The hidden costs of rising risk: Why proactive identification matters for health plans

Should we be surprised that three-fourths of Medicare Advantage plans posted an underwriting loss last year, representing a $5.7 billion⁽¹⁾ miss on the bullseye? Without a crystal ball, health plan leaders face a difficult balance of forecasting utilization and ensuring risk adjustment accuracy amid ever-shifting population dynamics. The natural movement within the population health risk pyramid⁽²⁾ continually challenges even the most sophisticated predictive models.

Challenges are:

• Rising risk escalation: Each year, 20% of a plan’s population moves from rising risk to high risk⁽³⁾, essentially doubling the number of high-risk members and the costs associated with them.⁽⁴⁾
• Cost impact: This shift is especially concerning given that high-risk members typically generate 10 times the cost relative to revenue.⁽⁵⁾
• Churn and data gaps: With average member churn around 10%, a 100,000-member plan can risk losing $120 million in reimbursements, while limited data from new enrollees weakens predictive risk adjustment. 

Without deeper insights on rising risk members and new enrollees, it’s a guessing game to pinpoint which ones will progress to high risk. To avert a repeat of last year’s underwriting loss, health plans must course correct now. 

Accelerating the cycle: From first contact to scaled insight

One way to course correct is by embracing AI-powered, member- and clinician-facing digital health platforms designed to gather more intelligent, timely, and action-oriented insights. These solutions are showing significant promise, beginning with “speed to scale” across a membership.

Speed to scaled engagement is critical. The faster a plan can deploy outreach and begin collecting data from a broad base of members, the faster it can surface meaningful trends and intervene. Instead of waiting months for claims data to reveal risk, digital engagement can identify emerging issues within days or weeks. This accelerated feedback loop not only drives earlier discovery of health risks but also strengthens member trust and participation—fueling even greater reach and richer data over time.

Surfacing health risks before it’s too late

Much of what drives cost and complexity within a population remains hidden until it’s too late, forcing plans to react rather than prevent. But what if we could identify costly, undiagnosed conditions far earlier than ever before? The data makes the case clear:

• Depression: More than 30% of moderate to severe cases go undiagnosed, inhibiting effective treatment and cost control.⁽⁷⁾
• Cognitive decline: 22% of age 65+ adults have Mild Cognitive Impairment (MCI), yet 92% of those remain undiagnosed – costing $25k-$35k per year per patient depending on severity. ^{(8, 9)}
• Hypertension: 26% of all seniors with Hypertension don’t know they have it, and 70% of seniors diagnosed do not have it under control.⁽¹⁰⁾

Claims data and traditional predictive analytics can’t close those gaps. What can: AI-driven health signals, captured directly via the member’s smartphone, without the need for any peripheral devices. 

• Vocal biomarkers can use just 40 seconds of speech for early and accurate pre-clinical detection of behavioral health and neurodegenerative conditions.
• Physiologic biomarkers analyze facial microvascular patterns for insights on blood pressure, heart rate, oxygen saturation, heart rate variability, and more.  

These innovations, when combined with insights on SDOH, medication adherence, activity, sleep, and nutrition, enable scalable, personalized engagement that helps plans uncover hidden risks and intervene sooner. Plans can now identify rising risk in real-time, not 6 to 12 months later when claims finally show up.

Automating workflows for better reach and results

Using smartphones and AI to capture and analyze member health data provides more reach and better results. As member engagement and data flow increase, automation drives real operational gains.  Features such as automated HRA digital deployment, automated risk stratification, AI-assisted care plan creation, and clinical decision support streamline workflows and amplify the impact of limited care management resources. 

Early insights fuels a continuous cycle of improvement: more accurate risk adjustment, improved reimbursement, deeper member engagement, and stabilization of the rising risk populations. Plans also gain earlier visibility into new member risk, reduced risk churn, and measurable cost containment.  

Discover and manage member risk before it becomes cost

The time to act is now. Most plans this time of year are gearing up to onboard thousands of new members in January, but with little to no insight into their true health needs. By the time claims reveal risk, costs are already climbing.

Percipio Health is offering a complimentary New Member Early Discovery to help plans:

• Uncover clinical, behavioral, and social needs that claims and encounters miss
• Identify high- and rising-risk members before claims come in
• Engage members immediately, using only their smartphone

This is a complimentary program that can help plans discover risk before it becomes cost and no longer wait for claims to tell the story. If you are interested in learning more, contact us today.

Citations

(1) Fierce Healthcare, 9-24-25, “Elevated utilization drove $5.7B Medicare Advantage underwriting loss in 2024”

(2) Advisory Board https://www.advisory.com/content/dam/advisory/en/public/ABResearch/Topics/Physician/Cheat-Sheets/Pop-health-cheat-sheet.pdf

(3) Wolters Kluwer, 3-4-20, “Getting the rising risk population under control.”

(4) Assumes a standard risk pyramid mix of 5% high, 25% rising, and 70% low risk.

(5) Journal of Managed Care + Specialty Pharmacy, 2-22-16, “What contributes most to high healthcare costs? Health care spending in high resource patients”

(6) Kaiser Family Foundation and Data Decisions Group, 11-21-23, “Increase Medicare Advantage revenue by reducing member churn”

(7) Cureus Journal of Medical Science, 8-14-22, “Prevalence and impact of diagnosed and undiagnosed Depression in the US”

(8) Alzheimer’s Research & Therapy, 7-22-23, “Expected and diagnosed rates of mild cognitive impairment and dementia in the U.S. Medicare population: observational analysis”

(9) The Journal of Prevention of Alzheimer’s Disease, 4-2-24, “Economic impact of progression from MCI to Alzheimer’s Disease in US”

(10) CDC National Center for Statistics, Data brief No. 511, October 2024, “Hypertension prevalence,awareness, treatment and control….”