We run the machine, they bring the medicine

Meridian owns the hospital outright and runs everything that is not the practice of medicine. The land, the building, the plant, the power, the water, the estate, the logistics, the automation, the measurement, and all procurement and supply. It funds this alone, from the proceeds it is already holding, and takes no outside equity.

Clinical management is shared. A joint clinical body, Meridian and partner hospitals together, sets the clinical standards, agrees appointments and job plans, and owns audit and mortality review. The consultants are the partners' people. They work here under a shared job plan and they remain employed where they are.

There is no government stake and there are no private investors. Land is bought rather than contributed and the licence route is followed rather than negotiated.

One sentence contains the whole argument. The group runs the machine and the partners bring the medicine, and because the medicine is never bought, nobody is taken from anybody.

The obvious structure is the reverse of this one. A hospital group keeps clinical management, because that is what a hospital is for, and contracts out the building to somebody who does buildings. Almost every private hospital is arranged that way.

It is wrong here for a reason this study has been assembling for twenty pages without naming it. Look at what the work has actually been about. Bed turnaround as real capacity. Walking distance priced in person years. Ring fenced theatre sessions. Safety stock as a safety control. Seven layers of supply verification. Power as a clinical requirement. Smart cabinets and lot level traceability. The queueing arithmetic that decides how big the place is.

None of that is medicine. All of it is the machine the medicine runs on, and it is the entire content of this group's expertise. Meridian does not have a faculty. It has an operating system, and an operating system is the thing that actually transfers between hospitals, because it lives in process, design and measurement rather than in people.

Which corrects something this study said earlier about its own product. The transferable asset is not the clinical standard. Clinical standards are embodied in clinicians and they walk out of the building every evening. The transferable asset is the substrate, and the substrate is exactly what the group should own outright and never share.

So the division is not a compromise about control. It is each party holding the thing it actually has. The group has a machine and no clinicians. The partner hospitals have clinicians and a machine that does not work. Neither of them is giving anything away.

The board paper calls the extraction charge decisive, and the brief weights it at twenty eight per cent for the same reason: it is the only criterion where a structure can change the sign of the outcome rather than its size.

Every previous answer to it was a repayment. Outbid the public hospitals, take their consultants, and then train more than you took, eventually, through an academy that does not yet exist. That answer is honest about the debt and the debt is still incurred, years before anything repays it, and the people harmed in the gap are not the people who benefit afterwards.

This structure does not incur it. A consultant working here under a joint job plan has not left their hospital, has not been outbid, and is not lost to the patients who cannot pay. Their institution keeps them, gains a share of the private revenue they generate, and gains access to facilities its own estate cannot provide. National clinician supply is unchanged, which is the honest measure, and the group has what it needs to open.

It is worth being precise about the limit. This does not add clinicians to the country either, so the academy obligation survives intact and should still be contracted. What it removes is the deficit: the group no longer spends years in a position it has to apologise for while waiting for a repayment mechanism to mature.

That is the strongest single feature of this proposal and the reason it scores nine rather than the two the wholly owned option scores. It also explains why partner hospitals come first and why they are given seats rather than shares: what they are supplying is not capital, and paying them in equity would be answering the wrong question.

One owner, one shared governing body, a set of institutional partners holding no equity, and two deliberate absences.

The absences are decisions rather than gaps. The instruction put government and private investors in the comparator set and both are declined here, for reasons that are in the brief rather than in anybody's prejudice. A public equity partner brings political risk and an objective that changes with an administration, and where the partner hospitals are themselves public institutions the group already has the public relationship without the entanglement. Outside investors bring money the group does not need and an interest in relaxing exactly the commitments the whole exercise exists to protect.

Declining both is only available because of a fact established on the entity page: the group sold its other hospitals and is holding the proceeds. A group that needed outside equity could not adopt this structure. That is a real limit on how generally this answer applies and it is stated again in the shortfalls.

Three parties, three different positions in the stack, and three different things being bought.

Every party holds a stake, including the ones supplying no capital, and that is deliberate rather than generous. A partner with no stake can reduce its sessions next year at no cost to itself, and the previous version of this proposal had exactly that weakness. A partner that has bought in has something to lose, and a hospital whose clinical supply rests on goodwill rests on nothing.

The partner hospitals take a preferential return at a modest fixed rate, ahead of Meridian in the queue and below it in rate. They are buying a yield and a relationship rather than a growth asset, and their stake can be funded out of the revenue share their own clinicians generate, which makes it affordable to an institution with no spare capital. Three at most, because beyond that the clinical body stops being a governing group and becomes a forum.

The state is the interesting one and it is where the structure earns its keep. It contributes what it actually has, which is land, the licence route and regulatory alignment, rather than money it does not have. In exchange it holds a real stake with real rights. And it does not extract its return.

That last provision is the most consequential sentence in this proposal. The state’s dividend is taken in treatment: a share of capacity, drawn as of right, for patients who are not paying out of their own pocket. The ethics paper concedes that a voluntary free care share is charity a board can quietly reduce in a difficult year and that the group grades itself on it. This converts it into a shareholder return. Withholding it is no longer a policy adjustment, it is failing to pay a stakeholder what it is owed, and the counterparty has standing, lawyers and an interest in noticing.

It also fixes something the brief called structural. Instruction 02 weights enforceability from outside at eighteen per cent precisely because nobody currently holds the group to anything. A shareholder whose entire return is the thing the commitments protect is the first party in this whole study with both the standing and the motive to enforce them.

Nobody else is admitted in the first round. No institutional equity, no diaspora placement and no clinician equity yet, for the reason the instruction itself gave: each of them adds a party with a direct financial interest in reducing exactly the return the state has agreed to take. Clinician equity remains attractive against the counter offer problem and it belongs in a later round, once the commitments are entrenched and a new shareholder cannot vote them down.

Shared clinical oversight is the hardest part of this arrangement and the part most likely to be fudged, so it is set out as a mechanism rather than as a principle. The thresholds below are negotiating positions. The mechanism is not.

The body governs by voting, and the vote does three things: it sets clinical direction, it appoints the Chief Medical Officer, and it is available on any clinical matter a member refers to it.

The weighting is the part that carries the argument, and it is deliberately not proportional to money. Meridian holds a fixed quarter of the clinical vote, granted by the constitution rather than bought, and its stake is excluded from the calculation entirely. The partner hospitals divide the remaining three quarters between themselves by the ratio of what each has invested. The group funds the entire hospital and holds a minority of the vote on how medicine is practised in it, which is the whole proposition stated in one sentence.

That is not modesty, it is the only arrangement that makes the partners real. Weight the clinical vote by capital and Meridian wins every clinical argument by having paid for the building, and the partners are advisers with a voting card. Excluding its stake is what converts a consultation into a governing body, and it is the price of the clinicians the group cannot otherwise get.

Where the partners tie, Meridian has the veto. That completes the shape: it cannot impose, because a quarter is not a majority and its money buys it no more of the vote, and it cannot have something forced on it on an even split either. Every change of clinical direction requires somebody to be persuaded rather than outnumbered.

The veto does one thing that has to be fenced off, and it is the hazard this site has been describing in other currencies for twenty pages. A motion to raise a clinical standard usually costs money, and the party that pays for the hospital can block it on a tie for reasons that are financial while the vote looks clinical. That is the ring fenced theatre session being removed by somebody reading a utilisation report, arriving in a boardroom instead of on a ward.

So the carve out. Where the Chief Medical Officer certifies that a motion is required on patient safety grounds, the veto does not apply and a tie carries it. That keeps the veto for direction, strategy and appointments, which is what it is for, and removes it from the one place it could do harm. It also gives the officer a specific power rather than a general one, which is the right shape for a safety mechanism.

The appointment of the officer is the load bearing vote and it needs a supermajority. The reason is practical rather than constitutional. The consultants working in this hospital are employed by the partner institutions, and an officer a substantial part of the body voted against has no authority over them at all. Legitimacy here has to be collective because the staff are.

And then the officer runs the hospital. Accountability cannot be held by a committee, and a structure that leaves it with one will discover that in front of a coroner. The resolution is not to give the group the appointment; it is to let the body appoint and the officer act. The vote confers the mandate, and the mandate is exercised by a person.

Which produces the two rows in the table that do the safety work. A clinical matter can be referred to the body, and the body advises, and the officer decides and records the decision against that advice. Anything happening now is not voted on at all. A hospital cannot convene at three in the morning, and an arrangement that implies otherwise is unsafe on paper long before it is unsafe on a ward.

One asymmetry is not negotiable and should be drafted first. Raising a clinical standard takes a majority. Lowering one takes unanimity of the whole body and a named clinician’s written justification on the record. That is the rule the standards page already sets, which is that objectives never fall and specifications are re-derived, given a voting threshold so that it survives the year somebody is under cost pressure.

The final row is the quiet one. Operational matters are not the body’s at all: the estate, procurement, the supply chain and the non clinical establishment are Meridian’s alone. The body sets clinical requirements and the group decides how they are met. Without that line the votes leak sideways into the machine, and the reason this structure works is that the machine has one owner.

Twenty five per cent sounds like a minority position and a veto sounds like a safeguard. Put the two together and work out which coalitions can actually carry a motion, and the arrangement turns out to be considerably stronger than either word implies. This is worth doing on paper before it is agreed, because none of it is obvious from the clauses.

A coalition carries a motion only by exceeding half of the vote outright, since an exact half is a tie and a tie is vetoed. Because Meridian’s quarter is fixed, that condition reduces to a threshold on invested capital that does not depend on the amounts at all: a group of partners must hold more than two thirds of the partner capital to act over the veto.

The consequence in the first row is the one to sit with. If the three partners invest equally, no two of them can carry anything. Two thirds of the partner capital produces exactly half of the vote, which is a tie, which Meridian vetoes. Equal stakes mean clinical direction against Meridian’s wishes requires all three partners unanimously, and Meridian plus any single partner can stop anything at all.

That is a much harder position than the constitution appears to grant, and if it is the intention it should be stated rather than discovered. If it is not the intention, the fix is in the stakes rather than the thresholds: partner holdings have to be deliberately uneven for a two partner coalition to be able to act.

Which gives the third row, and the opposite failure. A partner holding more than two thirds of the partner capital carries motions on its own, over Meridian and over the other two partners together. The cap of three partners was set to keep the body governing rather than debating; the distribution of stakes within that three decides whether the body has one voice in it or several, and that is a question about the capital structure rather than about clinical governance.

The practical instruction is short. Do not agree the voting rules and the stake sizes in separate conversations. They are the same conversation, and this table is what it looks like when it is held once.

The line is not clinical against non clinical, and drawing it there is the commonest way this structure fails.

Portering, cleaning and bed turnaround look like facilities and are capacity. A bed is not a bed again until somebody has cleaned it, and this study has already put a number on what that returns across a hospital. They sit with the group, which is what the group is for, and there is no contract boundary between them and the wards because there is only one company on that side of the line.

That is the quiet advantage of this arrangement over the version with an outsourced estate. There is no facilities contract. Nobody is measuring cleans completed while the hospital needs beds returned. Every operational function reports to one organisation whose entire expertise is flow, and the only boundary in the building is the one between the machine and the medicine.

Procurement deserves its own note. Buying at group scale, with verification, testing and a single accredited chain, is the only serious answer to the counterfeit medicines problem, and it improves with every hospital the group adds. Partner institutions buying through it is a benefit worth real money to them, and it is the kind of benefit that binds a partnership more durably than a revenue share.

Four rows are genuinely contested and the table says so rather than pretending otherwise. The sharpest is nursing establishment, which is a clinical safety decision that sits in the cost base the group carries. Both sides have a legitimate claim, this proposal gives it to the joint body, and the first difficult budget will test whether that holds.

A partner hospital is being asked to let its consultants work somewhere else, which on the face of it is a thing to resist rather than to sign.

What it receives is specific. Sessions for its clinicians in a facility its own estate cannot provide, which is a retention instrument for an institution that is losing people to emigration. A share of the private revenue those clinicians generate, flowing to the institution rather than only to the individual, which is the difference between a partnership and a moonlighting arrangement. Training placements and a route into the academy. Access to the group's procurement, which is worth real money and gets better as the group grows. And seats on the body that sets the clinical standard in a hospital built to a standard they could not otherwise afford to build.

What the group receives is the only thing it cannot obtain without harming somebody, plus the legitimacy that comes with a recognised institution's name on the clinical governance. The family testimony on this site says the barrier is evidence and accountability rather than capability. A hospital whose clinical standards are set jointly with an established teaching institution is making a claim a new entrant cannot otherwise make, and making it on day one rather than after five years of published results.

It is worth naming what this does to the trust ramp, because it is the second most valuable feature of the structure after the extraction charge. The board paper treats the slow ramp as a ground against the project: trust cannot be bought and takes years of published performance. Partnering with an institution that already has it is the one legitimate shortcut, and it is available immediately.

The criteria and weights come from instruction 02 and were fixed before this proposal was drawn. The weights are quoted; only the scores are ours, and they are a judgement rather than a measurement.

Read the rows. The first moves further than anything else in this study has moved a criterion, from two to nine, and it does so by removing the harm rather than by mitigating it. The second moves because the partner hospitals sit on the clinical body and they represent, imperfectly but genuinely, the constituency the ethics paper says has no standing anywhere.

Two rows move the wrong way and both are honest costs. Control of the operating standard is eight rather than ten, because the clinical standard is genuinely shared and that is a real transfer, even though the operating standard is held outright. And the exit drops to six, which deserves more attention than it usually gets: what a buyer inherits here is a hospital whose clinicians belong to somebody else, and the value of that asset depends on agreements the buyer must persuade the partners to renew.

Deliverability holds up better than the alternatives, and that is the pleasant surprise. There is no equity negotiation, no regulated raise, no government counterparty and no valuation of anything. The clinical agreements are difficult to write well and quick to sign, because nothing has to be priced.

Only one partner willing to engage. The structure needs more than one from the outset, because a hospital whose entire clinical supply rests on a single relationship has traded the extraction charge for a dependency, and the second is easier to see and harder to survive.

Partner institutions that will sign sessions but not shared governance. Sessions alone make this a staffing agency arrangement with better paperwork. The whole ethical argument rests on the partners owning the clinical standard their people work to, and a proposal that delivers only the sessions should be reported as having failed rather than as having partly succeeded.

A regulator that will not licence shared clinical governance. This is the technical risk that could end the structure and nobody has tested it. The licensing question is whether a single accountable clinical officer inside a jointly governed body satisfies the regulator, and it is a conversation to have before anything else.

And evidence that the group cannot fund it alone after all. The refusal of outside capital rests on the disposal proceeds covering it. If the capital requirement turns out materially larger, the choice is between outside equity, which the brief warns makes the enforcement problem worse, and a smaller hospital. On this analysis the smaller hospital is the better answer, and that should be said now rather than discovered later.

Four, and the first is the one that could genuinely harm somebody rather than merely cost money.

Shared clinical governance is comfortable in ordinary times and dangerous at the moment a patient dies unexpectedly. Accountability cannot be held by a committee, and an arrangement that leaves it ambiguous will discover the ambiguity in front of a coroner. The answer proposed is a single named accountable clinical officer, employed by the group, with authority that does not need the partners' consent to exercise, and the joint body setting the standards that officer is answerable for. Anything vaguer is not a governance structure, it is an agreement to discuss, and the board should treat this clause as the one that decides whether the rest of the arrangement is safe.