Jacob, a ten-year-old boy, is often in trouble at school. He finds it difficult to concentrate and has fallen behind in maths. He is also smaller than his classmates which can sometimes add to his low self-esteem.
Jacob has kidney disease. He and his family have many questions such as why, how and what next. They spend a lot of time dealing with the health system, investigating and managing his current health and wellbeing but also searching for answers about his future and the implications for their family.
Putting together the health and wellbeing puzzle that might answer these questions is one of the keys to good care. Today, more than ever, this is aided by sophisticated technology and IT systems that draw together knowledge, data and practice to better inform diagnosis and treatment decisions. While this is often described as precision healthcare, the time has come to take it further.
The key benefits of a learning health system are in the provision of unprecedented support for clinicians and better care for patients…
The Australian health system is well placed to leverage our current advances with precision healthcare to develop what the US Institute of Medicine calls a learning health system. One that doesn’t stop at just collating information—albeit in a highly sophisticated way— but learns, adapts and continually improves, in real time.
The key benefits of a learning health system are in the provision of unprecedented support for clinicians and better care for patients, including preventative healthcare; and the more efficient targeting of health expenditure.
Benefits at the bedside
To support a clinician answering Jacob’s questions, a system that is good at learning would not only bring together the many pieces of the puzzle, it would also continuously generate new or better knowledge from the inputs. Jacob’s history, past test results and the observations of practitioners are all in digital form, in the electronic health record. Then there are the referral notes from Jacob’s GP, plus medical imagining or biopsy results. There’s the latest research data and clinical guidelines to consider. Also key, is the self-reported health status and preferences of Jacob and his family. More recently we’ve gained the possibility of genetic testing which can aid in diagnosis and identifying possible treatments.
A learning health system can draw together all these facets and provide evidence-based, data-driven decision support tools for the clinician to inform their diagnosis and treatment plans. Importantly, a learning health system can also play a part in preventative health strategies by anticipating an individual’s future health needs or wellbeing issues. For instance, by bringing together personal and family medical histories via sophisticated machine learning, a clinician may be alerted to the possibility of an inherited condition or risk factors for a disease and thus advise screening for early detection.
In Jacob’s case, a learning health system could flag the need for monitoring to detect future hearing or eye problems that may emerge due to his particular kidney condition. Or it might prompt the family to consider genetic testing of other family members to identify related but previously undetected health issues and curate treatment plans.
Benefits for the system
A learning health system can also benefit the system as a whole by creating an environment where resources—drugs, tests, hospital beds, ancillary health services, GPs, specialists etc—are directed in the best way, in the best place and time. While we can celebrate the health and societal advances that mean we are all living longer, the flipside is we are surviving with more things wrong with us. We want people to have healthy ageing not just ageing. Yet half of all Australians have at least one chronic health condition such as kidney disease, asthma, diabetes, arthritis, cancer or heart disease. This places an enormous load on the health system which already costs $170 billion per year and rising.
By assembling knowledge, data and practice information in real time to support the best diagnosis and care decisions, a learning health system could help avoid the very real problem of providing too much care, too little care, or care where there is little benefit. Equally, care may be missed where it is needed most.
With a plethora of tests and treatments available, patients and clinicians alike would benefit from understanding the implications of each for their specific conditions and also in a broader context. Not all children with kidney disease for instance will benefit from genetic testing, even though it is available. A learning health system would support this kind of critical decision-making and help clinicians and patients make wiser choices.
How we get there
If Australia was to create a learning health system, of the kind described here, we would be ahead of the game. Despite widespread interest around the world, much of the work to date has been theoretical.
Recently however, my team and I received funding for a project to improve care for children and adolescents with rare cancers or kidney disease by developing a unique learning system designed to quickly and precisely match the best available treatments and approaches with the individual child’s health and wellbeing needs.
The sophisticated integration of data of all kinds using electronic health sources and artificial intelligence is believed at this point to be the most effective foundation for a learning health system and will underpin this project. It will be nothing less than creating a health system for the future.
We hope in this research to lay the foundation for leveraging sophisticated information technologies, data mining, machine learning and genomics to capture all the evidence a clinical team requires—in the time and place it is needed— to determine a patient’s unique and individualised needs. The research will be undertaken in collaboration with clinicians we already work with – such as our superb colleagues in the Zero Childhood Cancer Program led by the Children’s Cancer Institute and the Kids Cancer Centre at Sydney Children’s Hospital, Randwick (part of The Sydney Children’s Hospitals Network) and KidGen, the Australian Genomics Renal Genetics Flagship project.
For a young person like Jacob with inherited kidney disease, a project such as this could greatly enhance his quality of life and optimise his journey in the health system. It will also provide a model for other medical conditions with the potential to deliver considerable health, economic and social benefits.