Data Analysis Saves Lives
Albert Einstein died from it, and a Mr R on Auckland’s North Shore might well have too if it hadn’t been for an innovative data analysis project conducted by Precision Driven Health – a collaboration between Orion Health, Waitemata District Health Board and Auckland University.
Mr R was one of around 800 men and women identified through data analysis of patient records as a likely candidate for Abdominal Aortic Aneurysm (AAA) during a precision screening trial involving patients at the Coast to Coast Practice near Wellsford last year.
AAA is caused by a weakness in the wall of the main artery leaving the heart. Rupture is usually sudden, unexpected and fatal. It generally strikes in the 65+ age bracket.
A simple ultrasound scan was done by the Waitemata DHB AAA screening team once Mr R was deemed to be at risk. His abdominal aorta, which should have been 2cms in diameter, was instead found to have a bulge 6 cms wide.
Surgery was performed by the Auckland Vascular Service to repair the aorta in September and Mr R is now back home – thanks to data analytics.
Dr Peter Sandiford of Waitemata DHB worked with data scientists at Orion Health and Professor Greg Jones of Otago University to develop the algorithm that helped detect Mr R’s problem and put him on the path to a positive outcome.
The algorithm is based on their previous epidemiological studies that identified men and women most at risk of AAA from among thousands of patient records.
Dr Sandiford says the application of data science and machine learning, known as precision screening, has the potential to save more people like Mr R.
“AAA is the silent killer.” says Dr Sandiford. “Big data makes it possible to create precise criteria to select those most at risk to AAA, and in the future to other preventable conditions.
All of the patients identified during the precision screening trial were contacted and 632 took up the offer of an ultrasound. Thirty-six, including Mr R, were found to have AAA – a prevalence rate of 5.5% – almost exactly the rate that was predicted by the data analysis.
Most were ex-smokers with histories of high blood pressure and cholesterol. The algorithm performed well – 97% of AAA were found in the 29% of patients with a predicted risk of 2% or more.
Mr R, 78, was a typical candidate.
Maori patients at the practice were not included because they had already been screened in 2015 as part of a successful Waitemata DHB-funded pilot programme that showed Māori men and women are nearly three-times more likely to have AAA than non-Māori.
The pilot was extended to all Maori living in the Waitemata and Auckland DHB catchment in mid-2017 and will conclude in March.
The precision screen trial ended in early November 2017 and the Ministry of Health’s National Screening Advisory Committee is being supplied with results. Further work on IT systems and modelling will be carried out.
Orion Health Chief Executive Ian McCrae says there are also plans to develop the AAA screening tool further so it can be adopted by GP clinics throughout the country, and generalized to other medical conditions. Precision screening presents a huge opportunity for the health sector, where data science can help us to target health interventions to those with the greatest need.
The AAA project is one of three main areas of interest for PDH’s Project HOPE – Health Outcome Prediction Engine. The other two are in stroke recovery and patient-reported outcomes. HOPE is one of over forty research projects supported by Precision Driven Health, applying data science to health challenges.
Precision Medicine is the growing body of international research that is identifying and enabling the capture of data that is vital for the practice of precision medicine. This includes genetic data, device data, social and diet data. With its focus on keeping people healthy and out of hospital for as long as possible, precision medicine makes good economic sense.
As governments around the world grapple with the rising cost of healthcare driven by an ageing population, they are realising the importance of precision medicine. U.S. President Barack Obama has launched the Precision Medicine Initiative, while in Britain they have begun the 100,000 Genomes Project.
Precision Driven Health is a public-private research partnership committed to putting New Zealand and the forefront of this global movement.
Data from an increased number of sources has the potential to transform the way we deliver healthcare. Through better collection, analysis and utilisation of the data that we are already capturing for each individual, we will be able to make better decisions and deliver more personalised healthcare. Personalised care, or precision health, is a win for everyone – individuals receive the advice and treatment that is most likely to work for them, and less health resources are spent on ineffective care.
There are hundreds of new devices coming onto the market each week that collect data with the potential to improve our health. These devices are being sold to patients and clinicians with the aim of helping them make better decisions about health, but often it is difficult to fully utilise and understand the data they provide.
Precision Driven Health, one of the largest data science research initiatives to be undertaken in New Zealand, is aimed at providing world-leading research into the emerging area of precision medicine and personalised care. The research partnership aims to both improve healthcare for New Zealanders and enable commercial success for some leading New Zealand companies.
Data scientists often face what they perceive to be a dichotomy between either contributing to social good, or being commercially successful. I would argue that good data science needs to incorporate both in order to be sustainable. Data analytics can inform ways to deliver better healthcare, by identifying patterns in what risks people face, and what treatments are effective.
However, clinicians and consumers, facing an increasing volume of data from all directions, need tools that enable them to translate those findings into practice. This often comes in the form of technology products. Planners, funders, and consumers will only purchase these products if they deliver health benefits, and these health benefits will only be achieved if the product is commercially sustainable.
In order to deliver tailored healthcare to the greatest portion of the population possible, all of Precision Driven Health’s research needs to have a pathway to sustainable use, including commercialisation where appropriate. This is why our public-private partnership is made up of extremely important relationships between data scientists, healthcare providers and commercial parties.
The partnership model of Precision Driven Health enables three distinct groups to achieve their own goals.
- Academics taking part in breakthrough data science, who want to see their work being used in practice, get to work with people from the healthcare sector – the people who will actually be implementing it.
- Healthcare professionals directly influence product design, enabling them to make better decisions and therefore achieve their primary objective of improving health outcomes for patients.
- Commercial companies get to build the latest data science discoveries into their products, in a way that has been validated by their customers in the health industry.
Healthcare is a prime example of a field that is full of data that is not being used to its potential. Whether it is new genomic tests, or a wearable fitness advice, data needs to be collected, stored and interpreted in order to receive the potential health benefits. This creates all sorts of opportunities for data scientists, healthcare providers and commercial companies to join forces and work towards making personalised care a reality. It is essential to ensure that the findings from data science find their way into the hands of those who will benefit.
We ask a lot of our medical professionals, and Precision Driven Health aims to provide them with tools to make decisions more efficiently. For example, a GP sees a patient for fifteen minutes and may not have a complete medical history of the patient, however they are expected to make an accurate decision around treatment based on what they know and what they see in their patient, plus what they know about their ever-expanding field of expertise.
Decision support tools can give clinicians some context around what the data suggests, therefore allowing them to make better and more personalised decisions. Data driven health aims to help clinicians quickly understand the patient, the conditions, and interpret the information for that individual patient using the data they are presented with.
If clinicians are expected to make better use of data, they must be able to trust the companies who are providing the tools of their trade, and the science behind any recommendations that come from those tools. It is essential that the companies who make and sell these data driven health products are commercially sustainable, and that the analysis they are based on is scientifically validated. As health becomes more of a data science, those who deliver health benefits will gain the credibility that leads to a virtuous cycle of sustainable commercial success.
+64 27 656 3001
PDH is New Zealand’s unique health data science research partnership.