Enhancing care for persons with diabetes

Leveraging AI to gain deeper insights into persons with diabetes and their caregivers can be transformative. With the increasing number of influencers and the diverse sources of information on topics like sleep, nutrition, exercise, and alternative therapies, it's crucial to understand how these trends impact disease management.

As online health trends gain traction and influence decisions, both consciously and subconsciously, doctors face the challenge of staying informed and guiding individuals toward the right choices. In what ways can AI help bridge this knowledge gap and support healthcare professionals in offering optimal care?

Accenture recently conducted a study in partnership with a healthcare client, to demonstrate how AI can transform the vast plethora of online data into a wealth of insights to aid clinicians in understanding key trends and concerns.

Accenture’s Managing Director, Sonia Gupta, elaborated on the study, highlighting what is on the minds of patients and caregivers in the world of diabetes in Singapore.

To subscribe to the GovInsider bulletin click here.

Navigating online complexity with holistic AI models

Traditional healthcare systems generally address healthcare conditions via a range of policies, interventions, treatment protocols, and communication guidelines designed from a clinician perspective and may miss the current perspectives of those living with diabetes and caregivers on the ground.

The study created a holistic AI model that analysed 130,000 documents from different publicly available online sources, such as news portals, Reddit, Instagram, YouTube, PubMed, Twitter and Facebook to identify key online trends when it comes to managing diabetes.

“If a human had to physically comb through and understand 130,000 documents… it would have been impossible right?” Gupta explained.

“What if we introduced an AI model that holistically blends various AI algorithms, generative AI models, and machine learning to accelerate insights from both persons living with diabetes and their caregivers?

This approach could help us understand not only the known factors influencing their interactions but also the emerging thoughts and concerns shaping their perspectives on the disease.”

The model identified themes from 12 months’ worth of online materials in a couple of weeks, she explained.

Surprising findings

36 key themes emerged across the perspectives of individuals living with diabetes, healthcare professionals, and policymakers, with 18 of these coming directly from persons living with diabetes and their caregivers.

These 18 themes can be distilled to three main clusters, said Gupta.

Prevention and management of diabetes, managing the emotional and financial burden of caring for loved ones with diabetes, and overcoming lifestyle and logistical hurdles in adhering to medical regimens.

Significantly, the model revealed a key blind spot between the topics physicians discuss and the concerns of persons living with diabetes.

Importantly, it brought to light several major concerns among caregivers, such as the intricacies of providing support, the stress and strain of their role, the potential for burnout, and the necessity for resources and strategies to cope with stress and burnout related to caregiving.

“Healthcare professionals were not thinking about the caregiver’s role and their challenges. How do we make it easy as clinicians and government to better equip the caregiver from the emotional, financial, and physical challenges to improve the health outcomes of those living with diabetes?”

“That was one big area that was not as present as we thought it needed to be,” she said, noting that in a country like Singapore that has an ageing population, caregivers themselves tend to be older.

Some of the patient themes included concerns over diet choices; questions about traditional medicine including herbs and practices like Traditional Chinese Medicine (TCM) and navigating festive seasons with an abundance of food and social pressures, Gupta shared.

While these might be already on the radar of the physicians, the model can summarise emerging terms and ideas around these concerns. This will enable clinicians to communicate more effectively with individuals living with diabetes, she said.

To subscribe to the GovInsider bulletin click here.

Broader application for population health goals

While the model was a proof-of-concept, it has clearly demonstrated the potential to accelerate insights gathering to support Singapore’s population health goals. The proof-of-concept integrated Responsible AI practices, aligning with Accenture’s global Responsible AI governance principles.

In effect, a scaled model complemented with traditional methods could function as “an always-on engine that allows you to continuously understand the thoughts and concerns of persons with diabetes and their caregivers, allowing for a fresh approach to population health management,” she said.

Given the topic’s sensitivity, any future scaled model must proactively incorporate Responsible AI principles and practices.

Rather than relying on annual health research or reports, clinicians can utilize such a model to dynamically respond to and capture the wealth of untapped data emerging from common online spaces, providing “new and valuable perspectives to the ecosystem”, she noted.

This approach could also empower policymakers spot misinformation trends early and help clinicians in providing more personalised advice to persons with diabetes, as well as redesigning targeted outreach campaigns, Gupta shared.

Armed with these insights, healthcare workers could better engage a broader ecosystem of players as well, including caregivers and community members.

“What would change for clinicians is that you are proactively getting soundbites of what’s happening, what’s trending out there, and the macro themes that matter to complement dynamic patient engagement strategies.”

To learn more about this approach to population health management, do attend Gupta’s presentation on Thursday, October 3, from 11:30 am to 12:05 pm at Public Sector Day Singapore 2024, which is free to attend for members of the public sector. Register here to save a seat!