The Blind Spot in Workers’ Comp Risk Stratification: Behavioral Health

By Nancy Schaefer, Vice President of Sales & Customer Activation

Originally published in WorkCompWire

Risk stratification is not new in workers’ compensation. Carriers and payers have used it for years to route claims: flagging catastrophic injuries, watching for red-flag diagnoses, triggering nurse case management. What that infrastructure typically sees is physical risk: injury severity, body part, surgical versus non-surgical. What it typically treats as secondary is psychosocial risk. That gap is the blind spot, and it’s proving expensive.

Ask most claims organizations what determines a claim’s risk tier, and the model starts with the injury itself: diagnosis codes, body part, treatment guidelines. Reasonable inputs, but incomplete. A substantial body of research from pain science and rehabilitation medicine has identified a predictor most standard risk models don’t capture at all: a worker’s own expectation about whether they will recover. In one prospective study of workers on extended sick leave for low back pain, high recovery expectations were a strong, significant predictor of actual return to work at 12 months; job satisfaction was not.¹ A worker who believes they’ll be back in six weeks tends to be back in six weeks. A worker who believes the injury has ended their career often proves that belief is right, regardless of what imaging shows. Fear-avoidance and pain catastrophizing compound the effect. Neither shows up on a scan, both are reliable yellow flags that will extend claim duration.

What standard models miss

Claims with a behavioral health component run about 2.5 times more expensive than claims without one, according to NCCI research.² The Workers Compensation Research Institute has reached a consistent conclusion from a different angle: in a study of more than 131,000 physical therapy episodes, injured workers with unaddressed psychosocial risk factors showed measurably poorer functional recovery than those without.³ WCRI describes these as “yellow flags” that commonly prolong disability and delay return to work. This affects a relatively small share of claims, generally 5 to 7 percent, but that small share carries a disproportionate share of total cost, exactly the kind of claim a well-functioning risk model should be catching early and isn’t.

Why the gap persists

This isn’t the industry ignoring behavioral health; most organizations have some referral pathway. The gap is structural: behavioral health is typically a downstream referral, triggered after a claim has already become difficult, rather than an upstream input into the risk assessment itself. By the time a claim triggers that referral, the risk that made it complex has often had months to compound. Delays beyond 180 days roughly triples disability days compared to care that starts within 90, according to CorVel.⁴ The gap is also harder to afford now: injuries are taking longer to heal generally, averaging 80 missed workdays per Travelers’ 2026 Injury Impact Report,⁵ while behavioral health coverage obligations expand state by state. With 46 percent of psychologists reporting no openings for new patients⁶ and a 48-day average wait, ⁷ a downstream referral model can’t keep pace. Precision must move upstream, into the initial stratification itself.

What the blind spot costs

Consider a book of business handling roughly 10,000 claims a year, with an average claim cost of $15,000. If just 7 percent of those claims, 700 cases, carry a meaningful behavioral health component that goes unidentified early, the exposure adds up fast. Risk & Insurance reports that NCCI has found that claims with a mental health diagnosis run about 2.5 times more than claims without one.² Applied to a $15,000 claim, that puts the cost closer to $37,500, an additional $22,500. Across 700 claims, that’s roughly $15.75 million in incremental cost sitting inside the book, largely invisible until it shows up in results.

Now weigh that against the cost of catching it early. If assessment and targeted intervention averaged $4,500 per identified claim, the investment across those same 700 claims would total $3.15 million. Separate research published in JAMA Network Open found that an employer-sponsored behavioral health program generated roughly $1.90 in reduced medical spending for every $1 invested.⁸ Applied only as an illustrative benchmark, that would put potential savings around $5.99 million, a net benefit of about $2.84 million after program costs.

No carrier will hit that return exactly. The point is narrower than that: behavioral health risk already carries a measurable cost and not identifying it doesn’t make that cost go away. It just moves it somewhere harder to see.

What a complete model looks like

Closing the blind spot doesn’t mean treating every injured worker as a psychiatric patient. It means adding validated screening in a clinically meaningful way, Options include questionnaires tapping into return-to-work expectations, injury-related distress, and yellow flags like pain catastrophizing, then integrating with risk assessments already in place. Most claims need nothing more than monitoring. A smaller group needs a targeted intervention before the risk hardens. For the smallest group with the highest risk, a needs assessment within the first 4-6 weeks of the claim can evaluate which yellow flags pose the greatest threat to a timely recovery and appropriate intervention can mitigate recovery delays. Waiting months into the claim, hoping for resolution, results in poorer patient outcomes because recovery is already off track and a disability mindset has more time to solidify. Effective timing matters as much as the tools: screening at four weeks is early enough to catch yellow flags before delayed recovery patterns set in, but far enough along to separate natural recovery from a stalled one.

Case-level outcomes reflect the broader research when intervention is timely and focused. For example, an employee, a veterinary technician, fractured her ankle after being kicked by a horse and underwent surgical repair that went well, but recovery stalled: fear of movement led to inconsistent physical therapy, and her return-to-work timeline began slipping. Eight sessions of the Work-Focused Unified Protocol, a structured treatment for injuries complicated by fear and avoidance, addressed the psychological drivers directly. She returned to full duty in 80 days, against an ODG benchmark of 225 days, a 145-day acceleration.⁹ Same injury as many in workers’ comp, but the variable that changed the outcome was whether the risk model caught the driver early enough to act on it.

Where this leads

Risk stratification has always been about catching what matters before it causes poor outcomes for the employee and unnecessarily balloons expenses. Most models do this well for physical risk, not psychosocial risk, and that gap is where a meaningful share of avoidable cost currently sits, unmeasured. Closing it doesn’t require rebuilding claims management from scratch. It requires adding one input, a validated behavioral health signal, to a process already built to sort claims by risk. The tools already exist. What remains is recognizing the blind spot for what it is: not a missing program, but a missing input in a model the industry already trusts.

References

1. Opsahl, J., Eriksen, H. R., & Tveito, T. H. (2016). Do expectancies of return to work and job satisfaction predict actual return to work in workers with long lasting LBP? BMC Musculoskeletal Disorders, 17, 481. https://doi.org/10.1186/s12891-016-1314-2
2. Risk & Insurance. (2025, November 24). The current state of complex claims in workers’ compensation: Understanding the drivers of rising costs and duration. https://riskandinsurance.com/the-current-state-of-complex-claims-in-workers-compensation-understanding-the-drivers-of-rising-costs-and-duration/
3. Thumula, V., Liu, T.-C., & Lea, R. D. (2024). Importance of psychosocial factors for physical therapy outcomes (WC-24-17). Workers Compensation Research Institute. https://www.wcrinet.org/reports/importance-of-psychosocial-factors-for-physical-therapy-outcomes; Workers Compensation Research Institute. (2022). A primer on behavioral health care in workers’ compensation. https://www.wcrinet.org/reports/a-primer-on-behavioral-health-care-in-workers-compensation
4. CorVel Corporation. (2025). Mental health at the forefront: How early intervention and virtual care are transforming workers’ compensation. https://www.corvel.com/insights/mental-health-at-the-forefront-how-early-intervention-and-virtual-care-are-transforming-workers-compensation/
5. The Travelers Companies. (2026). 2026 injury impact report, as reported in Safety+Health Magazine. (2026, May). Injuries are becoming “more complex,” workers’ comp claims show. https://www.safetyandhealthmagazine.com/injuries-are-becoming-more-complex-workers-comp-claims-show/
6. American Psychological Association. (2025). 2025 practitioner pulse survey: AI in the therapist’s office. https://www.apa.org/pubs/reports/practitioner/2025
7. National Center for Health Workforce Analysis. (2024, November). State of the behavioral health workforce, 2024. Health Resources and Services Administration. https://bhw.hrsa.gov/sites/default/files/bureau-health-workforce/Behavioral-Health-Workforce-Brief-2023.pdf
8. Hawrilenko, M., Smolka, C., Ward, E. J., Ambwani, G., Brown, M., Mohandas, A., Paulus, M., Krystal, J., & Chekroud, A. M. (2025). Return on investment of enhanced behavioral health services. JAMA Network Open, 8(2), e2457834. https://doi.org/10.1001/jamanetworkopen.2024.57834
9. Ascellus Behavioral Health. (2026). Internal case data and claims analysis [Unpublished raw data].

Note: The dollar example in “What the blind spot costs” is an illustrative model built from the benchmarks cited above and applied to hypothetical assumptions (10,000 claims, $15,000 average claim cost). It is not published data from any specific carrier.