From 3M Health Information Systems
Exploring technology and the promise of AI in health equity – A HIMSS23 recap
I sat down with 3M Health Information Systems Chief Population Officer Dr. Melissa Clarke to talk about her recent experience at the 2023 HIMSS conference, including insight on health equity, the use of AI, new technologies and what the health care industry can look ahead to.
You participated in a HIMSS panel hosted by Amazon Web Service (AWS) during the conference. What insight was addressed and discussed?
The panel, Advancing Health Equity Through Powerful Technology Solutions, was moderated by AWS Chief Medical Officer Dr. Roland Illing. I was joined by Michael Cui, assistant professor, general internal medicine and associate chief medical informatics officer at Rush University Medical Center and Shawn Huang, chief AI officer at Elevance Health, for a great discussion. We dove into defining health equity, which to me, is all around leveling the playing field to allow everyone an opportunity to achieve their highest possible level of health.
We must remove structural barriers that perpetuate health inequities, and that is where we can bring technology solutions to bear. For Elevance Health as a payer, they are looking at doing outreach using technology to identify individuals with the greatest needs. Rush University is doing excellent work around standardizing the capture of social determinants of health (SDoH) data. Michael Cui described their Health Equity Care & Analytics Platform (HECAP), built on an AWS platform, that aggregates and harmonizes clinical and social data from different sources to provide insights that can be used to enhance care outcomes and reduce health inequities. It was striking when he said Chicago’s West Side residents, where Rush is located, have a 16-year lower life expectancy than residents of more affluent areas of Chicago.
I highlighted 3M’s efforts, including how we are creating time to care with our acquisition of M*Modal, using artificial intelligence (AI) and natural language understanding (NLU) to decrease the administrative burdens on physicians around coding, allowing more time to interface with the patient. This is critical because an engaged physician-patient relationship is a necessary component to closing health equity gaps.
That same AI is used to gather SDoH documentation no matter what clinician (e.g., social worker, community health worker, physician, case manager) documented it in the record and present it to the coder to capture Z codes. Increasing the richness of the available claims data to incorporate social risk and social concerns allows us to realize the complex interplay between social and clinical data.
Population health analytics also play a vital role. Now that we have that view of socio-clinical data around a patient’s complete data from the claims, we’re seeing uses of that in different ways. One is being able to assign patient panels right across an organization. One 3M HIS client, for example, assigns patient panels across its organization so that no single provider has a disproportionate number of patients with high clinical or high social risks.
We’re seeing the designing of payment systems, such as in New York State Medicaid program, which uses a value-based payment arrangement with Medicaid managed care organizations. Because the organization uses our 3M™ Clinical Risk Groups methodology, it can incorporate social risk into determining capitation payments. New York State Medicaid thus is assigning the additional resources needed to improve the health outcomes of patients with both adverse social and clinical risk.
We’re also incorporating these principles into our maternal health solution that can look at birthing individuals who have different levels of social and clinical risk. Health care organizations that collect race and ethnicity data to understand where there are health disparities, can overlay that onto our methodologies to identify disparate outcomes by facility or provider to identify where interventions are needed to improve maternal health outcomes.
What were some of your top observations and takeaways from the conference?
Far and away, the overarching theme mentioned in nearly every session I attended was the promise of AI in health care. Everyone was talking about ChatGPT, including its limitations and possibilities. My particular interest is health literacy, so the potential for AI to help make it easier for patients to understand the complex materials often put in front of them is a huge possibility.
Predictive modeling around disease states and disease progression is another component of AI that was mentioned. We know that bias sometimes gets woven into the clinical record and other data sources that power AI solutions. If you are presented with biased data, you will get biased output. We need to be cautious as we use AI in a predictive way.
Another area that was new to me was digital therapeutics, or apps that have been FDA-approved to use alongside or in place of traditional biologic therapy. They have the most applicability in the behavioral health space as treatments for anxiety, addiction or ADHD, to support the necessary skills to self-manage conditions. It’s an emerging area, but there’s a challenge in getting doctors to prescribe digital therapeutics because they are still unfamiliar with how the apps work and what they can do.
There was a strong presence at HIMSS surrounding health equity and using technology to address SDoH – it was discussed in many presentations throughout the event. People are doing innovative things, and many sessions focused on the experience and reporting of the results from various institutions and vendors attempting to achieve equitable care delivery using creative approaches.
Now that HIMSS23 has closed, what initial thoughts do you have on the future of health care?
From my perspective and focus on population health, I think we can look forward to the evolution of value-based care with technology increasingly being used to gather, standardize and risk adjust social risk data to enrich the clinical picture and set up payment arrangements that reward high quality health equity-informed approaches to care.
There is a proliferation of companies trying to bring to bear more patient engagement tools and more use of biometric data, so there’ll be more data points to consider in evaluating health outcomes. Digital equity will play a massive role because we need to ensure that all populations have equal access to broadband and devices, enabling them to reap the benefit of the advances in data gathering, patient engagement and digital therapeutics.
Melissa E. Clarke, MD, CMQ, is the chief population health officer, health care transformation and health equity at 3M Health Information Systems.