By Mike Hoxter

As value-based care (VBC) replaces fee-for-service, providers and payers face pressure to improve outcomes while controlling costs. Because VBC rewards outcomes rather than volume, system-wide risk stratification is emerging to identify not only the sickest patients but also those at future risk.

Too often, traditional stratification focuses on patients already in crisis, limiting its potential impact. This approach invests heavily with only modest improvements, missing opportunities for change.
The real opportunity lies with rising-risk patients. These are individuals whose conditions are beginning to worsen but who may still benefit from timely intervention. Reaching them earlier, while risks are modifiable, is the difference between bending the curve and paying for preventable admissions.

Seeing the whole risk picture
Population risk stratification (PRS) categorizes patients as high, rising, or low risk. Traditionally, these models relied almost entirely on claims and clinical data. But medical history alone rarely tells the full story. Social determinants of health (SDOH) – housing, transportation, food access, employment, social support – account for as much as 80% of outcomes. Ignoring them leaves organizations blind to the real forces driving patient trajectories, resulting in sub-optimal patient outcomes and avoidable, expensive utilization.

That reality is urgent in chronic disease management. Chronic illness is responsible for nearly three-quarters of deaths worldwide, and the number of adults over 50 living with at least one chronic condition is expected to double by 2050. When organizations miss the social and behavioral components of these conditions, they also miss the warning signs and the opportunities for earlier engagement.

Take a patient with type 2 diabetes. On paper, their condition may look stable. But if they face food insecurity, economic uncertainty, and unreliable transportation, their real risk could be much higher. By integrating SDOH into risk models, organizations can design and implement interventions that match the patient’s experience – before their condition escalates.

Making risk scores work in the real world
AI and analytics have pushed risk stratification far beyond categorization. Modern models draw on EHRs, claims, SDOH, and even wearable device data to forecast risk. Importantly, they also explain the drivers behind that risk and prescribe actions to mitigate risk.

U.S. Healthcare Homepage