The home is the new hospital. Since the pandemic started, home care has seen a renaissance as patients look to avoid care facilities where they could be exposed to Covid-19. Physicians too are increasingly turning to the Hospital at Home approach. A full 81% of doctors now prefer to send patients to home health agencies over nursing homes, up from 54% pre-pandemic.
The benefits of the approach are well documented, as we covered in our last blog post on the subject. But patients are still vulnerable at home, even if not from Covid-19. While remote patient monitoring tools can tell physicians about patients’ vital signs — heart rate, weight, blood pressure, blood glucose levels, etc. — care teams often lack visibility into the environmental, non-clinical factors influencing a patient’s outcome. We all know that these social determinants of health (SDOH) matter — but where do we start?
Social determinants of health should not be overlooked — it’s estimated they determine up to 80% of health outcomes. Does the patient live alone? Do they have access to healthy food or a pharmacy? Do they know how to take their medications? These are all important questions that home healthcare providers need to know the answers to — but unfortunately they usually don’t.
That’s not to say the industry hasn’t tried using data to better understand and address SDOH. But data alone will not solve the problem. I mean at the end of the day, more data really means just that….more data. The truth is not all social determinants of health are created equal or have equal impacts on outcomes. Clinicians and social workers need to be able to contextualize data to identify where they should focus on to succeed.
The second most common way to tackle SDOH is by generating some form of risk score. Risk scores, however, provide little insight on the modifiable risk factors driving a patient’s poor outcome. Without meaningful inferences and context on a patient’s lifestyle and living situation, home health providers are not empowered to change the patients’ outcome.
Fortunately for healthcare, our team here at Jvion has spent the better part of a decade aggregating, analyzing and extracting context from the sea of data out there. As a result, we can provide organizations caring for patients in the home setting with actionable, prescriptive recommendations that help eliminate barriers, close the health equity gap, and deploy home health resources more effectively.
How Does it Work?
The prescriptive insights available in the Jvion CORE™ are based on an Eigen Sphere approach. What that means is we look at thousands of data points on various risk factors: does the patient live in a food desert? Do they have social support at home? What is their digital fluency level? Are there environmental health hazards? How stable is their housing situation? We continue to add dimensions until all relevant variables are accounted for.
Across all these dimensions, patients end up clustered in groups of similar patients, known as Eigen Spheres (we love these at Jvion). As new patients are added, we can extrapolate a patient’s risk of mortality, healthcare utilization, or other adverse outcomes based on the outcomes of other patients within that Eigen Sphere. We can also extrapolate the interventions that will most effectively reduce their risk, and recommend those interventions for individuals at risk within a given Eigen Sphere.
The good news is that this can be done with incomplete, missing, or inconsistent data. When you look at thousands of data points, patients generally end up in the same Eigen Spheres even if a few of the data points are missing for an individual patient.
The result looks something like the image below. For each patient, Jvion identifies their vulnerability ranking — either as a category, score out of five, or percentile — and the top five drivers of their social vulnerability. For each driver, the CORE recommends actions care teams can take to address the patients’ social vulnerability.
These insights can be extended to the wider population. Where most SDOH solutions extrapolate the risk factors for the patient based on the characteristics of the population, Jvion’s solution works in reverse. Our insights on a selected population emerge from our insight on the needs of individual patients within that population. From there, we can identify the resources that will most benefit a given community, reducing waste and creating efficiency in community-based programs.
What We’ve Learned From This Data
By looking at millions of patients and their associated risk factors, our approach has revealed findings about SDOH that might be surprising to home health practitioners. These findings have been published in the American Journal of Managed Care, and are outlined here:
- The social determinant most associated with risk was air quality, which had a relative influence on risk more than twice that of the next determinant, income.
- Both air quality and income have a greater impact on risk than age, ethnicity, or gender.
- While age is still an important risk factor, ethnicity and gender are less associated with risk than socioeconomic factors like neighborhood in-migration, transportation access, and purchasing channel preferences.
The findings suggest that home health practitioners are missing important indicators of patient risk if they focus only on surface-level traits such as age, ethnicity, or gender. It also means that practitioners can have a greater impact on reducing patients’ risk by addressing air quality and income disparities, such as by installing air filters or offering financial support, respectively.
What Can You Achieve in Home Health with AI-enabled SDOH Insights?
Readmission rates are a common benchmark for success for home health providers. After all, the goal of Hospital at Home is to create an alternative to treating patients in the hospital. To this end, a peer-reviewed study published in Applied Clinical Informatics in 2020 found that Jvion’s SDOH insights helped a Wisconsin Hospital reduce readmissions by 25% within six months of implementing the Jvion CORE. Northwell Health, the largest healthcare provider in New York State, realized similar results.
Applying these insights can also have a measurable impact on healthcare costs, as we have shown by analyzing publicly available data from Oklahoma Medicaid.
After analyzing the influence of several social determinants including income, education, unemployment, environmental hazards, transportation costs, and access to food, Jvion scored each county by their social vulnerability from a scale of 1-6. The analysis found that 96% of the variation in per member per month costs were attributable to social vulnerability.
A closer analysis of the data revealed that by expanding access to food banks in counties where food security had a high impact on social vulnerability, Oklahoma could save $7.85 per Medicaid beneficiary per year by targeting food insecurity using Jvion’s SDOH insights. That’s nearly $10 million in savings per year statewide.
The Jvion team is here to support payers, providers, and other healthcare entities looking to tackle healthcare in the home setting. We’d love to connect with you and share how we can help you design and deploy care teams that can care for patients more holistically, reduce disparities, improve outcomes, and lower costs. Let’s talk!