Combining SDoH data with AI models gives healthcare delivery organizations the tools to contain the diabetes epidemic as well as personalize and improve population health.
Physicians and nurses have long known what has become increasingly obvious to hospitals, health systems and health plans — social determinants of health (SDoH) have an outsized impact on health outcomes. In fact, research suggests that medical care accounts for only 10 – 20 percent of health outcomes while the other 80 – 90 percent are attributed to demographic, environmental and socioeconomic factors.
It’s tempting to think that the impact of social determinants – food insecurity, transportation, financial literacy, loneliness and more – is concentrated within at-risk populations such as Medicaid recipients. Yet, at least two studies challenge this idea:
- 68 percent of patients have at least one social determinant of health challenge, according to a study of 500 random patients; 57 percent have a moderate-to-high risk for financial insecurity, isolation, housing insecurity, transportation, food insecurity and/or health literacy.
- A national study commissioned by Kaiser Permanente found 68 percent of adults had at least one unmet social need, and these needs span income levels. Not surprisingly, 91 percent of those with income of less than $25,000 report an unmet social need. More than half (54 percent) of adults with $100,000 – $124,000 in annual income report the same as do 40 percent of adults with annual earnings of $125,000 – $150,000 and $150,000+. Likewise, unaddressed SDoH impact all age groups; more than 80 percent of Generation Z (Gen Z) and Millennials, 74 percent of Generation X (Gen X), 53 percent of Baby Boomers and 43 percent of the Silent/Greatest generation say they have at least one unmet social need.
With the majority of adults reporting social needs, it stands to reason that identifying and mitigating social determinants of health will improve population health.
Want to publish your own articles on DistilNFO Publications?
Send us an email, we will get in touch with you.
Addressing Social Determinants of Health
Given the profound impact social determinants have on patient health outcomes, it makes perfect sense that addressing a patient’s housing, transportation and food needs reduces health spending. Research has shown this to be true.
Payers, in particular, lead the way with pilots and research studies demonstrating the effectiveness of managing patient social determinants. One good example is Geisinger Health System’s Fresh Food Farmacy program, which provides 15 hours of education about diabetes and healthier living followed by 10 free nutritious meals a week for diabetics and their families. It costs $2,400 per patient per year to operate the program, and early research shows an 80 percent reduction in overall health costs: from an average of $240,000 per diabetic member per year to $48,000.
Diabetes and Social Determinants
It is illustrative to look at the epidemic of prediabetes and diabetes and the role social determinant data could play. First, the numbers:
- 1 in 3 adults with prediabetes
- 90 percent Prediabetic adults who don’t know they have it
- 70 percent Prediabetic adults who will develop diabetes
- 30 million American adults and children with diabetes
- $327 billion annual costs of diagnosed diabetics
- $1 of every $7 healthcare spend on diabetes and complications such as amputations, strokes and kidney failure
Let’s also dive deeper into the issue of food insecurity. The Kaiser Permanente study found 48 percent of adults reported problems with food security in the past year. As one might expect, the incidence is even greater among people with lower incomes: 74 percent of those with annual incomes below $25,000 and 57 percent of people with annual income between $25,000 and $49,000 experience difficulty paying for food.
At the same time, research shows food insecurity increases the prevalence of diabetes as well as impairs the ability of diabetics to manage their disease and glycemic levels. Research published in Current Nutrition Reports, The Intersection Between Food Insecurity and Diabetes: A Review, “Food insecurity in North America is consistently more prevalent among households with a person living with diabetes, and similarly, diabetes is also more prevalent in food-insecure households.”
Social Determinants of Health in Action
The research demonstrates the United States is in the midst of a diabetes epidemic as well as the connection between diabetes and food insecurity. That means identifying the populations at risk of diabetes who also have difficultly paying for balanced meals can enable healthcare organizations to intervene earlier and target resources to these groups of patients.
Combining artificial intelligence (AI) models with social determinant data offers the distinct possibility that healthcare delivery organizations can do just that.
Increasingly, analytics companies are using innovative data science techniques to create models that identify who among the population without type 2 diabetes is likely to develop it in the next 12 months. Analytics and insights platforms are able to ingest this kind of output from AI models, combine it with population and patient-level data including food insecurity, housing, transportation and more, and integrate this information directly into the patient record. In other words, analytics plus technology enables healthcare organizations and their care managers to prioritize and work to engage the populations of patients who are food-insecure and at high risk of becoming diabetic in the near future.
As Americans become older and sicker, there is no doubt about the critical role SDoH data can and are playing in improving population health. Given the prevalence of chronic illness — about 50 percent of adult Americans have at least one chronic condition and 25 percent have two or more — and social determinants of health, it is encouraging to know health plans, hospitals and physicians increasingly have timely access to pertinent information about patient social barriers and the ability to use technology to scale the mitigation of these needs. Even for organizations that are not yet ready to invest in analytics and insights platforms that make it easier to access and act on patient demographic, environmental and socioeconomic data, there is research showing even a one-minute patient survey can reveal important social determinant information.
Date: October 11, 2019
Source: MedCity News