The drugs that cost the most are the drugs we know the least about — at least when it comes to pricing. Specialty pharmaceuticals, which account for a disproportionate and growing share of total drug spending in the United States, are systematically underrepresented in the pricing benchmarks that payers, regulators, researchers, and investors rely on to understand drug costs. The blind spot is not accidental. It is a structural consequence of how pricing data is collected and how specialty drugs are distributed.
NADAC
, the most empirically grounded pharmacy reimbursement benchmark, surveys non-specialty retail community pharmacies. Its methodology excludes specialty pharmacies by design. For drugs dispensed primarily or exclusively through specialty channels — which describes most biologics, many oncology therapies, and a growing list of orphan drugs — NADAC provides no acquisition cost data. A payer using NADAC to benchmark reimbursement for a specialty drug is using a ruler that does not measure the object in question.
The exclusion is methodologically defensible. Specialty pharmacies operate under different economic conditions than retail community pharmacies. Their acquisition costs, distribution arrangements, and margin structures differ substantially. Including specialty pharmacy invoice prices in a survey designed for retail pharmacies would contaminate the average without improving its accuracy for either channel. CMS chose specificity over breadth, and within its defined scope, NADAC performs well. But the scope leaves out the segment where pricing opacity is greatest and spending growth is fastest.
WAC covers specialty drugs — any manufacturer can publish a list price — but WAC for specialty drugs is even less informative than WAC for retail drugs. Specialty drugs are frequently distributed through limited distribution networks, where the manufacturer restricts which wholesalers and pharmacies can stock and dispense the product. The negotiated terms within these networks — acquisition cost, distribution fees, data-sharing requirements — are proprietary and may bear little resemblance to the published WAC. The gross-to-net spread for specialty drugs is often wider than for retail drugs, which means WAC tells you less, not more, about actual transaction economics.
ASP captures specialty drugs that are physician-administered, because the buy-and-bill model for Part B drugs applies regardless of distribution channel. But ASP covers only the Medicare Part B segment. For specialty drugs dispensed through specialty pharmacies to patients with commercial insurance or Medicaid — the majority of specialty drug volume — ASP provides no direct visibility. The AEI analysis of SSR Health’s methodology has noted that SSR’s volume and pricing estimates are weakest precisely in this segment, where dispensing data is fragmented and manufacturer financial disclosures may not disaggregate specialty revenue at the product level.
The practical consequence is that the most expensive drugs in the American formulary are priced using the least reliable benchmarks. A formulary committee evaluating a specialty drug for coverage must rely on WAC (which overstates cost), manufacturer-provided net price estimates (which are self-interested), and SSR or comparable analytics platforms (which acknowledge their own limitations for specialty products). The committee makes a coverage decision based on the best available data, which for specialty drugs is meaningfully worse than the best available data for retail drugs.
This asymmetry matters for drugs that straddle both channels. Budesonide, for instance, exists in formulations dispensed at retail pharmacies and in physician-administered settings. The retail formulations appear in NADAC. The physician-administered formulations appear in ASP. Neither benchmark captures the specialty pharmacy dispensing channel, where some budesonide formulations may also be distributed. Comparing costs across formulations requires bridging databases that cover different channels using different unit conventions — a reconciliation exercise that the existing data infrastructure does not support.
The limited distribution network model exacerbates the problem by design. Manufacturers that restrict distribution to a handful of specialty pharmacies gain control over patient support, adherence monitoring, and distribution economics. They also gain control over data. Acquisition costs within limited distribution networks are negotiated bilaterally and disclosed to no federal pricing database. The manufacturer knows the price. The specialty pharmacy knows the price. The payer negotiates reimbursement based on WAC or some other published benchmark that may not reflect the specialty pharmacy’s actual acquisition cost. The information asymmetry is structural.
Solutions are emerging in fragments. Some commercial payers have begun requiring specialty pharmacies to disclose acquisition costs as a condition of network participation, enabling cost-plus reimbursement models that mirror the NADAC approach applied to retail pharmacies. Some PBMs have developed specialty pharmacy networks with transparent pricing terms. A few state Medicaid programs have extended NADAC-like survey approaches to specialty pharmacies, though the smaller number of dispensing entities and the proprietary nature of distribution arrangements make survey response rates and data quality more variable.
None of these efforts constitutes a systematic solution. The federal government could extend the NADAC survey methodology to specialty pharmacies, but CMS has not indicated plans to do so. The data collection challenges — fewer pharmacies, more heterogeneous distribution arrangements, greater proprietary sensitivity — are real. But the alternative is a pricing landscape in which the fastest-growing cost category in pharmaceutical spending remains the one with the poorest data infrastructure.
The specialty pharmacy blind spot is not merely a data quality issue. It is an information asymmetry that advantages manufacturers and intermediaries at the expense of payers, regulators, and patients. When pricing data is thin, the party that knows the most — the manufacturer — has the strongest negotiating position. Improving data infrastructure for specialty drugs is not just an analytical priority. It is a market design question with direct implications for how healthcare dollars are allocated in the segment where they are growing fastest.
