There exists no public system that maps a drug’s price across WAC, NADAC
, and ASP in a unified, unit-normalized, continuously updated format. This is a remarkable fact about a market that accounts for over $400 billion in annual spending and is the subject of intense regulatory, legislative, and public attention. The United States has three major federal drug pricing datasets, published by the same agency — CMS — using different unit conventions, different update cadences, and different coverage scopes, with no crosswalk connecting them.
The healthcare data interoperability movement has spent billions of dollars and two decades building standards for clinical data exchange — HL7 FHIR, electronic health records, clinical data registries, quality measure reporting. The pharmaceutical pricing ecosystem, which operates in parallel, has received almost none of this standardization effort. No equivalent of FHIR exists for drug pricing data. No interoperability standard maps NDCs across pricing databases to a common schema. The infrastructure that would make cross-dataset comparison routine — rather than a bespoke analytical project for every organization that attempts it — has never been built.
The consequences are distributed and cumulative. Every payer that compares WAC to NADAC to evaluate pharmacy reimbursement adequacy builds its own mapping. Every analytics vendor that cross-references ASP and WAC for physician-administered drugs constructs proprietary unit crosswalks. Every state Medicaid program that reconciles NADAC-based reimbursement with AMP-based rebate calculations maintains internal reconciliation processes. Every researcher who compares drug prices across datasets in an academic study creates ad hoc harmonization methods, documents them in an appendix, and moves on.
The Daily Remedy email exchange illustrates the problem at the individual product level. The team identified a discrepancy in budesonide pricing across datasets during a routine audit and posed the question to an expert: what is the right methodology for reconciling unit differences? The expert’s response — that course-of-treatment rollups based on NCPDP billing units are the most commonly used approach, that the methodology requires substantial assumptions, and that an industry gold standard exists but is proprietary and imperfectly documented — describes the state of the art. The state of the art is: there is no standard, every organization figures it out independently, and the best available reference is a private vendor whose methods are undisclosed.
The cost of this fragmentation is impossible to calculate precisely but certainly large. Consider the labor: a data engineer at a health plan spending two weeks building a WAC-to-NADAC crosswalk for a subset of high-cost drugs. A pharmacoeconomics team at a manufacturer spending a month validating unit mappings for their product portfolio across federal datasets. An academic researcher spending hours on a crosswalk that will be used once, published in a supplement, and never reused. Multiply by the number of organizations that do this work and the frequency with which they redo it when NDCs change, formulations are updated, or dataset conventions shift. The aggregate investment is enormous, the output is duplicative, and the work products are proprietary.
The cost of errors is harder to observe but potentially larger. When an organization builds its own crosswalk and gets a mapping wrong — a unit conversion error, an incorrect NDC match, a formulation-level mismatch — the error propagates into whatever analysis the crosswalk supports. Reimbursement rates set against the wrong baseline. Cost-effectiveness models that compare drugs on different unit scales. Budget impact analyses that overstate or understate spending on a product because the pricing data was not properly harmonized. These errors are silent. They do not generate error messages or alerts. They sit inside spreadsheets and models, producing conclusions that are precisely wrong.
The technical requirements for a public reconciliation system are well understood. At minimum, it would require a master NDC database that maps every active NDC to each federal pricing dataset in which it appears, with explicit unit conventions, conversion factors, and effective dates. It would need to handle multi-source drugs where multiple manufacturers produce the same molecule at different prices. It would need to accommodate the different update cadences — NADAC weekly, ASP quarterly, WAC continuously — and provide historical data for time-series analysis. It would need documentation sufficient for any user to understand the unit assumptions, conversion methods, and known limitations.
CMS is the natural owner of this infrastructure, since it publishes most of the relevant datasets. The FDA could contribute by enriching the NDC directory with unit-level metadata that currently exists only in pricing compendia. NCPDP could extend its billing unit standard to include explicit crosswalk mappings to CMS unit types. None of these organizations has prioritized this work, because unit reconciliation is not a politically salient issue and does not compete for attention with drug price negotiation, insurance coverage, or clinical safety — the topics that generate legislative and media interest.
The vacuum has created a commercial opportunity. Private data vendors sell harmonized drug pricing databases at subscription prices that reflect the difficulty of the work. These products serve institutional clients — large payers, PBMs, consulting firms, pharmaceutical companies — but are inaccessible to smaller organizations, academic researchers, patient advocacy groups, and journalists who lack the budget. The information asymmetry that results from proprietary reconciliation is, in effect, a barrier to entry for drug pricing analysis. The organizations with the deepest pockets have the most harmonized data. Everyone else works with fragments.
A publicly maintained, freely accessible drug pricing reconciliation system would not solve every problem in pharmaceutical pricing transparency. It would not disclose net prices, resolve clinical non-equivalence between formulations, or eliminate the gross-to-net opacity that obscures actual transaction economics. But it would eliminate the foundational problem of comparing list prices across datasets that measure the same thing in different units — a problem that every user of drug pricing data encounters and that no authoritative source has addressed.
The infrastructure is missing not because it is technically difficult — it is not, relative to the clinical data interoperability challenges that the industry has already tackled — but because nobody has claimed responsibility for building it. CMS publishes the data. The industry uses the data. The gap between publication and use is filled with bespoke, proprietary, duplicative work that could be replaced by a shared public resource. The question is not whether the infrastructure is needed. The question is who will build it, and the answer, so far, is nobody.
