Position Statement: The Professional Record Standards Body (PRSB) on AI and Health Information Standards 

The Professional Record Standards Body (PRSB) recognises the transformative potential of Artificial Intelligence (AI), machine learning and natural language processing in reshaping health and care. Interoperability is key to enabling the potential benefits of AI to be realised more fully. It must be possible to exchange information between systems, to understand the context of the information being exchanged and information must be structured and coded in a way that it can be analysed. But interoperability has been a major challenge.

AIDespite the drive to structure and code data in a way that it can be used to support analysis for both direct care and secondary uses through the development and promotion of standards, we are some distance behind where we should be. Information is still commonly being shared in formats that require additional interpretation and coding. There is a huge opportunity to advance interoperability using AI to interpret and organise certain data held in clinical databases into standardised formats in suitable circumstances. This would enable better sharing of data between systems and make more data available for AI tools supporting clinicians in their practice for example for diagnosis, triage, clinical decision support and interpreting vast amounts of data from medical devices such as continuous glucose monitors augmented with other data from the electronic record.  

  

As we embrace this digital evolution, it is imperative that information standards evolve in tandem to ensure safety, accuracy, and interoperability. This position statement deals specifically with AI as it relates to our information record standards.  Our position is captured in the following points which will be used to guide our work: 

 

  1. Commitment to Safety and Accuracy: As AI algorithms become integral to people’s care, from diagnosis to treatment recommendations to support for daily living, the PRSB emphasizes the need for rigorous health and social care information standards. These standards must ensure the accuracy and reliability of data inputs and AI-driven outputs. 
  2. Interoperability: The PRSB champions the development and adoption of universally recognised standards. This facilitates safe and seamless data exchange between systems, ensuring that AI tools can function effectively across different platforms and care settings. The PRSB believes that AI can support interoperability by organising complex data against standard specifications. 
  3. Ethical Use of Data: AI applications rely on vast amounts of data. The PRSB upholds the need for standards to be implemented in a way that respects patient confidentiality, ensures informed consent, is transparent about how the data will be used in AI applications, supports professionals by providing clarity about accountability for AI outputs and clinician decisions based on those outputs and maintains data security. 
  4. Inclusivity: AI models should be trained on diverse datasets to avoid biases. The PRSB promotes standards that advocate for data inclusivity, ensuring AI tools are equitable and effective for all patient groups. 
  5. Personalisation and person-centred care: PRSB recognises that the experience of care is as unique as every individual.  The strength of our standards in promoting and encouraging personalised and person-centred care will not be lost or diminished through the introduction of AI. 
  6. Continuous Learning and Adaptation: AI in care is a rapidly evolving field. The PRSB commits to periodically reviewing and updating health and social care information standards to remain aligned with technological advancements.  This will include exploring how AI can help improve our processes to produce ever better information standard products. 
  7. Collaboration: The PRSB believes in fostering collaborations with technology developers, healthcare providers, and other stakeholders. Joint efforts are crucial to shape health information standards that cater to real-world needs while maintaining the highest standards of care. 
  8. Education and Advocacy: The PRSB and its members will play a pivotal role in educating health and care professionals about the significance of information standards in the realm of AI and related technologies. Moreover, we will advocate for the broader adoption of these standards across the sector.  Similarly, we have a role to play in educating the public about the relationship and impact of AI on our work in information record standards. 
  9. Evaluation:  PRSB will commit to regularly evaluating the impact of the ongoing spread of information standards on AI and vice versa. 
  10. Regulation: PRSB supports the development of a regulatory framework that ensures safety and efficacy of AI applications but does not stifle innovation and aligns with the nature of AI to improve and learn over time. 

 

Conclusion: 

As we stand on the cusp of an AI-driven revolution in health and care, PRSB is committed to remaining forward looking and open-minded to ensure that information standards and AI jointly deliver the most rapid advances and the best and safest outcomes for people. Our aim is to harness the potential of AI while ensuring the safety, integrity, and equity of the care people receive. 

Read more news
Read more blogs

CHAT theory also explicitly addresses five areas which if addressed systematically will help overcome stakeholder differences in pursuit of the common goal:

1. Understanding the artefacts that characterise the group and its activity.
• The artefacts might be clinical settings or the forms and templates used to capture and share information. During the pilot we heard about hard copy Dialog response forms; locally generated templates for collating information from different systems; letters and emails to GPs; images, poems or other non-text artefacts that service users might want to include in their ‘about me’ or care plan.

2. Understanding the multi-views of the group. Such groups are always a community of multiple points of view, traditions and interests. 
• Different participants in the group will have different roles and will bring to the group and their roles their own histories, language, and ‘rules’. During our Stocktake preparations and workshops we worked with psychiatrists, mental health nurses, occupational therapists, social workers, transformation leads and voluntary sector representatives, all professions and interests with their own language, approaches professional ‘rules’ but united in their interest in care plans, care planning.

3. Activity systems (like the ICSs) take shape and get transformed over periods of time. ‘Historicity’ is a term coined to express how the group’s problems and potentials can only be understood against their own history. 

 

• ‘We’ve always done it this way’, ‘that didn’t work before’, ‘it’s always like this’, ‘it wasn’t always like this’, ‘they are changing things again’, are all typical statements that often frustrate those charged with overseeing change initiatives. Without addressing the experiences that lie behind such comments you risk repeating mistakes of the past, alienating your stakeholders or just not understanding the real starting point for your transformation project. This is particularly the case for the implementation of the PCSP standard, the success of which will be largely reliant on point-of-care practices and information protocols as well as having systems which are user friendly and appropriately configured.

4. The central role of contradictions as sources of change and development. Contradictions are not the same as problems or conflicts. Contradictions are historically accumulating structural tensions within and between activity systems. Collectively addressing contradictions in how policy, practice, culture and technology interact will empower teams to find genuinely novel solutions for apparently intractable challenges, like interoperability and shared care plan/planning. 

This links to the fifth principle that:

5. the possibility of expansive transformations in activity systems. As the contradictions of an activity system are aggravated, some individual participants begin to question and deviate from its established norms. In some cases, this escalates into collaborative envisioning and a deliberate collective change effort. “An expansive transformation is accomplished when the object and motive of the activity are re-conceptualised to embrace a radically wider horizon of possibilities than in the previous mode of the activity.”

“If someone’s been working for a period of time, in a form of therapy, for instance, where there’s an agreed plan for the locality when they may present in an emergency situation … then it’s really important for us to be able to see that that information and be able to act appropriately according to that because, you know, there isn’t and there shouldn’t be a stock response to that. These plans are designed to be individualised and personalised.  All services should be giving that personalised approach to care wherever possible. And a standard such as [PCSP standard] definitely moves us closer towards being able to do that. [Not acting on agreed plans] is a key finding in terms of emergency responses over the years where responses have been inappropriate.” – Mental Health Nurse

“Following a particular plan that’s been put into place will result in much better outcomes and prevent the sort of poor outcome which would otherwise be leading to an unplanned hospital admission.” – General Practitioner

Successful teams have the functional skills to lead a task, benefit from diversity, and are led in a way that creates time and space for reflection; the ability to take stock periodically, of the task and of the way in which the team is engaged in delivering it. Your stakeholder analysis [HYPERLINK] should help you assemble the most appropriate team and identify how the team interacts and relates to other stakeholders like sponsors, services users, etc.

The variation in the size, both in terms of population served and numbers of constituent organisations, and of complexity, between Integrated Care Systems, precludes the possibility of any prescriptive guidance on the way in which this team is assembled.

Engeström’s expansive learning cycle of learning actions explains how there are 7 stages of learning actions;