My Device, Your Data: The Concept of BYOD
Harnessing patients’ own smartphones as part of a trial is a growing trend but the road to full adoption may be a little bumpy.
The concept of BYOD (bring your own device) in clinical trials isn’t new. Encompassing both patient- and clinician-reported outcomes, BYOD also offers the tantalizing potential of gaining that information in real time.
However, it is the use of patients’ devices to collect self-reported outcomes data that has seen the fastest growth in recent years, says Barbara Tardiff, Executive Consultant (Visionary Strategy Design and Execution) at Drug Development Informatics LLC and previously VP, Development Operations at Pfizer. “The discussion and impetus have really accelerated recently,” she says. “It isn’t mainstream yet, but it is definitely a reality.”
BYOD expert Kara Dennis, Managing Director of Mobile Health (mHealth) at Medidata, agrees that the concept has reached “an inflection point,” noting that one-third of the company’s mobile health trials include a BYOD or hybrid BYOD component. In these trials, subjects can use their own smartphone or tablet to complete study-related tasks or if they do not have a qualifying device, they can be provided with one.
Pharma companies are gaining confidence to explore the use of electronic Clinical Outcomes Assessments (eCOA) right across clinical research, says Tim Davis, CEO of digital engagement and data capture specialists, Exco InTouch. “Initially, we saw use of BYOD focused on large multinational trials with simple data points, however, more recently, we have begun to see use increasing across all phases and in other therapeutic areas with appropriate trial designs,” he explains.
The convenience factor
With benefits ranging from cost savings to more streamlined enrollment processes, fewer training and support queries and a heavy reduction in set-up activities, BYOD’s growth is hardly surprising. “Patients are not burdened with an additional, conspicuous device, and they benefit from the convenience and familiarity of using their own technology to interface with the trial,” says Davis. “This ready accessibility has been claimed to improve study participation, compliance and data accuracy.” BYOD offers scalability with the potential to service studies that would not be cost-effective for a fully provisioned approach, he adds.
According to Dennis, research showing how closely patients keep their phones to them and how often they check them supports the concept. “There is an ease-of-use benefit as well as a more seamless integration into a patient’s daily life.” Collecting data via a patient’s own device can boost quality, says Tardiff. “We are more likely to get data contemporaneously, it will be more complete and the patient will feel ownership of the data and its quality. It is important to take these things into consideration. Indicators suggest that data captured on an electronic platform is comparable to that captured on paper, and also from one electronic platform to another.” The economic impact cannot be ignored; BYOD can represent significant cost savings for trial sponsors when compared to device-provisioned trials. “There is the potential to reduce purchase or rental costs, handling, shipping logistics, inventory logistics and inventory management. The logistical challenges – ensuring there are enough phones at each site, for example – and the cost and effort associated with provisioning oversight are reduced when patients can bring their own devices,” she says.
Balancing the benefits are some concerns, which Tardiff places in two distinct categories; concerns around reliability (that data collected via patient-owned devices are valid and equivalent to those collected via traditional platforms) and technical concerns (especially around security and the applications or platforms used).
While the technical problems have largely been solved and do not represent a major roadblock, establishing the reliability of the data is a larger concern. “Data is the lifeblood of clinical trials and there is little point executing a study if the data isn’t going to be accepted and believed,” says Tardiff.
Trepidation is highest around endpoint data, particularly primary endpoints, although “there are some studies out right now that show people are beginning to grow more comfortable with the idea that you can accurately capture data with wearables.”
Privacy remains a concern, says Tardiff, who outlines a scenario where a trial sponsor could potentially view a patient’s location via his or her smartphone. If they saw that the person was in a hospital, for example, the sponsor could potentially send a message to ask if the patient is ok. “Some patients are more concerned about that than others – do you want somebody to know where you are? It can be a definite advantage but a subset of patients may be uncomfortable. That is something that needs to be worked through.”
However, knowing a person’s location could be of huge value in some studies, for example, where there is a mobility endpoint. “People carry their phones with them pretty much everywhere they go, so it’s a way of understanding how mobile somebody is, how much they move around or how often they leave the home,” she says.
Passwords, credit cards and the phone-less
Seemingly minor problems encountered during a trial can become major stumbling blocks, for example, if a subject changes their device midstudy or upgrades their operating system. “There are lots of pitfalls; with a BYOD trial, patients have to download and log into a mobile app. This sounds simple, but for some trials and some research populations it is not. Patients may not remember their iTunes or Google Play password, which can bring the subject enrollment process to a screeching halt. Research coordinators have told me that they are not an Apple Genius bar.”
[Regulators] all have smartphones, they have Fitbits, so they are interested and they want to be supportive but only if they feel they can fulfill their responsibility of ensuring the data is compelling.
Phone capacity is another issue, as BYOD trials can require subjects to download a new app or upgrade their device to the latest operating system, says Medidata’s Dennis. Trial sponsors certainly don’t want to tell people to delete photographs of their grandkids if they don’t have space for the latest upgrade. That said, there has seen a “significant reduction” in the number of research coordinators reporting such problems since 2014. “Every six months, you can see a marked improvement.”
A key challenge to the use of BYOD in a trial is a concern that using different devices and operating systems will affect the measurement properties of a validated assessment, says Davis. “Traditionally, the measurement properties of an assessment have been tested via usability and equivalence studies on a specific device type with a uniform screen size. With BYOD, a different approach needs to be adopted, whereby a minimum screen size requirement is specified and equivalence is proven using a variety of different device models and screen sizes.” This type of research is relatively new and will inevitably take time to gain acceptance in the eCOA community, he adds.
Finally, Dennis points out that not everyone owns a smartphone or device. “There are challenges associated with trials in therapeutic areas such as schizophrenia or Alzheimer’s disease, where subjects may not have their own smartphones or may not be familiar with the process of downloading and updating an app.”
A low regulatory hurdle
Last year, the FDA sought public input on the use of technology in clinical trials to see if its role as a regulator was limiting the application of tech in trials. According to Tardiff, the regulators “see the writing on the wall”. She says: “They all have smartphones, they have Fitbits, so they are interested and they want to be supportive but only if they feel they can fulfill their responsibility of ensuring the data is compelling.”
Such interest from the FDA is a clear indication of greater acceptance of and support for BYOD as an appropriate and beneficial means of eCOA data collection, says Davis. “However, while recognizing that there are potential benefits in the use of technology in clinical trials, the FDA wishes to learn more about the effects it has on patients, in particular, their acceptance of the methods and issues related to privacy and data protection.”
There is currently no guarantee that any data will be accepted as part of a submission, regardless of how it is collected, he adds. However, if the data collection is implemented correctly and in line with guidance there should be no reason why using a patient’s own device won’t be acceptable.
The proof of the pudding
All stakeholders will come on board “with time”, says Tardiff. “You have to do studies and prove the approach is robust. Ultimately, what’s going to convince people is seeing the data and seeing drugs studied using a BYOD approach getting approved.” “Ultimately, [pharma companies] are afraid of doing a study and then finding that the data is not acceptable – that it is not deemed trustworthy or there is no confidence in it because of concerns about how it was collected. The ultimate proof is that the data are accepted for a marketing registration,” she says, adding that the data does exist but is still quite limited.
Another, more insidious problem is company culture – while some may be early adopters, others may hesitate to make changes in their processes and approaches, including the use of wearable devices. “Some companies will only adopt something when literally everyone else has already done so,” says Tardiff.
While recognizing the enormous potential of BYOD, some level of device provisioning is likely to be required for the foreseeable future, says Davis. “This may be for the inevitable portion of trial candidates who do not own a suitable device or where device integration is required with next generation monitoring devices, such as activity meters and adherence monitors. The move towards eCOA and the addition of Bluetooth wireless technology across smartphone devices now enables integration of physiological readings into a diary and other outcomes reporting from patients. This can be done using BYOD in some circumstances, but leading-edge monitoring devices often use the very latest Bluetooth protocols, which means they only tend to be compatible with up-to-the-minute high-end smartphones. Sponsors need to recognize when to apply a BYOD approach and when provisioning would be a more efficient model.”
Hybrid BYOD could be where the near future lies, says Dennis. “This area saw the biggest growth in 2016 because sponsors don’t want to exclude patients based on what smartphone they have. It is a very interesting indicator that sponsors are willing to explore BYOD, as long as they have a back-up in place.”