An exploratory sequential mixed methods approach will be employed to address the research inquiries. The study will consist of three phases conducted in sequence. Phase I will involve a qualitative study to explore qualitative evidence aligned with study objectives 1–3. Phase II will focus on the development and validation of a questionnaire to investigate study objective 4. Lastly, Phase III will consist of a quantitative study using a survey conducted in three selected cities representing the Middle, East, and West of China (30), to achieve study objectives 5–7. Data integration will be achieved through a data-building approach, where both quantitative and qualitative data will be combined in the final analysis stage to enhance the understanding of complex phenomena, examine hypotheses, and deepen the findings (31).
The PRISMA model for integrated care
The analysis will be guided by the PRISMA model (Fig. 1) (32) as the conceptual framework. This model is designed for healthcare systems that are universal and predominantly publicly funded. It can be implemented at the linkage level, where independent organisations, such as healthcare systems, social service organisations, and community agencies initiate protocols and programs to facilitate referrals or collaborations to meet patients’ needs. The PRISMA model is more commonly implemented at a coordination level, which involves the development of mechanisms to manage patients’ complex needs while each organisation maintains its own structure but agrees to cooperate. Coordination between institutions is at the core of the PRISMA model. A multidisciplinary team of practitioners, led by a case manager, evaluates patients’ needs and provides the necessary care. Access to all healthcare services is facilitated through a single entry point, managed by the manager. This model not only facilitates care delivery but also continuously monitors resources to ensure the effectiveness and efficacy of services. In China, the long-term care system, including community-based care for older adults, has been shaped by policies, government funding, consumer needs, and market forces (33). Day care and temporary services are provided and supervised by community centers, which are government agencies aimed at relieving the burdens of family caregivers by offering basic nursing care, house cleaning, rehabilitation and counselling services (34). It is believed that the PRISMA model is suitable for generalisation to the universal healthcare system in China and can guide the analysis of qualitative evidence as well as the structure design and subscales of a new questionnaire.
Phase I: A qualitative study
During Phase I, a qualitative case study will be conducted using a purposive sampling method to recruit participants (35). The study aims to investigate the unmet healthcare needs, the acceptability of the PRISMA integrated care model, and the realistic needs of integrated care through a case manager among Chinese community-dwelling older adults. The PRISMA model will be introduced to participants along with a semi-structured interview guide, including probe questions, in simplified Chinese. In the context of realistic healthcare needs, it will refer to the current Chinese healthcare landscape (36). Three trained investigators will conduct interviews in the field.
Settings
The study will be conducted in the community health centers in Wenzhou, Zhejiang province, China. Wenzhou has a significant ageing population consisting of 19.3% of the total population, and is transitioning towards a super-ageing society (37). Community health centers in China play a pivotal role in the ongoing healthcare reform (38, 39) and provide essential medical treatment, prevention, rehabilitation, and long-term care services to local residents, particularly for the older adults. Participants will be selected from these centres based on their engagement with healthcare services (40). Eligible participants seeking healthcare within their communities will be identified and selected for recruitment.
Participants
Community-dwelling older adults aged 60 and above will be included. The purposive sampling method will be employed to recruit individuals. The inclusion criteria for eligible older adults are: 1) having at least one chronic condition, such as ischemic heart disease, stroke, COPD, or type 2 diabetes; 2) visiting hospitals or community healthcare centers more than three times a year for their chronic condition management. Exclusion criteria include unwillingness to participate or difficulty in communication due to linguistic challenges, cognitive impairments like dementia, or swallowing difficulties like dysphagia. The perspectives of adult children of eligible older adults, care administrators, and healthcare professionals within communities, including nurses, family doctors, and administrators of civil affairs, will be considered to provide a comprehensive insight into this topic.
Sampling process
Investigators will identify and approach potential participants from selected urban community healthcare centres. Participants will be informed about the study, the PRISMA model, and the interview process. Preferences for the interview method (face-to-face or online) will be discussed, and arrangements will be made accordingly. Participants will receive monetary incentives ranging from 50 to 100 CNY ($7–14).
Data collection
Data will be collected through in-depth interviews and focus groups conducted either face-to-face or online via WeChat, a widely used communication app in China that is similar in functionality to Facebook. WeChat has a significant user base among older adults in China, covering approximately 82% (41). Focus groups will consist of up to four participants. Participants will be required to sign a consent form, and the PRISMA model and the role of case managers will be explained using clear and simple language. During interviews, a semi-structured interview guide with probes will be used to elicit detailed responses. Interviews will be recorded in either video or audio format. A pilot interview with 2–3 participants will be conducted before the onsite investigation to refine the interview questions. Sociodemographic characteristics such as age, education, income, and the number and types of chronic conditions, will be collected. Additionally, the Chinese version of the Barthel Scale/Index (BI) will be used to assess the older adults’ levels of Activities of Daily Living (ADL). A total of 30 participants will be recruited, and the final sample size will be determined based on data saturation. The sample of open-ended questions and probes can be found in supplement file 1.
Qualitative data analysis plan
Raw data will be managed using NVIVO 11 or ATLAS.ti8. Trained investigators will transcribe the interviews verbatim, and an additional investigator will conduct a data audit. Thematic and coding analysis will be conducted using both inductive and deductive approaches, guided by the PRISMA model. The data will be analysed using the framework method, which consists of six steps: (1) reviewing transcripts and noting initial ideas about themes; (2) developing a coding system based on the transcripts; (3) conducting weekly team meetings to identify recurring themes; (4) comparing themes across participants through constant comparative analysis; (5) defining and naming themes; (6) compiling a written report of the findings.
Phase II: Questionnaire development and validation
The aim of Phase II is to develop and validate a questionnaire based on the PRISMA model and Phase I outcomes. This tool will be used to assess the levels of expectations for integrated care through case managers among older adults, examining both general and specific expectations, as well as their sociodemographic characteristics correlates.
Questionnaire design and measurement
The questionnaire will be divided into two sections: 1) sociodemographic characteristics, including age, income, number and type of chronic conditions. The BI Scale for ADL will be assessed using a combined tool; 2) items of expectations related to integrated care through case managers. This section is informed by qualitative outcomes from Phase I, with variables derived from coded data and scales based on identified themes. Questionnaire items are constructed using direct quotations from the qualitative study, aligning with the PRISMA model and emerging codes from Phase I. To quantify the level of expectations, a 5-point Likert scale will be employed, allowing for categorisation of responses into three groups: low, medium, and high expectations. This categorisation can be applied at both the item level and construct level, based on the scale's properties. The final questionnaire is anticipated to consist of approximately 50 items in the expectations section, with a completion time of around 30 minutes.
Questionnaire validity and reliability
The face and content validity of the items in the questionnaire will be assessed by a panel of at least six experts, including nurses, geriatricians, statisticians, healthcare case managers, social workers, and older adults. The feedback provided by the panel members will be used to calculate the content validity index (CVI) for each item included in the questionnaire (42). The relevance, comprehensibility, and comprehensiveness of the items will be evaluated using the COnsensus-based Standards for the selection of health status Measurement INstruments check list (43). To assess structural validity, a pilot study will be conducted.
The structural validity of the questionnaire will be assessed through a pilot study, employing Exploratory Factor Analysis (EFA) to examine the underlying factor structure. Factors with an eigenvalue greater than 1 will be retained, while items with factor loadings below 0.40 or significant cross-loadings will be excluded (44). Structural Equation Modeling (SEM) will be used to evaluate the model fit, ensuring that the data supports the hypothesized factor structure. Internal consistency of the questionnaire will be evaluated using Cronbach's alpha, with a threshold of 0.70 or higher indicating acceptable reliability (45). Additionally, the reliability of the questionnaire will be assessed using intra-class coefficients (ICC) to determine intra-rater test-retest reliability. A random sample of at least 50 participants will be selected to complete the questionnaire twice, with a one-month interval between administrations (43). An ICC value between 0.50 and 0.75 will be interpreted as moderate reliability, while a value of 0.75 or higher will be considered indicative of good reliability (46).
Phase III: A quantitative study through a survey
In Phase III, a survey will be conducted to examine the expectations of integrated care through case managers among Chinese community-dwelling older adults and their associated sociodemographic characteristics. The survey will use a validated questionnaire that has been developed in Phase II. The questionnaire will be provided in Chinese either through a web-based platforms such as wjx.cn or as printed hard copies. Online participants will need to provide their Informed Consent Form by clicking a response button, while onsite participants will be required to provide a signature. Additionally, a data integration approach will be applied to combine the quantitative and qualitative data from Phase I and Phase III. This approach aims to enhance the understanding of complex phenomena, examine hypotheses, and deepen the findings. Based on the outcomes, an evidence-based framework will be formulated to optimise healthcare delivery for Chinese community-dwelling older adults.
Setting
In China, there is an uneven distribution of population and regional economic development, with a gradually declining trend from east to west (47). The survey will be conducted in three major cities: Wenzhou, Taiyuan, and Hainan. Wenzhou, located almost at the center of China's eastern coast, has a significant aging population. Older adults in Wenzhou prefer to reside in their homes and communities and receive integrated healthcare services align with Chinese culture values (48). Wenzhou is known as a pioneer in China's private economy, with a thriving small and medium-sized enterprise sector. It is considered one of the economically developed coastal regions in the east of China (49). Taiyuan, the capital city of Shanxi province, represents the west of China. The average income in this area is slightly below the national average (50). The number of older adults aged 60 and above in Taiyuan has exceeded 21.9% of the total population, indicating that it has entered a super-ageing society (51). Hainan (island), China's southernmost province, is also experiencing rapid ageing. The number of local older adults aged 60 and above in Hainan has reached 1.46 million, accounting for 15.5% of the total local population (52). Due to its prosperous economy and pleasant living environment, Hainan has become a highly sought-after location for older individuals, particularly during winter migration from 27 different provinces across China (53, 54). The samples from Hainan are more representative of a large ageing population.
Participants and sample size estimation
In accordance with the sample in the qualitative study, the quantitative study will recruit older adults who meet the following criteria: 1) aged ≥ 60, 2) living independently in the community, 3) having at least one chronic condition, and 4) visiting a hospital or clinic more than three times a year to manage their chronic diseases. Individuals who are unwilling to participate, have language difficulties, or cognitive impairments will be excluded. Based on an expected 18.8% level of expectations of chronic disease management services among Chinese older adults (55), a 95% confidence level with a two-sided and 5% margin of error, the minimum required sample size was 235. However, a target sample size of 354 (118 in each city) is set with a response rate of approximately 80% for non-response and incompletion rates.
Sampling process
A stratified random sampling method will be used to ensure that the sample accurately represents the diversity of older adults. Six community medical centres will be randomly selected from three cities, with two centers chosen from each city. Initially, six enumerators (two in each city) will propose four communities and specialised geriatric hospitals in each city where the ageing population exceeds 21% of the community's total population. Two of these communities will then be randomly chosen to ensure an unbiased representation. Recruitment will be carried out through posters, flyers, announcements in medical centers, and various outreach methods such as online platforms and community events to maximise participation. Participants will be requested to provide demographic information, which will be used to stratify the sample based on age and presence of chronic diseases prior to randomly selecting participants within each stratum. This approach ensures that the sample reflects the diversity of older adults across different community medical centers and cities.
Data collection process
Onsite investigators, comprising of at least one investigator and one research assistant in each city, will be responsible for overseeing the data collection process. At these selected sites, investigators will screen older adults based on the inclusion criteria to determine their eligibility. Those who meet the criteria will then be approached and provided with detailed information of the study. Simultaneously, participants will be informed about the study's objectives and asked to provide their consent. Informed consent will be obtained through either signed forms for onsite participants or electronic consent for online participants. To facilitate accurate data collection, participants will have the option to provide their telephone numbers, which will enhance the study's credibility and support convenient recruitment for intra-rater test-retest. Collected data will be entered into a database and undergo a double-checking process by investigators and research assistants to ensure accuracy and completeness.
Measurement
At the beginning of the questionnaire, a description of the modified PRISMA model in simple Chinese language will be provided to explain the concept of integrated care through a single entry point by a case manager. Participants will be asked to repeat the PRISMA model to confirm their understanding of the fundamental concept.
Sociodemographic characteristics
Specific age and hospitalisation within the previous year will be collected as numerical variables. Other variables will be categorized as descriptive variables. For example, gender (male or female), health status (healthy, living with one chronic disease, and living with two or more chronic diseases), monthly income (\(\:<\)375 USD, 375–749 USD, \(\:>\)749 USD), education (no formal education, elementary school, middle school, high school, bachelor’s degrees or above), occupation (employed, unemployed, retired), insurance (No insurance, UEBMI=Urban Employee Basic Medical Insurance, URBMI=Urban Resident Basic Medical Insurance, and NRCMI=New Rural Cooperative Medical Insurance), number of children (no child, 1 child, 2 children, and 3 or more children), and living arrangements (alone, with a partner, with a child or children, with a partner and children, and living with others such as a housemaid).
Barthel Index (BI)
This section will include ten items to aid in evaluating and categorising the level of ADL among participants. The BI scale was developed to assess a patient’s self-care abilities in ten areas, including control of bowel and bladder. The patient is scored ranging from 0 to 15 points in different categories, based on their need for assistance, such as in feeding, bathing, dressing, and walking (56). The Chinese version of the BI Scale has been validated and extensively implemented in various clinical assessment settings (57, 58). The BI is conventionally divided into four levels of scores using the following divisions: 100 − 91 (complete independence), 90 − 61 (slight level of dependence), 60 − 21 (moderate dependence), and ≤ 20 (severe dependence) (59).
Expectation section
The variables in the final version of the questionnaire will be categorised as individual items and groups within the construct to measure participants’ overall and specific expectations of integrated care. A 5-point Likert scale ranging from 1 to 5 will be used to indicate expectations from low to high.
Data analysis for the quantitative study
The IBM Statistical Package for the Social Sciences (SPSS 26) will be utilised to input and analyse quantitative data. Data from online questionnaires will be automatically transferred to SPSS, while data from hard-copy questionnaires will be manually entered into the dataset. The mean expectation scores will be categorised into tertiles, representing the lowest, middle, and highest levels of expectations. Chi-square tests will be conducted to explore the relationship between sociodemographic characteristics and the different expectation categories for integrated care. A significance level of p < 0.05, with a 95% confidence interval, will be used to determine statistical significance. Multiple logistic regression models will be employed to analyse the association of independent variables with the levels of expectation, using the lowest tertile of expectation as the reference group. Independent variables with a p-value < 0.20 from univariable regression analysis will be considered for inclusion in the final multiple regression model. The presence of multicollinearity among independent variables will be evaluated using a tolerance threshold of < 0.4 or a Variance Inflation Factor (VIF) ≥ 2.5. The adequacy of model fitting will be evaluated, and Q-Q plots for normality, residual plots for linearity and homogeneity assumptions will be examined. Skewness and kurtosis statistics will be utilised to assess the statistical assumptions of survey item responses. A moderate normality threshold of 2.0 and 7.0 will be applied for skewness and kurtosis, respectively, in the assessment of multivariate normality (60).