Ambivalence in decision making for outdoor mobility among adult child carers of people with dementia: a study protocol for a randomized vignette study with message framing and eye-tracking

Abstract

Background

Independent outdoor mobility is crucial for the quality of life and well-being of people with dementia. However, adult child carers of older adults with dementia often face ambivalence when making decisions about restricting their parents' independent outdoor mobility, due to concerns for their parents and the need to balance autonomy and safety. This study aims to explore the influence mechanism of person-centered care values and reasons for/against independent outdoor mobility on ambivalence in decision-making regarding independent outdoor mobility, as well as the role of message framing in shaping caregivers' decision-making processes.

Methods

A randomized vignette study with eye-tracking will be conducted among 260 adult child carers of people with dementia (stage 3/4 dementia). Participants will be exposed to vignettes varying in lineage, dementia stage, message framing (gain/loss), and the outside environment (high/low ambivalence). Eye-tracking will objectively measure ambivalence during decision-making by recording the frequency of eye gaze and total gaze duration of participants. This study will also assess person-centered care values, reasons for/against independent outdoor mobility, and intentions of carers to allow people with dementia to go out alone by self-constructed scale from our previous modified Delphi study. Data will be analyzed using eye movement analysis with hidden Markov models and partial least squares structural equation modeling to test hypotheses.

Discussion

This study aims to develop a biopsychosocial model to explain ambivalence in independent outdoor mobility decision-making among caregivers. Findings will provide insights into how person-centered care values and message framing influence decisions, offering practical guidance for caregivers and healthcare professionals. The results may inform interventions to support caregivers in balancing autonomy and safety for people with dementia.

Introduction

Mobility, the ability to move freely, is identified as a key indicator of intrinsic capacity, a key concept that can be leveraged to achieve optimized healthy aging globally [1]. It comprises more than the function of walking, but also the ability to move around one’s environment, including using various forms of transportation [2]. Independent outdoor mobility (IOM) is crucial for independent living and has a significant impact on quality of life in old age [3]. Regretfully, one of the major consequences of dementia is a loss of independent mobility due to a gradual decline in spatial orientation and route planning [4]. According to the Global Deterioration Scale (GDS), people with dementia (PwD) at Stage 3 and Stage 4 would have to make decisions on whether they should go for IOM or not [5]. Stage 3 dementia is diagnosed when an individual is found to manifest regular memory and cognitive problems, which may cause word-finding difficulties and disorientation, but might not generally have a major impact on day-to-day functioning. When the PwD exhibit clear and visible signs of cognitive impairment and personality changes, they signify Stage 4 dementia, which is officially diagnosed as mild dementia according to the GDS [5]. PwD at this stage typically find trouble with organizing, planning, traveling, even in familiar places [6]. As a result, starting from Stage 3, decisions on confining PwD to smaller outdoor areas or even preventing them from venturing outside alone altogether were usually made by their carers due to safety concerns [7].

Imposing restrictions on outdoor mobility too early can negatively impact PwD, resulting in detrimental consequences, including the social, physical, cognitive, and emotional well-being of PwD. Limiting the IOM of PwD disconnects them from their communities, friends, and familiar environments, leading to profound feelings of loneliness, depression, and a diminished sense of self-worth, and that might inadvertently contribute to their social isolation and withdrawal [8]. Also, with the deprivation of regular physical activity, such as walking or engaging in light exercises outdoors, PwD may experience increased frailty, a higher risk of falls, and a decline in physical well-being [9]. For PwD, denying the chance to participate in independent outdoor activities robs them of their identity and autonomy [10], manifest in emotional distress and behavioral changes [11], and leads to increased irritability, restlessness, and challenging behaviors [12]. With over 55 million individuals living with dementia, it is crucial to reconsider the decision-making on IOM restrictions of PwD [2].

Along the caregiving trajectories, carers must make various decisions of IOM with full difficulties. Around forty percent of adult child carers participants reported having difficulties in making decisions on limiting IOM [13], which clearly shows that adult child carers are struggling with decisions related to limiting their parents’ freedom. Compared to adult child caregivers who reported less difficulties in decision-making, those with greater decision-making difficulties were more likely to suffer from depressive symptoms, anxiety, and greater stress [14]. Carers’ personal experiences in decision-making and identified ambivalence are key mechanisms. Previous research demonstrated that the decision-making process contains considerations of both gains and losses [15]. While some carers respect PwD’s autonomy but still installed tracking devices with GPS function, some locked the doors and restricted PwD to their homes. Such experiences are also associated with gender and lineage [16]. In this regard, the decision to restrict IOM is particularly fraught with ambivalence. The feeling of being torn between honoring the PwD's autonomy and prioritizing their safety is a key potential underlying mechanism in this decision-making process.

In home settings, caregivers faced autonomy-safety dilemmas, but in a more “unguided” way but loving context. Thus, family caregivers can refer to person-centered care (PCC) as a guiding principle. PCC, alternatively referred to as patient-centered care, is a sociopsychological intervention value and approach that is regarded as an evidence-based best practice in dementia care that is associated with high quality of life and well-being of PwD [17]. In other words, embracing autonomy and ensuring the safety of PwD are both PCC values; however, these two values create conflicts in forming ambivalence in the decision of IOM. However, only minimal attention has been paid to PCC values in the home settings. Also, there is no clear theoretical approach in existing studies to explain the ambivalence in decision-making among carers. According to the Cognitive Dissonance Theory, carers may employ various strategies, including supervised outings or utilizing assistive devices [18]. However, these strategies did not offer the best relief in difficulties in decision making and failed to help people understand the underlying mechanisms of ambivalence in decision making, as shown in prior studies [19].

Another key limitation of existing studies is the validity of the measurement approach on ambivalence in empirical studies. Subjective measures have been the prevalent method employed in the field of social sciences, including the measurement of caregiver ambivalence [20]. As cognitive ambivalence in decision making cannot be captured using subject measures, objective measures such as eye tracking are more promising and accurate. A recent study demonstrated that eye tracking traces the attentional dynamics in ambivalent decision situations [21]. The eye movement analysis with the switching hidden Markov model (EMHMM) approach can help to analyze related data in cognitive tasks, which paved the foundation for us to overcome the measurement validity limitation [22].

To sum up, this study seeks to shed light on the complex interplay between PCC values, reasons for and against IOM, and the resulting ambivalence experienced by caregivers might lead to ambivalence in the decision of IOM. Based on a previous literature review and a modified Delphi study by our team, we will adopt a randomized vignette study to investigate the specific influencing mechanism of adult child caregivers in the decision-making process regarding whether PwD should go out independently. We aim to establish a more detailed theoretical framework, incorporating reasons for (RF) and reasons against (RA) IOM, and the PCC value of caregivers together, so as to better explain the mechanism of decision-making ambivalence. Finally, we will also use eye trackers to measure the contradictory psychology of caregivers and gain a clearer understanding of the circumstances of decision-making on an objective level.

Objectives and research framework

This randomized vignette-based study aims to understand the PCC values held by adult child carers in making decisions in IOM in the context of dementia care. In this study, we propose a biopsychosocial model by integrating Behavior Reasoning Theory (BRT) and the Prospect theory (PT) to understand ambivalence in decision making for IOM among adult child carers of PwD. BRT is an emerging framework that outweighs the abovementioned theories. BRT establishes the relationship between values, reasons (RF and RA), global motives (attitude, subjective norm, and perceived behavioral control), intentions, and measures of user behavior [23]. A unique focus of this theory is on the reasoning construct, which is important in providing a better explanation for ambivalent decision-making processes. Based on our pilot study, the decision-making of IOM is the result of weighing between RF (e.g., autonomy) and RA (e.g., safety). BRT has sufficient explanatory power to support its construct development [24]. However, BRT does have drawbacks, as people are not always making rational choices.

To overcome this drawback, we proposed to integrate PT, a behavioral economic theory that explains how people make decisions under conditions of uncertainty and risk with BRT. It suggests that individuals do not always make rational choices based on objective probabilities and outcomes but are influenced by the way information is framed [25]. To test this integrated model, message framing is proposed since our team’s prior work showed that gain-framed messages can positively influence decision making [26]. Message framing refers to how information or messages are presented to individuals, emphasizing either the potential gains (gain-framed, e.g., improved well-being for carers and PwD) or potential losses (loss-framed, e.g., deteriorated well-being for carers and PwD) associated with a particular decision or action (e.g., allowing IOM). Below, we illustrate key constructs in detail, following the basic framework (Fig. 1).

Fig. 1.

Conceptual framework

Beliefs and values → reasons

Individuals’ processing of their belief and value information has a direct effect on the reasons people use to explain their anticipated behavior [25]. Theoretically, the decision alternative that has the most coherent explanation with strongly justified reasons should be selected. In this study, we refer values to PCC values, which are highly related to the RA and RF of decision-making of IOM.

Beliefs and values → ambivalence

PCC values are expected to have direct effects on global motives (i.e., ambivalence), without full mediation through reasons [25]. This argument is consistent with various psychological models that propose that automatic processing can circumvent deeper levels of thought activation [27].

Reasons → ambivalence

Reasons serve as important antecedents of global motives (i.e., ambivalence) [25]. Theoretically, people form a favorable evaluation of a given alternative when they have strong reasons that support and justify the alternative. In our context, adult child carers who have both reasons for IOM and reasons against IOM will experience ambivalence and ultimately affect their decision-making.

Ambivalence → intention

Global motives (i.e., ambivalence) are the primary antecedents of intention [25]. Ambivalence is the “torn” between positive or negative evaluation towards IOM. In general, ambivalence is hypothesized to have strong predictive power on decisions for IOM for PwD. Also, message framing is powerful in affecting decision-making. In our study, we will apply framing to test its hypothetical mediating role in the relationship between ambivalence and intention.

Research questions and hypotheses

By following the framework of decision-making of adult child carers, the general objective of this study is to uncover reasons that shape their perspectives on IOM for PwD. The internal influence mechanism will be discovered by identifying the PCC values that contribute to the decision-making of adult child carers, the mediating role of RA and RF in the influence of PCC values on ambivalence and examining the moderating effect of message framing on ambivalence and decision-making. The specific research questions are: How do PCC values influence ambivalence in decision-making for IOM for PwD? What PCC values contribute to the reasons for and against IOM for PwD? How do the reasons for and against IOM influence ambivalence in decision-making among adult child carers? How does message framing (gain vs. loss framed message) moderate the relationship between ambivalence and decision making for IOM for PwD? The following are the hypotheses of this study:

• H1: Adult child carers with higher PCC values will positively influence ambivalence.

• H2a: Adult child carers with more reasons for autonomy will positively influence ambivalence.

• H2b: Adult child carers with more reasons against autonomy will negatively influence ambivalence.

• H3a: Gain-framed message will positively moderate the relationship between ambivalence and the decision for IOM for PwD.

• H3b: Loss-framed message will negatively moderate the relationship between ambivalence and the decision for IOM for PwD.

Methods

Overview

Based on our research objectives, this study aims to better explore the relationship between PCC values, reasons for or against IOM, and decision-making ambivalence in adult child caregivers of PwD during independent outdoor mobility decisions. The research will recruit adult caregivers of older adults with dementia to participate in the study and conduct recruitment through a combination of online and offline methods. This research is affiliated with the University Grants Committee Council General Research Fund (No. 17611924).

As a part of a broader project about ambivalence in decision-making, this study is built on the findings from the first phase of the project. In collaboration with professional doctors, older adults with dementia, adult child carers, social workers, and other related populations regarding ethical considerations, we discovered the reasons for and against IOM held by various people, and the PCC value in family care. Methods of literature review, focus group, and modified Delphi study were used to create the basic findings, and insights from the previous studies were used to develop the basic needs for this study. Hypothesis, scales of reasons, PCC value, and relevant scenarios required for this survey will also be designed based on these previous findings.

Participants, selection criteria, and recruitment

Participants of this study are the adult child caregivers of PwD who meet the following inclusion criteria: (1) ≥ 18 years old; (2) providing care to a PwD parent ≥ 65 years old (Stage 3/Stage 4 dementia measured by GDS); (3) self-identified as primary caregiver; (4) having normal or correct to normal vision; (5) providing care for ≥ 10 h per week; (6) able to communicate in Chinese or English. People with these criteria will be excluded: (1) professional caregivers like nurses or social workers; (2) have cognitive/physical impairments affecting eye-tracking task completion.

The target sample size for this study is 260, which was determined using the pwr package in R, a statistical software tool for power analysis. Power analysis requires a minimum of 256 participants to detect a significant relationship between variables, assuming a 95% confidence interval, and 80% statistical power [28]. Balanced recruitment across four lineage groups (father-son, father-daughter, mother-son, mother-daughter dyads) will be used through stratified sampling. Recruitment will utilize the following methods: (1) Posting paper announcements in community centers, elderly service centers, and clinics, etc.; (2) Distributing electronic advertisements through social media platforms (e.g., Facebook), relevant organization newsletters, and emails to contacts from our research center's previous studies and community partners; (3) On-site recruitment at caregiver support events, meetings, and through our community partner, the 'Carer Space' station [29], a well-established hub and resource center for caregivers in Hong Kong; (4) Collaboration with third-party institutions, such as non-governmental organizations and hospitals. Participants who pass the eligibility screening will participate in the survey after signing the consent form and will receive corresponding rewards after completing the entire process. The recruitment for this study will commence in September 2025 and continue until the target sample size is reached. It is expected that 30 new research participants will be recruited each month, so the recruitment phase may last until June 2026.

Survey instruments

Demographics of characteristics of caregivers and PwD will be assessed through a demographic questionnaire. Demographics of caregivers include gender (male vs. female), year of birth, relationship with PwD, time (in years) of caregiving, weekly time (in hours) spent on caregiving, caregiving tasks, living arrangement, housing, self-perceived financial status (in terms of adequacy), highest education attainment, and self-rated health. Demographics of PwD will be collected by caregivers’ report, including relationship with caregiver, gender (male vs female), year of birth, year of dementia diagnosis, living arrangement, housing, highest education attainment, history of illness, history of fall, history of wandering, and history of loss.

We will also measure the use of care-related services by caregivers and the activities that caregivers and PwD engage in. The psychological and social conditions of caregivers will also be measured, which include anxiety, intergenerational solidarity, and caring responsibility. Additionally, the joint decision-making process and willingness of caregivers, and decision-making difficulties will be included in this survey.

The PCC values held by carers and their reasons for and against PwD’s IOM will also be measured as a possible influencing mechanism of decision-making. PCC values will be measured through adaptation of the 18-item Generic Person-Centered Care Questionnaire and the findings from previous Delphi study of our team [30]. Reasons for and against IOM will be measured by the Reasoning Inventory (RI) to be developed, which is also based on that study and a recent literature review on the variables [24]. To assess the reasons people provide for their behaviors, RI has been used in various studies, such as smoking cessation or exercise adoption, to understand the cognitive processes driving behavior change [31].

Outcome variables

The primary outcome variables of the trial are the ambivalence of decision-making of adult child carers. Secondary outcomes are the intention of carers to allow PwD to go out alone. Ambivalence will be measured using eye-tracking. After reading the scenarios, participants will be asked to respond to two questions: "Should the caregiver in the scenario allow PwD to go outside?" and "If you were the caregiver in the scenario, would you allow PwD under your care to go outside?" Participants’ frequency of eye gaze and total gaze duration over the ‘yes’ and ‘no’ options before their decisions will be calculated as measures of choice preference.

Intention will be measured based on BRT, which establishes the adapted constructs in global motives (attitude, subjective norm, and perceived behavioral control) [23]. This variable consists of three items: “I intend to allow PwD to go out alone in the next 3 months”, “I predict I would allow PwD to go out alone in the next 3 months”, and “I plan to allow PwD to go out alone in the next 3 months”. Each item will be rated on a scale from 0% (not at all possible) to 100% (extremely possible).

Vignette development and design

The textual vignettes were developed through a multi-stage process to ensure relevance and robustness. The initial design of the scene originated from the discussion results of four focus groups with four different groups of 27 people, including 8 dementia older adults, 7 staff members (3 social workers, 1 occupational therapist, 1 physical therapist, 1 personal care worker, and 1 rehab assistant), 7 adult child carers, and 5 doctors. We delved deeply into the realistic decision-making dilemma regarding restricting outdoor activities for PwD. Through a thematic analysis in our previous study, we provided a basis for the development of situational scenarios. The real conversations in these focus groups are directly integrated into the text to ensure ecological validity, with particular emphasis on retaining the specific terms used by the participants. This grounded research method helps us precisely generate scenarios that make the participants, as caregivers, feel involved while reading.

We constructed 32 different text scenarios based on findings, which systematically varied in four key dimensions: lineage of carer and older adult, the stage of dementia of the cared-for older adult, message framing, and the level of possible ambivalence. The vignettes will portray a home setting between a caregiver and a person with dementia (at stage 3 dementia and stage 4 dementia, respectively). The scene will portray a home setting between a caregiver (son/daughter) and a person with dementia. It will first describe the basic situation of a person with dementia (stage 3/stage 4 dementia), then this person with dementia will request to go out alone and describe the scene of going out and the environment of traveling (high/low ambivalence, depending on the familiar/unfamiliar destination the PwD is heading for). Then, the caregivers of the children put forward suggestions of agreement or opposition (gain/loss frame). A total of 32 scenarios will be developed with a unique combination of son versus daughter, PwD father versus PwD mother, Stage 3 dementia versus Stage 4 dementia, gain-framed versus loss-framed, and high ambivalence versus low ambivalence.

After the initial design of the scenarios, an expert panel will review all texts to ensure face and content validity. An expert panel consisting of 7 participants will be formed, including: 1 psychiatrist, 1 occupational therapist, 1 psychiatric nurse, 2 social workers, and 2 graduate caregivers. Experts will be recruited from our previous studies through online or offline recruitment.

Study design and setting

This randomized vignette study employs a 2 × 2 × 2 × 2 between-subjects factorial design to investigate how PCC values, reasons for/against autonomy, and decision ambivalence interact in dementia caregivers' outdoor mobility decisions. By applying eye-tracking methodology and randomized vignette design, this study would help establish a biopsychosocial model related to ambivalence and decision-making. The overall design is shown in Fig. 2. The experiment is primarily designed to be conducted in a standardized laboratory environment, which meets the essential control criteria: the place must be private, have consistent and adequate lighting, minimal environmental noise and distractions, and allow for the stable placement of equipment and participant seating as per the laboratory. Hence, as long as a private room with minimized environmental noise, it is considered as appropriate. We will choose the location based primarily on feasibility for maintaining experimental control, while considering participant needs. Under the condition of ensuring local experimental conditions, we will arrange personnel available for training at every location to guarantee the safety of the participants, the smoothness of the experimental process and the reliability of the results.

Fig. 2.

Research design

There are two sections of the survey before and after the randomized vignette study with eye-tracking. First, after the initial recruitment and signing the consent, participants need to be screened and matched according to the lineage and stage of dementia of PwD they are caring for. Matching lineage is a key design feature in this study based on two main reasons. Firstly, previous research has shown that the caregiving experiences, expectations, and decision-making process vary depending on the lineage of caregiver and PwD [32]. These differences may affect the degree of emphasis placed on issues of autonomy and safety, thereby influencing the outcome of the decision of IOM. Matching lineage aims to control and separate the factors that may affect the results during the experiment. Secondly, matching lineage can enhance the ecological validity and personal relevance of the scenario description [33]. When participants read scenarios that match their lineages, they are more likely to put themselves in the context of their own caregiving decisions, which helps to increase the accuracy of the experimental data in reflecting their responses to similar real-life dilemmas.

Second, participants proceed to the randomized vignette study section, and the research assistants should have completed the setup and basic correction of the eye tracker before the participants are invited to the experiment lab. They will be seated in front of a 22-in. computer screen (1,680 × 1,050 pixels) at a distance of 700 mm and instructed to position their head in a chin rest. Stimuli will be presented using a separate computer, while visual signals will be presented through the display PC. The lineage of the participants and the dementia stage of older adults being cared for will be matched with the basic information in the vignette. Participants will be presented with two vignettes, each depicting an outdoor scenario designed to induce high versus low ambivalence, respectively. The information framing (gain frame vs. loss frame) will be randomly assigned and displayed. After reading the scenario, participants need to make a ‘yes’, ‘no’ decision about permitting independent outdoor mobility of PwD.

During the study, eye movements will be continuously recorded with fixation detection parameters. Fixation detection will be performed using a peak velocity threshold of 60°/s and a minimum fixation duration of 40 ms. Immediately following each vignette, participants will also be instructed to respond with an external key device. The entire eye-tracking session lasts approximately 20 min.

Third, participants will be requested to complete the post-test measures, including nine parts, respectively: (1)Demographic information of caregivers; (2)Demographic information of dementia older adults being cared for (to be filled in by caregivers); (3)The use of care-related services; (4)Psychological and social conditions of caregivers; (5)Joint decision-making process and willingness; (6)Decision-making difficulties; (7)Decision-making intention; (8)Person-centered care values; (9)Reasons for and against IOM. This section requires about 40 min. The full protocol, requiring about 90 min per participant, is administered by trained research assistants following a detailed operations manual to ensure procedural fidelity across all sessions.

Statistical analysis

Descriptive statistics, including means, standard deviations, and frequency distributions, will first be calculated for all demographic variables (e.g., age, gender, education level) and primary measures (e.g., decision-making accuracy, response time, gaze metrics) to characterize the sample and check for normal distribution assumptions. These analyses will be used to characterize the sample composition to ensure representativeness and to assess the normality of data distributions to verify assumptions for subsequent parametric tests.

To quantify ambivalence during decision-making, this study will employ eye movement analysis with EMHMM, a data-driven approach that integrates spatial and temporal eye-tracking features. This method identifies latent states (e.g., fixation clusters) and transition probabilities between predefined regions of interest (ROI) corresponding to decision options. Predefined regions of interest over the decision choices will be used to quantify participants’ eye gaze transitions between the options during decision making using both spatial and temporal information of eye movement [22]. EMHMM parameters will be estimated for each participant, capturing the distribution of fixations across ROIs, transition probabilities between ROIs, and specific gaze strategies of different participants. Then, gaze transition consistency will be measured using the entropy of the models [34]. Lower entropy values reflect more predictable, stable gaze transitions, while higher entropy indicates erratic or exploratory patterns. This metric will be derived per participant and linked to the ambivalence of decision-making.

We will test hypotheses by conducting partial least squares path modeling or partial least squares structural equation modeling to examine the complex cause-and-effect relationship generated under our proposed conceptual model. Additionally, moderation analysis (on message framing effects) will be performed separately, using the “bruceR” package in R [35].

Ethical considerations

This study has received ethical approval from the Human Research Ethics Committee (HREC) of the University of Hong Kong (No. EA230649). From recruiting participants to data collection, storage, analysis, and reporting, thorough ethical considerations will be carried out. During the recruitment stage, the research content will be explained in detail to the participants, their consent will be obtained and confirmed one by one, and they will be asked to sign the consent form. At all stages of the research, participants' involvement in the study will be completely voluntary. Throughout the research process, participants will be informed that they can withdraw from the study at any time and are encouraged to contact the researchers at any time if they have any questions or concerns. The data collection process strictly adheres to the principle of confidentiality, and the identity information of participants will not be disclosed in the report. The data collected by the research institute will only be used for research purposes, and personal information will be kept strictly confidential.

During the investigation, research assistants will show the respondents a detailed informed consent form, which elaborates on the investigation content and the risks and benefits that may be involved in participation. Only after signing the consent form can the participants proceed with the subsequent research. After completing the entire study (including both survey sections and the eye-tracking vignette task), each participant will receive a gift card worth 500 Hong Kong dollars as a reward for their time and effort.

Discussion

This study aims to develop a biopsychosocial theoretical framework for explaining the role of ambivalence in decision-making among adult child carers who are taking care of PwD at Stage 3 and Stage 4. To the best of our knowledge, this is the first study to examine PCC values and decision making among adult child carers for PwD, which will contribute to a novel biopsychosocial understanding of the factors influencing decision making in the context of IOM for PwD. Applying eye-tracking methodology and randomized vignette design, the impact of message framing on ambivalence and decision-making would support establishing a biopsychosocial model for the first time in relation to ambivalence and decision-making. By establishing empirical evidence for the proposed theoretical model, new directions for future research on PCC value and decision-making among informal caregivers would be generated through this study.

The results of this study will help understand the decision-making behavior of caregivers and provide more support and assistance to various groups. For dementia carers and PwD, the results of this study will be translated into action-oriented tips that will be made online and accessible to all family carers, as well as PwD, to help them consider both the pros and cons of allowing PwD to go out alone, using the tips we provide. For the volunteers and the public, because not every older adult and/or caregiver can easily access online materials, we will develop materials and decision-making aids to train volunteers to be the enablers. At last, allied healthcare professionals such as social workers, psychologists, and counselors working in various professional and community services, who support PwD’s families and their caregivers, will benefit by designing evidence-based interventions based on the study’s findings.

Authors’ contributions

All authors have made substantive contributions to the manuscript. All have read, reviewed, and approved the final manuscript. B.F is responsible for planning and writing the protocol. R.L is responsible for writing and editing parts of the protocol. V.L, J.H, J.Y, H.T, J.T are responsible for planning the protocol, as well as commenting on and editing the protocol.

Funding

This work was supported by the Hong Kong University Grants Committee General Research Fund (GRF/HKU/17611924). We confirm that the funding agency conducted an independent peer review of this study as part of the funding approval process.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethical approval and consent to participate

This study was approved by the Human Research Ethics Committee at the University of Hong Kong (EA230649). An informed consent form will be obtained and confirmed from each participant individually regarding all parts of the study. Our study is not a clinical trial and therefore does not require registration.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.