2018 Autumn Abstract: Innovative Uses of Technology for Measuring Outcomes in Clinical Trials – (Parts 1 and 2)

Chairs:
(Part 1) Philip Harvey, PhD; Tiffany Farchione, MD
(Part 2) Richard Keefe, PhD; Mike Davis, MD, PhD

Technological solutions are increasingly being considered to improve efficiency and validity of outcomes measurement in clinical trials and to increase tolerability of assessments by participants.  These technological solutions are quite wide ranging.  At their most straightforward level, they replace standard paper and pencil performance-based assessments with computerized solutions, delivered remotely or in-person.  Solutions involving active and passive measurement technologies are also being used.  Active measurement solutions include ecological momentary assessment (EMA), which interacts with participants on either a time-based ( e.g., 9 AM) or event-based (e.g., sends alerts when leaving the house) level.  Passive measurement includes wearable devices such as latest generation actigraphy and home-based surveillance including motion sensing, computer usage, and activity patterns inferred from light and power consumption. 

Active and passive measurements can be applied to treatment recipients and informants/observers in clinical trials.  These strategies can also be used to prompt and monitor adherence to pharmacological interventions, which could increase the intrinsic validity of clinical trials.  They can also serve to eliminate recall failures and reduce response bias, by contemporaneous collection of outcomes data.  Thus, technology-based solutions have the potential to resolve many of the major problems in clinical trials.

With these solutions come challenges.  Technology based performance assessments need to manifest the same standards for reproducibility and stability as paper and pencil approaches previously approved by regulatory agencies.  Passive observations, although likely to be optimally tolerable to participants, still require psychometric characteristics that are consistent with signal detection and validity.  Within the same study using passive measures, remote cognitive assessments, or EMA, there can be multiple potential outcomes.  The sheer volume of data that can be collected with these devices requires sophisticated statistical approaches, because even if a single channel such as activity is selected as the outcome, devices can collect information on steps per day, hour, or minute, as well estimations of caloric expenditure and bursts of activity.  When this assessment is expanded to collection of additional passive measurement data such as sleep and heart rates, the complexity increases exponentially.

Clinical trial designers are faced with an array of choices when using technology.  These include design issues (use technology for all outcome assessments; alternate technology-based and in person assessments; use technology to deliver the equivalent of an in-office assessment) and selection of measurement strategies (e.g., active vs. passive; in person vs. remotely delivered).  Further, some recent trials involve using technology to deliver interventions as well as measure outcomes. These trials need to ensure that there is no conflict between the technology used to deliver the interventions and that used to measure the outcomes.

This full day symposium addresses a wide array of conceptual, practical, measurement, statistical, and regulatory issues associated with the use of new technologies to assess cognition, symptoms and functioning in clinical trials.  Presenters will represent the viewpoints of the pharmaceutical industry, American and European Regulators, device developers, and academia. An important additional feature will be a non-promotional discussion among device developers and the attendees regarding challenges faced in the process of developing assessment-related medical devices and validating their use prior to deployment in clinical trials.  This panel will be wide ranging and will cover the array of assessment strategies described above.