Simplifying the First Step into SDOH with AI

The following is a guest article by Brian Norris, Vice President & Managing Director of Strategic Consulting at MedeAnalytics

We’ve all heard the motivational proverb about the journey of one thousand miles beginning with a single step. The same goes for the first step in addressing a patient population’s social determinants of health (SDOH). Many factors comprise a patient’s SDOH profile, from transportation to air quality to the distance from their healthcare provider. The first step in the journey towards equalizing everyone’s SDOH can seem endless. There is no map, and forget about GPS. So which direction do you go? Which way do you turn for the best care? Does left or right make the most financial sense? Provider and payer organizations that embark on this journey will require assistance. Thankfully, with data, artificial intelligence, and the will to create a meaningful impact on people’s lives, the tools exist to guide these organizations towards the SDOH destination and create a roadmap for others to follow.

Knowing where to make that first step is challenging because SDOH factors go beyond the four walls of a doctor’s office and beyond a health system’s typical realm of control. Yet we know various SDOH factors determine between 30% and 55% of a patient’s health outcomes, so understanding these non-medical contributors — the conditions in which people are born, grow, live, work, and age — can significantly influence their wellbeing. Before taking that first step, it’s only natural to ask the following questions:

• Where can my healthcare organization impact our patient population with SDOH efforts? 
• What aspect of SDOH makes the most financial sense to address first? 
• What’s the return on investment? 

The answers to these questions are becoming more attainable, thanks mainly to the massive amounts of data already available within the healthcare industry and advancements in artificial intelligence.

Data

Claims data is one of the most useful datasets to assess the overall health of a patient population, and there’s a lot of it. Previous estimates suggest it’s somewhere in the petabytes. By combining claims trends with granular SDOH data and clinical insights, payers and providers can actuarially model and reveal the areas of most significant risk and the interventions with the greatest return on investment. Unfortunately, many healthcare organizations don’t know the value hidden within this data, especially when paired with predictive analytics and artificial intelligence. Taking this data and incorporating AI can help forecast real-world cost savings and better health outcomes.

AI

With the right analytics and AI tools, organizations can move from simply acknowledging SDOH to predicting which social risk markers will most affect utilization, cost, and outcomes.

For example, if data shows that some patients frequently use the emergency room due to transportation challenges or lack of accessible follow-up care, targeted interventions like admission, discharge transfer alerts, and proactive care navigation can reduce these unnecessary admissions. When integrated at the point of care, this insight allows clinicians to make more informed decisions and organizations to streamline operations. It also enables care teams to follow up more effectively after emergency room visits. Pairing with AI, payers and providers can begin to predict which patients are more likely to fall through the cracks and offer suggestions for organizations to make impactful changes.

The Path Forward

More payers and providers are becoming legally obligated to address the SDOH of their members, especially among organizations embracing value-based care models. As more of these health systems, employers, Medicare Advantage plans, and self-insured groups continue to grow, so will the need for these organizations to focus on SDOH.

The SDOH journey is long. But the first step doesn’t have to be difficult. Like any long-distance traveler, you need a plan. And that’s true for healthcare organizations. While it may not be a typical map, data and AI can provide clear direction for the path ahead.

About Brian Norris

Brian Norris, MBA, RN, FHIMSS, has more than 25 years of experience in healthcare, 20 of those years as an RN and 15+ years in analytics, value-based care, and product development. As Vice President & Managing Director of Strategic Consulting, Brian leads the MedeAnalytics consulting practice focused on helping clients advance their value-based care strategies, deliver phenomenal outcomes to their populations, and assist organizations across the healthcare vertical with their most pressing clinical, operational, and outcomes challenges.

MedeAnalytics recently partnered with Socially Determined and Mathematica to modernize the measurement, validation, and financial impact of SDOH interventions. By directly incorporating this information into an EHR, providers can make better SDOH-focused care decisions at the point of care.