A Hierarchical Framework for Selecting Reference Measures for the Analytical Validation of Sensor-Based Digital...

Area of Science:

• Biomedical Engineering
• Digital Health
• Health Informatics

Background:

• Sensor-based digital health technologies (sDHTs) are increasingly vital for scientific and clinical decision-making.
• sDHTs provide patient-relevant data, enhance access, reduce costs, and promote inclusion in healthcare.
• Existing scientific and regulatory guidance exists for evaluating sDHTs.

Purpose of the Study:

• To address inconsistent and insufficient evidence quality in the analytical validation of sDHTs.
• To propose a hierarchical framework for selecting appropriate reference measures in analytical validation.
• To codify best practices for maximizing the value of sDHTs in public health and medical product development.

Main Methods:

• Developing a hierarchical framework for selecting reference measures.
• Defining best practices for analytical validation of sDHTs.
• Evaluating algorithms that convert sensor data into clinically interpretable measures.

Main Results:

• Identified inconsistencies in the quality of evidence for sDHT analytical validation.
• Proposed a structured approach to improve the selection of reference measures.
• Outlined a method to enhance the fitness-for-purpose assessment of digital measures.

Conclusions:

• A standardized framework is crucial for robust analytical validation of sDHTs.
• Improved validation practices will enhance the reliability and utility of sDHTs.
• This framework supports the effective integration of sDHTs into public health and medical product development.