Field notes · The essay

The philosophy of cost containment.

By Dr Hossam Elkholy, physician and former hospital medical director on Egypt's Red Sea coast · Updated June 2026

Every claims desk eventually meets an invoice that reads further than the care went. The usual response is procedural: another form, another vendor, another percentage point of discount. This essay is about the other response, the one the discipline's own literature has been pointing to for sixty years. Cost containment is not a stack of techniques. It is a short list of principles about information, incentives and repeated games, and once you hold the principles, the techniques sort themselves into the useful and the theatrical.

Two papers that explain your loss ratio

In 1963 Kenneth Arrow, later a Nobel laureate, published the paper that founded health economics. His point was structural: medical care defeats ordinary market discipline because the seller knows vastly more than the buyer, about what is needed, what was done, and what it should cost. Seven years later George Akerlof's famous "market for lemons" generalised the law: wherever one side holds the information, price detaches from value, and the honest and the inflated trade at the same confident face value. Both men were writing theory. A claims desk reads the applied version every morning.

Notice what the two struck lines mean. Domestic cost containment, everything your company does at home, stands on referees: a tariff, a network contract, a regulator, a court that can be reached economically. The literature calls medicine an asymmetric market even with those referees present. A Red Sea tourist file is that same market with the referees deleted. Nothing about the desk's diligence changed; the floor under it did.

What the discipline actually built, and when

I hold the industry's own history in genuine respect, because each instrument in it was a real answer to a real failure. The sequence is worth knowing, because each step also names the assumption it cannot work without.

The masterstroke of the series is 1983. Diagnosis-related groups worked because they refused the invoice itself: pay the diagnosis a fixed, evidence-derived sum, and the incentive to grow the bill dies at the root. Note the philosophy, not the mechanism. DRGs are incentive design, not auditing. The lesson for any payer is that the deepest savings never come from arguing lines; they come from changing what the other side finds profitable to write.

The two illusions that eat budgets

Where the referees are missing, two well-documented cognitive traps do most of the damage, and both wear the costume of diligence.

The first illusion is the discount KPI. The moment "percent saved off billed charges" becomes the desk's target, the supplier obliges by raising the number the percentage is taken from. The metric improves every quarter while the cash cost climbs. Frederic Bastiat, writing in 1850, called this the difference between what is seen and what is not seen: the seen is the discount on the settlement sheet; the unseen is the baseline quietly indexed to your tolerance. It is the claims version of the management line often credited to Peter Drucker, that what gets measured gets managed: measure the discount, and a discount is exactly what you will be given.

The second illusion is the anchor. Kahneman and Tversky showed in 1974 that the first number on the table frames every number after it, even for experts who know the trick is being played. A headline invoice is not information; it is an opening position, written by the party who keeps your payment history. Negotiate downward from it and you are working inside the author's frame.

The five principles that survive

Strip away every instrument that needs a referee, keep what the sixty years of literature actually proves, and you are left with five working principles. They are the whole philosophy of this desk.

Sun Tzu is principle two stated twenty-five centuries early. The asymmetry Arrow described is not abolished by paperwork; it is abolished by a reader who has stood on the other side of it. I have written the documents your desk receives, run the operation that produces them, and priced the files your guarantees are measured against. That is not a credential; it is a position, and position is the thing the lemons market cannot defend against.

The repeated game

Robert Axelrod's tournaments on the evolution of cooperation proved what every billing office already knows by instinct: in a repeated game, reputation is the currency. Your claims relationship with this coast is not a series of independent negotiations; it is one long game in which each settlement teaches the other side what your paper is worth. Pay unread, and the lesson is recorded in your precedent file. Review visibly, pay conforming invoices promptly, dispute on the record and never by silence, and within a season the asks addressed to your desk fall on their own. Containment, done properly, is not an argument you win; it is a reputation you set.

Decided before, not after

The last conviction is about timing, and it is the one the checklist movement taught medicine itself. Atul Gawande's surgical checklists save lives not by adding intelligence but by moving a handful of questions to the moment before the irreversible step. A claims file obeys the same law. The expensive decisions, admit or treat as outpatient, operate or repatriate, guarantee with clauses or guarantee blank, are all taken in the first hours, and every one of them is cheap to influence before and ruinous to argue after. A desk that reads in September, authorises theatre on clinical findings within two hours, and reviews before settlement is not working harder than the desk that audits in February. It is simply standing earlier in the file.

The invoice is a text. The clinical record is its only honest critic.

The shelf behind this essay

None of this is private cleverness. It is the discipline's own canon, read with a clinician's eye and tested against live files. If your desk wants to go deeper, start here; every argument above has a spine on this shelf.

There is a line often credited to W. Edwards Deming: in God we trust; all others must bring data. On this coast the data is not a benchmark database, because no honest one exists. The data is the chart, the nursing notes, the medication record, the original-language file behind the translated summary. Bring those, read them beside the invoice with someone who knows how both are written, and cost containment stops being a percentage and becomes what the literature always said it was: the restoration of an informed buyer to a market that quietly removed one.