For property and casualty insurers and workers' compensation payers, utilization review (UR) sits at the intersection of cost containment, regulatory compliance and patient outcomes, making it one of the most operationally complex functions in claims management. As automation technology matures, many payers are under pressure to apply it to their UR programs. But doing so without the right guardrails can expose carriers to significant risk.
In this Q&A, Susan Doering, senior director of clinical operations, utilization review at Enlyte, explains how payers can navigate that tension. Drawing on real-world examples from California's evolving chronic pain program guidelines to the hidden risks of seemingly routine pharmacy requests, she describes a framework carriers can use to determine when automation adds value, when it creates liability and how to build governance structures that keep clinical judgment at the center of the process without becoming a bottleneck.
Payers are often told to automate as much as possible, but when it comes to utilization review there are some concerns in ensuring clinical expertise is part of the process. What's the best way to balance it?
Doering: "The most effective utilization review (UR) programs treat automation as a force multiplier for clinical expertise, not a replacement. The goal is to use automation technology as a tool to amplify and extend the impact of clinical professionals. It should never substitute for their judgment where it's genuinely needed.
When done correctly, automation reduces the noise and administrative burden by handling lower risk, conservative care options. As a result, clinicians can focus on quality, defensible outcomes rather than throughput. To ensure that it's properly balanced, automation should employ configurable rules and evidence-based logic with embedded checkpoints to escalate higher-risk cases to nurses and physicians.
For example, automated workflows can efficiently handle low-risk, guideline-compliant, conservative care options or treatments exempt from prior authorization per state regulations. Office visits and consultations fall into this category, too. These kinds of services don't require first-level (nurse) or second-level (peer) reviewer oversight to approve.
On the other end of the spectrum are intensive treatment approaches like chronic pain programs. These are a collection of services commonly billed under an unlisted CPT code (97799) with no standard fee schedule. These treatments are often too high-risk, costly, and absolutely require clinical oversight."
How do claim characteristics, such as severity, jurisdiction or pharmacy risk shape where automation actually helps versus where it creates problems?
Doering: "Not all claims benefit equally from automation. Straightforward, conservative care lends itself to it. Catastrophic or escalating claims do not, because the complexity of care requires clinical expertise and oversight.
Jurisdictional requirements are another critical layer. States with strict UR mandates, active audits and/or strong enforceability demand tighter controls and more clinical oversight.
Pharmacy and treatment risks are the third dimension. Even formulary medications can become problematic when providers prescribe non-standard dosages. Take ibuprofen, for example. Some providers are now prescribing a 300mg dosage, but some formularies have classified this as a non-exempt drug requiring prior authorization because there's no added efficacy over standard dosages and [it] carries a higher cost. Those requests shouldn't move through automated workflows without scrutiny.
Litigated claims are another category where automation falls short. Clinical expertise and oversight are critical to ensure documentation, rationales and guideline applications are well-explained and defensible should the case end up before a judge."
How should payers think about configurability with UR automation?
Doering: "Configurability is essential because UR is not a set-it-and-forget-it operation. Regulations change, guidelines are updated, new treatment patterns emerge, and automation rules need to keep pace.
Payers should be able to set automation thresholds based on their own philosophy, line of business and jurisdiction, supporting both permissive and restrictive models depending on risk tolerance. The ability to configure rules for specific triggers like duration of care, procedure type, or pharmacy exposure is critical.
Adaptability lays out the foundation for a program that can remain compliant and efficient."
Can you walk us through a real-world example of dialing automation up or down based on a specific risk, such as chronic pain management or a litigated claim?
Doering: "The chronic pain program example is really the central case study for a lot of payers, particularly in California. In June 2025, a state regulatory guideline update refined the criteria for California's programs, reinforcing just how rigorous and selective the standard should be.
As mentioned earlier, these programs are commonly billed under an unlisted CPT code with no set fee schedule. They involve a complex collection of services over multiple weeks, and the medical necessity criteria are highly specific. Auto-approving one of these without proper clinical review poses both clinical risks and major financial exposure. Programs of this nature may incur charges exceeding $100,000.
That's exactly where you need clinical expertise in the loop, not automation making the call."
How should payers build governance guardrails that protect clinical judgment without slowing down the review process?
Doering: "Governance guardrails work best when they're embedded in the workflow, instead of layered on top as an afterthought. When the confidence level on a treatment request isn't assured, the system should automatically route it to a clinician.
Clinical oversight and trend monitoring are equally important. As soon as health care providers prescribe non-standard treatments, it needs to be proactively captured in the rule configuration. You don't want to discover it after the fact through appeals or audits.
Ongoing feedback loops are what keep the guardrails current and effective."
For a payer trying to recalibrate their automation strategy today, where do you recommend they start, and what outcomes should they measure to know it's working?
Doering: "Start with the data, not assumptions. A strong recalibration effort begins by analyzing which reviews involve low-risk treatments that are consistently approved, where clinical time is being spent with limited added value, and which jurisdictions or treatment types are driving disputes or audits.
You'll also want to measure turnaround time, as some jurisdictions require decisions within two business days or 72 hours for expedited care. Others include cost per utilization review, medical savings per claim and audit and appeal outcomes.
Clinician productivity is also worth watching. When automation is working, utilization review physicians and nurses are spending more time on complex, non-guideline-supported treatments rather than routine approvals. When it comes together, payers see faster decisions, stronger defensibility, stable or improved outcomes, and clinicians applying their expertise where it truly matters."
Susan Doering, DC, is Senior Director of Clinical Operations, Utilization Review at Enlyte.
(Photo credit: Rawpixel.com/Shutterstock.com)
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.