Study setting
The study was carried out in Lishui county, Foshan, China. The county is located in the middle of Pearl River Delta, one of the most affluent rural regions in China. At the end of 2019, the population of Lishui county was 580,000, and the proportion of those aged over 60 reached 22.3% of the total population, higher than the national average of 13.0% [1].
Concurrently with the policy of long-term care reform in China, Lishui county wanted to develop a novel community care service for older people. Under this model, the county government is responsible for infrastructure construction and purchasing of service including daytime caring, domestic service, entertainment activities. The specific characteristic of this model is community rehabilitation. By incorporating community rehabilitation (physical therapy, exercise therapy, acupuncture and moxibustion, self-management education) in community care service, diverse service providers (rehabilitative professionals, social workers and grassroot government) are involved in this mode that can support people ageing in place. This enabled the cooperation between the rehabilitation department of our hospital, a 1000-bed tertiary hospital and the county government. Over one year of extensive consultation with representatives from 16 villages and 18 communities of the county, a novel community care service for older people was developed.
Study design
A cross-sectional survey was conducted in Lishui county of Foshan during the period of January to March 2021. People who were 60 years of age and above and living in Lishui county for over six months were included in this study. We excluded those who had difficulty in communication due to serious physical illness or psychiatric disorder.
Sample size determination
The sample size was determined on the assumed effect size of 0.50, referring the study of Liu, et al. [8]. To obtain reasonable estimates at 95% confidence level and 5% margin of error, a total sample size of 401 was needed, taking into account 5% non-response. This study included 413 senior residents allowing detection of significant differences with a power of 0.89 calculated by Gpower software.
Sampling procedure
Lishui county is composed of 16 villages and 20 communities. Prior to the sampling procedure, we hosted a meeting with heads of villages/communities and representatives from social work organizations on October 10th, 2020. We first explained the aims and target population of this community care service to each attendee. Then the healthcare program was introduced including adopted therapies, frequency of rehabilitation services and follow-up visits. After addressing their questions, they were encouraged to spread the information. Brochures regarding the community care were also given to them. A staged cluster-sampling method was adopted to select participants. First, 8 villages and 10 communities were randomly selected. From each village or community, 23 residents were randomly selected from the roster that includes residents over 60 year old provided by the county government. Systematic random samplings were used to select villages/communities. Then computer-generated random numbers were used to determine specific residents. The procedure of randomization was conducted by an administrative person not involved in subject recruitment. The researchers were blinded in the process of recruitment. Participants were asked to sign a written consent form before the beginning of the interview. For those who had difficulty in writing, written informed consent was obtained from their spouses/children. The oral informed consent was recorded when the older people or their family members cannot write.
Data collection
Data were collected using a structured questionnaire composed of three sections. The development of the questionnaire was based on Andersen- Nyman model of health services utilization [11]. After the pilot test on older people at community care centers, a manual was created to provide suggestions on how to ask questions so that older adults can understand. Eligible older people were contacted by the investigators with the help from the government staff. After explaining the research protocol and obtaining informed consent, participants were interviewed face-to-face at the community care centers. For those who preferred to be interviewed at their residence, we came to their homes. The data collection phase was completed with the help of 10 post-graduate students. They were trained for a day by the principal investigator covering research objectives, interview skills, quality control and research ethics. Approximately 15 minutes was spent on the total interview and each individual was reimbursed with a small gift after completion.
Variables and measures
Dependent variables
The dependent variable was the willingness to receive community care for older people measuring by the question‘Would you be willing to receive community care service?‘with the option yes or no. The interviewer first explained this novel community care service, then the interviewer asked the respondent whether he or she understood the concept.
Independent variables
The independent variables were classified into three categories according to Andersen-Nyman model [11]. 1) Predisposing variables, such as age, gender, marital status, education and having children; 2) Enabling variables such as living arrangement, health insurance, household income and walking time from current residence to the nearest community care center; 3) Needing variables, such as physical activity over the past two weeks, number of non-infectious chronic diseases and activity of daily living (ADL) disability.
Physical activity was assessed by the International Physical Activity Long Volume (IPAQ-L). It covers four domains with a summary score which can be categorized into three levels (low, moderate and high). The validity of Chinese version has been confirmed [12]. The ADL disability was measured by Katz Index Scale. It includes six items and each item has two response choices. If any answer was ‘dependent’, the seniors were categorized as ADL disability. This scale had good reliability (Cronbach alpha>0.80) and validity (content validity index, CVI>0.90) to assess functional status of older people in China [13].
Statistical analysis
The primary data was entered into Epidata 3.1, then analyzed by SPSS 20.0.Variables were presented as counts and proportions. Chi-square test was used to assess the differences in predisposing variables, enabling variables and needing variables between older people who were willing to receive community care service and those who were not. Multivariate regression analysis was used to identify factors that were significantly associated with willingness to receive community care for older people after controlling for possible confounders. Odds Ratios (ORs) with 95% confidence intervals (95% CIs) were calculated to measure the strength of association. Estimates with p-values less than 0.05 were considered statistically significant.
Ethics
Ethics approval was obtained from the Research Ethics Board of Southern Medical University. All participants provided informed consent before being interviewed.