By Dr Tom Foley, Dr Fergus Fairmichael.
Dr Graham Willis, Head of Research and Development, Centre for Workforce Intelligence
Dr Wai Keong Wong, Consultant Haematologist UCL
Professor Alan Cribb, Professor of Bioethics and Education KCL
Dr Fergus Fairmichael
Change in approach to healthcare
The learning health system would be a significant change from the current health system. In essence, it would remain an evidence based system but it would be using broader evidence with different sources to what we have traditionally used.
New Skill Sets
There will be varying scales to the LHS and different user groups will have different needs within a LHS. However, there may be some underlying generic skill sets and competencies that would enable the LHS.
These include developing an understanding of the system itself, so that a user may be able to ask a question of the system, receive a result and interpret this. As this result will be based upon probabilities there is a required skill set around numeracy and the understanding of potential bias. These skills would allow a greater understanding of the “fuzzy” answers returned to the user by the system.
Some but not all of these skills could be outsourced. Organisations will need a strategy that brings in new people but also upskills existing staff. This will require investment. University Hospitals Birmingham has a 50 strong data analytics team. Investing in these skills allows them to go down to the level of identifying doctors with abnormal prescribing behaviour so that they step in and provide targeted training. They have also employed people from the finance industry to bring their skills to health care problems. In the US, health care systems like Geisinger are already investing heavily in this and they offer expertise to other providers.
There will be a blurring of the divide between academia/research and clinical fields. Skillsets that traditionally lie within academia may need to be much closer to the clinician. In particular, understanding the data that needs to be recorded for the system to be able to provide insight. It is important that clinicians understand what the purpose of the system is, what the data is used for and the impact that it can have on improving healthcare.
There will be a need for clinicians to record data accurately and understand the consequences and potential impact of erroneous or poorly recorded data. Currently a clinician writes in the notes, it is used by whoever reads this in the care of that patient and it ends there. In the LHS, the author doesn’t know how that data will be used by the system and they need to input it with that understanding.
Role for informatics and analytics
There will be a growing role for informatics and analytics in the health system. Some models may bring in new teams to deal with this, others may develop the skillsets of its existing workforce to account for this need.
Skills and competencies are important, but organisations have to want to be a learning organisation and have to have an inclination and disposition to achieve this. There will be a need for an understanding of the potential benefits for practice. A set of softer skills will be required that are difficult to model when assessing workforce requirements.
Do we need more of the same competencies, or do we need a new set of competencies for the LHS, such as:
• Weighing evidence
• Working with outcomes
• Ability to input information into the system for future use
• Ability to see data as a potential source of knowledge
Interaction with the system
Need to develop an understanding that what is recorded could be used many times in many different ways that could influence the care given to others. Therefore there needs to be the skills to ensure that data input into the system is reflective of a patient or an encounter. This involves recognising data as a source of knowledge that can have an impact on the care of others as well. There is therefore a professional responsibility to understand what needs to be recorded and to ensure data accuracy.
It will also be necessary for organisations to understand how the data input interface will affect the data that is entered, for example difficult forms will be less likely to be filled in.
There may be changes in the doctor-patient relationship and with potential interactions through the system. Examples provided included patients taking greater ownership of their own data, collecting data from differing sources such as mobile and wearable technologies and setting agendas for upcoming consultations.
Coping with change
As the system will continue to evolve and develop with the continuous iterations of learning cycles, there will be the need for users to be able to cope with change. There should be opportunities for the workforce to keep abreast of changes and training opportunities for the workforce to develop the skills to use the new functionality of the system, so we have to train for change rather than train for a fixed system that may be in place in X years.
It is not just clinicians who will need new skills. It is all decision makers within healthcare systems. As data becomes ubiquitous, decision makers will have to interpret and use it. This will require:
• Data interpretation
• Critical scepticism
• Understanding of the validity of results
• Understanding of potential bias in data an users
• Understanding of probability and uncertainty
• Ability to cope with continuous change
• New understanding of information governance implications
• Understand the impact of what we record
These skills will have to be incorporated into clinical education and training but also into the discourse about leadership. Currently leadership is seen as a soft skill but if you start to interweave it with some of this hard stuff, it makes it all more robust.
This will change the way clinicians interact with patients. Patients will expect to have more access to their notes and even to add things to their notes. At Geisinger, patients can record an agenda in their upcoming appointments and there is research to show that this approach can improve outcomes.
There are a wide range of clinical informatics related roles that do not yet exist in sufficient numbers. There is also a big shortage of clinicians who understand the informatics and clinical work. They are really the only people who can make sense of the data.
The LHS is being used to identify unusual practice and then training can be targeted