MentalHealth4All: mapping and assessing existing multilingual resources in mental healthcare

The aim of this article is to illustrate the goals, process and results of the first stage of a European project: MentalHealth4All. Partners from 9 different European countries have worked together to design a multilingual validated repository targeted at third-country nationals with limited language proficiency (LLP TCNs) and healthcare providers, which contains resources to improve this population’s knowledge and understanding of how to access interlinguistic support in mental healthcare services in the main languages of the countries involved in the project. The present article provides a detailed overview of the assessment process of resources and its main results. Abstract


Introduction
Health care, and mental health care in particular, deals with delicate issues that create thresholds for anyone struggling to seek help.This is even more the case for those who find themselves in a foreign country and are unfamiliar with the local language or culture, while coping with mental issues, for example, because of what they experienced in their countries of origin (Miteva and others, 2022;Montemitro and others, 2021).MentalHealth4All 1 is a project within the AMIF (Asylum, Migration and Integration Fund) framework, which aims to remove or reduce language and cultural barriers to mental health care for third-country nationals in the European Union.
Mental health problems affect about 84 million people across the EU (European Commission, 2022).Refugees and other migrants are particularly at risk of developing mental health problems due to stressors they encountered before, during, or after their migration process.Prevalence rates for some mental health problems, such as posttraumatic stress and psychotic disorders, are higher among refugees and migrants compared to non-migrant populations (WHO, 2021).Even though many refugees and migrants need treatment for their mental health problems, they have great difficulties in accessing mental healthcare services, particularly those who do not speak the dominant language of their host country (Ai Ohtani and others, 2015;Brisset and others, 2014).Furthermore, if they succeed in gaining access to mental healthcare services, treatment often lacks cultural and linguistic sensitivity, resulting in inadequate care (Lázaro Gutiérrez and others, 2022).
In this paper, we will first present the main objectives of the project to then describe the first stage, which is the main focus of the article: the design of a multilingual validated repository, in order to map resources which are already at the disposal of third-country nationals who are speakers of vehicular or colonial languages (such as English, French or Spanish) or languages of lesser diffusion (i.e., languages that do not carry wide currency within a specific society) and who are looking for information on how to access (language) support in mental health services.The overall aim of this repository is to identify good practices as well as gaps in this type of information provision.

The MHealth4All project
In this project, 13 partners from 9 European countries developed, tested, and implemented platform in mental healthcare.This platform aims to reduce the negative impact of linguistic and cultural barriers in accessing mental healthcare services for refugees and migrants.The platform is based on evidence regarding the availability of high-quality language resources, stakeholders' needs, and barriers, and proven effective intercultural communication strategies.Refugees and migrants, their caregivers, interpreters, intercultural mediators, and mental healthcare professionals and organizations are the main stakeholders to benefit from this new resource.
The project is composed of a consortium in which all partners have complementary skills and expertise and ample experience in the field of migrant healthcare for LLP.The consortium spans 9 countries (the Netherlands as coordinator, Belgium, Germany, Italy, Lithuania, Poland, the UK, Slovakia, and Spain), ensuring a wide geographical spread across diverse European regions.The languages included in the multilingual digital platform will be Dutch, English, French, German, Spanish, Italian, Polish, Lithuanian, and Slovak.In addition, Arabic and Turkish will be included, as these represent languages spoken by the two largest migrant groups in Europe, viz.Moroccan or Turkish nationals or people of Moroccan and Turkish descent (Eurostat, 2023).Modern Standard Arabic is also well understood by nationals from the Middle East, e.g., Syria, ensuring a multiplier effect that extends beyond the countries included in our consortium and increases the platform's reach across Europe.Other languages for which reliable automated translation solutions are available will be added too (e.g., Russian and Ukrainian).The project started on 1 January 2022 and will run until 31 December 2024.

Resource gathering and classification
For the first stage of the project, which is considered an exploratory phase, partners performed an initial online search of existing resources meant to bridge language barriers in mental healthcare in their countries, which they had to report using Google Forms.Different types of resources could be collected, such as interpreting and language services, patient educational materials, and culturally sensitive mental healthcare services.
Inclusion criteria were broad at this stage and partners were encouraged to consider any resource addressing migrant mental care delivery, the organization of healthcare systems, translation, and interpreting (T&I) services, multilingual treatment possibilities, guidelines, policies and regulations.Additionally, partners contacted different organizations assisting migrant users or providing mental health services to the general population and asked for convenient reference resources that matched our research goal, identifying good practices as well as gaps by means of an assessment grid (see §3).
The following specific information was required for each resource: 1. Name of the resource or brief description.
As can be seen, some of these resources were created outside of the European Union, but they were still included in the repository due to their potential for applicability to other contexts and/or countries.
On the basis of this preliminary data gathering, a classification proposal was designed to organize a total of 471 resources (see table 2).Each resource was attributed a classification code, which included a general tag describing the topic addressed by the resource (i.e., MH, which stands for Mental Health); the target audience (i.e., P=Patient, D=Provider, "doctor" being used as a hypernym for the category of healthcare providers; B=Both); the country where the resource was created (e.g., SP=Spain, IT=Italy, UK=United Kingdom), and a classification number (e.g., MHP-POTC-1 → resource aimed at patients, created in Poland and providing information about therapy clinics/centres/specialists).

Assessment grid
To evaluate resources, the Delphi technique was chosen, as it is a well-established validated method to ensure good decision-making on quality matters (Moher and others, 2010).A group iterative process involving experts from the consortium from different specific fields of knowledge (psychology, psychiatry, psychotherapy, linguistics, languages, translation and interpreting, and intercultural communication) was followed, as suggested by Monsen and Van Horn (2007), to assess resources using a previously designed assessment grid.
As previously proposed by Santamaría Urbieta and Alcalde Peñalver (2021), the assessment grid was designed in two stages.First, a literature review process was completed considering relevant references in the field that were related to the objectives of this tool.As the main goal of MHealth4All is to provide a multilingual, culturally sensitive platform of high-quality mental health resources, our selection of pertinent literature to elaborate the assessment grid was performed within three research fields: (1) evaluation of resources in translation and/or interpreting (e.g., Nevado and others, 2020), (2) evaluation of resources aimed at improving communication between healthcare providers and (migrant) patients in (mental) healthcare (e.g., Sandín Vázquez and others, 2012), and (3) evaluation of digital and online resources (e.g., Leacock & Nesbit, 2007).These three topics were considered in-/exclusion criteria: all materials were first checked for compliance with either of these topics, and only those that passed this test went on to the following stage.For the second stage, we identified seven constructs that were adapted to fit our research aim and served to elaborate the first draft of the assessment grid, i.e., reliability, accessibility, target audience, current relevance, user-friendliness, and linguistic features (see table 3 for descriptions of the categories that were elaborated from the constructs).Subsequently, in an iterative process of expert meetings (involving experts from all consortium partners), the classification and its description were further refined (also with an eye for the inclusion of differences between countries to show that the selected resources are culturally sensitive in two ways, targeting the specific needs and responses of each national society).The final version of the assessment grid can be consulted in table 3.

93
In the following section, the procedure used to implement the grid is described.

Assessment waves and stages
To facilitate resource evaluation, specific instructions were issued to help assessors across project partners to navigate the assessment process, since each project partner was to set up a team of assessors.Depending on the size of the first batch of submitted resources by each country and based on an Excel document pooling all collected resources, the team coordinating this research phase assigned assessors a list of resources to evaluate.Assignment was principally meant to align a team's country to the country of origin of the resources it was to assess, but, in some cases, this was further expanded with other regions for which expertise was available within the team.This led to the following distribution of assignments: Netherlands (n=112), Belgium (n=71), Spain (n=80; Spain + Colombia), United Kingdom (n=57), Lithuania (n=33), Slovakia (n=27; Slovakia, Australia, Czech Republic), Germany (n=28; Germany, Canada, Luxembourg, South Africa), Italy (n=34, Italy, European Union and the United States) and Poland (n=29; Poland, Hungary, Unknown).AFFORDABILITY Describe whether all contents are free of charge; some contents are free, but you must pay to access others; or you must pay to use the resource.

ACCEPTABILITY
With own resources or consulting partners' contacts, assess linguistic quality (terminology, spelling, grammar; adequatetranslations, if applicable, etc.); cultural appropriateness for the target culture, and appropriateness for mental healthcare service.Still following the Delphi technique, an assessment was performed individually by different evaluators, who were asked to provide information about their profile (i.e., academic background, position, and other relevant details) and had to be identified and codified.Table 4 offers specific information concerning assessors' profile per country/partner.In general terms, it can be observed that most evaluators shared a background in Humanities, with a special emphasis on Philology, Translation, Interpreting or Applied Linguistics, whereas a smaller number of participants had pursued studies in Psychology and Communication Science.Concerning their job position, the pool of evaluators gathered academics at different levels (i.e., professors, lecturers, assistant professors, post-and predoctoral researchers).In a subsequent step, evaluators copied the scores obtained and pasted them in the "Resources" tab of the Excel spreadsheet, together with their evaluator code.This process was repeated until all the waves were completed.The program automatically calculated the final ranking and classified resources in different bands: band 1 (10-15 points, high-quality resource), band 2 (5-9 points, moderate-quality resource), band 3 (0-4 points low-quality resource).Depending on the final score, numbers automatically received a colour.
After the first wave of assessment, all resources received a score by at least one evaluator.As the total number of resources in band 1 was higher than 75, subsequent evaluators of waves 2 and 3 concentrated solely on the evaluation of all the resources in this band.In some cases, evaluators suggested that certain resources scoring fewer than 10 points should also be considered for inclusion (e.g., they were excellent resources that obtained lower score mainly because they scored low in linguistic availability).In such circumstances, the partners were advised to include a comment which allowed subsequent evaluators to assess those specific resources.
This process was repeated by different evaluators until the three waves were completed.The final step was to assess the acceptability and affordability of those resources that made it through to the final list of resources resulting from the previous two-stage procedure in the first, second and third waves (i.e., resources in band 1 and, in some cases, resources that could be considered for inclusion despite having obtained a lower score).
Two experts in psychology provided specific instructions on how to assess the status of a mental healthcare service by checking whether it was licensed/registered.This also included information on whether the professional title of a mental healthcare profession is legally protected.This is, for example, not the case in the Netherlands and Germany, where everyone is allowed to provide mental healthcare, even though one may neither be properly trained as a psychologist or a psychotherapist, nor have the correct credentials to do so.
To verify whether the mental healthcare service/organization/provider was trained and/or registered, partners were asked to check the website of the organization and look for information about the organization's staff team members with titles and/or logos of professional associations.If, after these checks, partners were still unsure whether the mental healthcare service was properly licensed, they were asked to e-mail the CEO/head of the organization to ask whether they were licensed/registered in a professional register.If these checks were insufficient to establish if the mental healthcare service was licensed, the resource was to be excluded from the database.
Similarly, two experts in translation & interpreting designed general guidelines to assess the cultural appropriateness of the resources for the target culture.This included considering the adequacy and adaptation of measures (e.g., pounds, kilograms, etc.), symbols, images, drawings, and cultural references (e.g., traditional food, religious practices, humor, jokes…), health beliefs (e.g., western vs traditional medicine), and sensitive or taboo topics that can be considered offensive in the target culture (e.g., sexuality).
Rather than being assigned to a single individual, acceptability and affordability were assessed by a group of evaluators that had to reach agreement on both criteria.
Unlike in previous assessment stages, specific scores were not assigned to these criteria, but a description of both was added to each resource in the online repository instead.

Final selection of resources
Considering the criteria previously identified, partners shared work at their convenience to complete the pilot wave, as well as the first, second and third waves.As these were implemented, the initial number of resources (n=471) was progressively reduced with each wave (see table 5).In case of discrepancies, evaluators were asked to reach consensus and, thus, decide whether a particular resource should be considered for evaluation in subsequent waves or, rather, omitted completely.

98
It is essential to highlight that, in some cases, resources were discarded because they did not meet the inclusion criteria established in the first stage (e.g., they only addressed healthcare in general, thus not specifically addressing mental health topics), whilst, on some other occasions, they were no longer available and, as such, were removed from the database.This is particularly relevant for some resources that were specifically created on a temporary basis for Ukrainian refugees.In a similar vein, links had to be regularly adjusted because some websites were down or updated.Furthermore, similar resources created by the same organization were sometimes merged into one to avoid redundancies.
Resources that were classified in band 1 after the third wave of assessment were consequently evaluated in stage 3.As previously described, in stage 3 a group of evaluators focused on assessing the items acceptability and affordability, which in some cases led to a further reduction of the number of resources to be included in the platform (see table 7).

Output
It is paramount that MHealth4All and its outputs are sustainable, and that the latter are made available to a large audience in a sustainable fashion.For that purpose, the main outcomes of the project, such as the database of resources, will be kept published and updated by consolidated institutions once the project finishes.The Amsterdam Center for Health Communication (ACHC) and the European Network for Public Service Interpreting and Translation (ENPSIT) have offered their website and staff for this purpose and will oversee this task.The online repository is designed into two different sections: one for healthcare providers and one for patients.These two main sections are subsequently divided into specific country pages for each of the participating partner countries.
The presentation of resources follows a template which serves as the basis for the interface, consisting in a browser operated both by filters and tags.The template includes the following closed fields, which will be searchable by means of menus: -Category (Patients, Health providers) -Subcategory (Therapy clinics/centres/specialists; Organisations/NGOs offering legal/social/psychological support; Training, education opportunities and materials; Guidelines, brochures or information sheets on psychological well-being; Glossaries/ dictionaries/terminological databases; Guidelines for providing culturally sensitive (psychological) care to migrants/refugees; Support hotlines; Apps/videos on mental health; Guidelines for providing psychological care to the general public; Information about access to (mental) health care services; Pictograms/translated questionnaires for the medical consultation; Guidelines for working with interpreters and/or intercultural mediators in mental health/socio-medical settings; T&I/intercultural mediation services).
-Country -Language -Availability (at local level, at regional level, at national level, worldwide) -Affordability (free of charge, not for free) The following fields are open: -Name -Link -Description (includes nature, purpose, focus, provider, and other comments included by researchers about the resources) It will be possible to browse the name and description fields after the tagging process.For this purpose, keywords will be extracted from their content using corpus analysis software.

Conclusion
The repository design stage of the MentalHealth4All project, a database of multilingual resources, has already been produced and has been available at https://mentalhealth4all. eu/ since January 2023.It currently contains 112 resources in 120 languages and dialects and The database will be complemented further with new resources developed within Mental-Health4All, such as sets of guidelines and multilingual videos.The use of the current database will also be monitored to check for quality and allow improvement, and at a later stage the database will be evaluated by patients and healthcare providers for its user-friendliness and usability.Future project plans include, apart from an updating schedule, a regular and progressive translation process of the whole interface into the most widespread migrant communities' languages.

TABLE 3
Assessment grid for mental health resources Stage

1. Exclusion criteria Stage 2. Items TOPIC Any of these should apply:
• Mental health, focusing specifically on migrants and refugees.•Guidelinesfor mental health providers.•Interpreting in mental health settings.Special Issue XIII -Present and future challenges for public service...: 84 -101 R.