The Medicaid Drug Rebate Program (MDRP) index we track carries 1,954,970 NDC-quarter rebate records spanning 2014 through early 2026 — and the count has grown every single year, from 74,049 records in 2014 to 184,027 in 2024 alone. That steady climb is not inflation in drug prices; it is the mechanical signature of a law that forces virtually every drug sold in the United States, generic or brand, through one rebate ledger. Across the full history, 93,184 distinct National Drug Codes from 1,464 manufacturers have been obligated to report a rebate — a footprint that covers most of the retail pharmacy shelf, not a narrow slice of it.
The mechanism is simple but has outsized consequences: under Section 1927 of the Social Security Act, any manufacturer that wants Medicaid to cover its drugs at all must sign a national rebate agreement with CMS and report pricing data every quarter, forever, on every covered outpatient drug. There is no opt-out. Generic and "noninnovator multi-source" drugs alone account for 1,521,530 of the 1,954,970 public MDRP records (about 78%) — more than three times the volume of brand-name innovator products (206,370 single-source, 227,070 innovator multi-source). Because Medicaid is the payer of last resort for tens of millions of low-income and disabled Americans, manufacturers cannot simply decline to participate; losing Medicaid coverage effectively means losing access to a state's entire Medicaid formulary.
Why "best price" reaches past Medicaid
The rebate formula itself is what creates spillover far outside the Medicaid program. For most brand drugs, the statutory rebate is the larger of (a) 23.1% of Average Manufacturer Price (AMP) or (b) AMP minus "best price" — the single lowest price the manufacturer offered any commercial purchaser that quarter, with narrow carve-outs (42 CFR § 447.505).[2] That "any purchaser" clause means a steep discount cut for one hospital system, GPO, or value-based contract can reset the Medicaid rebate baseline nationwide. Manufacturers respond by capping how deep any single commercial discount can go, because a rebate concession offered to win one health plan can silently raise the price Medicaid rebate the manufacturer owes on every other unit sold that quarter. This is the textbook "best price chilling effect" that pharmacy-pricing counsel and CMS itself have debated for two decades, most recently in CMS's 2024 MDRP final rule, which considered (and ultimately declined to finalize) a "stacking" provision that would have compounded 340B and commercial discounts into the best-price calculation.[3][4]
What the index does not show
The cms-mdrp fields are pricing-metadata only — Labeler, NDC, Drug Category, Year/Quarter — with no dollar rebate amounts, no AMP, and no best-price figures, because CMS treats those as manufacturer trade secrets and does not publish them in the public rebate agreement file. So we can measure the scope of the program precisely (which drugs, which manufacturers, how many reporting cycles) but not the size of the checks manufacturers cut; independent estimates from MACPAC and state Medicaid agencies put net Medicaid drug spending after rebates in the tens of billions annually, but that figure comes from budget data, not this index, and readers should treat it as directional. Line-extension drugs (reformulated brand products subject to an additional anti-evergreening penalty rebate) are a small but telling slice: 22,298 records, versus 1,285,955 flagged "not a line extension" — consistent with CMS's narrow, litigated definition of what counts.[5]
Why this matters: as CMS phases in its 2024 final rule and continues to weigh 340B/Medicaid duplicate-discount fixes, the rebate ledger's reach — nearly two million records, 93,000+ NDCs, 1,464 manufacturers — means any tightening of the best-price definition will not stay confined to Medicaid budgets. It will move through the same pricing chokepoint that already links generic-heavy dispensing, hospital 340B contracts, and commercial rebate negotiations into one interlocking system.