Workers' Comp Drug Screening Cuts Pharmacy Spend 18.4%, but Human Review Remains Central

What the Enlyte Report Found in Workers' Comp Pharmacy Spend

Workers' compensation claims administrators are increasingly using automated prescription-review tools to flag high-priced drugs, duplicate therapies and unsafe medication combinations. The software screens prescriptions against formularies, treatment histories, state requirements and lists of excessively priced or low-value products, then routes concerns to examiners or pharmacists for clinical review.

The financial scale of that approach was quantified in a July 31 Enlyte report. Clients that combined pharmacy benefit management, medical bill review and clinical decision support cut total pharmacy spending by 18.4%, according to the report. More than half of those savings, 58%, came from out-of-network prescriptions, even though those scripts made up a smaller share of volume. Savings tied to what Enlyte calls opportunistic products averaged $915 per prescription, nearly four times the average for other clinical interventions.

Those opportunistic products include private-label topical analgesics containing ingredients available in cheaper products and convenience packs that repackage existing drugs and supplies under new product codes, according to Nikki Wilson, senior director of clinical pharmacy services at Enlyte. Wilson said Enlyte's decision rules and clinical messages are developed and maintained by pharmacists, not artificial intelligence.

Gallagher Bassett and Sedgwick described a similar boundary. Automated systems organize cases, flag risks and summarize records, but drug authorization decisions remain human: claims examiners can override initial denials, and pharmacists contact prescribers before any medication is changed.

Why Enlyte, Gallagher Bassett and Sedgwick Keep Drug Decisions Human

What Enlyte's 18.4% savings number does and does not prove

The headline reduction in total pharmacy spending is substantial, but the report shows the gains are not evenly spread. Out-of-network prescriptions produced 58% of savings despite representing a smaller share of volume. That suggests the biggest opportunity lies not in broadly reducing all prescriptions, but in targeting specific high-cost, low-value categories. Treatments for opportunistic products averaged $915 in savings per prescription, almost four times the return from other clinical interventions, and nearly half of that opportunistic-product savings required no examiner intervention when automated decision rules were in place.

Still, this is a vendor-reported figure from Enlyte, based on clients that had already adopted its integrated tools. The direction is credible, but other workers' comp payers may not achieve the same percentage unless their review workflow is similarly integrated across pharmacy, medical bill review and clinical decision support.

The automation line: pharmacist-maintained rules, not AI approvals

Enlyte, Gallagher Bassett and Sedgwick all drew the same operational boundary: software can organize information and recommend action, but the final call on a medication stays with a person. Enlyte's Wilson said its rules are built and maintained by pharmacists, not artificial intelligence. Gallagher Bassett's Leah Sharp said her organization does not allow pharmacy benefit managers to automatically decide drug authorizations; its tools give real-time guidance based on state rules, clinical evidence, health-risk scores and client policy. AI is used for record summarization, not for approving or denying a drug.

Sedgwick's Andrew Newhouse added that even routine substitution of an FDA-approved generic is not an AI decision: a pharmacist contacts the prescriber, and any therapy change remains a clinical decision. This shared position suggests the workers' comp market is adopting AI for efficiency while consciously keeping drug determinations in human hands.

Polypharmacy is where the clinical stakes are highest

Sedgwick's tools sort cases by severity and push potentially dangerous polypharmacy cases to the front of the review queue. The rationale is straightforward: when an injured worker takes several medications, one drug can increase or reduce the effect of another, making the whole regimen hard to balance. Early flagging of interactions or duplicate therapy gives a pharmacist or claims professional a reason to review the entire file, even if the system itself cannot change the medication.

For employers and payers, this is the part of automation with both cost and patient-safety benefits: it speeds up review of risky cases rather than simply denying expensive items.

Claimant attorneys' caution points to future disputes

California Applicants' Attorneys Association president Tiffany Speers said the claimant bar is skeptical of automation and values safeguards. Her view, that automation can summarize information but a person must analyze it before a decision, outlines the likely area of conflict if automated denials are perceived as replacing individualized review. Payers that preserve a visible human override and document the clinical basis for a decision will be better positioned in disputes.

Where Claims Teams Can Target the 18.4% Savings Opportunity

For claim operations and insurance buyers, the Enlyte report points to specific review priorities rather than a blanket push into more automation.

• Target opportunistic products first. Enlyte found average savings of $915 per prescription for these products, nearly four times other clinical interventions, by identifying private-label topical analgesics and convenience packs that repackage existing ingredients. A prior-authorization rule that flags these product codes can capture the same category of waste.
• Review out-of-network scripts before broad formulary cuts. Out-of-network prescriptions generated 58% of total savings while representing a smaller share of volume. Concentrating clinical review there is likely to produce more savings per case than across-the-board restrictions.
• Ask vendors whether their decision rules are pharmacist-maintained. Enlyte's rules are built by pharmacists, and Gallagher Bassett explicitly does not let its pharmacy benefit manager automatically decide authorizations. Contract terms should specify that any automated denial can be overturned by an examiner and that final therapy changes require prescriber contact.
• Use severity sorting for polypharmacy. Sedgwick moves high-risk multi-medication cases to the front of the review queue to catch interactions and duplicate therapy. A similar queueing rule can prioritize patient-safety reviews without delaying standard claims.
• Expect scrutiny from claimant attorneys. The applicants' bar is skeptical of automation and treats it as information summarization, not decision-making. Keeping a documented human analysis on every denied or changed prescription is the practical safeguard against disputes.