Medicaid Drug Payment Spread

This page compares Medicaid pharmacy reimbursements against what pharmacies paid to acquire the drug, using State Drug Utilization Data State Drug Utilization Data (SDUD) is filed with CMS by every state Medicaid agency under the Medicaid Drug Rebate Program and published quarterly at data.medicaid.gov. Each row is a state × NDC × quarter: units reimbursed, prescriptions, and total amount reimbursed. This page uses the 2025 annual file with fee-for-service and managed-care utilization combined. Amounts are pre-rebate, so they measure what the pharmacy was paid, while federal and supplemental rebates settle separately and never enter the file. Suppression applies at ten or fewer claims, which thins less frequently prescribed drugs in smaller states. and Average Drug Acquisition Cost National Average Drug Acquisition Cost (NADAC) is CMS’s survey of what retail community pharmacies pay wholesalers, published weekly as a per-unit rate by NDC. It is a national benchmark. The survey covers retail pharmacies, so it prices a 340B or specialty-pharmacy channel only by approximation. Rates move through the year; each drug code is priced at the rate in force at the end of the quarter it was billed in. datasets. Nationally, the gap has barely moved over eleven years, but just over half the states have shifted: 27 of 51 show a sustained change in level larger than their own quarter-to-quarter variation explains. The median change over this period for those 27 states is 6.8 percentage points.

By Kevin Contino · September 2026 · SDUD, NADAC and MDRP data from data.medicaid.gov, 2015–2025.

State and national reimbursement trends

The national spread increased slightly from — over cost in 2015 Q1 to — in 2025 Q4. NADAC’s data coverage drops from 86% of reimbursed dollars to 75% over the same period, so the basket of drugs priced within that total is somewhat diminished. Holding the drug set fixed removes the apparent decline entirely.

Quarterly spread, all states pooled. The fixed set holds the drug list constant across all 44 quarters. The comparator takes every drug matched to a NADAC rate in each quarter, so its membership changes as it goes. A state overlay uses that state’s own fixed set, which makes its trend readable but not its level comparable to another state’s, because a small basket is mostly long-lived generics carrying the widest percentage margins.

States by dollar-weighted spread (2025)

Includes all drugs a state reimbursed in 2025, not only the 24 charted below. High-cost specialty products such as GLP-1 agonists and injectable biologics carry most of the dollars and are reimbursed close to acquisition cost, and sometimes under it, which pulls these totals well below the median claim line in the headline. Oregon at −7% turns on a handful of high-volume National Drug Codes (NDCs) paid far under cost, while its typical drug code is reimbursed at +26% of acquisition cost.

This measure covers a state’s entire formulary and runs from a minimum of −7% in Oregon to a maximum of +37% in South Dakota.

Spread by state and by drug

This grid shows the 24 highest-reimbursement drugs nationally by state. Color runs from blue where a state reimbursed below what pharmacies paid, through grey at acquisition cost, to orange where it reimbursed above. Cells drawn as a small inset square carried under $500K in reimbursement, and hatched cells are gaps in the data. Hover over any cell for payment and comparison details.

Across all 538,969 state × drug-code lines the median spread is +93.5%. This runs well above every dollar-weighted figure here because most lines are cheap generics carrying a wide percentage margin on a few cents.

Geography

Select one drug at a time or the dollar-weighted total for the top 24 by reimbursement. The scale matches the one above.

Method

SDUD 2025 annual file (fee-for-service plus managed care utilization combined), aggregated to state × NDC, joined to the NADAC rate used at the end of the quarter each line was billed in. Drug names are the FDA product name from the MDRP product-list file. Package-size and dose-strength variants of the same brand are summed together. Spread = reimbursed dollars − (units × NADAC rate); percent spread is that gap over the acquisition-cost total.

• Unit basis checked, not assumed. SDUD “Units Reimbursed” and NADAC “Pricing Unit” are both NDC-billing-unit based (each / mL / gram) and should align per NDC by construction. Where they do not, the ratio of reimbursement to acquisition cost runs to absurd values. The worst case found was 2,402,312×. Drug codes whose national median ratio falls outside 0.1–50 are excluded: 750 of 48,232 codes, 0.09% of dollars. A second screen catches a drug code that prices normally nationally but is miscounted in one state. Where a state’s median ratio for a code sits more than tenfold from that code’s national median, the pair is dropped: 15,111 state-code pairs, 1.7% of dollars. Most are IV solutions and injectables, where one state reports bags and another millilitres.
• 92.4% of reimbursed units matched a NADAC rate by total volume. Unmatched NDCs (compounds, some brand-only products, discontinued drugs) are dropped from the analysis.
• Half of SDUD dollars are national-total rows. CMS publishes rolled-up national totals coded XX in the same file. Because national totals include figures suppressed at the state level, they are 7.7% higher than the sum of visible state totals in 2025.
• A negative spread is not a finding about a state’s fee schedule on its own. The deepest cells here are single NDCs where reimbursement fell far under the national acquisition benchmark — a 340B-acquired product, a supplemental-rebate arrangement, or a data-entry artifact all produce the same shape in this file.

Prior and parallel work. 46brooklyn Research publishes a Medicaid Heat Map built from the same two files at the same grain — state Medicaid reimbursement against NADAC, by drug and quarter — which they term markup rather than spread. My analysis was built independently and reaches the same two observations they report: that expensive generics do not deflate as fast as their acquisition cost, and that states pay very differently for the same drug at the same time. The measures here that go beyond a cross-section — the fixed-drug-set index, the structural-break test, and the per-state pass-through coefficient — are not part of that map as published.

Known reporting gaps

SDUD is filed by state, and sometimes contains filing discontinuities such as dollars missing against normal units or units restated against normal dollars. Uncorrected, these show up as interruptions in reimbursement per unit. Where a quarter’s reported dollars are incomplete, it is excluded from the figures here. Quarters in which the apparent interruption is a shift in prescribing are left as filed. These quarters are identified by comparing each state’s reimbursement per unit against the state’s trend, then testing whether individual drug codes moved with the aggregate. Of the 2025 figures on this page, only Wyoming is affected.

Source: Medicaid.gov SDUD, NADAC, and MDRP Drug Products files, downloaded 2026-09-19. What each drug is prescribed for is condensed from the INDICATIONS AND USAGE section of its FDA-approved label, read from the openFDA drug label API on 2026-09-10. State outlines are the US Census cartographic boundaries as published in us-atlas, projected offline.