New Consent and Policy Solutions - for Better Trust and Consumer Control

MedInfo 2023– the 19th world congress on Medical and Health (8-12 July 2023, Sydney) – is a prestigious international event that will bring together thousands of digital health leaders and practitioners at the forefront of healthcare.

Best Practice Software is pleased to present a new approach for computable consent and policy solutions at this event – as an enabler for establishing better trust and consumer control in an increasingly interoperable and AI-aware digital health ecosystem.

Best Practice Software is working on this approach to allow consumer consent expression in a way that would support their integration with clinical and administrative workflows in the increasingly interoperable digital health ecosystem. This ecosystem involves many existing and future health service delivery partners, as well as new providers specialising in analytics and AI solutions, as we presented earlier in [1]. The ecosystem also requires giving consumers more control over how their health data are used, for the primary or secondary purpose, the latter of which are research and clinical trial applications.

The approach has been in the making over the last year or so [2], as a result of the recognition of Best Practice Software for a need to better support future digital health expectations of clinicians and consumers. These expectations are in terms of helping with increasing their trust in the use of new digital health services, including CDSs and generative AI, while ensuring that the existing and future regulatory and medical principles are respected. The aim is to develop an approach that can best accommodate new and changing rules coming from the regulatory, legislative, and organizational governance structures.

The essence of this approach consists of the ability to represent consent and other policies in terms of computational rules that, on one hand, closely reflect natural language expression of policies while, on the other hand, can be flexibly integrated with interoperability solutions, such as HL7 FHIR, as shown in the figure below. These computational rules are based on the latest scientific and technical proposals for the expression of the concepts of obligations, prohibitions, permissions, delegation, authorization, and broader accountability concepts while leveraging relevant ISO and HL7 standards [2].

At MedInfo23, Best Practice Software will demonstrate how the concept of informed consent is positioned as part of the broader, ‘computable policy’ layer and how it can be used with the HL7 FHIR standard while harmonized with the information security policies – to ensure confidentiality, integrity, and availability of healthcare data and systems. This will be done through a clinical trial use case.

The significance of this approach is broader when considering an increasing number of new AI solutions, both general and domain-specific generative AI solutions. Their applications in the medical domain will no doubt require meeting privacy, ethics, safety, and other regulatory policies, and the ability to express related policy rules in a computable fashion facilitates faster, more predictable, and reliable digital health systems for both clinicians and consumers.

Best Practice Software will be pleased to share this approach with the international digital health community at MedInfo23 and discuss further improvements with interested colleagues. In the meantime, Dr. Zoran Milosevic and Dr. Frank Pyefinch would be available for early discussions, and they can be contacted at zoran.milosevic@bpsoftware.net and Frank.Pyefinch@bpsoftware.net.

References
[1] Dr Frank Pyefinch, Dr Zoran Milosevic, The New Generation of Practice Management Systems – From Interoperability to AI, BP Evolution Newsletter, Feb 2022, https://bpsoftware.net/new-generation-of-practice-management-systems/

[2] Milosevic Z, Pyefinch F, Computable consent – from regulatory, legislative, and organizational policies to security policies, Proc. of the EDOC 2022 Conf, Lect. Notes in Computer Soc., Springer, p.3-18