Do You Know Which GPO You're Actually Buying Through? These Two Companies Didn't Agree Either — and Now It's in Federal Court.

Two lawsuits, one arbitration, three unsealed contracts, and a status report due to a federal judge by August 17. The whole story, sourced to the docket, in ten minutes.

Start with a question that sounds simple: who was Metz Culinary Management's non-healthcare GPO sponsor in early 2024?

Alliant Purchasing ⚖️ Active Litigation says the answer was Pandion Optimization Alliance ⚖️ Active Litigation — and cites Premier, Inc. ⚖️ Active Litigation's own published roster, which listed Pandion. Premier says the answer was OMNIA Partners ⚖️ Active Litigation, which it asserts had become the sponsor in September 2023 through an assignment, months before Pandion even acquired the sponsor entity in question.

Two of the most sophisticated organizations in group purchasing, looking at the same network, could not agree on who held one food-service company's sponsorship. That dispute — over a fact most people in this industry would assume is a lookup, not a lawsuit — now sits inside two federal cases and a parallel arbitration, with the parties due to tell a federal court by August 17 whether their settlement talks will end all three at once.

If organizations at the center of the network can litigate over who a customer's sponsor is, it is worth asking how confident anyone downstream should be. That is why this case matters beyond the parties — and why GPO Intel has collected all 47 public court documents from both cases — complaints, answers, the three governing agreements, the transfer policy, the briefing, the transcripts — into a free, searchable library. This article is the map.

Everything below comes from public federal court records and public SEC filings. The claims in the complaints are allegations; the defendants deny wrongdoing; no court has ruled on the merits of anything.

The $800 million sale that didn't quite say goodbye

In June 2023, Premier announced the sale of its non-healthcare group purchasing business to OMNIA Partners for approximately $800 million in cash — expressly an estimate, subject to post-closing adjustment. (Alliant's pleading, citing Premier's SEC filings, puts the figure at approximately $680 million; the defendants deny that amount in their court filings. The numbers coexist because announced deal prices true up after closing.)

The agreement itself, publicly filed with the SEC since the week it was signed, shows how the price was built: 14 times trailing adjusted net administrative-fee revenue. That is the only arm's-length valuation multiple ever made public for a GPO business, in a market with essentially no comparable transactions.

The price build. Base purchase price of $689,759,330 as stated in the agreement, computed from adjusted net administrative-fee revenue of $49,268,524 × 14. The stated figure differs from the arithmetic product by $6, and the revenue base reflects defined exclusions. What was included, what was excluded, and how the price adjusted after closing is in our companion analysis.

The mechanics matter as much as the number. Premier gathered its non-healthcare contracts — members, sponsors, channel partners — into a newly created entity, Non-Healthcare Holdings, LLC (the entity holding Premier's former non-healthcare GPO contracts, acquired by OMNIA Partners in July 2023), and OMNIA bought the entity. The sale closed in July 2023.

Premier's own announcement-day language described what followed: the two companies would have aligned growth incentives and would economically benefit from non-healthcare members' continued purchasing through Premier's supplier contracts. The sale did not end Premier's economics in non-healthcare. Members kept buying on Premier's contract portfolio, accessed through NHH, with the administrative-fee stream divided among Premier, NHH, and sponsors under formulas the public versions redact.

The divestiture therefore left: one contract portfolio, two companies with economics riding on it, and a network of sponsors who needed a new agreement with the buyer to keep growing the business they had built.

According to the complaint, OMNIA's chief executive told Alliant it had to sign a new agreement with NHH to keep growing its non-healthcare business — and Alliant, "with no alternative" in the pleading's words, signed in January 2024. The defendants deny Alliant had to sign. Either way: that agreement is now one of the most consequential documents in the industry, because litigation made it public.

Days later, a customer went to market

Days after Alliant signed, Metz — a food-service operator with accounts on both the healthcare and non-healthcare sides — put its GPO relationship out for RFP. It chose Alliant in February 2024. Alliant submitted a sponsorship transfer request in March.

Premier approved the healthcare transfer. The non-healthcare transfer was refused in writing: moving Metz's non-healthcare accounts "from Pandion to Alliant" was not permitted under the Recruitment and Retention Policy as applied to non-healthcare entities.

Metz required all of its accounts under a single sponsor. The transfer never completed. The complaint values the lost relationship at approximately $70 million in additional revenue to Alliant — a figure that, at typical administrative-fee rates, implies more than $2 billion in underlying purchasing volume, so it is best read as the purchasing volume flowing through the relationship rather than fee revenue. Either way, commercially material.

Underneath it sits the sponsorship-identity dispute from the top of this article: the refusal named Pandion as the sponsor Metz was moving from. Whether Pandion even held that sponsorship is contested in the filings.

In January 2025, Alliant sued OMNIA and Premier — declaratory judgment, breach of contract, tortious interference, breach of fiduciary duty, civil conspiracy, unfair trade practices under Kentucky law, with treble and punitive damages sought — and filed an arbitration demand the same day. A second lawsuit followed in November 2025 after new transfer policies were issued mid-litigation. The defendants deny wrongdoing across the board.

The contracts became public

Who gets to read GPO paper depends on the layer. Members see their own agreement — and in the non-acute world, where administrative-fee shareback is often absent, many GPOs post their participation agreements publicly. (GPO Intel is assembling a library of them.) The sponsor layer is different. The agreements between a national network and the regional GPOs and sponsors that route members to it — where the fee splits, the primacy commitments, and the exit terms live — are negotiated, confidential, and never posted. A career on the GPO side means signing hundreds of member agreements in the acute world, thousands in non-acute, without ever reading another network's sponsor paper. It surfaces through essentially one channel: litigation. Litigants almost never let it happen.

Here, after a sealing dispute that ran through spring 2025, OMNIA and Premier agreed to withdraw their sealing motions, and in June 2025 Premier filed three agreements on the public docket with limited redactions: the 2015 Premier sponsor participation agreement, the 2024 NHH sponsor participation agreement, and the channel partnership agreement that structured the divestiture. The transfer policy was already public as a complaint exhibit. For the first time in memory, the industry can read a national network's operative sponsor agreement — and its predecessor — side by side.

Five findings, each a door into the full analysis:

1. The 2024 agreement contains a clause that lets a sponsor compete with its own network. Where a Premier program contract is not competitive on pricing relative to volume — as determined by the sponsor itself, no benchmark, no arbiter — the sponsor may source and sign its own supplier deal, and NHH commits that the supplier won't owe Premier administrative fees on those purchases. In a business whose economics run on admin-fee capture, that clause zeroes the capture out on the redirected trade. Conditions apply, and no court has tested it. The 2015 agreement has no equivalent in its public text — this arrived with the 2024 template.

2. The non-compete is narrower than assumed. The 2024 agreement bars making another non-healthcare national GPO "primary or preferred" without written consent. It says nothing about ordinary multi-GPO relationships, nothing about category-level arrangements. Whether that leaves a sponsor a real path to a rival — or a paper right the network's other rules can close — is a live question the record doesn't yet resolve.

3. The exit door closed between 2015 and 2024. The 2015 template let either party walk on twelve months' notice. The 2024 template states it is intended to be noncancelable except for cause or change of control — and on a for-cause termination by the GPO, the sponsor owes liquidated damages computed by formula: trailing average fees times years remaining. Any sponsor on this template can model its own exposure with numbers already in its files.

4. The transfer rulebook is public — and being rewritten mid-lawsuit. The Recruitment and Retention Policy that decided the Metz refusal is in the record verbatim. Industry participants can judge for themselves whether the refusal's reading of it holds. The second lawsuit concerns the new policies issued mid-litigation — Alliant alleges the replacement contains provisions "aimed directly at Alliant"; the defendants say they had every right to issue it. The replacements are sealed. Every sponsor in the network now operates under rules rewritten during a lawsuit about the rules.

5. What stayed sealed maps where the money is. Across all three agreements, the redactions follow one pattern: fee terms and term lengths. Everything about how the money works is public except the numbers. What the parties kept protecting tells you what they consider most valuable.

What August 17 decides

The court-ordered mediation on April 15, 2026 was unsuccessful — the parties' own word. Direct talks resumed; in May the court stayed both cases at the parties' joint request. By Monday, August 17, they must report on the status of those discussions, which — in their own representation to the court — may resolve both lawsuits and the arbitration together.

If the talks succeeded: three proceedings end in one agreement, and the settlement terms, unlike the contracts already filed, would not automatically become public. If they're continuing: the holding pattern holds. If they failed: the dismissal motions return, the injunction track resumes, and the question of court versus arbitration comes back with it.

Whatever the report says, the documents remain public.

What the record can't answer

Was the 2015-era exit right ever actually exercised? How do sponsors in practice handle a customer with accounts on both sides of the healthcare line? What does a transfer request look like from inside the room?

The record supplies the paper. People who operated under these agreements supply the meaning — and that is what GPO Intel's Expert Network exists to capture. If you have run a sponsor, managed a category, carried a bag for a manufacturer, or sat on the health-system side of these relationships, we are asking short questions only people like you can answer — asynchronous, screened before publishing, attributed on your terms (named, title-only, or anonymous), never about your current employer.

Read the full 47-document library — free, with Ask GPO Intel able to answer questions against the full set.

Join the Expert Network — short questions, answered on your terms.

Get the August 18 analysis the morning after the report lands.

These are public federal court records. Allegations are unproven. All defendants have denied wrongdoing. GPO Intel provides these documents for research purposes and takes no position on the merits of either case.