Monday, August 22, 2016

‘Too much weight placed on FDA approval of wearables,’ says UK health research expert

In a rapidly evolving digital landscape, the disruptive innovation of mobile technology, applications and so-called ‘wearable’ technology could be a complete game-changer for the healthcare sector. However, the potential for wearables and mobile apps to be used in clinical trials is raising all kinds of questions about legality, privacy and accuracy.

Matthew Simmons, Head of the Drug Development Unit at the Sarah Cannon Research Institute, believes that too much focus has been placed on the need for U.S. Food and Drug Administration (FDA) approval of these devices and applications.

“Clinical trials are based on analyses of often large numbers of patients, where trends and comparisons between groups are more important than an individual’s data. If you look at, for example, the heart rate on a particular day – the same margin for error applies to everyone, making it a ‘precision vs. accuracy’ problem that can be solved by statistics,” said Simmons, who will be a panellist on disruptive innovation at this year’s Partnerships in Clinical Trials Europe conference.

In fact, the FDA last year announced it would only try to regulate devices making specific medical claims, or having some risk involved with its use. This is why well-known consumer health wearable Fitbit, which has collected over two billion minutes of data over the past four years, will not face FDA scrutiny. Despite controversy over Fitbit’s ability to accurately monitor heart rates and sleep cycles, and the availability of other FDA-approved wearables on the market, it continues to be the wearable device of choice for clinical trials.

Background data collection 

Like many others, Simmons sees the future of wearables as a great potential tool for the collection of what he terms “ambient” real-time data for clinical trials.

“Like ‘white coat hypertension’, the very act of actively measuring something has the possibility of changing the observation, but where the data is collected in the background with no additional effort, there is the possibility of not only more, but better quality data” he explained.

As he points out, a ‘smart’ pill bottle sending an alert to the patient to let them know they have missed a dose is better for a clinical trial in terms of compliance than collecting retrospective data at monthly clinic visits and finding out that a dose was missed the previous week.

‘Gamification’ and the risks of wearable data 

Of course, the collection of accurate background or patient-generated data often relies on the patient being incorruptible – something that strikes fear into the heart of any medical researcher.

“There is increased variability inherent with activities occurring outside the doctor’s office; how do we know that the patient hasn’t put the Fitbit on his wife when she walks the dog to ‘play’ the system and improve his score?” Simmons asks. “You can’t lock it on his wrist like some sort of prisoner’s tag, but perhaps more tamper-proof options will become available, such as self-adhesive patches where the circuit is broken if removed.”

The ‘gamification’ of health data collected from consumer devices means researchers will have to first consider whether patients might try to cheat, either to make their ‘scores’ look better to themselves, or to compete with others.

“Gamification is often cited as a great driver of user engagement, but are points and badges the correct motivator in this scenario? And do they motivate all people in the same way? Many reports suggest not” he explains. “Score a point for correct participation in the trial itself by all means: attend the visit, return your unused meds, fill in the questionnaire, but perhaps not for things that would affect outcome measures – like ‘how many steps can I do today’?”

‘My team is not a helpdesk’

Another downside to the adoption of mobile apps and wearables for clinical trials is the increasing drain on time and resources it imposes on the researchers themselves.

“Already my clinical research site team have to work with multiple systems, platforms, portals and passwords, many of which seem incapable of talking to one another in any meaningful way,” said Simmons, adding that the development of standards would be critical for the use of new technologies to reduce additional workload burden on site staff. This additional workload can also include training the end user, i.e. the patient.

As Simmons states quite frankly, his team are researchers, not a helpdesk for patients with app problems.

“We need to be careful that ‘wearables’ (as a broad descriptor) do not shift resource burden from sponsor/contract researcher on to the site, as already seen with electronic data capture and risk-based monitoring implementations, unless accompanied by additional funding.”

Privacy and data ownership 

An FDA report published in July indicates that ‘large gaps’ exist in the regulation of health apps and wearable technology when it comes to consumer data protection.

Source: Wearables in mHealth Technology Survey Report by SCORR Marketing
Many popular wearables and apps are not covered by regulation, leaving consumers with little control over how their data is used. Simmons believes that although people should have access to their own data and be able to understand how it is used, popular concerns over data protection may be overemphasised.

“People are happy to click ‘accept’ at the end of an unread 20-page license agreement for a shopping app or online game, potentially sharing all sorts of data with all sorts of people, but claim to have massive concerns regarding personal health data,” he said, adding that as long as no personally identifiable data is transmitted or stored remotely, the practice should not be objectionable.

The massive amounts of data generated by background apps and wearables leads to another difficult question – do we actually know how to make the best use of it? Simmons says that part of the value of so-called ‘big data’ is the ability to make connections between sets of data to find correlations that the researcher might not have expected to find in the first place.

“This is somewhat contrary to clinical trial methodology where all analyses must be predefined,” he said. “Big data gives us the ability to find answers to questions we may never have thought to ask.” 

Matthew Simmons will joined over 120 other industry leaders speaking at Partnerships in Clinical Trials Europe 2016 in Vienna. Find out about the 2017 event at

No comments: