The technology behind avoidable acute care reduction
We predict which Medicaid patients are headed towards preventable hospital and ER visits - and deliver the right intervention before it’s too late.
$50 billion in avoidable acute care costs, every year
40% of hospital and ER visits in Medicaid are preventable. But traditional risk models only identify patients after they're already high-cost. The challenge is proactively identifying which patients are heading towards crisis, who will respond to intervention, and when to deliver it.
40%
of hospital and ER visits are preventable
$50B
in avoidable acute care costs
Source: AJMC (2024). DOI: 10.37765/ajmc.2024.89630
Our models
The Waymark SignalTM Suite
Signal for
Rising Risk
90% accuracy predicting future avoidable hospital/ER visits
Signal for
Duals
80% accuracy predicting future avoidable hospital/ER visits for dual-eligible populations
Signal for Quality
Improvement (HEDIS)
85% accuracy in predicting who needs outreach to close HEDIS gaps
Signal for
Maternity Risks
55 days earlier identification of high-risk pregnancies
How it works
Beyond risk scores: predict, prioritize, intervene
1. We find the right patients before it’s too late
Waymark Signal identifies rising-risk patients headed toward avoidable hospital and ER visits
90% accuracy predicting future avoidable ER/IP visits for Medicaid patients
Most Likely to Benefit Model predicts which rising-risk patients will respond to intervention
49% additional reduction in acute care events compared to using rising-risk scores alone
2. We help care teams prioritize when and how to intervene
Time-to-Event Model predicts the optimal moment to intervene
81% accuracy identifying patients who will have an acute event within 30 days
Next-Best-Action Model recommends what intervention to deliver
20% reduction in acute care vs. standard care management approaches
3. We get smarter with every interaction
Built on outcomes from 30.6M Medicaid patients
The largest Medicaid-specific dataset used for predictive modeling — incorporating claims, behavioral health, pharmacy, and social determinant data.
Continuously calibrated to local populations
Reinforcement learning models help us continuously optimize patient outcomes.
Sources
- Nature Scientific Reports (2024). DOI: 10.1038/s41598-023-51114-z
- npj Digital Medicine (2025). DOI: 10.1038/s41746-025-01797-7
- npj Digital Public Health (2026). DOI: 10.1038/s44482-025-00003-5
- JMIR AI (2025). DOI: 10.2196/74264
Proven through peer-reviewed research
Our technology and outcomes have been independently validated, demonstrating measurable clinical and financial impact.
Clinical impact
48% reduction
in avoidable hospitalizations
20% reduction
in avoidable ER visits
35% improvement
in HEDIS gap closure vs. traditional methods
Financial impact
$2,347 in savings
per member per year
3:1 ROI
return on investment in first year
3x improvement
in cost prediction over traditional models
Sources:
- NEJM Catalyst (2024). DOI: 10.1056/CAT.24.0060
- npj Digital Medicine (2025). DOI: 10.1038/s41746-025-01797-7
Waymark Compass
Better insights for care teams, more support for patients
Our physician-supervised AI platform delivers predictive insights to care teams and engages patients directly between visits, helping them act on health risks before they escalate.
