Healthcare organizations entering risk-based contracts often don’t adequately consider how difficult it will be to stratify the risk of patients who will be treated under the contracts, and as a result, they don’t get the outcomes that they envisioned.
The reason is simple, Chilmark Research contends in a new report. Only 10 percent of outcomes are driven by medical care, 20 percent of outcomes are driven by genetics, and 70 are based on individual behavior and social context, says Jody Ranck, an analyst at Chilmark and lead author of the report.
Further, behavioral and social data help clinicians see the barriers that patients face, such as not being able to walk in the neighborhood each day because of high crime rates or the inability to pay for medications.
Risk stratification was developed by healthcare payers to introduce fairness into physician compensation based on patient severity, says Ranck. Now, new models of risk stratification focus not just on triaging high-risk patients but on what to do to keep them from using excessive amounts of medical services.
That’s a big change in approach, because physicians have been paid for triaging—incentives were such that doctors provided routine care and the onus was on the patients to follow their recommendations—if patients didn’t follow instructions, they just returned for more care, and physicians received an additional payment for that care encounter. That won’t wash in a new era of accountable care, where reimbursement will be based on quality, not the volume of services.
Accountable care requires access to real-time clinical data, patient-reported data and health assessments that can be fed into an analytics program, according to Ranck. That’s different from traditional data sources based on claims data and patient health risk assessment forms.
However, getting behavioral and social data into EHRs is difficult, Ranck acknowledges. But there are start-up companies emerging that could solve that problem over the next five years.
One of the newer vendors, Forecast Health, collects 4,000 data elements on patients, such as transportation options, finances, lifestyle factors and social media activity. Another vendor, Scio Health, uses claims, clinical, census and ZIP code data to understand risk well enough to intervene and reach out to patients.
Provider organizations can use these data to identify patients that should be called by a nurse to see why they are not adhering to their care plan; for example, if the barrier is transportation, a provider might decide to provide transportation services to pick up a patient for care. The data also can show which patients best respond to phone calls, texts or emails, as well as their literacy levels, and receive personalized messages with scheduling options for appointments.
As these patients are being identified and contacted, risk stratification can show if patients have a pattern of not showing up for appointments, resulting in subsequent hospitalizations, Ranck says.
“You need a 360-degree view of high-risk patients and a strategy. What is the context of the patients, and how can we customize a care plan to keep them healthy? It’s intelligence gathering and transferring that intelligence into an actionable intervention.”
For instance, analytics can show that most falls happen in certain kinds of apartment buildings, and providers can use this intelligence to find ways to reduce falls, which could lower hospitalizations.
With social factors traditionally being a barrier to getting care, the job of improving access to care has fallen on social services agencies, Ranck notes. Leaving the job entirely to such agencies isn’t sufficient in an accountable care era. “It was always someone else’s job, and now it is the physician’s and hospital’s job to augment traditional social services.”
Consequently, providers need to focus on the highest-risk patients under their risk-based contracts, then use predictive analytics to find the next level of high-risk patients that could transition to become high utilizers of services, Ranck says. “That’s the holy grail of predictive analytics—finding out who they are.”
(This article appears courtesy of our sister publication, Health Data Management)
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access