Implementing cognitive training in primary health care for older adults with MCI: study protocol for a type 3 hybrid effectiveness-implementation cluster-randomized factorial trial

Abstract

Background

Dementia prevention is an urgent global health priority in aging populations. Mild cognitive impairment (MCI) presents a key window for intervention, where cognitive training can effectively slow progression to dementia. There remains a need for implementation strategies that activate primary healthcare providers (the first point of contact) to deliver such interventions. This study aims to evaluate a multi‑component implementation strategy (enhanced training, financial incentives, and leadership engagement) to promote the delivery of cognitive training for elderly with MCI by family doctor teams in primary care.

Methods

This study employs a hybrid type 3 effectiveness-implementation cluster-randomized trial. Guided by the Multiphase Optimization Strategy (MOST) framework, a 2 × 2 × 2 factorial design is used to examine the main effects and interactions of three intervention components (enhanced training, financial incentives, and leadership engagement), thereby generating empirical evidence for optimizing the implementation strategy. It will be conducted across 16 community health centers (CHCs) in Beijing, targeting family doctor teams as the primary intervention units. The active intervention will last 6 months, followed by a 6‑month observational follow‑up to assess sustainability. The primary outcome is the initiation of cognitive training delivery by primary healthcare providers (change from non‑delivery to delivery). Secondary outcomes include providers’ self‑efficacy, patients’ cognitive function, beneficiary coverage, and implementation‑related measures.

Discussion

This study will inform the integration of cognitive training into routine primary care, support the design of community-based dementia prevention strategies, and contribute to promoting healthy aging. The findings will provide actionable evidence for policymakers engaged in active dementia prevention and control globally.

Trial registration

The study was registered with the Chinese Clinical Trial Registry ChiCTR2600116108. Registered 05 Jan 2026.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12877-026-07967-w.

Background

The world is facing a significant demographic shift towards an aging population. By 2030, one in six people globally will be aged 60 years or over, with 80% of older individuals projected to reside in low- and middle-income countries by 2050. Advancing age is associated with an increased likelihood of experiencing multiple coexisting health conditions, such as hearing loss, musculoskeletal disorders (e.g., back pain, neck pain, and osteoarthritis), diabetes, depression, and dementia. These trends pose substantial challenges for health and social systems worldwide in adapting to this demographic transition. In response, the United Nations Decade of Healthy Ageing (2021–2030) aims to reduce health inequities and enhance the well-being of older adults, their families, and communities [1]. One of its four core action areas explicitly emphasizes the provision of person‑centered integrated care and responsive primary health services for older populations. Driven in part by socioeconomic considerations, high‑income countries are increasingly shifting from institutional care models toward community‑ and home‑based care [2]. This transition places greater demands on primary healthcare systems, requiring family physicians, community nurses, and public health professionals to assume broader roles in delivering coordinated, accessible care to older adults in their homes.

Mild Cognitive Impairment (MCI) is defined as a progressive decline in memory or other cognitive functions that does not significantly interfere with activities of daily living and does not meet the diagnostic criteria for dementia [3]. As a high-risk, preclinical stage of dementia, MCI represents a critical window for intervention and prevention [4]. Effective intervention at this stage can delay disease progression in approximately 30% of patients. Taking China as an example, the vast number of older adults with MCI living at home currently lack access to timely and effective interventions. Without intervention, over 50% of individuals with MCI will progress to dementia within five years, resulting in substantial socioeconomic costs [5]. Cognitive Training, a non-pharmacological intervention aimed at enhancing cognitive function, involves adaptive, difficulty-graded exercises targeting domains such as attention, memory, and logical reasoning. It has become an important approach for preventing and managing cognitive disorders [6, 7]. The Chinese Guidelines for Cognitive Training (2022 Edition), following the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) framework [8] and the evaluation principles outlined in the Chinese Guidance for the Development and Revision of Clinical Practice Guidelines (2022 Edition) [9], assigns the highest grade (Level 1 A) to the recommendation that “cognitive training can improve overall cognitive function in patients with mild cognitive impairment” [6].

Given the limited specialist and inpatient resources in medical institutions, the gatekeeping role and accessibility of Primary Health Care (PHC), and the generally lower economic status of the older population, PHC providers, represented by community-based general practitioners, are widely recognized internationally as the most accessible, economical, and feasible avenue for older adults with MCI to obtain health services [10–12]. However, due to practical constraints such as limited frontline manpower, heavy workloads, and cost considerations, cognitive training interventions in primary care settings in developing countries remain severely inadequate, lacking systematic implementation and sustainability.

Several key national policy documents in China, including the National Medium- and Long-Term Plan for Proactively Addressing Population Aging, the Healthy China 2030 Plan, and the Healthy China Action Plan (2019–2030), explicitly emphasize strengthening effective interventions for conditions such as Alzheimer’s disease and related dementias. These guidelines specifically call for community-based general practitioners to organize routine cognitive training sessions for older adults diagnosed with mild cognitive impairment (MCI). Moreover, since 2009, the national Basic Public Health Services program has incorporated elderly health management (such as free health check-ups, cognitive screening, and chronic disease management) into its standard provisions.

Given this policy backdrop and existing service infrastructure, China serves as a relevant and informative case for exploring how cognitive training interventions for home-dwelling older adults with MCI can be delivered within a primary health care context. Findings from such research could help delay the progression to dementia in this population and offer practical insights for other developing countries seeking to strengthen their dementia prevention and response systems.

In this study, cognitive training interventions for older adults with MCI can be considered an evidence‑based practice (EBP). However, most EBPs, despite substantial investment in their development, take an average of 17 years to be translated into routine clinical practice, and only about half are widely adopted [13, 14]. Implementation science has emerged to promote the application and dissemination of EBPs in real‑world settings. The focus of implementation science is not to verify the efficacy or effectiveness of an intervention itself, which is the aim of traditional randomized controlled trials (RCTs), but rather to identify factors that influence the implementation of EBPs and to use this information to develop implementation strategies (where the strategy itself acts as an intervention) in order to accelerate and improve the translation of evidence into practice [15–17]. Implementation science employs various theories, frameworks, and models to understand and explain factors affecting implementation processes and outcomes, and has established a range of research design paradigms [17, 18], which provide a sound conceptual and methodological foundation for this study. Moreover, implementation strategies themselves constitute complex interventions, composed of multiple components such as financial incentives, training and education, and delivery modalities. A key current challenge is how to design implementation strategies consisting only of effective components under real‑world resource constraints, so as to effectively enable primary care providers (PCPs) to deliver cognitive training interventions.

Literature review

A systematic search of PubMed, CNKI, and Wanfang databases conducted prior to this study revealed no randomized controlled trials focusing on implementation strategies for cognitive training interventions towards MCI older adults [19]. Evidence regarding the impact of different intervention components on promoting the provision of cognitive training intervention by community-based general practitioners remains highly limited. Two systematic reviews published in 2021–2022, which focused on intervention strategies for the assessment and management of MCI patients in PHC settings [11, 20], highlighted that while numerous technical tools and effective interventions exist for older adults with MCI, high-quality evidence on how to promote the implementation of these evidence-based practices is lacking. This further underscores the necessity of the present study.

Current research primarily consists of randomized controlled trials (RCTs), systematic reviews, and meta-analyses examining the effects of cognitive training interventions on cognitive function in older adults with MCI [7, 21–28]. While these studies exhibit some heterogeneity in design and population, their conclusions are largely consistent: cognitive training interventions demonstrate a significant positive effect on the cognitive function of older adults with MCI. Regarding evidence-based implementation strategies for dementia-related interventions, Lourida et al. (2017), in a systematic review of 88 studies [29], identified key strategies for promoting implementation: education and training (94%), strengthening relationships among key stakeholders (69%), and evaluation systems (46%). Economic incentives were used in 15% of the included studies to facilitate implementation. The review also noted that 70% of the studies were conducted in residential care settings, highlighting a lack of research on implementation strategies within primary health care (PHC) contexts. Similarly, Gitlin et al. (2020) [30]summarized barriers to implementing evidence-based dementia interventions in health systems, including uneven evidence application, premature implementation, workforce shortages, failure to identify the most suitable evidence-based practice (EBP), and unsupportive external environments. Aligning with several findings from Lourida et al., this study also emphasized the importance of employing scientific research designs to evaluate the effectiveness of implementation strategies for promoting EBP uptake in practice [30].

Study aim

This hybrid type 3 effectiveness‑implementation trial uses a cluster‑randomized factorial design to evaluate a three‑component implementation strategy (enhanced training, financial incentives, leadership engagement) for promoting MCI‑focused cognitive training by family doctor teams within primary care setting. Its aims are to: (1) assess the strategy’s effect on provider practice behavior change, and (2) identify which component and combinations most effectively increase the delivery of cognitive training towards MCI older adults (MCI-CT intervention). The results will clarify active ingredients and interactions to optimize implementation and support the integration of MCI management into routine primary care for dementia prevention.

Methods

Study setting

This study will be conducted in community health service centers (CHCs) across Beijing. As the foundational providers of primary healthcare in their districts, for instance, CHCs in Beijing recorded a total of 92.458 million outpatient and inpatient visits in 2024 [31]. The CHCs also serve as the principal platforms for delivering China’s National Basic Public Health Services, a role solidified since the launch of the new healthcare reform in 2009. Services for older adults include free physical examinations, management of four major chronic diseases (hypertension, diabetes, coronary heart disease, and stroke), and dedicated elderly health management. Since 2016, the family doctor contract service has been implemented, with older adults as a key priority group typically signing contracts with family doctor teams. This entitles them to basic public health services as well as personalized service packages tailored by individual CHCs.

However, cognitive training services for older adults with MCI (aged ≥ 60) are not yet routinely available in Beijing’s communities. This service gap presents an ideal setting for the present study. Building upon the existing basic public health service infrastructure also offers a natural and feasible pathway for translating research findings into practice. The study design is therefore well aligned with local routine care contexts.

Overview of the study design

This study applies a hybrid type 3 effectiveness‑implementation, cluster‑randomized factorial trial, guided by the Multiphase Optimization Strategy (MOST) framework [32]. The MOST framework consists of three phases: preparation, optimization, and evaluation. The preparation phase of this study has been completed. The current protocol is designed specifically for the forthcoming optimization phase and is reported following CONSORT [33] and SPIRIT [34] guidelines (see supplementary file 1 and 2).

The preparation phase involves selecting and defining the intervention components that constitute the overall implementation strategy, all of which will be tested during the optimization phase. This selection was informed by a systematic review (currently under peer review) and two rounds of focus group interviews [19]. Guided by the COM-B model (Capability, Opportunity, Motivation–Behavior) of behavior change and taking into account practical resource constraints, three intervention components were ultimately identified: enhanced professional skills training, financial incentives, and leadership engagement. The corresponding conceptual framework is presented in Fig. 1.

Fig. 1.

Conceptual framework of the study

The optimization phase involves applying a 2 × 2 × 2 factorial experimental design to examine the main effects of each intervention component or level as well as the interactions among them, thereby generating empirical evidence to optimize the implementation strategy. Table 1 shows the factorial design of this study.

Table 1.

The 2 × 2 × 2 factorial design of this study

Group	General professional and policy training	Intervention components			Implementation strategy
		I1: enhanced Professional skill training	I2: financial incentive	I3: leadership engagement
Group 1	Yes	No	No	No	I0
Group 2	Yes	Yes	No	No	I1
Group 3	Yes	No	Yes	No	I2
Group 4	Yes	No	No	Yes	I3
Group 5	Yes	Yes	Yes	No	I1*I2
Group 6	Yes	Yes	No	Yes	I1*I3
Group 7	Yes	No	Yes	Yes	I2*I3
Group 8	Yes	Yes	Yes	Yes	I1*I2*I3

Eligibility criteria

Aligned with the practical context of primary healthcare in China, where services are predominantly delivered by family doctor teams, this study targets the family doctor team as the primary unit for intervention implementation. A typical team consists of a physician, a nurse, and a public health professional (some teams may also include volunteers or family doctor assistants). The inclusion criteria for participants are as follows: (1) Employed at one of the sampled community health service centers; (2) Currently providing contracted family doctor services; (3) Willing to participate in the project and adhere to its required activities. The target population for health management by the family doctor team is older adults aged 60 and above with MCI within the jurisdiction.

Patient and public involvement

Building upon the barriers identified in the systematic review, family doctors from community health service institutions were invited to participate in discussions to define and finalize the intervention components. This participatory approach was adopted to enhance the relevance and feasibility of the components within routine primary care practice.

Intervention

Based on a completed systematic review and guided by the COM‑B model, semi-structured interviews were conducted with stakeholders including health administration departments, primary healthcare institutions, and community family doctor teams. The interviews aimed to identify implementation barriers and inform the development of targeted implementation strategies. Ultimately, a core intervention (provided to all participating groups) and three distinct intervention components (I1, I2, I3) were defined (Table 2).

Table 2.

Intervention components

Intervention	Specific contents	Type
General professional training and policy requirements training	Seminar: • Policy implications and requirements for implementing cognitive training for elderly individuals with MCI in community settings • Professional skills training for cognitive function screening and cognitive training in MCI populations • Community-based health management protocol for home-dwelling elderly individuals with MCI	I0 (Comparator)
Enhanced professional skill training & support	• Targeted expert lectures conducted quarterly, specifically at the 3-month and 6-month marks of implementation • WeChat communication groups will be established by institution to provide real-time expert support for profession-related challenges encountered during the intervention, leveraging the geriatric specialist resources of the leading hospital within the medical consortium	I1
Financial incentives	Financial incentives provided based on the quantity and quality of MCI cases managed by family doctor teams, at a rate of 25–40 RMB per cognitive training record • Base payment (25 RMB/record): Granted for completing a cognitive training record, which includes administering AD-8 and MoCA screenings, and documenting health management services such as cognitive training interventions, reminders, and follow-ups • Performance-based bonus (up to 15 RMB/record): Awarded after review based on the detail and accuracy of the recorded information as well as the frequency of cognitive training sessions conducted	I2
Leadership engagement	Directors/Deputy Directors of community health service centers issue administrative directives during monthly regular meetings, requiring family doctor teams to carry out MCI cognitive training interventions	I3

The intervention will be implemented over a period of six months, followed by an additional six-month follow-up survey. This follow-up phase is designed to assess the sustainability of the original intervention after the project-related support has been withdrawn—an outcome of significant policy relevance.

Adherence to intervention protocols (Fidelity)

Team members, in collaboration with the upper-level hospital (Beijing CYL Hospital) of the medical consortium, monitor the process and quality of intervention implementation. This includes collecting cognitive training service records on a monthly basis and providing timely clarification of questions arising during the project, so as to ensure the fidelity of the intervention.

Enrolment (Recruitment)

In January 2026, the research team collaborated with the leading hospital of the Beijing Medical Consortium to recruit community health service centers. This was followed by further coordination with institutional leaders to identify family doctor teams eligible to participate in the project. Eligible family doctor teams that met the inclusion criteria subsequently provided written informed consent (Table 3).

Table 3.

The time schedule of enrolment, implementation and assessments

 d day, m month; I1: Enhanced professional skill training & support; I2; Financial incentives; I3: Leadership engagement

Outcomes

The information regarding outcome measures is presented in Table 4. The primary objective of this study is to promote a behavioral shift among community family doctor teams from rarely conducting to actively implementing MCI cognitive training (MCI-CT) interventions. Therefore, the penetration rate among implementation outcome indicators was selected as the main outcome measure. For analytical purposes, this variable is transformed into a binary variable labeled “MCI-CT intervention delivery behavior of community family doctor teams,” where delivery is coded as 1 and non-delivery as 0.

Table 4.

Outcome measures, data sources and data collection time points

Domain	Measure	Data source	Time
Primary healthcare provider:
Penetration	Number of family doctor teams conducting MCI-CT interventions/Total number of participating family doctor teams × 100%	Provider survey; health service records	T1-T3
Self-efficacy	The General Self-Efficacy Scale (Chinese version of GSES) [35]	Provider survey;	T-1-T3
MCI Participants:
Cognitive function	MoCA-B [36]	Patient survey	T-1-T3
Reach	Number of elderly individuals with MCI who received the intervention	Health service records	T1-T3
Implementation:
Cost	Costs related to implementing the intervention, covering personnel, materials, meetings, etc	Cost log sheet	T2, T3
Acceptability	The extent to which primary healthcare providers perceive the delivery of MCI‑CT interventions (e.g., content, complexity, time required) as agreeable or practicable	Qualitative interview	T-1, T2, T3
Appropriateness	Reflecting primary healthcare providers’ perceptions of the suitability, relevance, or compatibility of the MCI‑CT intervention	Qualitative interview	T-1, T2, T3
Maintenance	The degree to which community general practitioners believe the MCI‑CT intervention can be maintained, stably delivered, or institutionalized within their setting	Health service records; Qualitative interview	T3

T-1 = Baseline (prior to intervention), T1 = 3 months after the first day of intervention, T2 = 6 months after the first day of intervention, T3 = 12 months after the first day of intervention

Assessment

Surveys will be conducted at four time points: baseline (pre-intervention, T‑1), 3 months after the start of intervention (T1), 6 months (T2), and 12 months (T3). The primary modes of data collection include: questionnaires for providers or patients, qualitative interviews, and health service record abstraction. Specifically, the Chinese version of the General Self-Efficacy Scale (GSES), used to estimate providers’ self-efficacy, was derived from Wang CK et al.(2001), with a Cronbach’s α of 0.87 [35], while the cognitive function of patients is evaluated by the Chinese Version of Montreal Cognitive Assessment Basic (MoCA-B), derived from Chen KL et al.(2016), with a Cronbach’s α of 0.807 [36].

Harms

No anticipated harm to participants is expected from the project intervention.

Sample size

For the primary outcome (whether community PHPs actively initiate MCI‑CT interventions), the following assumptions were used: α = 0.05, power = 80%, control group event rate = 5%, intervention group event rate = 30%, and an intra‑cluster correlation coefficient (ICC) of 0.03. Randomization was performed at the level of community health centers, with an average of 10 general practitioners enrolled per center and an estimated attrition rate of 15%. An individually randomized design would have required 104 participants. After adjusting for the design effect due to clustering (DE = 1.27) and the factorial structure (2 × 2 × 2), 16 community health centers will be recruited with 10–12 practitioners enrolled per center, yielding a total sample of 160–192 practitioners. The 16 centers will be completely randomized across the eight experimental conditions of the factorial design, with two centers assigned to each condition. Statistical verification shows that this design provides between 82 and 88% power to detect the primary main effect, which satisfies the study requirements.

Randomization and blinding

The 16 community health centers (CHCs) were first grouped into 8 pairs based on size and geographic location, with each pair corresponding to one of the eight factorial intervention arms. A unique random sequence was then generated for these pairs using a random number generator. An independent researcher, not involved in data collection or analysis, assigned each pair to a specific intervention arm according to this sequence, thereby ensuring randomization and minimizing selection bias. A flow diagram for the randomization and intervention allocation is shown in Fig. 2. Each CHC’s group assignment was printed on a card, sealed in an opaque, numbered envelope, and securely stored. After obtaining informed consent, the corresponding envelope was opened by the CHC to reveal its allocation, after which the intervention was initiated.

Fig. 2.

Study flow chat. (I0: General professional training, I1: enhanced professional skill training, I2: financial incentive, I3: leadership engagement)

CHCs were randomized as clusters, with interventions implemented according to the trial protocol across eight experimental groups. However, because the interventions were delivered primarily by family doctor teams, these teams remained single‑blinded to group assignment throughout implementation. While they were inevitably aware of the intervention components applied within their own group, they were not informed of the specific allocation scheme or the identity of other trial arms. The single‑blind condition was maintained rigorously during data collection and analysis. Unblinding occurred only after completion of data analysis, at which point group allocations were disclosed. The effectiveness and feasibility of the interventions were subsequently evaluated objectively based on the trial outcomes.

Data management

Research staff collected the aforementioned outcome measures using a combination of paper-based questionnaires and/or an online mini-program developed specifically for this purpose. All collected data were exported and categorized in Excel files, which were stored on separate disks and hard drives. During questionnaire administration, research staff promptly identified and corrected any errors, thereby effectively controlling data quality. This approach also eliminated potential mistakes arising from double data entry and verification, ensuring data accuracy. All questionnaires were archived sequentially by identification number after completing the required data entry and verification processes, and are retained for future reference.

Data analysis

Quantitative data analysis

This study will adhere to the intention‑to‑treat (ITT) principle for handling missing data [37].

Descriptive statistics (means, standard errors, proportions, etc.) will be used to summarize the baseline characteristics of participants. Differences in outcomes across intervention groups will be examined using analysis of variance (ANOVA) for continuous variables and chi‑square tests for categorical variables.

In line with Collins’ recommendations for estimating effect sizes in factorial designs guided by the MOST framework, the three intervention components (I1, I2, I3) will be effect‑coded as + 1 (component present) and − 1 (component absent) [32]. The general statistical model is formulated as follows:

where Y represents the outcome measure; I1, I2, I3 denote the three effect-coded intervention components; the terms I1*I2, I1*I3, I2*I3, I1*I2*I3 represent two-way and three-way interactions; ~ reflect the main effects of the components;~ represent the interaction effects. The interaction effects are the primary focus of this analysis.

Logistic regression or generalized linear models will be used depending on the distribution of the outcome data. All models will adjust for baseline participant characteristics (gender, age, professional title, years of practice, education level).

Qualitative analysis

Qualitative analysis will focus on the acceptability, appropriateness, and sustainability of implementing the intervention. Interview transcripts, verified by the research team, will be independently coded by 2–3 researchers based on the interview guide. Special attention will be paid to respondents’ explanations regarding the reasons underlying perceptions of acceptability, appropriateness, and sustainability. New themes emerging during coding will be discussed and finalized through team consensus. Because the Consolidated Framework for Implementation Research (CFIR) framework includes constructs related to these outcomes, the qualitative findings will be further synthesized with post‑implementation CFIR‑based interview data. The analysis will be conducted using MAXQDA 2018 software (VERBI Software, Berlin, Germany).

Data monitoring

The Academic Committee of the corresponding author’s institution is responsible for overseeing the overall conduct of the study, ensuring data security, and verifying adherence to ethical standards. The research team will collaborate with experts from the leading Hospital of medical consortium to conduct monthly supervision of the intervention implementation at each participating institution and provide on-site feedback.

Discussion

Against the challenging trend of a growing global population of older adults with cognitive impairment [12], this project, grounded in the perspective of implementation science and guided theoretically and methodologically by MOST, explores implementation strategies for community general practitioners to deliver cognitive training interventions to elderly patients with MCI in PHC settings. It further examines the implementation effects of different intervention components within these strategies in real-world contexts. The findings hold significant reference value for achieving policy goals aimed at actively preventing and controlling dementia in the aging population.

The strengths of this study primarily focus on three aspects: First, while existing research has predominantly validated the effectiveness of cognitive training in clinical settings [3], establishing a solid evidence base (which also serves as the evidence-based practice for this study), there is a lack of implementation research evidence on how to deliver such interventions in real-world scenarios. Second, this study concentrates on the PHC setting, where family doctor teams act as health gatekeepers, offering favorable cost-effectiveness attributes [38]. The anticipated research outcomes are expected to facilitate the integration of cognitive training services within primary healthcare institutions, contribute to the development of community-based dementia prevention and control strategies, and ultimately help reduce healthcare expenditures. Third, this research holds considerable reference value for other developing countries facing rapid aging and underdeveloped PHC systems by using China as a case study. Since the launch of its new healthcare reform in 2009, China has consistently prioritized strengthening the primary healthcare system, although a mandatory first-contact service at the primary care level has not yet been fully established.

This study anticipates two main challenges: First, as the primary observation units are PHC providers, most communities currently lack comprehensive information on the cognitive function of their elderly residents. It will be necessary to incorporate MCI cognitive screening (using tools such as MoCA-B) into existing elderly health check-ups or health management programs. The project team will provide detailed guidance on MCI screening and management protocols during the basic intervention phase to ensure that PHC providers are competent in implementation. Second, the study employs a factorial trial design with three intervention components distributed across eight groups, which is more complex than the traditional two-arm RCT design (intervention vs. control). This increased complexity necessitates stringent quality control throughout the implementation process.

Abbreviations

MCI

Mild cognitive impairment

GRADE

Grading of Recommendations Assessment, Development, and Evaluation

PHC

Primary health care

EBP

Evidence‑based practice

RCT

Randomized controlled trial

PCP

Primary care provider

CHC

Community health center

MOST

Multiphase optimization strategy

COM-B Model

Capability, opportunity, motivation–behavior model

MCI-CT

Cognitive training towards MCI older adults

ICC

Intra‑cluster correlation coefficient

ITT

Intention‑to‑treat

CFIR

Consolidated framework for implementation research

DSMB

Data safety and monitoring board

Authors’ contributions

YS led the original conception, study design and preparation of study protocol. MX, LP, and PF contributed the development of the intervention and training. DJ contributed substantially to the operationalization of implementation strategies and data collection procedures. LH and WS substantively revised the manuscript. All authors contributed to refining and finalizing the study protocol. All authors read and approved of the final manuscript.

Funding

This study is supported by the National Natural Science Foundation of China (72304282), following a rigorous peer-review process. It is also supported by the Capital’s Funds for Health Improvement and Research (CFH2024–2G–4311) and the Beijing High-Level Innovation and Entrepreneurship Talent Support Program Leading Talent Projects (G202513010). The funding sources had no role in the conceptualization, design, manuscript preparation, or the decision to publish.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of the Institute of Medical Information (No. IMICAMS/K01/26/HREC). During recruitment, researchers will distribute informed consent forms to participants to ensure that they are fully aware of the study. Their participation in this study is voluntary and they have the right to leave at any time and without reason. All participant data are de-identified to ensure confidentiality.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.