What will it take to get digital tools into regular use in mental health services, and what could researchers do to help make this work?
The answer is simple. Do better research and research the right things. But how?
The following is the text of a speech delivered by Mark Brown to Pathways to progress: Next steps for digital tech in mental health services at Lumen United Reformed Church & Community Centre in London on 27th November 2019
I’ve been asked to speak to you today in an effort to answer the question: “What will it take to get digital tools into regular use in mental health services, and what could researchers do to help make this work?”
You’ll be really happy to know there’s a really simple answer which it took me ages to come up with: make digital tools in mental health that work and then make sure they get to people who would benefit from them.
Job done. Easy peasy. But, of course it’s not as simple as that. What makes a mental health tool work is a surprisingly vexed question. What a mental health system actually is isn’t as simple to define as you might think. The concept of people who might benefit from digital tools in mental health is more murky than you might imagine.
Who gets to decide the future?
To put my cards on the table early on I should state my position. I believe deeply in the potential for digital technology to make people’s lives better. I am also deeply disturbed by the power being bestowed upon a new form of techno-feudal system where cleverness is taken as evidence of goodness and where real changes are made to real people’s lives without their permission or even knowledge by people who are less grounded in an understanding of ethics and the public good than your average UK secondary school child.
I am worried that digital technology has the potential to become the underpinning of a new enclosures act where what was once common space or common good is transferred to the perpetual ownership of a small number of rich techno berks. I’m also at the same time hopeful that the digital future involves choices not inevitabilities and that we can make a future where technology helps people to have better mental health.
I believe in good technology, but the question for all of us, researchers included is: what makes good technology? An atom bomb or an autonomous battlefield robot or functional facial recognition with a low level of false positives and negatives is an amazing technical achievement but I don’t think that makes it ‘good’.
In technology related to health it’s entirely possible to make a product that everyone who uses it loves, everyone who pays for it thinks is value for money but which also makes life worse for people not using it. We are not in any position to oppose a digital future in mental health, but it’s our job to make sure it’s not a shit one.
The worst possible future is the one we’re living in now: the one where people who need help with their mental health can’t get the things they want or need to help them to have a life they want to lead. For over a decade now, digital mental health has been promised as the future, back when today was the future. We’re not there yet, so it’s up to us to take action now to bend the arc of events as it stretches out into future decades.
Digital tools are already in mental health services
When it comes to getting digital tools into regular use in mental health services, they already are for some people in some places. Services like Big White Wall, Kooth, SilverCloud, Beating the Blues and Oxford VRs Gamechange immersive VR are in use in places in the UK. They’re in use because commissioners looked at outcomes and value for money and balanced them against budgets and capacity and possibility for implementation. They’re there because someone bought them in. The companies are there and still in game often through a mixture of private, charitable and public investment. Some of the providers of digital solutions to the NHS are spin outs from academic or health organisations. Some are private companies.
To a great extent, the drivers to-date for digital mental health within the UK have been internal ones, whether internal to a particular trust or CCG, internal to the health service itself or internal in the sense that policy has decreed that digital mental health should be present. One of the holy grails at this level has been to broaden effective treatment of mental illness, distress or trauma without the requirement to greatly increase spending on bricks-and-mortar and flesh-and-blood infrastructure. If we can find effective digital tools and services, then we can do more of what we need to do for less. Our current secretary of state for health and social care, Matt Hancock, is an enthusiast for this quest as well he might be as the first health secretary the UK has ever had that was also an app himself.
The challenges of this approach are numerous. Frontline health workers are very suspicious of top down technology. It changes their working life, challenges their sense of the boundaries and clarity of their profession and it often doesn’t feel like it ‘belongs’ to them. It’s something they interface with, something sold to them as a done deal, something they don’t feel like they can really alter in practice once it is implemented. In big top down digital there are three possible states of acceptance: full acceptance, cheeky work-arounds and full rejection.
In health terms we talk about how digital tools need to be better implemented by services, when what we really mean is they should be better integrated and accepted into daily practice. Both members of the public and staff are often worried that a digital tool means shunting something important away into a kind of digital siding like a kind of ‘take three of these and call me in the morning if you get worse’. Effective integration requires work far before the shiny digital thing arrives, to spend time with staff at all levels to explore and understand how, when, why and where this new thing will interface with what they already do.
At present digital tools in mental health are held in a kind of paradox. If they are powerful enough in that hey have a measurable effective element, they are practice changing enough to really change how treatment might be delivered and as such are most likely to present a perceived threat to existing structures. They become a new kind of specialism to be integrated into a team or service and as such are most likely to face resistance.
In contrast, those considered to be a nice add-on which may or may not change anything are easier to run in parallel but are also unlikely to be fully integrated because they are considered to be more in the realm of ‘stuff we might suggest to our service users, like going for a walk or joining a gym. They may or may not be promoted or recommended by staff, but they won’t necessarily be considered core to their treatment or support. Culture exists in healthcare. Without changing how staff view digital tools, they will always be considered marginal or someone else’s job.
One of the challenges for something like the NHS apps library was that while it was possible to develop a way of assessing ways in which an app would cause harm to potential users it was harder to assess what good an app might do patients.
Greater claims need far greater evidence, so apps with greater potential to actually affect someone’s health or wellbeing are far harder to recommend than apps that do something less intense but which also runs far lower risk of causing harm. Hence the proliferation of apps that teach techniques validated in other mediums of delivery and of things like mood diaries.
To solve this we need better research.
How do we build external pressure for change?
To date, external pressure to provide digital services has been slim and where it does exist it relates more to users of services hoping digital tools can be used as a way of making existing services function better. Many of these demands have been around the use of digital tools to provide better communication, better out of hours care and easier booking and tracking of treatment. A meaningful digital tool to many people who use services might be a secure way of texting or emailing services that guarantees a response. Or of adding to or amending existing records or correspondence. Or of accessing something like anonymous chat, counselling or support without recourse without waiting for Monday morning to come around.
The language of ‘do more for less’ infects nearly all discussion of digital tools in mental health and the people that it lands with hardest are people worried about getting the care and support they need. If you really want therapy and someone to talk to, ‘do-more-for-less digital’ sounds very much like you’ll be getting less of what you need. It’s another example of disempowerment and feeling like the future is being done to you.
People are aware of cuts to public services. They’ve lived through them. They’ll have lost friends during the duration of austerity. Firing a shiny service at them as if from a cannon saying ‘we’ve bought this for you’ as it hits them square in the chest won’t make them feel better. The work of beginning early with the idea of digital as a vital integrated part of a whole service doesn’t just apply to staff.
Coproduction and codesign can provide a way out of this impasse where both communities work together to build new tech, but that runs into the problem that the ‘doctors who code’ movement runs into: it’s really really hard to build really good tech that has a big potential to change people’s lives. It requires specialism as well as good intention.
Often the debate and the pressure digital services comes from early adopters. According to GP Online, the online GP consultation app GP at Hand has: “In the two years since the video consultation service was given the green light to begin registering patients outside the south-west London area where its host practice is based, it has added more than 63,500 patients to its list.
Nine out of 10 patients who have signed up since November 2017 are aged between 20 and 39 years old — and this group make up 86% of the practice’s total patient list — leading to accusations of cherry picking, which GP at Hand denies.”
The healthy white affluent tech bros and wannabee tech bros define what gets built and for whom.
I listened to a session in Utrecht last week by Tobias Gantner, a German Doctor and founder of HealthCare Futurists who also runs hackathons. He was talking about one of the vital elements of their work being getting people excited about the possibilities of technology. To that end they’d been involved in setting up a mobile health tech expo, a group of exciting looking tents and seating areas that could be set up in a shopping centre or other public space where people could come to play with new technology, watch videos and have conversations. The idea was to give people a chance to touch the possible new futures at scale.
In mental health, such a cool roadshow illustrating what we could make happen with digital technology and explaining its implications, limitations and possibilities could help to remove the mystery, fear and systems speak from digital mental health tech and show people who have the largest stake in mental health care — that is, the people who bloody need it — what could be possible if we had the will to make it happen.
Researchers need to actually research tech
So what is the role of researchers in helping the digital tech we need and want to happen in mental health? Again, I’ve got a simple answer: do better research. And to do that, I’d suggest, requires a bit of rethinking of what researching digital tools means.
One of the major challenges in both understanding and researching mental health apps is the difficulty in escaping from the mental shapes our understanding of a problem or a task forces upon us.
We are challenged by our essential conservatism in conceptual understanding of what digital interventions might do or be. We are trapped in the land of the skeuomorph, a land where an object or feature imitates the design of a similar artefact made from another material. When we are looking at digital tools in mental health we are often look to design things that retain the shape of another treatment, service or interaction. When we wish to study a digital tool or service and we treat it as if it were directly comparable to another non-digital tool. Neither captures what digital might do.
A digital tool is more complex than that, and so must our research be. We make a mistake thinking that because a digital tool exists for a medical purpose, our only role is to evaluate its clinical efficacy by deploying it as if it were a medicine or therapy. I’ve seen far too many research papers that barely describe what the digital intervention they are evaluating actually is. The app or intervention is described only in terms of what it does or the device it is on. That’s not good enough.
Research into digital tools needs to be as vibrant and detailed as the design of digital tools
Modern digital design includes far more disciplines than just a person in a room writing code. It draws on psychology, art, ethnography, statistics, linguistics, deployment, business, medicine, physical engineering, health science, politics, economics, law… The physical and code elements of a digital tool create an almost infinite range of possible objects and functions.
Research into digital tools need to combine a similar array of research and interrogatory skills. In mental health digital research we, to the best of my knowledge, don’t have many, or any, taxonomies to identify, classify and group different forms of digital interactions. Is a pull to refresh the same as a button? We all know it isn’t. One is addictive, the other probably less so. Is using a touch screen to tick a box a different experience from typing yes?
I’m picturing something as beautiful and vibrant and magical as Victorian sketch books filled with sketches of corals and flower stamen and fish gills. I’m picturing something as detailed as the system for classifying folk stories. I’m picturing researchers getting out there and classifying apps and services in the wild like they were playing pokemon go.
Until we have a research language as detailed as the language of design we are just testing washing machines for Which? magazine rather than contributing to the wider growth of human health. Research needs to tell us what to build next as well as tell us what to use now.
Research into digital tools needs to bring together the same mix of skills and disciplines that go into making digital tools.
Why is this important? The combination of interactions and processes and design decisions in a digital intervention are the same as the chemical makeup of a medicine. We wouldn’t trial a new sedative and describe it as ‘the blue one’ but I read too many digital papers (and press releases and fulminating articles) that leave out details like this when it comes to digital. If we don’t want digital to become an increasingly privatised, Intellectual Property defined walled garden of knowledge we have to make sure that research isn’t just useful to health service decision makers but to people who actually want to build and design tools and make them better.
The arrival of ubiquitous always on digital devices with constant connectivity and a range of sensors has created conditions where digital tools are less something that happens on a screen and more a complex singular event. Where someone is, what they are doing before hand, what they did afterwards, what device they’re using, what their expectations are all of these have a bearing on what a digital tool does for someone. With digital tools that are used in people’s lives rather than in controlled conditions our research is like a lab test of a medication rather than a field test. We’re asking: did this work in the artificial way we trialled it rather than in the real world conditions in which it is actually intended to be used?
I love David C Mohr’s work for exploring the use of mico randomised control trials, ecological momentary assessments and the use of sensor and interaction data to understand more fully what is going on when someone is actually interacting with an app.
With digital, we both have the capacity and the obligation to understand how different users differ from each other, rather than just how they differ from a control or a comparison group. We need to know what happens when someone uses a digital tool, not just how they measure upon our proxy for effectiveness.
We must stop surrendering the ground of making the future to tech bros
And beyond that, we must stop surrendering the ground of defining what and who is considered valuable to companies that have business models that are antithetical to the delivery of universal services. The business model of GP at Hand isn’t the same as the business model of a regular GP practice. One wants maximum sign-ups, the other wants to improve the health of the entire community it serves. This doesn’t mean that GP at Hand isn’t a great service taken on its merits, but there are broader questions to be answered about how it fits in with the broader goals and aspirations of a national health service.
We need to study digital tools as events in services and understand what their implications are for those services. And to study them as events in human communities, too.
Who and how something is funded matters in evaluating its value to those that need it. The role of publicly funded research into digital tools for mental health should be to maximise the public good that digital tools create. Not all benign interventions come free of malign consequences.
The best possible future is one where we have a range of digital tools and interventions that can be provided at a cost to individuals and communities which we can collectively afford; a future where we know who is most likely to benefit from using a particular digital tool or intervention; one where the development of and profit from these digital tools does not amount to nationalising the risk of their implementation while the profit from their operation is privatised and one where progress, development and new knowledge can be built upon rather than dying when an app, tool, service or company dies.
It might sound like I’m being scathing both towards the developers of digital tools and towards those who research them, and to an extent I am. It’s because I can see a future where the people I really care about still don’t get the future they deserve.
And that’s where you come in.
@markoneinfour
