To be clear, the story of industrialised health care, particularly in the hospital, is one of significant reductions in mortality and increased life expectancy. But the model we’ve built—relying on a finite clinical workforce, centralised services, and episodic intervention—is not designed for chronic conditions requiring life-long management.
Enter artificial intelligence: the next generation of software applications that mimic human analysis and pattern recognition. There’s no shortage of health care applications for AI. But for chronic disease management, it could be a game changer.
In recent years, there’s been a rapid development of applications focused on this area. Here, we’ve profiled a few of the applications at the forefront of using AI to diagnose, treat, and manage chronic conditions.
Measuring and managing chronic pain
Twenty percent of Australians live with chronic pain that is often debilitating and hard to diagnose clinically. PainChek® assists in detecting and monitoring chronic pain in patient populations that are not able to communicate their pain to carers and clinicians. PainChek® is the world’s first pain assessment tool that has regulatory clearance in Australia and Europe.
Through the use of AI and facial recognition, PainChek® monitors facial muscle movements in individuals who can’t self-report pain and provides a quantifiable score of the pain they’re experiencing. The analysis and recording of patient pain is completed on a smartphone app and the history is stored in the cloud to enable coordinated care delivery and management.
PainChek® has been clinically researched and proven as a reliable and valid assessment of pain in patients with dementia. The technology has been used at a number of aged care facilities in Australia, leading to reduced resident pain levels through better assessment and more accurate treatment.
Virtual nurse assistants being used across the NHS
Health organisations such as the UK’s National Health Service (NHS) and the University of California, San Francisco (UCSF) have implemented AI-powered nurse avatars, developed by Sensely. These avatars cover a wide range of content modules, including system assessment, health information, and wellness and chronic care, and are available in over 30 languages.
In the US, an integrated payer-provider is using Sensely to help Chronic Heart Failure (CHF) patients self-manage and reduce preventable hospital readmissions. Newly discharged CHF patients install the Sensely app on their phones and are given a Bluetooth-enabled scale and blood pressure cuff. Each morning, patients receive a notification to complete a check-in routine, with Sensely avatar “Molly” guiding patients to record weight and blood pressure data.
Sensely then calculates a risk assessment for each patient and provides clinicians with timely information to trigger appropriate interventions. In a trial, the platform resulted in a 75% decrease in 30-day readmissions and a 66% reduction in patient monitoring costs (compared to the labour costs of a non-Sensely routine).
Early detection of chronic disease escalation
Over 1.2 million Australians live with diabetes and one-third present with signs of diabetic retinopathy, eye damage that impacts vision.
It’s easily treatable if detected early, but diagnosis typically requires the expertise and equipment of an ophthalmologist. Many patients put off these checks until their vision begins to deteriorate, at which point treatment becomes much more difficult.
To address this problem, an Australian-based research team developed Dr Grader as a simpler way to detect diabetic retinopathy, allowing the check to be performed by GPs. Dr Grader uses deep learning, an artificial algorithm that analyses multiple layers of raw data, to compare retinal scans to a recognisable data set of diseased eyes.
The technology was successfully trialed in Western Australia in 2017 and 2018. Ophthalmology diagnostic imaging company TeleMedC has licensed Dr Grader and is currently implementing the technology across South East Asia.
These are just a few of the many technologies aiming the power of AI at chronic disease management. But health care is never short of new technologies. And AI in particular carries sizeable hype and hope. Counterintuitively, though, the pace at which health care adopts AI will be directly proportional to our risk tolerance. Health care leaders are fully aware of clinical risks, but it’s an entirely different problem when a machine makes the error.
Still, as the proliferation of new AI applications continues, it's clear the use of AI in health care is more a question of "when" rather than "if’".