The MindKind Study has launched in the UK- update and insights so far

We’re excited to announce that the MindKind Study (formerly known as the Global Mental Health Databank) has launched in the UK and is about to go live in South Africa and India.

As we move into the data collection phase of the study, here’s a quick update to share some of our recent learnings about working with young people, ensuring equity between sites, and some of the inevitable technical hurdles we’ve encountered.

MindKind Study Logo

About the project

For those who haven’t seen our previous updates, the MindKind Study is a 2-year feasibility study commissioned by Wellcome’s Mental Health team as part of their strategy to enable research into what works for whom, and why, in the anxiety or depression of 14–24 year olds.

The aim of the MindKind Study is to test whether it’s possible to build a sustainable and fair databank that can collect survey data alongside passive sources of data such as location and social media use. These types of data have the potential to unlock research into new types of prevention and intervention for depression and anxiety. But it is vital we develop ways to collect and share that data which work globally, enable cross-discipline collaboration, and empower those contributing data about their lives to have a voice in the process.

The work is being led by Sage Bionetworks alongside a consortium working across UK, India and South Africa. Professional Youth Advisors coordinate panels of lived experience advisors across the three countries who are collaborating directly with a range of researchers from multiple disciplines. Together they have designed a study with two key activities:

1. The development of a prototype databank which has been designed to collect data related to sleep, bodily movement, positive activities and social connectivity. The data will be collected using multiple governance models including direct download of the data, a sandbox model, participant voting on data use, community review panels, and professional review panels in order to demonstrate the efficacy of different approaches.
2. A deliberative democracy exercise with 150 remote participants exploring the issues and tensions participants perceive about data governance models, both those being tested in the MindKind study and models that aren’t yet technically feasible.

Through these activities we hope to learn:

• How to bring together different forms of passive and active data related in a form that is useful to researchers, and empowering to data contributors.
• How to embed lived experience into global research.
• What technical, regulatory, ethical and cultural factors need to be accounted for to ensure data can be collected globally.
• The relative efficacy and acceptability of different data governance approaches.

Some of our recent learnings

Lived Experience and Youth Engagement

• Communicating complex topics
  The area of data science can be complex and riddled with topic-specific jargon, which can hinder the development of a well-rounded understanding for non-specialists such as the lived experience advisors. Co-designing the material and sessions used to explain key concepts such as data governance models with the youth advisory panels has been a significant benefit. This enabled the team to create develop educational material that is consistent and engaging across different groups, for example using visual explainers and different animal identifiers for different governance models, along with models for structuring the study meetings and communications in ways that shares information in the right way and uses the platforms that are most appropriate.
• Qualitative research questions may be able to provide opportunities for more authentic, more expansive feedback from youth.
  Deliberative democracy approach has been selected because it works best with people who do not begin with deep-seated opinions and builds capacity building into the process.
• Closing the feedback loop
  After providing insights and sharing knowledge on different aspects of the project there was no systematic way for youth to know if and how their feedback had been incorporated. These concerns prompted the development of a feedback mechanism where full-time youth advisors can easily share insights with project leads, who can then provide their feedback on how information was or was not used. Subsequently, these uses can be discussed with panel members to ‘close’ the feedback loop.
• Retention of study participants through connection with researchers
  It is anticipated that there will be a very low retention rate for participants in the MindKind study and any subsequent databank that may be built. Sage Bionetworks have hypothesised that enabling more participatory forms of research that create meaningful connections between youth participants and researchers as opposed to the pure provision of data would be one way of increasing retention. The Professional Youth Advisors are therefore being further engaged in an additional piece of research to understand what forms of infrastructure could be built into the databank to enable this.

Equity between sites and team members

• Technology access
  Youth do not have equitable access to technology across the three countries. This means that the pilot, which requires access to a smartphone, will not be able to reach a representative sample population in each country. The technology requirement biases sampling towards wealthier people particularly in India and South Africa. As such, we will not be sampling a similar distribution of youth in each of the countries.
• Connectivity
  Data & broadband issues have been a challenge for youth engagement. Youth panel meetings have had to use a range of communication approaches in different regions order to accommodate a range of access needs. Paid for data plans for youth panelists and study participants in South Africa have proved to be essential in order to get the level of engagement we need. However, it also means that those participants will be incentivised to engage with the study in a different way to the other two regions.
• Multi — language research & diversity
  The research for this feasibility study is being conducted in English. This has the consequence of limiting the diversity of potential participants in India and to a lesser extent South Africa. However, translation of research material bears the risk of propagating inequities as a result of the process by which text is translated/who is doing the translating. Plus, ‘appropriateness’ of phrasing for a particular culture is subjective based on who is consulted; socioeconomic status and other factors are at play.
• Publications and authorship
  There has been a need to address the intersectional needs of early career researchers and researchers from lower resource settings in defining the consortia’s approach to authorship for formal publications. Early career researchers need to have their contributions to the project recognised through the authorship of papers in order to further their careers. In addition, researchers from lower resource settings often face bias in publishing in Western journals. However, the work has been very collaborative across the three sites and there is a need to ensure the science isn’t diluted across too many papers. The team have been working to collaboratively understand and respond to these dynamics across the team.

Technical and data insights

• Data equity
  In any databank that may be eventually built off the back of the MindKind Study, Wellcome will need to address the issues of scientific equity inherent in the access and use of data. Stepping away from punitive or restrictive measures (e.g., embargoes), how do we build a novel scientific community structure that empowers and enables researchers from lower income contexts in ways that ensure insights from a global mental health databank are not are not ‘vacuumed up’ by researchers from high income contexts? Considerations around training and capacity building need to be incorporated into further work.
• Data sovereignty & international data sharing
  There is a risk that the Indian team may have their regulatory approval denied due to the fact that the study data will move outside of Indian territory. This is indicative of a growing geopolitical trend towards data sovereignty being highly politicised. There are multiple potential outcomes to this trend including data nationalism (data being restricted from leaving national territory) and data regionalism (data moving around regions that share similar data regulations eg GDPR) each of whih will impact research of this nature in different ways.

1. Engineering vs scientific processes
   There was a push towards the development of the research protocols in tandem with the submission for regulatory review and technology development. You cannot submit to regulatory authorities until you have developed a research protocol, which means that there’s only so much tech development can be done before protocol is submitted. Also, once the protocol is submitted the need to stay aligned with the submission. These dependencies limit the natural iteration and optimisation cycle of mobile development for research.
2. Challenges of collecting unified demographic data across regions
   There is no global standard for measures of socioeconomic status. The team have reflected that they spent too much time exploring regional-specific differences in the demographic variables that should be included for the different sites. They have stated that if they were to do this again, they would want to consider a minimum set of demographic variables to collect.
3. Assessment of active ingredients in the real-world setting
   Determining how to assess active ingredients in a real-world setting without overwhelming participants with too many surveys, and without being an intervention has been an ongoing challenge.

The MindKind team is already developing rich and valuable insights into the crucial field. We’re trying to share as much as possible as it happens so keep en eye on Wellcome Data for further updates.