09 Sep 2019
Session Block—Standards and Applications 15:20 - 15:40

This presentation will demonstrate how Activinsights has used open standards (OASIS COEL & Kantara Consent Receipt) to build privacy-by-design infrastructure for the collection of lifestyle data using wearables. While this presentation will demonstrate a specific and in-market application, we will show how the lessons are transferable to any Internet of Things (IoT) devices automatically authoring behavioural data.

Activinsights has 10 years' experience in the measurement of behaviour with a precision, raw data wearable (GENEActiv) and the analysis of this data with increasingly sophisticated algorithms capable of identifying a very wide range of personal behaviours. These technologies are being used globally in research applications such as national cohort surveillance programmes and clinical trials. 

However, in non-research applications such as the management and prevention of long-term conditions (e.g. diabetes, obesity, heart disease, frailty), this level of data collection is unnecessarily intrusive and has high privacy risks. For these applications, a wearable with an alternative design approach (Activinsights Band) uses edge computing techniques to minimise data source by strictly controlling the data types the device can record and report. These data types are specified within the OASIS COEL data model - a hierarchical taxonomy of everyday living.

This novel approach pseudonymises all information about patients at source, ensuring that only the clinician holds the directly identifying personal data. This pseudonymised data can then be shared with other service providers in the ecosystem - the mechanism by which we process lifestyle data in parallel to the physiological health and medical data of health service providers (e.g. NHS). Data are transferred by the wearable using short-range wireless technologies through a hub application to a COEL-compliant storage facility. The data can then be queried by service providers with appropriate authorisations and permissions.

The record of patient notice and consent is captured in the OASIS COEL architecture, as a specific event coded using the Kantara Consent Receipt and with the option of attaching consent references to all behavioural events. The standardisation of data types and coding provides complete data portability.

We will describe how minimising data at source provides the greatest gains in creating a dialogue with patient about privacy. This combination of privacy enhancing technologies with transparent governance and clarity about purposes forms the backbone of achieving informed consent.

Finally, we will discuss some of the future opportunities and challenges, including providing patient-centric data access, directly addressable IoT devices and the best approaches to further anonymisation for safe data sharing.



Discussion not started yet.