91³ÉÈ˰涶Òô

Consumer digital interventions can be used to support the delivery of safe and effective healthcare for a range of physical and mental health conditions. However, not all interventions are developed to the same quality, and publicly available health apps can be potentially harmful. Risks include harmful rather than therapeutic instructions, inaccurate biomedical calculations, and unauthorised sharing of health data.

Various accreditation standards have recently emerged, however these largely rely on accurate disclosures by developers – which may not always be accurate. This project will explore and evaluate new approaches to support the objective assessment of the quality and safety of digital interventions, supporting approaches to assess risks quickly and at scale.

Aims

The overall aim of this project is to investigate how digital tools and data analytics can be used to improve consumer and clinical confidence in the use of digital health interventions. This will be achieved through these specific aims:

1. To develop and pilot automated data capture techniques, to reliably capture multi-modal data from publicly available interventions.
2. To assess the concordance between information derived from the objective data, and other cited information (e.g. developer disclosures, expert assessments).
3. To identify new risks associated with emerging consumer health technologies.

Design

The first part of the project will scope software frameworks and map their capabilities to support appropriate data collection use cases. A proof of concept will be developed to validate multimodal data collection, e.g. text based descriptions, image-based screen recordings, and raw network traffic.

A validation study will be conducted using a sample of highly ranked or clinically endorsed digital health interventions available to the public. The sample may include targeted physical or mental health conditions. Data collected will be synthesised and compared against other official documentation, e.g. regulatory intended use cases, privacy policy declarations, and expert opinions.

A qualitative sub-study may be undertaken to understand consumer and clinician attitudes to the findings, or which aspects of intervention quality should be prioritised.