Buy the faculty, then build the school

Build in Lagos. Configure the hospital around the four specialties where Nigerian money already leaves the country, which are oncology, orthopaedics, nephrology and cardiology. Staff the opening by outbidding, because there is no other way to open a hospital in a market this short, and select those first appointments for their ability to teach as much as for their ability to practise. Then build a training academy on the same site, tiered so that the fastest programmes repay inside the first year while the clinical ones are still running, and produce the rest of the establishment yourself. Choose the parcel on power, expansion and campus land rather than on price. Commit nothing at any gate until the whole hospital has been run against that gate's constraints.

The order of those sentences is the argument. The staffing sequence decides the building, the building decides the parcel, and the parcel decides the city. Lagos wins the scoring below, but it wins because of the academy rather than in spite of it: the universities, the teaching hospitals and the placement capacity are there, which is what makes the second half of this strategy possible at all.

It is worth saying what this is not. It is not a health tourism recommendation. Health tourism means persuading foreigners to fly in, which would put Meridian into competition with Bangkok, Istanbul, Dubai and Delhi from a standing start on their established terms. The proposition here runs the other way: give Nigerians who currently fly out a reason to stay. The paying customer and the catchment then become the same population, which is the most useful single fact in this decision.

Nigeria has roughly one practising doctor for every 3,600 to 4,000 people, and around 16,000 doctors have left in the past five years. That shortage is the binding constraint. Capital is not, and demand is not, and mistaking either of them for the constraint is how this decision goes wrong while looking successful.

The instruction sets out two ways to staff a hospital in a market this short. Hire locally at private rates, which means outbidding the public hospitals for the same scarce consultants and moving clinicians from people who cannot pay to people who can. Or build the hospital around bringing Nigerian clinicians home, which is slower and is the only version that raises the number of doctors in the country.

It leans toward the second and is right about the arithmetic. It is wrong to present them as a choice, and the error is worth naming precisely because it is a common one: it compares two strategies at the same moment in time when they do not occupy the same moment.

Nobody senior returns to a department that does not exist. Every one of the things that brings an experienced clinician back, a unit that works, equipment that is present, colleagues at their own level, a career that goes somewhere, is a property of a hospital that has already been running for years. Return is a year five strategy being asked to do a year zero job, and a plan that depends on it for the opening is not a plan, it is a hope with a spreadsheet attached.

So the two are a sequence rather than an alternative, and the sequence has a third term that neither of them contains. Outbidding opens the building. Return becomes possible once the building is worth returning to. And the academy is the thing that changes the total rather than redistributing it, which is what the statement was actually asking for and did not find.

The instruction puts the objection to outbidding about as strongly as it can be put. The building opens, the figures are excellent, the total number of doctors in Nigeria has not moved by one, and the catchment is measurably worse off at the exact moment the press release says the opposite. It then observes that a group which treats goodwill as a resource it is not allowed to plan with has no business solving its staffing problem by draining somebody else’s.

All of that stands. This recommendation does it anyway, and the difference between doing it and doing it honestly is the whole of this chapter.

It is a debt. For a period measured in years, this hospital subtracts clinicians from the public system in its own catchment. The academy is the repayment, and the repayment is real but it is late, and the people harmed in the gap are not the people who benefit after it. No amount of good intent changes the sign of that number while it is negative.

Which means the only defensible version of this strategy is one that publishes the number. A baseline of practising clinicians in the catchment, taken before we open. The running total, published annually, including every year it is negative and especially those years. And a crossover date stated in advance, so that missing it is visible to somebody other than us. That is the difference between taking a debt and pretending there was never a loan.

There is one honest mitigation and it is smaller than it sounds. Outbidding does not only move consultants out of public hospitals; it moves some of them out of private practice, out of part time arrangements, and in a few cases out of the departure lounge. The proportion matters and it is measurable, and if it turns out to be large the harm is smaller than stated here. Claiming it before measuring it would be exactly the kind of convenient assumption this study spends ten pages objecting to.

This is where the two halves join, and it is the part that makes the sequence a strategy rather than two separate decisions taken in different years.

The consultants Meridian outbids for are not only its clinical staff. They are the faculty. They are the people who will teach the first nursing cohort, supervise the first technicians, examine the first residents and set the standard that everything downstream copies. Which changes what the group is buying at the opening: not the ability to run a hospital for five years, but the ability to reproduce how it runs a hospital indefinitely.

That should change how the first appointments are made. Select for the ability and the willingness to teach alongside clinical standing. Write teaching sessions into the job plan rather than into the aspiration. Build the academic ladder at the same time as the clinical one. And accept that this narrows the pool and raises the price, which it does, and which is a cost worth paying because the alternative is a hospital that has bought six excellent years and no way of producing a seventh.

The usual objection to a training academy is the lead time. A doctor is six years away before residency begins, which is a long way past the horizon of any board discussion, and it is the reason the idea is normally raised and dropped in the same meeting.

That objection only holds if the academy is a medical school. It is not, or rather it is a medical school last. Read the second column of the table: the programmes that repay fastest are the ones nobody argues about, because the staff they produce are the ones nobody counts.

Portering, housekeeping and stores are weeks of training, and this site has already shown what they hold. A bed is not a bed again until somebody has moved a patient and somebody else has cleaned it, and the arithmetic on the people pages turns that turnaround into real beds returned across a hospital. Ward clerks and records are weeks to months, and they decide how much of a patient's stay is spent waiting on a document. Technicians are one to two years, and they decide whether the equipment produces results or service tickets, which is separately the largest reason a senior clinician gives up on a hospital in this market.

So the academy starts producing inside the first year, in exactly the roles where the group's operating standard is most easily taught and most rarely bothered with. Nursing follows at three years, which is the programme that most determines what the hospital is actually like. Medicine comes last, in partnership with an accredited institution, and it is the only line that addresses the national shortage head on.

Built in that order the academy is not a decade long act of faith. It is a thing that pays for part of itself early while the slow programmes run, and the early tiers are also the proving ground for the teaching itself: a group that cannot reliably produce a housekeeper to its own standard has no business claiming it will produce a surgeon.

Meridian builds hospitals, matures them, sells them at best in class, and does it again. Everything in the group is therefore either a thing that is sold or a thing that is kept, and almost nothing has been explicitly designed as the second.

The academy is the second, and that single property makes it the most valuable thing in this recommendation. The hospital goes to a buyer. The people, the programmes, the curriculum and the faculty relationships stay with the group and staff the next site. The first hospital stops being a one off asset and becomes the place where the operating standard was manufactured, and the second one opens with a trained cohort instead of a recruitment campaign.

It also resolves something the entity page asserts and does not really explain. Meridian's product is stated as demonstrated, transferable operating performance rather than buildings. Transferable by what mechanism, though? Not by documentation. Every organisation has written its standard down and every organisation has watched the standard fail to travel, because what actually travels is people who were taught it. An academy is the transfer mechanism, and without one the word transferable in that sentence is a claim with nothing behind it.

And it changes the conversation at the exit. A buyer's technical adviser will pull turnover and agency spend first, which is where a hospital that spent its staff shows up. A hospital with its own pipeline shows up in the same numbers from the other direction, and the seller can point at the mechanism rather than at a run of good years. That is a better answer in that room than any figure on its own.

The criteria, the weights, the threshold and the four comparators were all fixed in the instruction before any of this was scored. What follows is the scoring against them.

Lagos clears the threshold and the margin is wider than it would have been before the academy entered the criteria, which is worth being suspicious about rather than pleased by, since the weights were set by the same people who like the answer.

The defensible part is that it wins on a different line than the obvious one. Not on outbound spend, which is the crowd pleasing number, but on whether a training academy can actually be stood up: universities to affiliate with, teaching hospitals, placement capacity, and a labour market a cohort can be recruited into. Abuja closes most of the gap if the training weight is cut, which tells you exactly where this recommendation is load bearing. Accra scores better on regulatory route and worse on the size of the flow it would capture.

The scores are marked schematic and the marking should be taken literally. They show what the instrument does and which criteria dominate it. They are not a finding, and they will move when parcel, tariff, placement and workforce data are collected for each candidate, which is work that has not been done and is not being claimed here.

This answer belongs to the first gate, which is land, and that is why it is written now rather than better and later. Once an option is taken on a parcel the jurisdiction is fixed, the catchment is fixed, and the ceiling on everything after it is fixed. Analysis of staffing strategy delivered past that point can only describe what has already been bought.

The academy pushes hard into the second gate as well, and it is the consequence of this strategy that is easiest to miss. Teaching space, simulation, a library, examination rooms and some accommodation are floor area, and floor area is decided at massing. An academy that is agreed in principle and not drawn at massing stage is an academy that does not get built, because by the time anybody looks for the space it is a car park with a planning consent on it. It belongs in the brief at gate two, not in a later phase.

So the answer is deliberately coarse where precision would be false and specific where a decision is imminent. It names a city, a staffing sequence, a specialty concentration and the requirement for campus land, because those four have to be settled before a parcel can sensibly be chosen. It does not name a parcel, a bed count, a cohort size or a capital figure, because those belong to the second gate and inventing them now would produce exactly the confident wrong number this study keeps warning about.

Four things, written now rather than after the event, and specific enough that they could actually occur.

An accreditation route that turns out to be closed. The whole repayment argument depends on a private group being able to have its programmes accredited and its students placed. If that is not achievable in this jurisdiction on a reasonable timescale, the recommendation collapses to outbidding with no repayment, which is a strategy this document would not defend, and the answer becomes a different country rather than a different plan.

A crossover date that cannot be made to arrive under pessimistic assumptions. If the modelled running total stays negative across the plausible range of cohort sizes and attrition rates, then the academy does not repay the debt and the honest conclusion is not to take it.

Parcel data that closes the power and area gap elsewhere. Lagos loses points on infrastructure cost and on what the plant compound plus a campus takes out of a constrained urban site. A parcel elsewhere that is cheap to power and has genuine room for both could outweigh the margin without any weight being changed.

Outbound flow that turns out not to be addressable. Money leaving is not the same as money that would have stayed had the service existed locally, and the family testimony on this site makes a strong case that most of the barrier is evidence, accountability and ownership of the case rather than capability. If that holds, capability alone captures very little of the flow, and the demand weighting in the instruction is too generous.

Five weaknesses, at the same size as the recommendation and in the same document, because a shortfall written down is something somebody can be given to close and a shortfall left out simply stops being anybody's job.

The first two are the ones to read twice. This strategy takes a real harm up front and promises to repay it later, and both the harm and the repayment are things this group would be measuring about itself. That is the weakest structural feature of the whole recommendation, and the only answer to it is to publish the running total from a baseline set before opening, annually, in the years it looks bad as much as the years it does not.