m-hotopf-2

Predict, pre-empt, prevent

Research company Gartner, Inc. projects that 275 million wearable electronic devices – from sports wristwatches to smart garments – will be sold worldwide in 2016, generating revenue of USD 29bn (~€26bn). Making use of the explosion of data this will bring is RADAR-CNS (Remote assessment of disease and relapse – Central Nervous System), a major new research programme being supported by the Innovative Medicines Initiative (IMI) which will assess the potential of shop-bought wearable devices and mobile technology to help treat three significant brain disorders:

1)        Depression, a mental health condition which affects some 350 million people worldwide and is the leading cause of disability;

2)        Multiple sclerosis (MS), a chronic, disabling disease which affects two million people globally and causes disability, problems with balance and co-ordination, fatigue, sexual dysfunction and spasticity – to name just a few symptoms; and

3)        Epilepsy, a non-communicable brain disorder which affects approximately 50 million people and is characterised by recurring seizures.

The programme is being led by King’s College London, UK, and Janssen Pharmaceuticals NV, and brings together 24 organisations across Europe and the US in a consortium which draws from such diverse fields as engineering, data analytics and clinical research.

Improving care delivery

“There are essentially two questions we’re trying to address,” says Matthew Hotopf (who co-leads the project with Vaibhav Narayan from Janssen), a professor of general hospital psychiatry and director of the NIHR Maudsley Biomedical Research Centre, UK: “Can smartphones and wearable devices provide clinically useful information about a person’s current disease state, and can they predict if that patient is going to have a relapse of some description?”

The programme has been inspired by patient surveys which have repeatedly highlighted the need to predict when relapses will occur and to improve the treatments that are available to stop them from happening, as well as the sometimes arbitrary nature of clinic-based care.

“There’s quite a lot of redundancy in terms of how frequently people are seen in clinics and when, and clinicians are often forced to rely on an individual patient’s account of how they’re doing to write their medical reports or make any kind of judgement about whether they need to adjust medication etc. – information which can oftentimes be unreliable and inaccurate. If you compare epilepsy patients’ reporting of seizures in the last month with an objective marker of seizures, the results are very different, for example. So we’re trying to give the clinician a better idea of the patient’s current state,” Hotopf says.

The hope is that phone and wearables data will provide a more complete, accurate picture of the patient’s condition, thereby alerting the clinician to any change which requires medical intervention before the patient’s health deteriorates or a relapse occurs.

“It can be very difficult to actually catch the point at which a patient with depression is starting to have a problem,” explains Hotopf. “But if you can detect something like a change in sleep pattern or activity level, then you might be able to step in early and prevent them from having a full blown relapse.”

To this end, RADAR-CNS will collate data from already available, inexpensive technology – things like accelerometers and voice sensors on smartphones or wristwatches – collected as the patient goes about their daily life.

“That will give us information about an individual’s circadian rhythm, and from that we can work out how much sleep a person is getting and how active they are, which can be pretty informative in terms of a disorder like depression or MS.”

GPS data will meanwhile allow clinicians to map where people are and what they’re doing there, which will allow them to make inferences about a person’s sociability, etc. Hotopf is also hopeful that movement sensors will be able to help clinicians detect, reasonably reliably, whether someone is having an epileptic fit.

“A change in someone’s voice, i.e. slurring, can be an early indication of something like a possible MS relapse, so we’ll be using smartphone voice sensors, too,” he adds. “We’re also trying to look at cognition, which will probably involve memory or cognitive tasks, and symptoms, so we’ll ask people to complete brief questionnaires from time to time.”

Putting the patient first

Feedback will naturally be vital to the programme’s progress. One of several clusters of activity is as such dedicated to evaluating the user experience. This will see a series of focus groups launch over the next three to six months in which patients with each of the conditions will be asked about what they’d find helpful, which are the most important symptoms to target, and what the constraints might be in terms of things like privacy, etc. They’ll also provide input as to how best to engage patients with the technology itself.

“We’re putting a lot of effort into the user experience and patient involvement,” says Hotopf. “If we do reach the conclusion that these technologies are useful, then we want to know that we’ll be able to actually implement them in a sensible way and that the patients will make use of them – asking the right questions during our research will be crucial.”

A second workstream is meanwhile focusing on clinical care pathways, i.e. what are the critical time points at which a clinician would actually want to receive this sort of information?

Elsewhere, a separate cluster of workstreams is busy on one of the first steps of the programme: working out what the system requirements will be. RADAR-CNS will make use of already accessible hardware, but deciding which devices will be used (“We’re trying to set something up which is device agnostic so that we can ‘plug and play’ as we go forward.”) and developing the necessary analytics platforms to make sense of the data being generated will provide a real challenge.

Beyond that, says Hotopf, “the real transition point will come next autumn when we start the three main cohort studies actually trialling the use of these technologies.”

Looking ahead

Although early days (the programme only launched in April), there is already hope that it’s results could be rolled out to help treat other chronic disorders – like dementia, for example. “Where someone is living in supported accommodation, it would be easy enough to track how well they’re functioning and identify if they’re going to need extra support at a certain point in time,” Hotopf says. “There are also a lot of ways that you could build interventions – e.g. prompts about drug adherence – off the back of the technology.”

For the moment, however, the programme is focusing exclusively on epilepsy, multiple sclerosis and depression, three debilitating conditions which come with sizeable health and economic burdens. Depression alone is estimated to affect one in five people during their lifetime and carries with it an annual cost to Europe of some €92bn, yet the significance of the problem is still going unrecognised. If successful, RADAR-CNS could provide a vital first step in making a reality the improved care and treatment that’s so urgently needed.

 

Professor Matthew Hotopf

Co-lead

RADAR-CNS

www.radar-cns.org

This article first appeared in issue 20 of Pan European Networks: Science and Technology, available here.