In late September 2024, Hurricane Helene flooded a Baxter International plant in North Cove, North Carolina. Most hospital executives had never heard of the town. Within weeks many of them were rationing IV fluids.
According to the American Hospital Association, that one facility produced around 60% of the IV fluids used in the United States. Hospitals across the country started reviewing which patients truly needed an infusion, switching some to oral hydration and, in places, postponing elective surgery. Baxter restored the plant to pre-hurricane production levels in less than five months. For those months, a storm in the Blue Ridge Mountains was setting operating schedules in hospitals hundreds of miles away.
Seven months earlier, a different kind of crisis had hit from a different direction. On 21 February 2024, a ransomware attack took down Change Healthcare, the UnitedHealth Group subsidiary that processes an enormous share of American medical claims. An American Medical Association survey found that 80% of physician practices lost revenue from unpaid claims. Practices that had never been hacked could not get paid.
Neither crisis started in a hospital. That is the lesson most crisis-leadership training still misses.
We train for the wrong kind of emergency
Hospital emergency planning grew up around events that walk through the front door: a mass-casualty incident, a fire, a surge of flu patients. The drills reflect that. Command structures are activated, beds are cleared, staff are called in.
Those drills matter. But the crises that have tested hospital leaders hardest in recent years mostly arrived sideways. A single supplier fails. A software vendor is compromised. A payments clearinghouse goes dark. The building is fine, the staff are present, and the organisation still cannot function normally.
These events share three features that make them harder to lead through than a trauma surge.
First, they are slow and ambiguous at the start. Nobody knows on day one whether the outage will last a day or two months. Second, the hospital does not control the fix. Leaders are waiting on a manufacturer, a vendor or an insurer. Third, the damage is distributed. Every department is affected a little, so no single clinical leader owns the problem.
A trauma surge rewards fast, decisive command. A dependency failure rewards something different: early rationing, honest communication about uncertainty, and the discipline to plan for the long version while hoping for the short one.
Ration early, even when it looks like overreaction
The hardest decision in a supply crisis is the first conservation order. Restrict IV fluid use or delay elective cases on day three, and you will be accused of panic if supply recovers by day ten. Wait until stock is visibly low, and your options have already narrowed.
The only way to keep options open is to act before you have certainty. That is uncomfortable, because hospital governance rewards decisions backed by data, and early in a crisis the data does not exist yet.
One practical fix is to agree trigger points in advance. If days-on-hand for a critical product falls below a set threshold, specified conservation steps begin automatically. It takes the decision out of the realm of personal judgement at the worst possible moment, and it protects the leader who makes the call.
A trigger in practice
Take a mid-sized hospital that normally holds about three weeks of IV fluids. The supplier announces a disruption with no firm restart date. The board wants to know what happens next.
A trigger plan answers that before anyone panics. Below fourteen days on hand, the pharmacy reviews standing orders for maintenance fluids and pushes oral hydration wherever it is clinically reasonable. Below ten days, IV medication delivery shifts from small-volume bags to other methods where safe, and each elective list is screened for its fluid demand. Below seven, elective procedures with high fluid use are deferred, with clinical leads deciding which ones. None of these thresholds is magic, and each hospital will set its own. The point is that they were agreed in a calm room by people who knew the clinical trade-offs, not improvised in a corridor.
The arithmetic also rewards early action more than intuition suggests. Cutting daily use by a fifth at day twenty-one of stock buys about five extra days. Cutting it by the same fraction at day five buys barely one. Conservation is a compounding investment: the earlier it starts, the more time it buys for the supplier to recover.
The second-order problem: everyone rationing at once
There is a catch that hospital-level planning cannot solve alone. When every hospital responds to a shortage by ordering more, the shortage gets worse. Rational stockpiling by each buyer produces collective hoarding, and the institutions with the best purchasing teams or the biggest contracts end up protected while smaller and rural hospitals run short.
Manufacturers usually respond by putting customers on allocation, rationing supply according to historic orders. That is fairer than a free-for-all, but it bakes in yesterday’s pattern of demand, whatever today’s clinical need. A hospital that has recently grown, or that serves a population with unusually high need, can find itself short through no fault of its own.
This is why supply resilience is partly a public problem. Regional coordination, shared visibility of stock and agreed clinical priorities across hospitals can prevent the scramble. Few health systems have built that machinery for anything other than pandemics.
Cash is a clinical resource
The Change Healthcare attack exposed something clinical leaders rarely consider: how many days a hospital or practice can operate without incoming revenue. For many small practices, the honest answer turned out to be: not long.
A cash crisis becomes a patient safety crisis quickly. Suppliers want payment. Payroll does not wait. Services get cut, not because anyone decided they were unnecessary, but because they are the easiest to stop.
This is where the gap between clinical and financial leadership becomes dangerous. Crisis teams are usually clinically led, with finance consulted later. In a dependency failure, the finance director may be the person who knows first how bad things are about to get. They belong at the table from the first hour.
The same logic runs in reverse. A clinical leader who knows which services generate cash quickly and which run at a loss can help protect the ones patients cannot do without when money is tight. Without that conversation, cuts fall on whatever is easiest to pause, which is rarely the same as whatever matters least. A cash crisis managed only by finance ends up making clinical decisions by accident.
The case against redundancy, and where it holds
It is worth taking the other side seriously. Consolidated supply chains exist because they are cheaper, and cheaper supplies are not a trivial benefit when hospital margins are thin. Holding extra stock ties up cash and warehouse space, and some products expire. Paying a premium to a second supplier raises costs every year to insure against an event that may not come for a decade. A board that spends heavily on redundancy everywhere will be poorer, and a poorer hospital also harms patients.
That argument holds for most products. Nobody needs a second source for every item in the catalogue. The defensible position is targeted resilience: identify the small number of products and systems whose failure would stop core clinical services, and spend on redundancy there, and nowhere else. IV fluids, a handful of critical drugs, the electronic health record, the payments pipeline. Everything else can stay lean.
The harder problem is that individual hospitals cannot fix concentration upstream. If one plant makes most of a country’s IV fluids, no amount of hospital-level planning creates a second plant. That requires purchasers, governments and payers to reward manufacturers for diversified, resilient production, usually by paying somewhat more for it. Until buying decisions value reliability as well as unit price, the market will keep producing single points of failure.
Communication when you do not know the answer
Staff in a slow crisis tend to fill silence with rumour. Leaders often stay quiet because they have nothing definitive to say.
A better rule is to communicate on a fixed schedule whether or not there is news. A short daily update that says “here is what we know, here is what we do not know, here is what changes for you today” does more for morale than a polished statement a week later. It also surfaces problems. Frontline staff will tell leaders about workarounds, shortages and risks that never appear in a dashboard, but only if they believe someone is listening.
Map your single points of failure before they map themselves
The most useful exercise a hospital board can commission is not another mass-casualty drill. It is a dependency map. Which products come from a single plant? Which clinical systems run on a single vendor? Which revenue streams depend on a single clearinghouse? What would happen if each went dark for thirty days?
The answers are usually alarming, and often there is no quick fix. Healthcare supply chains have been consolidated for good reasons, mostly price. Diversifying suppliers costs money. Maintaining manual backup processes for digital systems costs staff time. Neither shows up as a return on investment until the day it is the only thing that works.
That is the real trade-off in crisis leadership. Resilience looks like waste in normal times. The leaders who build it will be second-guessed every budget cycle, right up until the week it matters.
Who needs to do what
For hospital boards, the task is to ask for the dependency map and the trigger plan, and to review them annually the way they review financial risk. For chief financial officers, it is to know how many days of cash the organisation can operate without claims income, and to arrange a line of credit before it is needed rather than during the crisis. For chief information officers, it is to keep downtime procedures practised, not just written: a paper-based fallback that nobody has used in five years is a document, not a plan.
For group purchasing organisations and national buyers, the question is whether contracts can reward suppliers for multiple production sites and transparent reporting of their own dependencies. For regulators and payers, it is whether they are prepared to pay for resilience in the products they depend on most. None of these actors can solve the problem alone. All of them can make it less likely that one flooded factory or one compromised server dictates operating schedules across a country.
Over the next year or two, the signals to watch are practical ones. Do hospital boards start asking for supplier concentration data as routinely as they ask for financial statements? Do purchasing contracts begin to include resilience terms? Do cyber-incident exercises start including the finance team and the payer relationships, not only IT? Where those things happen, the next dependency failure will still hurt, but it will not surprise.
The takeaway
The next major crisis in your hospital is more likely to begin in a factory, a data centre or a payments network than in your emergency department. The leaders who handle it well will not be the ones with the best command structure. They will be the ones who knew their dependencies, set their triggers in advance and told their staff the truth early.
What dependency in your organisation worries you most if it failed for a month? We would like to hear how you are planning for it.
