Background: Late-life depression is common, modifiable, yet under-treated. Service silos and human resources shortage contribute to insufficient prevention and intervention. We describe an implementation research protocol of collaborative stepped care and peer support model that integrates community mental health and aged care services to address service fragmentation, using productive ageing and recovery principles to involve older people as peer supporters to address human resource issue. Methods/design: This is a non-randomised controlled trial examining the effectiveness and cost-effectiveness of the "Jockey Club Holistic Support Project for Elderly Mental Wellness" (JC JoyAge) model versus care as usual (CAU) in community aged care and community mental health service units in 12 months. Older people aged 60 years and over with mild to moderate depressive symptoms or risk factors for developing depression will be included. JoyAge service users will receive group-based activities and psychoeducation, low-intensity psychotherapy, or high-intensity psychotherapy according to the stepped care protocol in addition to usual community mental health or aged care, with support from an older peer supporter. The primary clinical outcome, depressive symptoms, and secondary outcomes, self-harm risk, anxiety symptoms, and loneliness, will be measured with the Patient Health Questionnaire-9 (PHQ-9), Self-Harm Inventory, Generalized Anxiety Disorder 7-item scale (GAD-7), and UCLA Loneliness 3-item scale (UCLA-3) respectively. Cost-effectiveness analysis will assess health-related quality of life using the EQ-5D-5L and service utilisation using the Client Service Receipt Inventory (CSRI). We use multilevel linear mixed models to compare outcomes change between groups and calculate the incremental cost-effectiveness ratio in terms of quality-adjusted life years. Discussion: This study will provide evidence about outcomes for older persons with mental health needs receiving collaborative stepped care service without silos and with trained young-old volunteers to support engagement, treatment, and transitions. Cost-effectiveness findings from this study will inform resource allocation in this under-treated population. Trial registration: ClinicalTrials.gov NCT03593889. Registered on 20 July 2018.
Keywords: Older adults; Depressive symptoms; Collaborative stepped care; Peer support; Prevention; Early intervention; Effectiveness; Cost-effectiveness
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1186/s13063-022-06122-1.
Depression is one of the most common yet under-recognised chronic conditions among older people. In Europe, for example, a meta-analysis found an overall prevalence of 12% of depression among older people [[
Prevention and early intervention are effective for late-life depression and can reduce suffering and societal costs [[
Despite the promise of prevention and early intervention for old age depression, there are multiple challenges to its implementation as a routine service. First, system fragmentation is a common barrier in mental health care in older age, as multiple service sectors and providers are often involved [[
These challenges can theoretically be addressed using concepts of collaborative stepped care, task shifting, and productive ageing.
Collaborative care refers to a systematic approach that integrates different mental health professionals and/or other care extenders that has been shown to be effective in improving outcomes of patients with depression [[
Stepped care such as IAPT also allows for task shifting, a possible solution to the challenge of the shortage in the professional workforce [[
We have designed a collaborative stepped care and peer support mental health service model for the identification, prevention, and early intervention of depression in older people in Hong Kong to synthesise this theoretical and empirical knowledge. In contrast to existing collaborative stepped care models, this new service focuses on prevention and early intervention, and hence the integration of services takes place at the community level, with collaboration between two types of service (i.e. aged care services and community mental health services) operated by different organisations. The new service also has a peer supporter programme based on the principles of task shifting and productive ageing. We aim to investigate the effectiveness and cost-effectiveness of this new service in an implementation study (Jockey Club Holistic Support Project for Elderly Mental Wellness, JC JoyAge).
We developed the collaborative stepped care model through a co-creation process involving multidisciplinary researchers and service providers (frontline and management) from community aged care and mental health care services. In the finalised model, preventive intervention services will be provided by trained social workers and peer supporters from aged care or mental health service units according to level of risks, symptom severity, and intervention response. A community mental health service unit and a community aged care service unit will collaborate to deliver the service for step-up and step-down cases. Fig. 1 summarises the stepped care model in JC JoyAge. We have developed a clinical protocol and guidelines for coordinated care among community mental health and aged care services for elderly mental wellness, based on ageing and mental health theories and the grounded feedback and practice wisdom from the social workers.
Graph: Fig. 1 Collaborative stepped care and peer support services flowchart
A 256-h theoretical and clinical training course is developed to equip project social workers with the competence to provide mental health services to older people. The training course consists of both theoretical sessions and practical sessions. The theoretical sessions involve three modules: (
Peer supporters are recruited through non-governmental organisations (NGOs) in the community. The inclusion criteria for peer supporters are (
Based on local experience in detecting and engaging older people at risk of depression, effective outreach programmes will be delivered in the format of general health screening in the neighbourhood area, and events on topics of concern to at-risk older adults (e.g., chronic pain, sleep). Trained peer supporters will undertake home visits to detect and engage hidden cases. This will be in the format of generic health promotion or other pragmatic approaches to lower mental health service stigma and barriers to help-seeking. Depression risk assessments will be provided as part of these activities to detect these otherwise hard-to-reach, hard-to-engage older adults.
Participants will be recruited through community outreach and open referral. Those who meet the eligibility criteria will be asked to give informed consent to participate in the study. Upon doing so, they will undergo a clinical intake and assessment interview conducted by a clinical social worker. The assessment battery will include instruments that measure depression, suicidal risks, and other outcomes (please refer to the programme evaluation section for more details). The initial intervention provided for participants will be determined by their depression level indicated by their scores on the Patient Health Questionnaire (PHQ-9) [[
Trained social workers and peer supporters from aged cared or mental health service units will provide collaborative stepped care to clients according to their level of risks, symptom severity, and intervention response. According to the project clinical protocol, social workers will perform professional needs and risks assessment to ensure triage and programme selection. Table 1 and Fig. 1 outline the collaborative stepped care workflow. Adherence to the service model is monitored through monthly case meetings, on-demand clinical supervision, and an electronic data input system. Any deviation from the service model, e.g., unmatched service to the client's intake scores, will be flagged for the supervisor's review and approval before processing the client's data.
Table 1 Service care model and collaboration between service units
Client group Criteria Intervention Service unit(s) Service period At Risk PHQ9 ≤ 4 or with risk factors Selective prevention: engagement activity + PS Community aged care 2–9 months Prodromal 5 ≤ PHQ ≤ 9 without score on item 2 or 6 Indicated prevention: psycho education + PS Community aged care 3–9 months Mild 5 ≤ PHQ ≤ 9 and scored on item 2 and 6 Indicated prevention: group psychotherapy + PS Community mental health care 3–9 months Moderate 10 ≤ PHQ ≤ 14 Individual psychotherapy + PS Community mental health care 6–12 months Moderately severe and above* PHQ ≥ 154 Care as usual Traditional mental health service, HA N. A
*Beyond the scope of the current project, mainly referral for appropriate services PHQ Patient Health Questionnaire, PS Peer Supporter, HA Hospital Authority
Older people with milder conditions ("vulnerable" state), operationalised as a PHQ-9 score of 4 or below, and having one or more risk factors of old age depression, will receive a 4-week "selective prevention" group (group size: 6–8), a mental health information package after completing the group, and a review at week 6–8. Peer supporters will lead the group with social worker supervision at the aged care service unit, and the group content covers wellness topics tailored to the person's concern (e.g. coping with insomnia, exercising). Participants will be discharged at week 8 if they exhibit improvements in depressive symptoms or have reduced risk. If they show no change, they will enrol in another 4-week group, followed by a review until they reach the service limit (9 months). If participants show increased depressive symptoms at review, they will be stepped up to receive indicated prevention.
Older people at the prodromal or mild stage of depression will receive structured low-intensity psychoeducation followed by a review at aged care service units, delivered by trained social workers and supported by peer supporters. Older persons who are at the prodromal stage, operationalised as having a PHQ-9 score of 5 to 9 but do not have negative mood or thoughts (as indicated by items 2 and 6 of the PHQ-9, respectively) will receive a 4-session psychoeducation group, with an approximate group size of 10 people, on one of the following topics (pain, sleep, stress, physical activities, and bereavement) based on their needs. Older persons at the mild stage of depression, operationalised as having a PHQ-9 score of 5 to 9 with the presence of negative mood and thoughts, will receive 6-session low-intensity psychotherapy such as cognitive behavioural therapy (CBT), problem-solving therapy (PST), and reminiscence therapy in groups of 6 to 8 persons. After the review at weeks 10 to 12, participants will be stepped down to receive selective prevention if they exhibit improvements in depressive symptoms. If they show no change, they will enrol in another round of indicated prevention, followed by a review until they reach service limits (9 months). If participants show increased depressive symptoms at review, they will be stepped up to receive treatment.
Older persons with moderate depression, operationalised as having a PHQ-9 score of 10 to 14, will receive 8 sessions of CBT, either in individual or group format, provided by the trained social workers at mental health care units, followed by a review. After the review at weeks 10 to 12, participants will be stepped down to receive selective or indicated prevention if they exhibit improvements in depressive symptoms. If they show no change, they will enrol in another round of treatment, followed by a review until they reach service limits (12 months). If participants show increased depressive symptoms at review, they will be referred to existing services. Peer supporters will be matched to individual older adults to walk them through the process and provide regular follow-up for 1 year maximum. Older persons who have a PHQ-9 score of 15 or above will be referred to existing specialist mental health services in the community. Those with suicidal risks will be referred to a hospital, mainly through the Fast Track Clinic for Elderly Suicide Prevention Programme.
This study aims to (
We use a non-randomised controlled trial to compare older adults at risk of or with sub-threshold symptoms receiving JoyAge service (intervention group) with those receiving CAU (control group). The intervention study will be conducted in four districts (out of 18) in Hong Kong. Participants residing in the four piloting districts will be recruited through six partner NGOs to the intervention group by social workers, and members of the same NGOs residing in other districts will be recruited as control subjects by researchers to increase the compatibility of the two groups. No randomisation is applied due to the clinical nature of the study and several feasibility considerations, for example, a new collaborative workflow between different service units, workforce training, and the high possibility of intervention and control group participants to interact outside treatment groups. Data collection will take place at three time points: baseline, discharge (applicable to the intervention group only), and 12-month follow-up from baseline. Since the JoyAge service is stepped care with peer support, the length of the service will depend on individual client's progress, and it could vary from 2 to 12 months. To promote participant retention, we will inform participants from both groups about the approximate date of the follow-up assessment. We will also send out postcards to the CAU group in special holidays, e.g., Chinese Lunar New Year, to remind them about the study since they do not receive service from the JoyAge project. Ideally, a follow-up interview will be conducted face-to-face at the client's convenience and preferred venue; however, a telephone interview will also be an option if the client prefers so. Each client will be contacted initially by phone to arrange a follow-up interview, and a reminder call will be made 1 day prior to the interview. If a client cannot be reached by phone after three calls, social workers of the NGO where the client is recruited will step in and check the client's status via home visits or other means. For an overview of the schedule of enrolment, interventions, and assessments, see Table 2.
Graph
This study has been approved by the Human Research Ethics Committee (HREC) of the University of Hong Kong (Reference number: EA1709021). The study is registered with the US National Institute of Health (NIH) ClinicalTrials.gov (NCT03593889) on 20 July 2018. The trial protocol fulfils the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guideline [[
There are no greater anticipated negative effects in psychological interventions in JoyAge services, such as strong but temporary feelings of distress and deterioration, than care as usual [[
During the course of this study, a participant who does not respond to intervention positively after nine months will be referred to existing services. Participants may withdraw from the study at any time point. Social workers and peer supporters will regularly follow up with the clients and perform risk assessment; the frequency may vary between weekly and monthly depending on the clients' needs and social workers' clinical judgement. If at any time during the study, there are concerns about a person's suicidal risk, or if the person is assessed as having moderate or higher risk of suicide, referral will be made to a hospital or Fast Track Clinic for Elderly Suicide Prevention Programme [[
Graph: Fig. 2 Consolidated Standards of Reporting Trials (CONSORT) flow chart of JoyAge trial
The intervention study will be conducted in community setting in four districts (out of 18) in Hong Kong. In each district, a community mental health service unit and a community aged cared service unit will collaborate to provide a coordinated service to older adults at risk of (the prevention group) or having mild-to-moderate level of depressive symptoms (the treatment group). All service units are NGOs, and there are six NGOs participating in this study (listed in ClinicalTrial.gov as collaborators). The prevention group, comprising adults aged 60 years or older who are in a vulnerable state (exposed to increased risk factors, such as recent spousal death, disability, chronic illness) or who have sub-syndromal depressive symptoms plus risk factors, such as disability, chronic illness or living alone, will receive preventive intervention at community aged care centres. The treatment group will comprise two subgroups: those with mild to moderate sub-threshold depressive symptoms without immediate risks who will receive care at aged care service centres and those with moderate-to-severe sub-threshold depressive symptoms who will receive care at the community mental health service centre. Older adults in this programme will be able to move up or down the stepped care levels and between the different types of service according to needs and risks. They will receive assertive preventive care and treatment with a clear goal for recovery. Across levels of care and services, peer supporters will establish rapport, guide and support older adults in the programme and offer longer-term engagement (up to one year within the programme including follow-up).
Participants in the control group will be recruited from the community aged care units and mental health service units of the six partner NGOs in districts other than the four piloting districts of Hong Kong, and they will receive care as usual (CAU). CAU in community aged care units includes health education, case management, social and recreational activities, and other tangible support like meals and laundry services etc. CAU in community mental health care units includes casework counselling, clinical psychological services, and referral to the community psychiatric services of the Hospital Authority for clinical assessment or psychiatric treatment. These services are all district-based, and there is a limited collaboration between the two service units. Continuation and provision of the intervention model after the trial is subject to funding opportunity. If the government adopts the JoyAge model as part of the regular service, it will be available to all eligible older adults in Hong Kong; if no additional funding can be secured, there is no plan for post-trial care.
The inclusion criteria for participants are (a) aged 60 years or older; (b) at risk of depression, or have sub-threshold depressive symptoms based on assessment by social workers; and (c) able to give informed consent to participate. The exclusion criteria are (a) known history of autism, intellectual disability, schizophrenia-spectrum disorder, bipolar disorder, Parkinson's disease, or dementia/significant cognitive impairment; (b) (temporary exclusion criteria) imminent suicidal risk; and (c) difficulty in communication.
Based on previous work in stepped care for depression, we separate the sample size estimation for preventive care and intervention groups. For preventive care, meta-analyses of 47 depression prevention trials revealed an effect size of Pearson's r=0.15 from pre-to-post and r=0.11 from pre-to-follow-up comparisons [[
The primary outcome is change in depressive symptom, operationalised as the difference in total scores of the PHQ-9 between the baseline and 12-month follow-up, mean difference scores between the intervention and control group will be compared. For intervention group participants, additional discharge time-point data will be collected for clinical but not research purpose. The validated Chinese version of PHQ-9 will be used [[
Secondary outcomes will be assessed at the baseline and 12-month follow-up for both groups, and at an additional discharge time-point for intervention groups for clinical service. Changes at 12-month follow-up from the baseline will be used for analysis.
Self-harm risk will be assessed using eight items adapted from the Self-Harm Inventory [[
Anxiety will be indicated by the total scores measured using the Generalized Anxiety Disorder 7-item scale (GAD-7) [[
The UCLA Loneliness Scale (UCLA-3) [[
The EQ-5D-5L [[
A locally adapted short version of the Client Service Receipt Inventory (CSRI) [[
The Hong Kong Montreal Cognitive Assessment 5-Minute Protocol (HK-MoCA 5-Min) [[
We will follow the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) statement [[
We will follow the National Institute for Health and Care Excellence (NICE) guideline [[
Costs of the intervention itself will be estimated. Intervention costs in JoyAge include costs for social workers, peer supporters, training costs, programme expenses, administration costs, and costs for office space and IT equipment and software. Due to the differences in preventive care (selective and indicated) and treatment, we separately estimate the cost for these two groups. The costs for these two groups are assumed to be varied in the costs for social workers, who are the main service providers, while the remaining cost categories, or called general resources, are assumed to be the same for all the participants. A time survey will be developed to collect the time used among social workers for preventive care and treatment groups. This will allow us to calculate the cost associated with staff time required to provide preventive care and treatment interventions. The cost of general resources will be divided by the number of participants to provide a per-participant cost. In addition to the intervention cost, the total cost calculations will include expenditures of health and social care utilisation collected through CSRI and be based on a societal perspective. We will set the time-horizon at one year, therefore costs will not be discounted. Its uncertainty will be graphically represented in the ICER plot using the bootstrap method. The cost-effectiveness acceptability curves will be plotted. Sensitivity analyses will be undertaken to ascertain the robustness of the findings.
Results of this study will be published in peer-reviewed scientific journals that have open-access options and presented at local, regional, and international conferences in relevant disciplinary fields. Topics suggested for presentation or publication will be circulated in the research team and the person making the suggestion may be considered as the lead author. Substantive contributions to the design, conduct, interpretation, and reporting of this study's data will be recognised through the granting of authorship. No professional writers will be used. Additional dissemination will be through the JoyAge website, social media platforms (Facebook, Twitter, and YouTube), and traditional media outlets including local newspapers, radio, and TV media.
The JC JoyAge project is the first large-scale study in the world to investigate the effectiveness and cost-effectiveness of a collaborative stepped care and peer support programme for the prevention and treatment of sub-threshold depressive symptoms in community-dwelling older adults.
We hypothesise that the JoyAge model will be more effective and cost-effective compared with the CAU. In the primary outcome, we hypothesise that JoyAge group participants will show greater improvement than CAU. To elaborate on depressive symptoms, we expect the indicated prevention element in the collaborative stepped care model will be more effective than CAU in preventing depression among those who are at risk or have mild depressive symptoms at baseline, and the treatment will be more effective than CAU in treating depressive symptoms among those who have moderate or above sub-threshold depressive symptoms at baseline. In addition, we expect that the ICER in the collaborative stepped care model will be lower than the cost-effectiveness threshold suggested by NICE [[
This paper details a protocol that aligns services from different units to address the fragmentation of the existing mental health system, provides a range of services to participants based on needs and changes in needs, and expands the community's caring capacity by engaging peer supporters through productive ageing activities. The evidence produced from this protocol along with future research may help to guide the reform of current mental health services for elderly mental health, and build a sustainable long-term community with higher caring capacity.
First, the study design is non-randomised and non-blinded due to several feasibility considerations. JoyAge is a complex intervention model that introduces a new collaborative workflow between two different service units, i.e., age care and mental health service units, provides standardised training to all service providers at the same time, and promotes mental health to increase public awareness at the community level. Therefore, centre level is not an appropriate unit of randomisation and stepped wedge design is not a viable option, and we choose district as a unit of comparison. In addition, community aged care centres are open to all community older adults age 60 years and over, and provide both organised programmes for members to participate and free space for them to mingle, for example, a corner to watch TV together and a canteen to have meals and chat. Therefore, randomisation at the individual level may result in contamination of the intervention effect. To sum up, a randomisation trial in the JoyAge context will need to consider the manpower, fund, stakeholder coordination, and operational limits at a whole district level, with a total elderly population of 1.27 million [[
Second, the data analysts will not be blinded. Due to the uneven sample size of the two groups and the unmatched time-points, that intervention group have three time-points, it is difficult for the data analysts to process missing data without knowing the experimental design. Third, the present study evaluates the effectiveness and cost-effectiveness of the JoyAge model as a whole instead of its individual elements, which is a complex intervention programme composed of difficult elements. It is difficult to delineate the effects of individual elements and investigate the mechanism of change with the current study design. Future studies with triple-blinding randomised design are needed.
At the time of submission (November 2020), the trial is in the follow-up stage. Recruitment and baseline data collection began in 2018 and was completed in December 2019.
The authors acknowledge the contribution of multidisciplinary researchers and service providers (frontline and management) from community aged care and mental health care services in the service co-creation process.
TSYL and GHYW led the conceptualisation of the study and participated in designing the study, developing the intervention, and application for funding. TL participated in implementing the intervention, planning the data collection, and writing and revising the manuscript. DKYL participated in implementing the intervention, planning the data collection, and revising the manuscript. SL participated in designing the cost-effectiveness part of the study. WK participated in developing piloting and implementing the intervention. LS participated in implementing the intervention and planning the data collection. All authors read and approved the final manuscript. SSKT provided expert advice in study development and training support, particularly in the area of peer support. SMN provided expert advice in study development and training support. PWCW provided expert advice in study development and training support, VWQL provided expert advice in study development and training support. JT provided expert advice in study development and training support. DFKW provided expert advice in study development and training support, particularly in the area of cognitive behavioural therapy. WCC provided expert advice on intervention development and clinical protocol development. RYKK provided expert advice in study development. The authors read and approved the final manuscript.
This study is funded by The Hong Kong Jockey Club Charities Trust. The funder has no role in the trial design, manuscript writing, or decision making for publication.
The dataset collected and/or analysed during the current study is available upon request.
Ethical approval was received from HREC at the University of Hong Kong (Reference number: EA1709021). All amendments to the protocol are approved by HREC at the University of Hong Kong and amended on the trial registries. All participants need to provide informed written consent.
Not applicable.
The authors declare that they have no competing interests.
Graph: Additional file 1. SPIRIT checklist.
Graph: Additional file 2: Table S2 : All items from the World Health Organization Trial Registration Data Set (SPIRIT item 2b).
• CAU
- Care as usual
• CBT
- Cognitive behavioural therapy
• CONSORT
- Consolidated Standards of Reporting Trials
• CSRI
- Client Service Receipt Inventory
• GAD-7
- Generalized Anxiety Disorder-7
• HA
- Hospital Authority
- HK-MoCA 5-Min
- Hong Kong Montreal Cognitive Assessment 5-Minute Protocol
• IAPT
- Improving Access to Psychological Therapies
• ICER
- incremental cost effectiveness ratio
• IMPACT
- Improving Mood–Promoting Access to Collaborative Treatment
- JC JoyAge
- Jockey Club Holistic Support Project for Elderly Mental Wellness
• MDD
- Major depressive disorders
• NGO
- Non-governmental organisation
• NICE
- National Institute for Health and Care Excellence
• NIH
- National Institute of Health
• PHQ-9
- Patient Health Questionnaire-9
• PS
- Peer Supporter
• PST
- Problem-solving Therapy
• QALYs
- Quality-adjusted life years
• TREND
- Transparent Reporting of Evaluations with Nonrandomized Designs
• UCLA-3
- UCLA 3-item Loneliness Scale
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
By Tianyin Liu; Dara Kiu Yi Leung; Shiyu Lu; Wai-Wai Kwok; Lesley Cai Yin Sze; Samson Shu Ki Tse; Siu Man Ng; Paul Wai Ching Wong; Vivian Wei Qun Lou; Jennifer Yee Man Tang; Daniel Fu Keung Wong; Wai Chi Chan; Ricky Yu Kwong Kwok; Terry Yat Sang Lum and Gloria Hoi Yan Wong
Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author