Medical costs per claim are rising again. The Workers Compensation Research Institute reported in October 2025 that after several years of relative stability, medical payments per claim increased 5% in California, 6% a year in Wisconsin, 7% a year in Delaware, and 14% in a single year in Pennsylvania. Utilization, medical prices, and fee schedule updates are all pushing in the same direction. When the C-suite sees that trend line, the question is not whether to respond. The question is where the response comes from.
Training is usually not the first-place people look. This line item sits next to travel and software as a discretionary line, and discretionary lines tend to get cut first. That is a mistake, because the decision that determines whether a claim’s medical spend stays reasonable or drifts upward is made by an adjuster and their support system, not by a fee schedule. Fee schedules control price. They do not control whether the treatment plan submitted for approval matches what the evidence says should happen next. That judgment call is trained into that user, or it is not.
The mechanism: guideline-concordant care changes the outcome, not just the cost
A 2021 study published in PLOS ONE examined nearly 60,000 acute low back pain claims in the California workers’ compensation system between 2009 and 2018. The researchers compared lost workdays for workers who received only guideline-recommended interventions against workers who received only non-recommended interventions. After adjusting for other factors, workers who received guideline-concordant care had 11.5 fewer lost workdays than those who did not, a difference the study found statistically significant at well beyond conventional thresholds (Gaspar et al., PLOS ONE, 2021). Workers who were prescribed an opioid, a treatment protocol that guidelines generally do not recommend as a first-line intervention for acute low back pain, were out of work a median of four days longer than workers who were not.
A separate study in the Journal of Occupational and Environmental Medicine quantified the same relationship in dollars. It found a statistically significant trend between adherence to ACOEM guidelines for initial management of low back pain and lower claim costs, with medical and total costs trending down by roughly 353 dollars and 586 dollars respectively for each unit of improvement in the guideline compliance score. No outlier high-cost claims appeared in the groups with the best guideline compliance.
Put those two findings together and the plan to pursue is clear. The relationship to be followed is not that guideline adherence is not limited as a nice quality marker. It is that a treatment plan drifting away from the evidence-based medicine protocols tends to run longer and cost more, and this is easily identified in a way that is visible early, before the claim becomes an outlier. That fully trained and educated individual has to be able to see that drift when the request for authorization crosses their desk. That is a teachable skill. It is not a personality trait, and it is not something a fee schedule update can substitute for.
Where the judgment call actually happens
A skilled claim file handler who cannot distinguish a guideline-concordant treatment plan from one that has quietly expanded past what the mechanism of injury supports will approve what is in front of them, because the alternative is guessing without a framework. That is not a criticism of the adjuster. It is a description of what happens when clinical judgment is expected without clinical training to support it. The fix is not to make adjusters into clinicians. It is to give them enough grounding in how evidence-based treatment guidelines work, and enough comfort recognizing when a request for authorization needs a second look, that the decision in front of them gets made with the right frame of reference.
This is where the initial and ongoing education spend stops being a benefit and starts being a control. A trained claim file handler catches guideline drift in week two. An untrained handler catches it, if at all, when the claim is already six months old and the file has become expensive to unwind. The 11.5-day gap in the California study and the per-unit cost trend in the JOEM study are not abstractions. They describe the difference between those two claims.
Measure the outcome, not the seat time
Most training programs report attendance and satisfaction scores. Neither number notes the ROI of this aspect. If the much-needed education spend is going to survive the next budget cycle, it needs a scorecard built from the numbers that the finance section already tracks:
- Average claim duration and lost workdays by injury type
- Reopen rate
- Share of treatment requests approved that match evidence-based guideline recommendations
- Reserve accuracy at 90 and 180 days
- Medical cost trend per claim, tracked against the WCRI benchmark for your state
Set a baseline before a training initiative, then track the same five numbers afterward, tied to the training calendar so movement can be traced to specific content. This is the same discipline a technology purchase is held to. A tool spend deck usually measures activity, logins, tickets closed, time saved. A competent spend deck should measure outcomes because that is where its value actually shows up. Against a medical cost trend running 5 to 14% a year, even a modest improvement in decision quality pays the program back several times over.
The practical takeaway
If education is up for review in next year’s budget, do not defend it as a benefit. Hold it to the same standard as any other cost-control investment. Ask what specific claims outcome each training dollar is meant to move, set a baseline, and measure the result against it. A program that cannot answer that question deserves the scrutiny it gets. A program that can answer it belongs in the same conversation as any other lever you pull to manage total cost of risk.
We spend a fair amount of time at WorkCompCollege.com on exactly this question, how to connect professional education to the claims outcomes that finance is already tracking. If you are building the case for a training budget and want a framework for measuring impact rather than attendance, that is the kind of thing the resources there are built to help with.
