Recommendations for Defining and Reporting Adherence Measured by Biometric Monitoring Technologies: Systematic Review

Area of Science:

• Clinical Trials
• Digital Health
• Wearable Technology

Background:

• Suboptimal adherence to procedures or interventions often leads to clinical trial failure.
• Biometric monitoring technologies (BioMeTs), including wearables, can capture adherence to digital therapeutics and data collection.
• BioMeTs offer opportunities to identify adherence determinants and develop strategies to maximize adherence.

Purpose of the Study:

• To describe methods and definitions for capturing and reporting adherence using BioMeTs.
• To identify gaps in current BioMeT adherence reporting.
• To provide recommendations for minimum reporting requirements and definition consistency for BioMeT adherence data.

Main Methods:

• Systematic review of studies published between 2014-2019 using BioMeTs outside clinical settings.
• Inclusion criteria: quantitative, nonsurrogate, sensor-based adherence measurement reported.
• Extracted data on study design, participants, BioMeTs, and adherence definitions/units; categorized adherence as continuous (duration/measurements) or categorical.

Main Results:

• 100 out of 940 manuscripts met eligibility criteria, describing 110 BioMeTs.
• 30% of studies failed to report sensor-based, quantitative adherence measurements.
• 37 unique adherence definitions were identified; uniformity increased with data resolution (continuous time-based definitions used in 92% of cases).

Conclusions:

• Recommend reporting quantitative, nonsurrogate, sensor-based adherence data for all BioMeTs.
• Advocate for clear descriptions of sensors, algorithms, and validation data for BioMeT adherence.
• Prioritize reporting primary adherence data as a continuous variable, with categorical definitions supported by validation.