The Manufacturer-Pays Model for Planning
The manufacturer pays surgical planning model has four shapes. Two hold up under US and Turkish law, and the intuitive one, free to the surgeon, does not.
Key takeaways
The manufacturer-pays surgical planning model answers a real problem: preoperative planning time is not reimbursed, so the surgeon is not a viable payer. But "the manufacturer pays" is not one model, it is four different shapes, and they do not carry the same legal weight. The shape most people reach for first, a neutral third-party planner offered to surgeons at no charge and funded per plan by an implant maker, is the hardest to defend wherever a federal healthcare programme or a public-sector physician is involved. Two shapes hold up: the manufacturer licensing the software as a customer under its own brand, and the hospital or surgical facility paying per case. This article sets out all four and why we structure our own agreements the way we do.
Updated 9 September 2026. An earlier version of this article described a single manufacturer-pays structure with per-plan access funded for surgeons. That framing was too narrow and, in some jurisdictions, wrong. It has been replaced with the analysis below.
Start with why the surgeon cannot be the payer
Follow the money and the first thing you find is a wall. In the United States, the CPT Category III codes covering this kind of work (1030T to 1035T) carry no assigned RVU, which means planning time is not separately billable. Payers have generally treated AI-assisted planning as investigational. So the surgeon who spends thirty minutes on a plan absorbs that time against a schedule that pays for operating, not for planning.
The patient's pocket is not the answer either. Patient-funded personalised implants have been tried at scale and stayed small.
That leaves two candidates: the institution that captures the theatre-efficiency value, and the manufacturer whose implant the plan is built around. Both are legitimate starting points. What matters is the structure.
The four shapes
| Shape | Who pays | Where it stands |
|---|---|---|
| A. The manufacturer builds or buys its own planner | The OEM, for its own product | Universal in shoulder arthroplasty. Every major maker has one. Not a licensing model |
| B. The manufacturer licenses a third-party engine as white-label | The OEM, as a customer: development fee, annual licence, services | Holds up. This is a decade-old, ordinary software supply relationship |
| C. A neutral platform, paid per case by the facility | Hospital or ambulatory surgical centre | The cleanest shape. The payer is the party that captures the value |
| D. A neutral platform, free to the surgeon, funded per plan by an OEM | The OEM, to place the tool in front of a surgeon | The problematic one. See below |
Shapes A, B and C are ordinary commerce. Shape D is where the trouble sits, and it is the one that sounds most attractive when you first sketch the business.
Why the intuitive shape is the hard one
Under the US Anti-Kickback Statute (42 USC 1320a-7b(b)), remuneration offered to induce referrals or the purchase of an item reimbursed by a federal healthcare programme is a criminal matter. Free software of value to a physician can be remuneration. The regulatory safe harbours at 42 CFR 1001.952 are narrow and were not written for this fact pattern.
The relevant test is old and clarifying. The OIG's position distinguishes software with no independent value apart from the service it supports from software that has real value to the physician. A neutral planner that a surgeon would otherwise pay for has independent value. An OEM-branded planner that only plans that OEM's implants largely does not, because it is part of the product offering rather than a standalone benefit.
Two things people assume will save them, and do not:
A fixed fee does not cure it. In United States v. Practice Fusion (D. Vt., 2020, resolved with a $145M deferred prosecution agreement), the sum paid to the software company was a fixed amount, not per click or per prescription. It was charged all the same.
Being the software vendor rather than the manufacturer does not cure it. Practice Fusion is the case worth reading closely for anyone in this business, because the party charged was the software company, not its pharmaceutical customer. The theory was that clinical decision support had been shaped to serve the paying customer. That is the precise risk a planning company takes on when its output could be said to favour whoever funds it.
This is a description of a legal structure, not an allegation about anyone. Other companies operate free tiers lawfully by design: one navigation vendor publishes a no-charge planning service explicitly positioned as marketing material, with no exchange of consideration. Free to the surgeon is not automatically prohibited. It has to be deliberately built, and per-plan funding by a manufacturer on a platform that presents itself as neutral is the version that does not survive contact with counsel.
Outside the United States the analysis changes, and not always in your favour
The AKS bites where a federal healthcare programme is involved. Where none is, it does not apply, and each market has to be assessed on its own terms rather than assumed to be permissive.
Turkey is stricter, not looser, and this is under-discussed in English. The medical device sales, advertising and promotion regulation works from a closed list of what may lawfully be provided to healthcare professionals, and a planning tool given free to a surgeon is not obviously on it. More seriously, a physician at a public or university hospital may be treated as a public official, which moves the question out of administrative law and into the Turkish Penal Code bribery provisions, where sentences run to years rather than fines. Any structure intended for Turkey has to be cleared against that before it is offered, not after.
The UK Bribery Act reaches conduct by associated persons with a long extraterritorial arm, and its corporate offence has no adequate-procedures defence in the newer senior-manager attribution route. Across the EU, the MedTech Europe Code of Ethical Business Practice governs what members may provide to healthcare professionals. None of this makes manufacturer funding impossible. It makes shape D specifically difficult.
What we do instead
Two structures, and we say which one we are in before commercial discussion starts.
White-label licensing, where the manufacturer is the customer. The maker licenses the planning engine and presents it under its own brand, integrated with its own implant library. Payment is a development fee, an annual licence and services, in the ordinary shape of enterprise software supply. There is no pretence of neutrality on the surface: the product is the manufacturer's product, planning the manufacturer's implants, and no one calls a shoulder maker's own planner a kickback. Scope, field of use, branding and IP ownership are contract-defined.
Institution-pays, per case, on the neutral platform. Where Salnus is the brand in front of the surgeon and the engine stays independent of any implant catalogue, the payer is the hospital or surgical facility. The party paying is the party capturing the value, which is also the structure that is easiest to defend and easiest to explain.
The rule underneath both: manufacturer money and platform neutrality do not go in the same product. The engine can stay technically neutral so a second customer can license it and the output stays implant-independent, but a product funded by one maker and presented to surgeons as neutral is trying to be two things at once. Keeping them apart is what makes both shapes work.
We also do not describe our software in terms of implant pull-through or return on a manufacturer's investment, in a contract or in marketing. The plan is computed on the patient's anatomy and remains independent of who paid for the software. Component fitting happens afterwards, from the library the surgeon selected.
What to ask before you sign anything
If you are a manufacturer, a hospital or a software company negotiating one of these, the useful questions are structural:
- Which of the four shapes is this, in writing?
- Is any federal healthcare programme, or any public-sector physician, in scope in any target market?
- Who is the payer, and is the payer the party that captures the value?
- Does the paying party's identity change the software's output in any way that could be documented?
- If the software is free to a clinician, what is the deliberate legal basis for that, and is it written down?
- Is compensation tied to anything volume-based or outcome-of-purchase-based, in either direction?
None of this is legal advice, and every one of these questions belongs with counsel qualified in the relevant jurisdiction. It is the framework we use to work out which conversations to have.
Bottom line
The manufacturer-pays model is real, and the reasoning behind it, that unreimbursed planning time makes the surgeon a poor payer, is sound. What does not follow is the version most people sketch first: a neutral planner given free to surgeons and funded per plan by an implant maker. The shapes that hold are the manufacturer buying software as a customer under its own brand, and the institution paying per case for a neutral platform. Salnus is Research Use Only, so any engagement begins as a pilot and co-development relationship rather than the purchase of a cleared product. If you are working out how your company should structure digital planning, see how Salnus works with manufacturers.
Reviewed by the Salnus biomedical engineering team.