In recent years, the U.S. Food & Drug Administration (FDA) has encouraged the use of wearable devices to support consumer health, and, in particular, to support drug development. This encouragement has included the creation of programs to accelerate approvals of such devices.
Nevertheless, given that widespread use of devices in human experimentation is still relatively new, many in the field still have questions about which devices can be used in drug development and when. Here, we discuss the regulatory landscape and explain how both so-called consumer-grade devices and medical devices fit into it.
At Koneksa, we help clients develop new drugs get to the right patients faster by deploying innovative technologies that measure patient outcomes in the real world. This means introducing novel technologies such as digital sensors or mobile apps into the regulated space of clinical trials.
Sometimes, people assume that the means we use only "medical-grade" devices and apps cleared or approved by regulators. These devices are unlike the “consumer-grade” devices with which the public may be more familiar, in that they’re developed to comply with specific regulations and are marketed differently.
However, it’s a misconception that medical grade devices are an absolute requirement for clinical trials. The FDA has clarified in a number of public forums that digital technologies can be used in human experimentation regardless of their regulatory clearance status. The CTTI recommendations on mobile technologies are an example of this stance. (These were developed in collaboration between the FDA, academia, and industry.) They state:
“In stressing the need for a specification-driven approach to mobile technology selection, CTTI strongly recommends that a technology’s regulatory status not be the sole driver in the sponsors’ decisions about which mobile technology to use. As such, while many mobile technologies are marketed and sold as consumer products, they may still be useful as data collection tools in clinical investigations.”
The regulatory path for devices in the U.S. is a risk-based system. Medical devices are categorized into three groups:
Class I. Low risk. These include consumer healthcare devices, such as fitness trackers, aimed at promoting a healthy lifestyle.
Class II. Medium risk. These include non-invasive, non-life-support devices, such as a wrist-worn actigraphy device that could measure movement or certain sleep parameters. Many clinical-grade wearable sensors fall into this category.
Class III. High risk. These include invasive or life-support devices, such as a cardiac stent, implanted insulin pump, or a companion diagnostic test that could provide information about whether a patient should receive a certain medication.
For a company to market a medium-risk / Class II device in the U.S., it must demonstrate that it is equivalent to an already-marketed “predicate” device; define its intended use (what it will measure); and specify what its approval will apply to (via a product code).
However, just because a device is cleared to be marketed does not mean that its intended use and context of use automatically render it fit for a clinical trial. The intended use and context of use (COU) must be revisited and adjusted to the objectives and endpoints of the particular study in question.
Here is an example of why such an evaluation is critically important: In a recent situation, a cardiovascular patch was deployed to assess its utility in a hospital setting. Some of the data was collected in patients with atrial fibrillation (AFib). The ECG waveform morphology and heart rate interval are different in patients with AFib than in patients who do not have this medical condition - and, as a result, the device algorithm rejected large % of the data (BMJ 2015 May 19;5(5):e006606).
In cases where device-derived data is intended to support labeling claims, 510(K) cleared devices could be a good place to start, as some assurance of device performance has been generated for the purpose of the regulatory clearance. However, while it may be a good starting point, it may not be sufficient to determine whether a device of interest is appropriate for its intended use in a trial.
This means that both medical-grade and consumer-grade devices may be used in human experimentation, if they are evaluated properly for COU. This evaluation might include a dedicated validation study, or it might include an examination of research. Either could provide information to help determine whether a device could be appropriate for the purposes of a particular study.
Consumer-grade devices have indeed already been successfully deployed in clinical trials. For instance, in oncology, they provided meaningful data indicating that decreased steps per day were associated with worse health outcomes (J Radiat Oncol Biol Phys. 2017 April 01; 97(5): 1061–1065; NPJ Digital Medicine (2018) 1: 27).
There is particular interest in consumer-grade devices in cardiology R&D; one recent opinion can be found here. Consumer-grade devices can be particularly useful in capturing certain parameters of mobility, such as step counts. (They are, however, not intended to define a specific state of health, as required for diagnosis, prognosis, response to treatment, etc.)
Furthermore, consumer-grade device data may be useful to aid general practices of care management, because data can provide confirmatory evidence for a subjective measurement (e.g., patient perception vs. sleep quality measurement).
When making a decision about which devices are appropriate to use in a specific clinical trial, the concept of “fit-for-purpose validation” is often used. An example of such an evaluation is in Clin Transl Sci (2019) 12, 247–256. This describes how a device-based measurement can be evaluated to be deemed for particular study objectives. These objectives may include internal decision-making, such as go/no go decisions on advancing a drug candidate into the next phase of development; regulatory submissions supporting labeling claims; or a peer-reviewed journal publication. (It is possible that, in the future, consumer-grade devices can be used for regulatory purposes; however, we are unaware of current examples.)
In summary: The choice of a device for a clinical trial depends upon what it is measuring, how, and for what purpose. Internal decision-making, external publication, and regulatory submission are all important steps in the process of bringing new treatments to patients. And all of them can benefit from the use of carefully chosen tools. It’s important not to ignore consumer-grade devices in these conversations: they’re inexpensive, widely available, often developed with a great deal of attention to user experience, and can offer benefit to researchers.
For additional information on this topic, please see these publications:
Izmailova ES, Wagner JA, Perakslis ED. Wearable Devices in Clinical Trials: Hype and Hypothesis. Clin Pharmacol Ther. 2018 Jul;104(1):42-52. doi: 10.1002/cpt.966.
CTTI Recommendations: Advancing the Use of Mobile Technologies for Data Capture & Improved Clinical Trials
Clinical Trials Transformation Initiative
Coravos A, Goldsack JC, Karlin DR, Nebeker C, Perakslis E, Zimmerman N, Erb MK: Digital Medicine: A Primer on Measurement. Digit Biomark 2019;3:31-71. doi: 10.1159/000500413