By Laura Khurana and Karin Beckstrom, ERT
There is increasing interest in incorporating data from wearable devices (activity and sleep trackers, heart rate monitors, smartwatches) into clinical trials. This is mainly due to the advantages sponsors, sites, and patients can recognize from their use. Sponsors can gain additional data to better understand a treatment’s effect on the participant. Investigative sites can use the data to monitor the patients’ health and to gain insight into the patients’ engagement. And, patients can potentially see metrics about their life and health they wouldn’t otherwise have access to.
As existing wearables mature and as new products enter the market, the possibilities for their use in pharmaceutical research may be endless. By integrating varied data streams captured from wearables with electronic Clinical Outcomes Assessment (eCOA) data (symptom frequency and occurrence, quality of life, etc.) captured by the patient on handheld and/or web interfaces, sponsors may soon be able to gain complete insight into the patient experience during clinical development.
However, there are many factors trial sponsors must consider before including wearables in their clinical trials. Here we present some of the scientific and operational considerations.
Data from wearable devices can best serve the scientific objectives of a trial when they are:
Fit for purpose. Trial sponsors should let the decision to use wearable data be driven by their ability to support the trial’s endpoints (or serve as independent endpoints themselves), rather than letting the choice of data collection technology drive the scientific questions of the trial.
In context. Data about sleep patterns, activity level, and physiological measures such as heart rate are only meaningful if you know the context in which this data was collected. Integration of wearable technology with eCOA enables the real-time capture of context surrounding a clinical event.
Accurate and validated. Data from wearables that were originally created for commercial use may not be subject to the same standards of accuracy as medical devices vetted by the FDA. For example, a recently filed class-action lawsuit alleges that Fitbit® misrepresents the accuracy of the heart rate monitor in its wearable devices – this lawsuit was filed after university researchers attempted to validate the Fitbit monitor against an electrocardiogram (ECG) and found an average difference of 20 beats per minute.i Other efforts are underway to validate wearable or smartphone data with existing COAs, including an app developed by Roche that uses smartphone sensor data to measure symptoms of Parkinson’s Disease and attempts to predict patient scores on the Unified Parkinson’s Disease Rating Scale (UPDRS).ii
Beyond the scientific considerations of wearable devices, there are several operational aspects a trial team needs to consider when adding a wearable into the protocol, including:
Prior to looking at any devices, sponsors need to determine what data is needed. There are hundreds of wearables on the market but you can quickly narrow the scope by knowing exactly what data you want to collect and for how long you want to collect it. An important consideration is if the data will be submitted to regulatory agencies, and if any regulatory precedent has been established for their use in clinical research.
When people refer to wearables they often mean activity meters but the wave of innovation in wearables means there are options beyond just activity, including many that have multiple sensors. For example, Vital Connect has a wearable patch that collects single-lead ECG, heart rate, RR interval, respiratory rate, temperature, body posture, fall detection and activity including steps.
Similarly, you can select the form that best meets your trial population’s needs. Maybe a wrist worn device works fine but perhaps a shirt like Hexoskin offers would be a better fit for your trial participants. Researchers are also working on minimal, tattoo-like health monitors that could be used in the future.iii
Part of the patient centric movement in clinical trials is to understand the impact that trial requirements have on a participants’ life. This is especially important for wearables, as an uncomfortable device or one that requires extensive set up can lead to loss of data and non-compliance.
Data from some wearables can be quite extensive, so it’s important to have a plan about what is a useful collection interval. Collecting a data point for every step may be more than you need, an alternative might be steps per minute or per day.
The possibilities for wearables in clinical research are endless, but not all will meet the objectives of a given clinical development program. Having plan that addresses the scientific and operational considerations of wearables and optimizes the integration of data sources such as eCOA will lead to a smoother roll out, a higher likelihood of success, and potentially significant improvements in efficiency and accuracy during new drug research and development.
Laura Khurana, MPH is a Senior Scientific Advisor, ERT
Karin Beckstrom is a Product Manager, ERT Innovation Lab