While burnout is well known among clinicians, “patient burnout” is an equally critical yet silent problem within the chronic disease population. Managing chronic conditions, especially diabetes, is complex and overwhelming. Nonadherence impacts 12% of the US population with Type 2 diabetes; 42% of patients with chronic conditions manage at least two or more.
Dr. Tejaswi Kompala, head of cardiometabolic clinical strategy at Teladoc Health, says AI is rethinking the way healthcare approaches chronic conditions like diabetes. She believes AI can help by:
-
Personalized interventions to improve patient engagement and lower A1c levels in diabetes management.
-
Predict and identify patients at risk for uncontrolled outcomes one year in advance.
-
Collaborating with healthcare providers and coaches to improve the patient experience.
We interviewed Kompala to hear her insights on these topics and to get examples from her own practice as an endocrinologist on how AI is helping to treat patients holistically, focusing on the key pillars of cardiometabolic health for sustainable results: nutrition, activity, sleep, and stress.
Q. Describe what you call “patient burnout” within the chronic disease population. And how it is, as you say, a “silent issue.”
A. Clinician burnout is a serious and well-documented problem, but patient burnout is often overlooked. Although it is rarely discussed, patient burnout has serious health consequences – and in my own practice, I see it increasing.
A contributing factor is the increasing prevalence of chronic diseases. About 42% of Americans suffer from at least two chronic conditions. This only adds to the complexity that every patient already faces during their care journey.
Managing these conditions is an ongoing challenge that requires significant commitment, involvement and resources and time to fully adhere to their medications and treatment plan. Juggling the demands of managing a chronic condition is stressful and often takes a toll on one’s mental health. In fact, patients with diabetes are two to three times more likely have more difficulty with their mental health than people without.
These circumstances can make it even more difficult to successfully adhere to a treatment plan. – when patients do not take their medications or follow directions as prescribed – is a major problem and can arise from the daily stressors of life, compounded by complex dosing or treatment schedules.
It’s a lot to keep track of: remembering to take medications, making lifestyle and behavioral changes, and keeping up with medical appointments. And don’t forget that patients often live with these conditions for the rest of their lives, which can also contribute to feelings of overwhelm and burnout.
In my practice, we see chronic conditions increasing in younger populations, which can exacerbate these challenges. We are asking patients to be involved in their health for decades. – not just for a few weeks or months.
Q. How can artificial intelligence in general help prevent burnout in patients?
A. Managing chronic conditions is a burden, and it’s common for patients to experience periods of low motivation and burnout. Artificial intelligence can play a powerful role in keeping people engaged during those ebbs and flows.
In recent years, we have begun using new AI applications to proactively identify patients at risk for uncontrolled diabetes. Our predictive models use data such as blood glucose levels, medication refills, food logging, and other health signals to identify those at risk and increase program engagement. Engagement is a key factor in improving health outcomes and can help address patient burnout.
Additionally, AI models like these help identify patterns in patient behavior that indicate potential signs of burnout or nonadherence, such as when patients have missed multiple doses or skipped recording their activities.
These insights help patients stay on track and avoid falling through the cracks of the healthcare system. Ultimately, it helps clinicians allocate their resources and time more effectively to provide additional support where patients need it most.
Q. How can AI predict and identify which patients are at risk of uncontrolled outcomes a year in advance?
A. Predictive modeling has the potential to transform diabetes care by proactively identifying a patient or population at risk for uncontrolled outcomes up to a year in advance. This is especially important because uncontrolled diabetes can lead to serious and devastating complications in the long term, such as nerve damage.
Using predictive modeling, we can analyze over 100 pieces of personal health data in real time to identify who is at risk and target characteristics that may impact their risk status based on their personal health journey.
The data is analyzed regularly, taking into account a range of inputs, from the frequency of glucose measurements to medication dosages and even a person’s engagement with educational materials in the app. Historically, medicine has taken a reactive approach – respond after a complication. But new applications of technology give us the opportunity to potentially respond and intervene before a complication.
Once you know who is at risk, clinicians can respond to help patients get back on track. By moving from a reactive to a proactive approach, you can provide more effective, personalized care. These tools enable timely, personalized interventions to prevent complications, improve outcomes, and better manage costs for employers and health insurers.
Q. How can AI work synergistically with healthcare providers and coaches to improve the patient experience?
A. I like to think of it as collaborative intelligence versus “artificial” intelligence. As clinicians, these tools help us leverage data and insights in ways that make us smarter, more effective, and more efficient. By using data in, on, and around each person, AI can help enable more personalized care.
Emerging technologies will play an increasingly important role in extracting meaningful insights from data in the future. But ultimately, it is the people themselves (coaches and clinicians) who must leverage the trusting relationships they have built with patients to put those insights into practice and achieve better outcomes.
These insights indicate which engagement methods are effective for specific target groups and where additional interventions are needed to engage patients in a targeted manner during the care pathway.
Targeted interventions with access to caregivers or human coaches can make a difference in achieving sustainable behavior changes, such as helping patients continue to take their medication, exercise regularly and eat healthily. Ultimately, these tools help us be more human and provide more compassionate and empathetic care.
Q. Can you provide some insights and examples from your own practice as an endocrinologist on how AI helps in the holistic treatment of patients?
A. In my own practice, I see the greatest potential of AI in providing insights that allow me and my colleagues to provide much more personalized care, supporting individual behavioral change day in, day out.
For example, I spend a lot of time working with patients to implement behavior change, but often I only have a few touch points with them throughout the year. Digital tools help us implement our care plans and can help us intervene at the right time if a patient goes off track.
Managing diabetes is a marathon, not a sprint. It requires many lifestyle and behavioral changes that must be sustained over time to achieve better outcomes. AI plays a role in connecting and engaging my patients outside of our regular appointments to ensure they are adhering to their medications, exercising regularly, and eating healthy to achieve A1C lowering, weight loss, and mental health.
Follow Bill’s HIT reporting on LinkedIn: Bill Siwicki
Send him an email: bsiwicki@himss.org
Healthcare IT News is a publication of HIMSS Media.