WC Physician Dispensing Markups of 16,000% Distort Medical Severity Trends

When a single unit of diclofenac sodium costs $0.10 through a PBM-contracted retail pharmacy but $16.42 through a non-retail workers' compensation dispenser, the pricing actuary faces a severity distortion that standard medical trend models were not built to handle. That markup exceeds 16,000%. The WCRI's 2026 National Inventory of Workers' Compensation Prescription Drug Regulations (report WC-26-04, published February 25, 2026) catalogues formulary adoption, physician dispensing rules, and PBM regulations across all 50 states and the District of Columbia as of January 1, 2026. Roughly half of U.S. states still prohibit insurers from directing injured workers to contracted pharmacy networks, creating a structural cost driver that must be isolated and quantified in loss cost filings before it contaminates the overall medical severity indication.

The Physician Dispensing Cost Structure

Workers' compensation is the only major private insurance system in which many states bar mandatory contracted pharmacy networks. Medicare, Medicaid, and commercial health plans all route prescriptions through negotiated PBM contracts that enforce formulary compliance, generic substitution, and therapeutic equivalence. In approximately half of WC jurisdictions, the injured worker can obtain medications from any dispensing source, including the treating physician's office, with no copay, no deductible, and no financial incentive to seek a lower-cost alternative.

This structural gap creates the conditions for opportunistic pricing. An April 2026 Claims Journal analysis by Enlyte VP Brian Allen documented the specific markups that result. Beyond the 16,400% diclofenac sodium example, the lidocaine product Lidopro carries a $34.20 per-unit cost through non-retail dispensers versus $4.79 through PBM contracts, a 614% markup. Private-label topical products show even more extreme pricing: Zylotrol bills at $2,425 per prescription against a $9 OTC equivalent, Trubrexa at $2,400 versus $27, and Lidocan at $2,890 versus $90.

From tracking this market segment over several years, the pattern is consistent: physician-dispensed pricing runs two to three times the contracted PBM rate on average across all drug classes. But the distribution is heavily right-skewed. The mean masks a fat tail of private-label topical analgesics where markups exceed 10,000%. These outlier products represent just 3.9% of total WC prescription volume, according to Enlyte's 2025 Workers' Comp Drug Utilization Report, but account for 20.5% of total WC drug spend. That concentration creates a severity distortion that standard trend regressions will overweight unless the actuary explicitly decomposes the pharmacy component.

Why Standard Medical Trend Models Miss This

Most WC pricing actuaries select a single medical severity trend factor, applied uniformly to the medical loss cost component. NCCI's annual State of the Line report tracks overall medical severity growth, which reached approximately 4% for calendar year 2025, alongside a 2% decline in lost-time claim frequency. The NCCI prescription drug inflation analysis shows that changes in prescribing patterns, particularly the sustained decline in opioid utilization from over 50% of claims in 2012 to under 30% by 2021, have offset price increases and helped moderate overall medical costs.

But that moderation is an average across dispensing channels. The pharmacy component of medical severity contains at least three sub-populations with fundamentally different cost dynamics:

Retail pharmacy via PBM: Subject to formulary controls, generic substitution requirements, and negotiated fee schedules. Trend selection here should reflect CPI-Rx or PBM contract escalation rates.
Physician-dispensed medications: Subject to state fee schedules where they exist, but otherwise priced at AWP (average wholesale price) or higher. Trend selection must account for both price-per-unit inflation and the migration of volume toward higher-cost private-label products.
Mail-order and specialty pharmacy: Typically managed through PBM specialty networks with prior authorization requirements. Trend reflects specialty drug cost inflation, currently running above general pharmacy inflation due to biologic and gene therapy approvals.

Blending these three channels into a single medical trend factor masks divergent cost drivers. A state where 40% of pharmacy spend flows through physician dispensing channels will show a materially different pharmacy severity trajectory than a state where physician dispensing is restricted to fee-schedule parity, even if the underlying injury mix and treatment protocols are identical.

Three-Step Framework: Isolating Physician Dispensing in Loss Cost Indications

The following framework translates the physician dispensing problem into a quantifiable adjustment for rate filings. Patterns we have observed across multiple state filings suggest that carriers who decompose the medical trend produce more defensible indications and face fewer regulatory questions about trend selection.

Step 1: Decompose the Medical Severity Trend

Split the medical severity trend into three sub-components: hospital/facility, professional services, and pharmacy. Use NCCI split data where available, or the carrier's own payment-level triangles coded by service category. Within the pharmacy component, stratify claims by dispensing channel (retail pharmacy via PBM, physician-dispensed, and mail-order) and calculate a dispensing channel severity differential.

The differential is computed by comparing average cost per prescription across channels, controlling for drug type and injury classification. For example, select claims with the same ICD-10 diagnosis code and the same NDC drug code, then compare the per-prescription cost when dispensed at retail versus when dispensed by the treating physician. This apples-to-apples comparison strips out case-mix effects and isolates the pure channel markup.

Enlyte's data shows that physician-dispensed pricing consistently runs two to three times the PBM-contracted rate on average. For topical analgesics specifically, the differential is far wider. The actuary should compute this differential separately for each major therapeutic class (topicals, opioids, non-opioid pain management, anti-inflammatories) because the markup varies substantially by drug category.

Step 2: Calculate the Excess Dispensing Factor

The excess dispensing factor (EDF) quantifies the pure markup effect, net of any legitimate clinical differences between dispensing channels. The EDF is defined as the ratio of physician-dispensed average cost to retail-pharmacy average cost for therapeutically equivalent medications:

EDF = (Avg Cost per Rx, Physician-Dispensed) / (Avg Cost per Rx, Retail PBM)

Apply the EDF as a multiplicative adjustment to the pharmacy severity trend. If the overall pharmacy trend is 5% annually, but the physician-dispensed channel is growing at 12% while the PBM channel grows at 3%, and physician dispensing accounts for 35% of the state's pharmacy spend, the blended pharmacy trend should reflect the weighted channel mix rather than a single undifferentiated rate.

The EDF also serves as a diagnostic. An EDF above 2.0 for a state signals that the physician dispensing channel is meaningfully inflating the pharmacy component. An EDF above 5.0 on specific therapeutic classes (topicals, compounded medications) indicates the fat-tail problem described above, where a small volume of prescriptions drives a disproportionate share of pharmacy spend.

Step 3: Model Prospective Severity Reductions for Formulary States

For states where formulary adoption or physician dispensing restrictions are pending or recently enacted, the actuary must model the prospective severity reduction. WCRI research estimates a 15-20% reduction in prescription drug costs when physician dispensing is effectively restricted. Use this range as a Bayesian prior, then credibility-weight it with the carrier's own pre/post data where available.

The credibility weighting follows standard actuarial methodology. If a state enacted physician dispensing restrictions two years ago and the carrier has 18 months of post-reform claims data with full development, the carrier-specific experience deserves substantial credibility weight. If the restriction is prospective (effective next year), the WCRI 15-20% estimate carries more weight, tempered by the specific regulatory details of the state's implementation.

This three-layer decomposition prevents the physician dispensing effect from bleeding into the overall medical trend and distorting the rate indication. It also creates an audit trail that regulators can follow: each component of the medical severity trend has a documented data source, a defined methodology, and a clearly stated assumption about channel-specific cost dynamics.

State Regulatory Variation and Credibility Implications

The WCRI 2026 National Inventory (authored by Karen Rothkin and Terence Cawley) catalogues the full spectrum of state approaches to physician dispensing regulation. The variation is substantial, and it directly affects how pricing actuaries should credibility-weight loss cost indications across their multi-state books.

States with fee-schedule parity or effective restrictions: California, Arizona, Georgia, South Carolina, Tennessee, Pennsylvania, and Illinois all impose fee-schedule parity requirements that limit physician dispensing reimbursement to the same rate as retail pharmacies. These states show measurably lower pharmacy severity relative to permissive states. The actuary filing in a fee-schedule parity state should reflect the lower pharmacy component in both historical experience and prospective trend, and should not apply the EDF adjustment since the regulatory structure already controls the markup.

States with minimal or no physician dispensing: New York, New Jersey, Massachusetts, Texas, Utah, Wyoming, Montana, and Ohio (state fund) either prohibit or severely limit physician dispensing. These states provide the cleanest baseline for pharmacy severity trends because the dispensing channel distortion is largely absent from the data.

Permissive states without fee-schedule controls: In these jurisdictions, physician dispensing can account for 20-45% of total WC drug spending, per the WorkersCompensation.com analysis. The EDF adjustment is most critical here, and the actuary should consider state-specific adjustment factors rather than applying a single national pharmacy trend.

For carriers writing across multiple NCCI jurisdictions, the credibility challenge is real. A state with 200 lost-time claims per year does not produce enough pharmacy-specific volume to estimate a stable dispensing channel differential. The solution is to group states by regulatory regime (restrictive, fee-schedule parity, permissive) and estimate channel differentials at the group level, then apply state-specific adjustments only where credible volume supports them.

Loss Development and Reserving Implications

The physician dispensing effect does not just inflate point-in-time severity; it also extends the claim development tail. A 2014 peer-reviewed study of Illinois WC claims (White et al., published in the Journal of Occupational and Environmental Medicine) analyzed 6,824 indemnity claims and found that physician-dispensed claims showed significantly higher pharmaceutical costs, more prescriptions per claim, and extended disability durations compared to matched controls. When opioids were dispensed by the physician specifically, medical costs were 78% higher, indemnity costs were 57% higher, and days off work were 85% higher after adjusting for age, gender, attorney involvement, and injury complexity.

These findings imply longer development tails for the pharmacy and indemnity components of claims involving physician-dispensed drugs. The pricing actuary selecting loss development factors should consider whether the claim population underlying the development triangle has a changing mix of physician-dispensed versus retail-pharmacy claims. If physician dispensing volume is growing as a share of the book, the historical triangle may understate ultimate severity because the more recent, less-developed accident years carry a higher proportion of the longer-tailed physician-dispensed claims.

Similarly, the reserving actuary selecting IBNR tail factors should account for the possibility that physician-dispensed claims develop more slowly. A carrier that recently entered a permissive state, or one that has seen a shift in its provider network toward offices that dispense, may need to lengthen its development assumptions in the pharmacy and medical components.

Florida: A Natural Experiment for Pricing Actuaries

On February 25, 2026, Florida's First District Court of Appeal ruled in Publix Super Markets, Inc. v. Department of Financial Services that physicians may not dispense medications directly to WC patients under Chapter 440, holding that the statute's "absolute choice" provision applies to pharmacies and pharmacists, not to physician offices acting as dispensing sites.

The Florida Insurance Council and the American Property Casualty Insurance Association estimated in their amicus brief that ending physician dispensing will save WC insurers approximately $43 million over five years. WCRI research cited in the proceedings showed that commonly dispensed drugs carry substantial markups in physician channels: Vicodin at $1.41 per pill physician-dispensed versus $0.52 at retail pharmacy.

For pricing actuaries, Florida now provides a controlled natural experiment. Carriers with multi-year Florida WC books can measure the pre-ruling pharmacy severity trend, then compare it against post-ruling experience as physician dispensing volumes decline. The difference quantifies the actual severity reduction attributable to the dispensing channel, providing carrier-specific data to calibrate the WCRI 15-20% estimate.

The framework for using this data in a rate filing is straightforward: measure the pharmacy severity per claim in the 24 months before the ruling, measure it in the 12-24 months after the ruling takes full effect, and express the difference as a severity adjustment factor. Apply standard credibility weighting based on claim volume. If the carrier's post-ruling experience shows a 17% pharmacy severity reduction with 500+ claims, that observation deserves high credibility weight and can serve as the basis for adjusting pharmacy trend selections in other states considering similar restrictions.

Why This Matters for WC Pricing Actuaries

The zero cost-sharing structure of workers' compensation, no copays, no deductibles, no coinsurance, eliminates claimant price sensitivity entirely. This distinguishes the WC pharmacy trend driver from analogous problems in health plan pricing, where member cost-sharing provides at least partial demand-side discipline. In WC, the only constraint on pharmacy costs is the regulatory structure: fee schedules, formulary requirements, dispensing channel restrictions, and PBM contract terms.

When those constraints are absent, the market produces $2,425 prescriptions for products with $9 OTC equivalents. The pricing actuary who folds this distortion into a blended medical trend factor will either over-indicate (if projecting the inflated trend forward without adjustment) or under-indicate (if using a national average that dilutes the state-specific physician dispensing effect). Neither outcome produces an adequate rate.

The three-step framework described above, decompose, calculate the EDF, and model prospective adjustments, provides a structured approach to a problem that will intensify before it resolves. As of January 2026, only about half of states have effective controls on physician dispensing. Legislative and judicial action is moving toward restriction (Florida's ruling, the Massachusetts PBM licensing regulation requiring WC carrier disclosure, and ongoing WCRI advocacy), but the timeline is measured in years, not quarters. Until then, the distortion sits in the data, and the pricing actuary must decide how to handle it.