Using AI to Amplify Care for Patients With Chronic Disease

A critical obstacle to chronic disease management is that the patients most likely to benefit from primary care aren’t necessarily the patients who see their providers. Digital health helps providers support patients outside the traditional 4 walls of the doctor’s office.

I vividly remember the scene in the back office of my practice: clinical notes to the left of me, unsigned prescriptions to the right, and I was stuck in the middle looking at new quality metrics for our patient population. My team poured over who had well-controlled diabetes, who had missed their flu shot, and who hadn’t had their eyes checked that year. While this is now a familiar process for many clinicians, back then we felt like we were doing something different, something that would make a difference to patients with chronic disease. For the first time, we were using data to determine patients’ needs and deploying our clinical resources to meet those needs—before the patients suffered complications or sought care themselves.

A trend emerged. When I looked across the list of patients with gaps in care, there were some patients I recognized, and many others I didn’t. How could it be that I was unaware of some of the most complex patients in my population? I looked at my appointment availability to see if I had any free consultation slots to see these patients, but my calendar was completely filled with my regular patients. I came to understand a critical obstacle to chronic disease management: the patients most likely to benefit from primary care aren’t the patients who come see us.

The promise of digital health management

Digital health management, supporting people outside the 4 walls of care delivery by using technology, offers the potential to extend clinical resources, with some encouraging results. Through this approach, more patients have been served with the same clinical resources across multiple clinical specialties. As many of these patients are capable of self-managing (at least some of the time), digital health management makes it easier and more effective to do so by virtually connecting patients and clinicians through digital experiences, such as a smartphone app. These patients would otherwise have to be seen through primary care visits, diverting attention and resources from more complex patients. However, digital health management can engage complex patients, including the elderly, in the care management context and deliver improvements in cost and outcomes.1

We know that digital health management helps clinicians reach more patients, more effectively, with the same resources. But how can organizations make sure they offer digital health solutions to the right patients, those who are most likely to benefit? That’s where artificial intelligence (AI) comes in.

Using AI to uncover patients with the highest risk

There has been a lot of talk about AI in healthcare, and it is not unreasonable to conclude that much of it is hype. But there have been huge tangible advantages in a subset of techniques called deep learning that can produce super-human results in very focused tasks such as image recognition or language translation. These techniques are more A than I: they are loosely influenced by how the brain works, are implemented with computers, and take advantage of huge amounts of data and computational power.

But despite these advances—and what popular culture may tell us—there’s no reason to expect that sentient robots will be replacing clinicians en masse any time soon. Instead, the intersection between these AI techniques and digital health management presents a unique and powerful opportunity to amplify the work of clinicians.

When organizations deliver digital health management solutions to patients, such as mobile apps, they can generate an entirely new dataset on patient activity, patient-reported outcomes, and more insights. It can also be combined with data from the health system such as claims data and electronic medical record data. Through machine learning, organizations can then harness that data to prioritize members based on real-time needs, generate intervention alerts, and recommend follow-up actions with their providers.

This technology ultimately creates a positive feedback loop: because patients receive timely, personalized support, they engage with clinicians more often. This generates more data, which in turn gives care teams the insights they need to provide the right clinical intervention to the right patient at the right time. Plus, this approach helps the organization as a whole learn which interventions are most effective and where clinical resources can be most effectively deployed.


For clinical services to scale to meet the needs of all the patients they serve, they will need both technology that extends the reach of clinicians and data science solutions to focus that technology on those who are most likely to benefit. I anticipate that this will allow clinicians to give more attention to the people who would not otherwise come to see them—the patients with the most complex needs, who would most benefit from the bespoke, compassionate care that that only people can deliver and technology can help us amplify.


1. Panch T and Goodman E. Technology, Seniors, and Sense Making. Am J Manag Care. 2016 May;22(7 Spec No.):SP250-SP251.

Related Videos
Christopher Cannon, MD
James Howard, MD
Dr Chitra Lal
Dr Debra Boyer
Dr Milena Pavlova
Related Content
© 2023 MJH Life Sciences
All rights reserved.