Study design and objectives
The Indigo-Local feasibility pilot study aims to:
- Develop a community-based public awareness programme intervention (Indigo-Local) that involves training community health workers (or similar cadres of workers) and mental health service users, alongside a media campaign, designed to: i) reduce stigma and discrimination, and ii) increase referrals of people with mental health conditions for assessment and treatment.
- Implement and pilot the Indigo-Local intervention in a small feasibility (proof-of-principle) platform activity using a pre-post mixed-methods study design in seven sites in five LMICs, to evaluate procedures for a subsequent fully-powered study comparing the clinical and cost-effectiveness of Indigo-Local in: i) reducing stigma and discrimination amongst trained community health workers (or similar cadres of workers) and service users, and ii) increasing mental health service uptake.
Setting
The Indigo-Local feasibility pilot study is being carried out in seven sites in five LMICs [31, 32], i.e. two sites in China (Beijing and Guangzhou), two sites in India (Bengaluru and Delhi National Capital Region), Ethiopia, Nepal and Tunisia. See Table 1 for further details about the study setting/location for each of the seven sites. The study sites have been selected based on accessibility, appropriateness and feasibility, and where possible entail a distinct region or neighbourhood.
The Indigo-Local intervention is being implemented in community settings within the seven study sites, such as in public spaces or community facilities. The training elements of the intervention are being conducted either within health, community, private or work spaces as appropriate, depending on the local contexts. For ethical reasons, mental health services need to be in place in the settings in which the Indigo-Local intervention is being implemented, given the likely stimulation of and anticipated increase in help-seeking.
Table 1 around here
Participants
A wide range of stakeholders will be involved in the Indigo-Local feasibility pilot study in each of the seven sites. This may include local key stakeholders such as health service leaders and/or members of service user organisations, community health workers or similar cadres of workers, (mental) health and/or site staff, service users and their caregivers. Table 2 shows an overview of the different study activities that each of the participant groups are involved in.
Table 2 around here
Inclusion and exclusion criteria
All participants of the Indigo-Local feasibility pilot study will be at least 18 years of age and have to freely consent to participate. We will review mental capacity to consent where a concern is raised, but seek to respect preference of the service user in all cases. For all groups, sampling further aims to achieve adequate sample variability with regard to gender and age group of participants. Further details about participant eligibility are outlined below in the section on the key components of the Indigo-Local intervention.
The following groups of people are excluded from the study, where relevant: health workers who do not have appropriate government credentialing for their cadre or who have any known professionalism infractions or legal infraction revoking professional licensure. This is screened by site research teams when approaching potential participants, and involves for example only recruiting participants from legitimate and licenced health care facilities, where staff are required to meet appropriate regulations and professional standards. We are also excluding anybody who is at risk of a psychiatric emergency, who may not be able to provide consent, or who may not be able to perform the intervention and research activities.
Recruitment
All participants of the study will be sampled purposively by each of the site teams. All participants will be identified and approached by either the implementing partners in the country sites, or by the local health service leaders or similar key stakeholders, to engage them to participate in the study. Where possible, contact regarding the study will be conducted by an impartial third-party individual (i.e. not the participants’ clinician [for service users] or staff managers [for health workers], but instead, for example, a recruitment officer, research assistant, PhD student, or clinic administrator, depending on site resources).
Sample sizes
Sample sizes will vary between the seven study sites, depending on feasibility and the local resources available, as well as the size of the site – see Table 1 for further details. We plan to recruit a minimum number of ten community health workers (or similar cadres of workers, depending on the local context) and service users in total for training in each of the seven sites. If possible, of the total number of participants recruited for training, a minimum of 15-20% should be service users in each of the sites (and the rest community health workers or similar cadres of workers). All trained community health workers and service users should ideally be involved in the quantitative evaluation of the Indigo-Local intervention, and a sub-set of them are taking part in the qualitative evaluation per site (depending on site feasibility). In addition, two to three people will receive the ‘Training of Trainers’ training in each of the seven study sites, and between five and 20 participants will take part in the stakeholder group workshop per site (depending on local feasibility).
Since the Indigo-Local study is being conducted on a proof-of-principle feasibility basis, sample sizes for the quantitative evaluation elements are not guided by power calculations, but by a minimum of 30-50 participants. The intention is not to formally test for pre-post differences in the sample, but we will instead examine the effect size and direction of change, which could guide the sample size for a future full-scale study. Further evaluation data will be collected through qualitative means, for which the sample sizes outlined are appropriate.
Indigo-Local intervention
Principles guiding the Indigo-Local intervention
See Box 1 for the principles guiding the Indigo-Local intervention as its ‘essential ingredients’, based on the Amaudo Mental Health Awareness Programme in South-East Nigeria[29] and other work since then [14, 36, 37].
Key components of the Indigo-Local intervention
The key components of the Indigo-Local intervention are outlined below. Each of these key components will be carried out in each of the seven study sites.
Training of Trainers
The plan for future Indigo-Local interventions is for the ‘Training of Trainers’ (ToT) to be conducted for five days residential, whereby master trainer(s) are trained to train people to conduct the community health worker / service user training. This ToT training should include a direct (e.g. a service user provides a ‘lived testimony’ in person) or indirect (e.g. showing a video of a person talking about their experiences) contact element with service user(s). However, since this is a feasibility study with small sample sizes and since the teams in each site are mental health stigma experts with prior knowledge on the topic, in this study an online ToT programme is being carried out, in which the Indigo-Local research leads train site teams to conduct the community health worker / service user training in around one day through a series of online training videos and seminars.
A minimum of two to three people should take part as recipients of the ToT training in each site. These participants would be expected to have mental health knowledge, since the training they are being trained to deliver is focused on the methods for delivering messages about mental health conditions and mental health services, rather than teaching much about mental health/conditions. Recipients of the ToT are taught to train the community health workers (or similar cadres of workers) and service users to share mental health related messages in community forums (e.g. community meetings), for example to give advice about the location and availability of mental health services (including opening times), referral methods, follow-up and monitoring of service users in the community, and the costs involved. The training also includes a brief overview and materials to understand effective implementation strategies for the intervention.
Stakeholder group workshop
A stakeholder group workshop is being conducted for the duration of half up to one full day in each of the study sites. In each site between five and 20 participants are joining, including relevant local stakeholders, such as health service leaders, members of service user organizations, local community groups or NGOs, community workers, health staff, service users, traditional healers, religious leaders etc. Local health service leaders are selected based on the following characteristics: they should hold a leadership role at their institution within health services in the site, ideally within mental health services (or have a good working knowledge of mental health issues). Any other local stakeholders should be people or groups who advocate and manifest the interest and will of mental health service users in the community, or who are engaging or supporting people seeking mental health care.
The aims of the stakeholder group workshop are to: (a) bring all key stakeholder groups together to establish the project team, build relationships, and ensure buy-in from the beginning; (b) advise on the local context, training needs and the local media landscape; (c) review, refine and adapt the training materials and translate them into the local language (where needed/appropriate) – for consistency and fidelity, the material templates have been developed centrally (based on the materials used in the Nigeria study, provided by Amaudo [29]), which allows for sharing of evidence-based practice; however, these materials are being adapted by each of the sites to cover local cultural beliefs and specific issues related to the area of intervention; (d) plan and define the media strategy and clarify its messages; (e) help in planning the training, including identifying which cadres of workers to train – it is crucial that this is done carefully to maximise the efficacy and retention of those trained, and involves defining in advance what is expected, post-training, of the trainees (e.g. to hold community forums, to identify and refer patients in their community etc.); and (f) help in planning the implementation of the intervention, including refining details of the intervention to match local services, resources and needs, and deciding on the most appropriate way(s) to raise awareness in the community. The stakeholder group workshop builds on detailed formative work already completed previously in study sites as part of the Indigo Partnership [31, 32].
Training of community health workers and service users
Community health workers (or similar cadres of workers) and service users are being trained over a minimum of two days (for resource-limited settings) up to an ideal maximum of five days in each of the sites. Training could be conducted over successive days or in separate blocks over a few weeks, depending on feasibility and the local context within sites. At least ten participants in total per site will be trained, within or near their local communities.
Community health workers or similar cadres of workers who are trained are selected based on the following characteristics:
- they should be well-respected members of the local community;
- should know their communities well and be intimately familiar with community cultural perspectives;
- should be familiar with community education and mobilisation;
- should be part of existing cadres of personnel if possible, for instance Accredited Social Health Activists (ASHAs), female community health volunteers (FCHVs), government officers, faith-based group leaders etc.
Careful choice of such workers was found to be crucial for good results, coordination and sustainability during the previous Amaudo programme in Nigeria [29].
Eligible mental health service users to be trained alongside community health workers (or similar cadres of workers) can include any person seeking care from and using a mental health service. We expect to involve people with a range of diagnoses from common mental illness (depression, anxiety) to more severe mental illness (bipolar disorder, psychosis) or harmful substance use. These service users who are being included as recipients of the training should be able, willing and feel safe to discuss their own experience of living with a mental health condition as well as their own mental health service use. Ideally this should be somebody from the local community, though service users from elsewhere can be involved if necessary (recognising that for some, speaking in their own community may pose greater challenges or risks). In sites where this is deemed to be appropriate and beneficial service users’ caregivers may also be involved in the training.
The training is facilitated by the recipients of the ToT within or near their local communities, who train the community health workers (or similar cadres of workers) and service users (and possibly their caregivers where appropriate). Community health workers and service users should ideally be trained together to reinforce the social contact element of the training (in that case, both groups will likely need to be briefed before and debriefed after the training), but if this is considered not to be possible or good practice in sites (e.g. because of power dynamics, social hierarchies etc.), the two groups could also be trained separately. In addition, sites identify service users in the community (or who at least have the same language/culture), who could contribute to and co-facilitate the training by providing a ‘lived testimony’. If such direct in-person contact is not possible during the training, then the social contact element could also be done through indirect contact, for example video or online materials that could have been developed previously (e.g. Time to Change Global [36-38] or other locally relevant materials).
The training content includes mental health and stigma, awareness-raising, i.e. how to spread messages of mental health (services) in the selected community, and how to conduct outreach and referrals (for which the pathways will be contextualised by sites). See Table 3 for further details on the training content. Sites have the flexibility to culturally adapt it and complement it with contextually relevant information from other sources.
Table 3 around here
Implementation of intervention
The trained community health workers (or similar cadres of workers) and service users will then implement an intervention (i.e. locally contextualised awareness-raising activities/engagement) in the local community within each of the sites. This can be embedded within their usual role/ community engagement activities. The exact awareness-raising activities vary across sites depending on the local context, but may include community contact activities, speaking to community groups (e.g. faith locations, women’s groups, youth groups etc.), or at events or locations such as markets.
Supervision meetings / booster trainings
Supervision meetings for the trained community health workers (or similar cadres of workers) and service users will take place every two to three months, with brief booster trainings after three to six months and six to 12 months (if feasible in sites). Process data, for example on their level of activity in regard to mental health awareness-raising, could be collected as part of these sessions. Ideally these supervision meetings and booster trainings will be conducted by the same people who conducted the initial training.
Local awareness-raising media campaign
A media campaign is being conducted over a minimum of a one-month period (ideally longer), which starts at the same time as the training of the community health workers and service users. The format and messages of the media campaign depends on what is feasible and appropriate within each of the sites, but may include posters, flyers, newspaper articles, social media (WhatsApp, Facebook, Instagram, Twitter etc.), announcements or jingles in local radio or television etc. At least two different media outlets should be used in each site – see Table 1 for further details on this for each of the study sites.
The media campaign is being developed by the local site teams according to the local context. The content of the campaign is framed and phrased as such that it will aim to help increase public knowledge and improve attitudes and awareness around mental health conditions, and inform the community about the availability of mental health services. The messages are linked to services and to the content of the training activities (e.g. myth-busting, information about available services etc.). The campaign messages should include an (indirect) social contact / ‘lived experience’ element, such as a video or an interview with persons with lived experience (for which there are good examples available [36, 39]).
Evaluation of Indigo-Local intervention
The evaluation of the Indigo-Local intervention will be conducted as a feasibility (proof-of-platform) pilot study using a mixed-methods design. This will involve quantitative evaluation of stigma outcomes; quantitative evaluation of mental health service utilization rates (where feasible in sites); a qualitative evaluation; a process evaluation; an implementation evaluation; and an evaluation of implementation costs. These aspects are each described further below. An overview of these evaluations along with the time points for their assessment are provided in Table 4 (adapted from the SPIRIT flowchart; a populated SPIRIT checklist is provided as additional file [40-42]).
Table 4 around here
Quantitative evaluation of stigma outcomes
This involves pre vs. post assessment of quantitative scales to measure stigma and discrimination (in terms of knowledge, attitudes and (intended/expected) behaviour) amongst the community health workers (or similar cadres of workers) and service users who receive the training, using the following quantitative questionnaires:
• Changes in knowledge about mental health conditions: The ‘Mental Health Knowledge Schedule’ (MAKS) [43] will be completed by the trained community health workers and service users. The MAKS has 12 items, which are each scored on a five-point Likert scale, with higher scores indicating higher levels of knowledge.
• Changes in (intended/experienced) behaviour:
- The ‘Reported and Intended Behaviour Scale’ (RIBS) [44] will be used to assess changes in intended behaviour by the trained community health workers. The RIBS contains eight items across two sub-scales, which are rated either as ‘yes/no’ response or on a Likert scale, with higher total score indicating higher willingness to interact with a person with lived experience of a mental health condition.
- The shortened version of the ‘Discrimination and Stigma Scale’ (DISCUS) [45] will be used to assess changes in mental health service users’ experience of stigma and discrimination. The DISCUS has 11 items, which are rated on a four-point Likert scale, with higher scores indicating higher levels of discrimination. Service users who take part in the Indigo-Local training component will complete the DISCUS.
• Stress: The shortened 2-item version of the Stigma Stress Scale [46], for completion by the service users who take part in the training. Higher scores indicate higher levels of stress due to stigma, with total scores ranging between -6 and 6.
• Changes in attitudes towards people with mental health conditions: Social Distance Scale (SDS) [47, 48], for completion by the trained community health workers (or similar cadres of workers) and service users. The SDS has 12 items, which are each rated on a six-item Likert scale, with higher scores indicating greater social distance. This scale is optional rather than obligatory for sites.
All scales have been validated and used in earlier studies [43-48]. They have already been adapted and translated by the site teams locally as part of the formative work within the Indigo Partnership [31, 32].
All scales will be completed at several time points (see Table 4). As a minimum, these data will be collected immediately before (Time 1) and after (Time 2) the community health worker / service user training. If feasible in sites, at least one further follow-up point will be included, ideally at three or six months (Time 3). Further follow-up assessment time points (e.g. at the time of the booster training sessions at six or 12 months) are optional depending on feasibility within sites (Time 4).
Quantitative evaluation of mental health service utilization rates
Where feasible in sites, this will be conducted to test the difference on mental health service utilization rates of the Indigo-Local intervention. In sites where this is feasible and appropriate, routinely collected quantitative data will be used to assess (at site-level) the following (or similar/related/proxy) outcomes:
- Total number of ‘new referrals’ to mental health services by the community health workers who participated in the training (e.g. by comparing to one-year pre-intervention);
- Total uptake of mental health services, including total number of service users seen by mental health services (and % change), and new referrals to mental health services (and % change);
- Contact coverage (defined as service utilization taken from the programme records divided by the total population in need of services taken from prevalence surveys of the disorder), where feasible, i.e. where adequate data is available in the scientific literature for the site about the number of people who require mental health services (to act as denominator of contact coverage) [49].
If feasible, routine data should be collected (retrospectively) on a monthly basis for one year before the Indigo-Local intervention is implemented, and then on a monthly basis for a minimum of one year after the intervention is implemented (to assess the long-term impact of the intervention and also the impact of the supervision meetings/booster trainings).
Where feasible, data will be collected on previous referrals of patients, as well as on referral pathways / how referrals are made, for example referral by community mental health workers, self-referral following the media campaign etc.
Qualitative evaluation
A qualitative evaluation will be conducted to complement the quantitative findings, i.e. to obtain in-depth qualitative data from community health workers and service users on the potential effectiveness of the Indigo-Local intervention in terms of stigma (knowledge, attitudes, behaviour) reduction, mental health service utilization rates (including referral rates), and the impact of the intervention amongst participants who received the training to deliver the intervention. The following will be explored qualitatively: (1) ways to improve the training; (2) changes in stigma, including possible explanations for changes in the quantitative outcomes/lack thereof, based on the directions of change observed; (3) information around possible changes in mental health service utilization rates; (4) other outcomes not covered by the quantitative measures, including any possible negative, unintended consequences. This will be done through focus groups and/or semi-structured interviews, ideally immediately after the training (i.e. Time 2) and/or at the end of the intervention period (i.e. Time 5); the data collection approach will be selected based on feasibility and appropriateness in each study site.
Process evaluation
In addition, a process evaluation will be conducted at site-level, to record the exact implementation details of the Indigo-Local intervention in each of the sites. For this, process indicators (e.g. information on how many times the community health workers / service users are involved in awareness-raising activities and the types of activities, attendance details of training etc.) will be collected using a specially-developed Excel file.
Implementation evaluation
Implementation of the Indigo-Local intervention will be evaluated at site-level with members of the seven site research teams. Semi-structured interviews will be carried out by the Indigo Partnership project coordination team with the research teams in each of the implementing sites at a minimum of one time point post-intervention. These interviews will collect information on the site teams’ implementation experiences and perceptions of the facilitators and barriers to implementation. These interviews will be framed around an established implementation strategy framework, the updated Consolidated Framework for Implementation Research (CFIR) [50, 51].
Data for this will be analysed descriptively. Patterns in these data will be explored across and within sites, based on data of what types of implementation strategies were used, and how many strategies were reported to be used. Data on implementation facilitators/barriers will be synthesised narratively, guided by content analysis and thematic analysis principles.
Evaluation of implementation costs
A cost analysis will be undertaken that will estimate the quantity of resource inputs and costs associated with intervention implementation activities across the seven study sites. This will draw on data supplied by local site leads who will complete a costing pro-forma designed specifically for the Indigo pilot evaluation. This asks for quantitative information on staff time inputs, local pay rates and financial expenditures recorded against key implementation activities. The design of the pro-forma has been informed by an activity-based costing approach to assessing the cost implications of implementing health programmes, as outlined by Cidav et al [52].
Estimates of total implementation costs and costs related to broad categories of implementation activity will be presented by study site. Costs will be presented in both local currency values and in US dollar purchasing power parity (PPP) adjusted values using appropriate PPP conversion factors published by the World Bank [53].
Data management
REDCap [54] will be used for entry of quantitative data, with response fields for all items (including respondents’ socio-demographic characteristics, site characteristics, and outcome variables). In each site a member of the local research team is identified, who is responsible for local data collection and data entry. The coordinating team at King’s College London will then export data from REDCap for data checking and cleaning.
Data analyses
The suitability of the measures will be examined, for instance for their distribution, and ceiling and floor effects. This is in line with the aims of this being a feasibility (proof-of-principle) pilot study.
For the quantitative data analyses, descriptive summaries such as total scores and simple counts will be performed, which will then be compared at the different time points, as well as the % change before and after the intervention is implemented (using chi-square tests). Primary and secondary outcomes will be analysed using mixed effects linear or logistic or Poisson regression models depending on the data type accounting for clustering due to repeated observations at three time points (Times 1, 2 and 3) in each site. Regression results will be pooled across countries using random effects meta-analysis, with a test for heterogeneity of regression coefficients being summarised using the I2 statistic [55]. All data analyses will be conducted with the use of STATA 17.
For the qualitative analyses, focused framework analysis based on the themes in the topic guide will be carried out, with some thematic analysis principles also applied with further bottom-up codes generated by sites where applicable and site teams identifying select key illustrative quotes to enrich the data. The focus groups and/or semi-structured interviews will be audio-recorded and transcribed verbatim before being translated into English (where appropriate) and then analysed.