Dual-eligible patients — those who are enrolled in both Medicare and Medicaid —end up getting readmitted to the hospital 30-days after discharge more often than those patients that are not dually eligible, per a recent study by Yale. But, the Yale study also found that these disparities for dual-eligible Medicare patients did not directly correlate to differences in social determinants of health (SDOH), even when adjusted for community- and state-level social and health service availability factors.
The study used census-level data to probe the reasons for the 30-day readmission rates of 2.5 million U.S. adults 65 years or older to see what contributed to the increase in readmission rate for some patients. The final sample for the study included 898,395 who were admitted for heart failure, 475,444 patients who were admitted for acute myocardial infarction, and 1,214,282 who were admitted for pneumonia, of whom 17.4%, 13.2%, and 23.0% were dual-eligible patients, respectively.
With the Rise of AI, What IP Disputes in Healthcare Are Likely to Emerge?
Munck Wilson Mandala Partner Greg Howison shared his perspective on some of the legal ramifications around AI, IP, connected devices and the data they generate, in response to emailed questions.
The rationale was that if community factors are the cause of the trend in readmissions then they need to be addressed. By contrast, if the higher rate of readmissions were due to factors beyond SDOH, then it’s possible that hospitals may have control over the factors and could thus take steps to improve these outcomes, according to the study.
Census data the study relied on included several SDOH — ethnicity, race, and cultural context; socioeconomic position; social relationships; gender; and residential and community context. However, additional SDOHs exist beyond the scope of the study and what the census includes. For example, health literacy, social support, educational attainment, to name a few.
One company that weighed in on the Yale effort believes that it has has AI-powered technology that utilizes a broader range of SDOH beyond the census-level data used for this study. Suwanee, Georgia-based Jvion claims its data can provide more comprehensive clinical decision-making capability to clients.
“There are many more underlying factors at play that can not be accurately measured or accounted for by census-level data alone. Other factors, such as digital fluency and social isolation, are also immensely influential social determinants that can’t be captured at the community level,” said John Frownfelter, Jvion’s chief medical officer, said in an email. “Being made aware of something like social isolation enables the provider to refer the patient to support group services, which can help improve long-term health outcomes; while digital fluency can determine how much, if any, remote or telehealth services should be employed (a critical insight especially post-Covid).”
A Personalized Approach to Medication Nonadherence
At the Abarca Forward conference earlier this year, George Van Antwerp, managing director at Deloitte, discussed how social determinants of health and a personalized member experience can improve medication adherence and health outcomes.
Frownfelter noted that Jvion’s technology looks at patients and their SDOH risk factors to identify overlooked disparities that, when understood and addressed, have been shown to reduce readmission rates by up to 20%, he said.
“Moving forward, we need to couple these census-level findings with more granular, individualized data to understand patients with the precision necessary to identify and address inequity in all its forms,” Frownfelter urged.
It’s not that this data is not available, he said, contending that the data needs to be compiled in a meaningful way. Jvion’s technology aims to harness AI to aggregate existing data about theses additional SDOH and compile and analyze it to help translate that information into actionable information.
“The directionality of AI is very important. The insight provided by AI at the community level is especially relevant with the heavy investments that healthcare providers are making to improve communities and to solve the challenge of health inequities. Since zip code data is not specific enough to guide programmatic resources, we need to get down into the smaller community level, or block group level, as we refer to it here at Jvion,” Frownfelter said.
“For example, if we know an entire zip code is deficient in access to primary care providers and we are opening a new clinic, the precise neighborhoods where the worst access is may be far from where that new clinic is built, even within the same zip code. Large geographic areas are far too heterogenous to assume that any intervention within the zip code is ‘good enough,'” he added.
In addition to analyzing the data, AI helps congregate the data, especially helpful during the current provider shortage, Frownfelter added.
If the disproportionate readmission rates for dual eligibles can be attributed to other social determinants, hospitals can possibly take actionable steps to advance equity. They can also explore avenues to improve the quality of care transitions at discharge, according to the Yale study.
Photo: South_agency, Getty Images
