Predictive analytics become essential as high-cost claimants surge

The financial gravity of healthcare is increasingly concentrated in a small but rapidly evolving segment of patients: high-cost claimants. For employers and health plans, this group represents one of the most urgent financial risks in benefit management.

New therapies, changing oncology paradigms and an expanding drug pipeline, especially for rare diseases, are reshaping the definition of high-cost members. At the same time, advances in machine learning-based risk prediction models enable organizations to be more proactive in planning benefit strategies and care coordination for these individuals.

The growing concentration and volatility of high-cost claimants

The foundational challenge remains stark: Healthcare costs are highly concentrated. According to 2025 reporting, approximately 1% of members generate about 30% of total healthcare spending, underscoring how a small fraction of populations drives disproportionate financial risk.

At the same time, the definition of "high cost" is shifting upward with more members reaching $1 million or even $2 million in annual claims cost. What is new (and particularly challenging), though, is volatility. Only about 21% of high-cost claimants persist year over year, meaning the individuals driving costs today are unlikely to be the same individuals driving costs next year.

This unpredictability has profound implications for employers and plan sponsors trying to forecast budgets, design benefits, and deploy care management programs. Traditional approaches that focus only on current high-cost members are increasingly inadequate.

Read more:  How procurement services can lighten healthcare spend

Cell and gene therapies: A new cost paradigm

Few healthcare developments illustrate the changing cost landscape more clearly than cell and gene therapies (CGTs). These treatments, often designed as one-time, potentially curative interventions, dramatically alter both clinical outcomes and financial models.

And prices for these therapies are staggering. CAR-T therapies, widely used for certain cancers such as leukemia and lymphoma, frequently carry list prices ranging from approximately $373,000 to over $500,000 per treatment. Factoring in biopsies and testing, cell harvesting and prep, hospitalizations, outpatient follow-ups and complications, total episodic costs often exceed $1 million.

In some cases, gene therapies, which are largely used for rare disorders, but increasingly being explored for more common chronic diseases, even exceed these levels. One recent therapy launched with a price of over $4 million per dose, highlighting the upper boundary of emerging treatment costs. 

For employers and payers, these therapies present a paradox. They can replace years of chronic treatment costs and potentially cure the disease, but they also introduce extreme, episodic financial shock that is difficult to predict and manage.

This shift from recurring costs to high-impact, low-frequency events fundamentally alters actuarial assumptions. Stop-loss coverage, reinsurance, and risk-pooling strategies are being reevaluated, but even these mechanisms are under pressure as CGT utilization grows.

Read more:  3 ways women are still struggling in the workplace, and what benefits could help

The expanding drug pipeline intensifies future risk

Cost challenges are not limited to therapies already on the market. The pipeline of new medications suggests that high-cost claimants will become more numerous and more complex.

Specialty drugs now account for a dominant share of innovation, representing approximately 75% of the drug development pipeline, according to Pharmacy Times. Oncology is a major driver of this expansion. Projections suggest oncology-related spending could nearly double on a per-member basis over a short time horizon, reflecting both new treatments and broader indications. 

This pipeline-driven growth is compounded by broader macro trends. Specialty drugs, though used by fewer than 5% of patients, already account for at least half of pharmaceutical spending, a share that continues to grow. 

For employers and health plans, this means the future is more uncertain, as new therapies create entirely new categories of high-cost claimants.

Oncology screening gaps: A hidden driver of future costs

While innovation is pushing costs higher, gaps in preventive care are quietly amplifying long-term risk, particularly in oncology.

Cancer screening rates remain uneven across populations. An American Cancer Society screening survey showed that 31% of adults are more than a year behind on recommended cancer screenings, often due to cost concerns, administrative barriers, or lack of awareness. Additionally, cancer diagnoses are rising in younger populations. One in five new colorectal cancer diagnoses are for patients under 55, and breast cancer is now the most common cancer diagnosed in young adults, driven by estrogen-receptor positive (ER+) cancer. Unfortunately, these populations are historically left out of screening coverage or tend to have higher gap rates than older individuals.

These gaps have significant downstream implications too. Deferred, delayed, or missed screenings increase the likelihood that cancer is detected at later stages, where treatment is more aggressive, prolonged, and expensive. The President's Cancer Panel has emphasized that insufficient screening leads to avoidable disease burden and higher treatment intensity, with both clinical and economic consequences. 

At the same time, the rapid evolution of oncology therapeutics is changing how payers think about screenings altogether. With more high-cost, specialty therapies available, gaps in screenings can be a predictor of future high-cost claims and interventions and gap closures, a strategic way to help manage prospective costs.

Read more:  Healthcare costs are forcing employers to reconsider their benefits strategy

A new framework: Predicting the "rising cost" population

Against this backdrop, employers and health plans are increasingly adopting predictive analytics to identify current and future high-cost claimants.

New machine learning models demonstrate a shift in healthcare toward prospective, personalized risk identification through analyzing a wide range of variables — including prior utilization, demographics, disease burden, and social determinants — to predict which members are most likely to experience significant cost increases. 

Unlike traditional risk adjustment approaches, these models answer a set of forward-facing business questions:

• Who is at risk of becoming high-cost next year?
• When will that cost occur?
• What conditions or utilization patterns are driving the risk?
• How much is that individual expected to cost? 
  

From a technical perspective, these models can incorporate hundreds of features across medical, pharmacy, and eligibility data, enabling a far more granular understanding of risk.  
From a business perspective, their value lies in enabling earlier intervention and prospective budgetary planning. Rather than reacting to members after they become high-cost, organizations can identify individuals on an upward cost trajectory, formulate a plan to manage their costs, and deploy interventions such as care management, site-of-care optimization, medication management, or targeted navigation.

Machine learning is moving from optional to i mperative

The broader healthcare analytics landscape reinforces this shift. Machine learning models have demonstrated increasing effectiveness in predicting clinical events, such as hospital readmissions, which are closely tied to cost outcomes. These models leverage complex relationships across data types and can achieve higher predictive accuracy than traditional approaches, enabling earlier and more targeted interventions. In the context of employer health and payer strategy, this capability is becoming essential for several reasons:

1. Budget planning in a high-volatility environment
With high-cost claimants changing year to year, retrospective analysis cannot support reliable forecasting.

2. Managing emerging therapies
Cell and gene therapies, along with pipeline drugs, introduce episodic, high-impact costs that require proactive identification.

3. Aligning clinical and financial strategy
Predictive insights enable organizations to align care management with financial priorities — targeting the right members at the right time.

4. Supporting benefit design decisions
Insights into rising-risk populations can inform decisions on stop-loss thresholds, specialty drug coverage, and preventive care incentives.

The strategic imperative for employers and health plans

The convergence of these trends — cost concentration, therapeutic innovation, preventive care gaps, and predictive analytics — marks a turning point in risk management for employers and health plans. They are moving from a reactive posture, where costs are explained after the fact, to a proactive model, where risk is identified and managed before it materializes.

In this new paradigm, the success of employer-sponsored healthcare will be defined by how effectively managers can predict costs and remain nimble in their benefit strategy. In a world where a single gene therapy can exceed $1 million, where oncology innovation is accelerating, and where screening gaps quietly drive downstream costs, the ability to identify individuals at rising risk is no longer optional.