Providers and payers are increasingly aligned in the journey to improve clinical outcomes at lower total cost of care. While there are numerous strategies to achieve this goal, they all come down to improving the allocation of resources: The right treatments, in the right setting, and the right support to the right patients at the right time. Effectively, our societal challenge is not to spend more but to spend it smarter.
Interestingly, the key to changing the value equation looks to not be in clinical domains, but in social ones. Studies from the CDC, Robert Wood Johnson Foundation, and the Kaiser Family Foundation, among others, all show that the biggest influence on an individual’s health status is not the clinical care they receive, but the social, behavioral, economic, and environmental conditions in which that person lives or the social determinants of health (SDOH).
Gaps in care knowledge
Despite the knowledge that social factors are disproportionately important, the vast majority of technology and data investments—electronic health records (EHRs), population health systems, claims data, diagnostic tools, etc.,—are focused on clinical elements. The key to success and the piece that is often missing, are SDOH. When a whole-patient view is available to providers early in their engagement with the patient, it enables more complete understanding of the challenges ahead and thus more efficient targeting of interventions, services and programs to achieve and sustain wellness. In a recent Geisinger Health report, a program for diabetic patients demonstrated that changing the nutrition for a subset of ’food insecure’ patients resulted in over a two point drop in A1C levels (every 1-point drop in A1C levels corresponds to a more than 20% decrease in chance of death and serious complications from diabetes). The cost of the program is approximately $2,200 per patient per year, with the payer-side benefit reaching $6,000 per patient per month—a greater than 30-times financial return.1
In an ideal world, providers would have unlimited time to spend with each patient and no clinical, social, economic, or behavioral element would be missed. However, the current realities of healthcare delivery preclude providers from gathering and synthesizing a comprehensive patient view. EHRs excel at capturing clinical insight and integrating them into care pathways and treatment flags, but they are short on sociodemographic and behavioral elements that critically modulate treatments and follow-ups for maximum impact.
Closing the gaps for value-based models
With hospitals, ambulatory services, and physicians increasingly interconnected under local networks and operating within newly designed value-based reimbursement models, the financial risk is clearly rising, and success requires greater sophistication. This leads to a series of critical challenges going forward:
- How do we augment the data currently available to the growing network of providers in a way that presents a more complete and common view for all?
- How do we gather social determinant insight about the patient early in the health relationship without interviewing every patient one-by-one?
- How do we prioritize patients for various programs, recognizing their finite capacity, and consistently gather feedback on their impact relative to clinical and sociodemographic challenges for better targeting over time?
What if you could gather a social determinant risk profile of a patient without the detailed interview? What if you could identify the 10 to 15% of patients with complex social context amid thousands of attributed lives? A key answer to these questions is to leverage data and technology that gathers the missing sociodemographic data and process the information to deliver an SDOH risk profile specific to the individual patient. These risk measures augment existing clinical insight to further optimize care management programs for the patient. All of this is done in the background so that the clinical teams are presented only the aggregate risk measures, the individual SDOH factors, and specific recommended workflow suggestions. The technology does not lead to data inundation, but rather insight and integration across clinical teams.
The solutions are built to augment clinical measures and systems, allowing care teams to more accurately assess the health of its patient population and to separate the risk drivers for more targeted intervention approaches. Such solutions consistently demonstrate higher predictive accuracy than models that use only clinical data in a variety of care applications such as readmission risk and appointment no-shows.
To build a highly effective predictive model, a solution needs to:
- Compile a diverse patient data set, one that spans geography, patients, and conditions. Within this diversity is the potential to unlock multifactorial relationships.
- Access third-party data about the patient, household, neighborhood, and environment while also understanding and accounting for the limitations of any such external data.
- Have the ability to convert the experiential information into stable and calibrated predictive models.
While the efficacy and accuracy of the model are a part of achieving a more complete patient view, it is the capability to use the insights to change the patient experience for better outcomes that creates the true value.
Reference:
- Feinberg, Andrea T., Slotkin, Jonathan R., Hess, Allison, Erskine, Alistair R. “How Geisinger Treats Diabetes by Giving Away Free, Healthy Food” Harvard Business Review 19 Dec. 2017
