By Dr Tom Foley, Dr Fergus Fairmichael.
Dr Gregory Maniatopoulos, Senior Research Associate, Institute of Health and Society
Dr Graham Willis, Head of Research and Development, Centre for Workforce Intelligence
Dr Jonathan Richardson, Chair of Informatics Committee, Royal College Psychiatry
Ms Anna Lee, Project Manager Patient Safety and Human Factors, Health Education England
Professor Alan Cribb, Professor of Bioethics and Education KCL
Ms Amanda Hunn, Engagement and Policy Manager, Health Research Authority
As Learning Healthcare Systems evolve, there will be a need to revisit and adjust the ethical framework within which clinical practice and research takes place. We will also need to examine the information governance issues again.
There was discussion regarding the different models of consent and how they may apply to the learning health system. These included, implied consent as part of being a patient within a learning organisation, opt in and opt out models, fully informed consent and streamlined consent processes. There are huge tensions between pragmatics and principles with regards to consent. There is a need to encourage engagement and participation with the system, and therefore, there is a desire to avoid a complex lengthy consent procedure that may discourage participation. On the other hand, there is a need for transparency and to ensure that patients are aware of the uses to which their data may be put. It was felt that, as a minimum, there should be an opt out system with the opportunity for regular review.
The changing nature of the Learning Health System (LHS) was also explored with the expectation within the group that the system would constantly evolve as the learning cycle continues. For this reason, it was also felt that the type of consent required from patients may also need to vary depending on the potential use cases for the data. In particular it was highlighted that there may be difficulties defining consent when the possible uses of data are broad and do not remain static. Therefore, it was felt that patients should be kept abreast of the uses of data regularly, with options to remove consent easily if necessary.
There was an underlying feeling that patients and the public should be involved in the process of determining the type of consent required within the LHS. This engagement with the public was also considered to be key, regardless of which model of consent would be used.
It was highlighted that the methods for engagement should be considered and that different populations may require different approaches. For example, those who tend to be most engaged are often those with chronic conditions who are accustomed to managing their own health.
There was discussion about the most appropriate medium of engaging the public, including, information provided during the registration process, media campaigns with posters and leaflets, advertisements and websites. It was felt that there were significant interfaces of care that would provide opportunities for the public to be informed of the uses of their data. The overall aim should be to engage through a number of methods so that the concept becomes common knowledge.
It is important to highlight the benefits to the public and to stress that the purpose for using this data is to improve health outcomes and improve the quality of services provided. It is also crucial to reassure the public about who would and would not be able to see and use their data as well as what the potential uses are.
It is important to demonstrate the benefit to patients at a local level. There are current examples of rapid learning projects such as the Department of Health’s friends and family’s feedback scheme which allow near real time improvement in services. This type of improvement is clear for the public to understand and may benefit the public acceptance of a LHS
There is a need to build public and professional trust in the system. Recent programmes, such as care.data, have failed to engage the public and as a result of this, there has been a backlash against it. There is a need to frame learning health systems appropriately so that there is an understanding about the aims of improving care.
Often the public struggle to understand that the doctor may not know the answer and that these gaps in knowledge may still need to be addressed. This may present a significant cultural gap to overcome.
It was felt that the LHS should also demonstrate improvement in safety. This may help build confidence in the system. The ability to rapidly learn from clinical incidents would be a powerful selling point for the LHS and this is something that a lot of providers would also buy into. If an organisation can demonstrate that it is committed to learning and has a strong history of that happening, then people will be more likely to come on board with the concept of the LHS.
There was concern around the use of free text and narrative. This form of record is particularly relevant for social services and mental health, where there is significant depth and meaning recorded in narrative form. It was felt that this was a potentially rich source of information that should not be neglected, however the use of this may give rise to information governance difficulties. In particular, this focussed around the difficulties in anonymising this information. This information may also have third party references who have not given consent for their information to be shared.
There was discussion around the difference between clinical care provision and research. Historically, these two areas have had clearly demarcated boundaries and differing standards for consent and use of data. It was felt that the LHS will blur these boundaries as learning from routine data becomes more common place. It should be emphasised that this is an extension of an evidence based system, with a broader scope and different types of evidence than have been used in the past.
There also needs to be an understanding from users of the system that what is recorded may influence other patient outcomes. At present, we do not necessarily learn, information is largely used for the benefit of the individual patient’s care. This would represent a significant change for healthcare professionals.