may be that the spouse is the person most familiar with the elderly individual, and the choice of life care can be more flexible. The married elderly people can meet their most important life care and spiritual comfort needs with the help of their spouse [36,37]. Due to the influence of traditional Chinese ideas, they are more willing to accept informal care services from their families. In addition, economic ability is a factor that attracts substantial attention in the relevant studies on pension choice. Most of the social pension services are currently paid services, which means that economic level is a limiting factor for the elderly who choose the social pension. Many studies [38,39] have shown that the elderly with higher economic status are more likely to choose the social pension because of their strong ability to pay for pension services, and this is further confirmed by this study.

Self-care ability largely determines the medical service needs of the elderly: the worse their capability for self-care, the greater their medical service needs. Losing the capacity for self-care is almost an inevitable life course for everyone. As a social public service to improve the quality of life of the elderly, CMCP service has been gradually accepted by the public [40,41]. CMCP endowed institutions are able to provide the elderly with more professional care services and with medical resources as well. This is particularly important for the elderly in poor health, and is an important reason to promote the choice of social endowment. Therefore, this study found that older people with poor self-care ability were more willing to choose CMCP institutions for the elderly. A unified long-term care insurance system urgently needs to be established by Chinese policymakers to reduce the financial burden of the elderly.

Consistent with previous results in other countries [42,43], this study found that the elderly generally have a higher demand for life care and medical care than in the past, thus indicating the importance of China’s efforts to promote the CMCP development strategy. Among them, the elderly with higher age and education level, who are non-remarried, and have better economic status and worse health condition have a higher demand for a range of CMCP services. Compared to the urban elderly, the rural elderly people have more serious needs, which are mainly reflected in daily care, medical care, and psychological counseling. In addition, moderately and severely disabled elderly people tend to choose medical care, health examination, and health education, among other medical care services. These results suggest that we should focus on elderly people in rural areas who are living in empty nests or alone and in poor health, and meeting their service needs such as daily care, medical care, and spiritual comfort.

However, there were several limitations to this study. First, the data were a cross-sectional survey, and this limited the interpretation of the results, making it hard to draw causal conclusions. Second, the question measuring the preference for CMCP and the exclusion criteria for the elderly under investigation (inability to communicate effectively) may have a selective bias in this study. Third, the models explain a low proportion of the variance, suggesting that other factors aside from those included in our analysis influence the various aspects of CMCP needs. Fourth, the participants were sampled from a single city in China. The results may reflect specific characteristics of that city, or at the very least introduce a rural–urban bias into the interpretation of the data. Therefore, longitudinal studies should be conducted to assess policy trends and key stakeholders.