By Dr Tom Foley, Dr Fergus Fairmichael.
Mr Joshua Rubin is the Program Officer for Learning Health System Initiatives at the Department of Learning Health Sciences, University of Michigan Medical School. Mr. Rubin is a former Executive Director of the Joseph H. Kanter Family Foundation, a non-profit organization working towards the realization of the LHS vision. He is a founding member of the Interim Steering Committee of the Learning Health Community.
Need for a learning health system
Costs: Healthcare costs continue to rise and, in the US, now account for 17-18% of GDP.
Workload: Currently, health IT creates tremendous burdens on doctors and nurses. We want it to work in a way that the process of learning and recording happens seamlessly.
Rare Diseases: Healthcare systems are fragmented enough that even the largest systems would need to share information with others to study the rarest diseases.
Healthcare systems have not historically tracked outcome measurements: For example, only around 3% of cancer patients are enrolled in trials. It can therefore be assumed that we are not systematically learning from the other 97% of patients.
Medical Errors: Some studies suggest that there are 400,000 preventable deaths per year in the US. We cannot afford not to learn from this.
The recording of outcome measures is becoming more common place
• Patient reported outcomes are becoming increasingly important and there is a growing understanding that bringing the patient’s experience into the equation is key to improving care.
Outcome measurement could be mediated through:
• patient engagement
• mobile health
• personal health records (where the patient has control of their own health records)
The Learning Health Community
In order to achieve a learning health system we need to make sure every type of stakeholder is represented. The Learning Health Community (http://www.learninghealth.org/) grew out of a Joseph H. Kanter Family Foundation summit of around 80 different stakeholders which was used to develop a set of core values for the learning health system, currently around 70 different organisations, representing vendors, pharma companies, patient advocacy groups, public health groups, a lot of previous rivals have come together to become part of this community.
The Learning Health System is a network of networks. One of the biggest challenges to the development of the Learning Health System is ensuring that networks can work together and have the capability to communicate with each other. For this to happen, there has to be the right incentives for exchange of data.
Completing the learning cycle
Getting information into the system and disseminating it afterwards is an increasing challenge. Currently systems rely on dissemination through traditional mechanisms with paper publications. This can be very slow. It currently takes around 17 years for knowledge to transfer from “bench to bedside”. It is increasingly difficult to keep up with the current rate of knowledge creation. Simultaneously the rate of knowledge generation is accelerating, however knowledge is like a drug and has to be disseminated in the right way, the right dose and form so that it can be digested and used appropriately.
A learning health system will need to be able to serve different use cases e.g. comparative effectiveness research, epidemiology, surveillance for adverse effects from treatments.
Areas that require further funding to encourage development
• Standards are essential to enable learning from data.
• Public education and privacy concerns need to be addressed.
• Legal/policy/economic/social/ethical considerations need to be addressed. How these very different areas will fit together should be examined further.
“We are waiting for the killer app – we just want to create the infrastructure to make that possible”
Over the next 5 years we may begin to see:
• Increasing efforts to learn from data
• Improvement in user interface/user friendliness
• Development of outcome measurements
• Sharing data for other purposes (secondary uses)
• Emergence of interoperability
• More acceptance from patients to share data, with the recognition of the incentive of improving healthcare
Looking beyond this towards 10 years
• Further innovation
• Change in how healthcare is paid for
• Increasing financial pressures on healthcare
• The ageing population will provide further challenges
• Standards will develop that everyone will need to buy into for them to be successful
• How to get different disciplines working together to enable a LHS
• More focus on inequalities.