Speaking from HIMSS18 in Las Vegas, Ian McCrae talks about Amadeus Intelligence, our new service using machine learning.
Imagine trying to drive a car with an icy windscreen, navigating down a windy road with very limited visibility. There are some clear patches to catch a glimpse of the path but despite being a skilled driver, you still can’t be certain of the road ahead. Predicting for future patients is a similar endeavour for healthcare providers and clinicians, who have limited information about patient and population health.
As annual healthcare expenditure continues to rise globally, the need for innovative solutions that keep people healthy and out of hospital becomes more vital. The ability to predict the needs of a patient’s care – whether in the community or in a hospital – would prove highly valuable to hospitals and other providers, so they could intervene early. Orion Health and Precision Driven Health, a public private partnership between New Zealand District Health Boards and University of Auckland have conducted ground-breaking research using machine learning and its potential to gradually clear the road ahead, and simultaneously save the healthcare sector billions of dollars.
Some of the costliest challenges for healthcare organisations are avoidable readmissions, operational inefficiencies and wasted resources. $1 trillion in the U.S. is wasted each year on costly administration and avoidable hospital readmissions. An ageing population and patients with chronic or long-term conditions are a significant contributor to the resources spent. Machine learning has the potential to provide clinicians with valuable insights that allow them to manage long-term care patients much more effectively.
In the United States, there are an estimated 24 million people with diabetes, and a further 57 million people who are pre-diabetic, putting them at risk of developing more serious life-changing chronic conditions. The ability to predict patients’ risk of developing these health issues has great benefits for healthcare organisations. It allows for early intervention, giving people the opportunity to make lifestyle changes that can prevent them from developing diabetes, heart disease or stroke. Using machine learning models for early diagnosis will help prevent these patients from needing costly and long-term treatment, as well as keeping them healthy and out of hospital.
In the U.S., the expense of long-term care services from detection of a condition, through treatment until the end of these patients’ lives, is approximately USD$200 billion. With machine learning models that can predict long-term conditions, clinicians may be able to act early enough to prevent the condition from exacerbating. Adapting deep learning technology to this area of the healthcare industry could save USD$150 billion over the next 10 years. Both the number of diabetics and the associated costs are expected to double in the next 25 years, and without these predictive advancements, healthcare organisations will struggle to bear the cost.
Machine learning is perfectly suited to Orion Health’s existing product set. Using Orion Health’s Amadeus Platform and Rhapsody to ingest and manage multiple data types in large volumes and Amadeus Analytics to create cohorts based on existing and new insights, Amadeus Intelligence applies machine learning to the data on the platform to enhance precision and prediction. It works by training machine learning models to produce more accurate predictive analysis, giving clinicians the tools they need to make accurate decisions at the point of care.
Unique to Orion Health, Amadeus Analytics and Amadeus Intelligence service connect seamlessly to its Amadeus Coordinate, care coordination tools including Care Plans and Pathways and Care Coordinator Activity Management which create actions and workflow to support identified patients.
This is only the beginning of machine learning’s impact on healthcare. As the industry shifts towards precision medicine and personalised care, services like Amadeus Intelligence are helping to bridge gaps for clinicians, allowing them to tap into the vast amounts of data from entire populations to treat and manage a person’s health.
Click below to learn more.
