Who decided that?

It is a Tuesday afternoon. A nine-year-old arrives in your emergency department, seriously unwell, and over the next six hours your hospital does everything a paediatric intensive care unit would do — intubation, ventilation, lines, inotropes, a consultant at the bedside and a nurse who does nothing else — until the retrieval team lands and takes the child four hundred kilometres away.

Somewhere in those six hours a registrar asks the question that every registrar eventually asks. Why don't we have a paediatric intensivist? Why is the equipment in three different cupboards? Why am I told I can't do something I have done a hundred times? And underneath all of them, the real one: who decided that?

The answer is almost never "management, out of meanness." It is usually a document, written years ago, that you have never read and could find in about four minutes. Here are the main ones.

Role delineation: the document that explains most of it

Every clinical service in every public hospital carries a formally assigned level. Not a description — an assignment. In New South Wales the instrument is the Guide to the Role Delineation of Clinical Services, now in its seventh edition; other states have equivalents under other names. The levels run from No Planned Service up through Levels 1 to 6, ascending in complexity, and they set out the minimum workforce, support services and infrastructure required to deliver a service safely at that level.

This single concept answers a startling proportion of the questions trainees ask. Your hospital has what it has because of what it has been delineated to provide.

The same framework covers paediatric medicine and surgery for children, and the highest level — the specialist children's hospital — is the only one that includes paediatric intensive care. Worth knowing too: services with no designated paediatric level at all are still required to hold documented arrangements with the paediatric retrieval service. The retrieval relationship is a requirement, not an informal courtesy between friendly consultants.

Why you can't just do the thing you can do

Three words get used interchangeably and mean entirely different things. Getting them straight is worth more to you than almost anything else in this piece.

• QualificationWhat you have achieved. A degree, a Fellowship. Permanent, portable, yours. It says what you were assessed as able to do.
• RegistrationNational and regulatory — the Medical Board of Australia through Ahpra. It grants legal authority to practise anywhere in Australia and recognises your specialty. It exists to protect the public from the unqualified.
• Scope of clinical practiceOrganisational and local. This employer, this facility, these procedures, reviewed at least every five years. It exists to match a particular person to a particular environment.

The national standard defines scope of clinical practice as delineating the extent of an individual's practice within a particular organisation based on their credentials, competence, performance and professional suitability — and the needs and the capability of the organisation to support it.

That last clause is the whole thing. Scope of practice is a function of two variables, and only one of them is you. The NSW policy says it without any diplomacy at all:

There is no obligation upon a public health organisation to provide support or infrastructure for any given clinical service simply because the practitioner has the credentials, skills and experience to perform that service.

NSW Health PD2019_056, Credentialing and Delineating Clinical Privileges

Which is bracing, and also correct. A surgeon credentialed for an operation that this hospital cannot support — no appropriate ICU level, no after-hours imaging, no one to manage the complication at three in the morning — is a hazard rather than a resource, however good they are. Clinical privileges are facility-specific because facilities differ. When you are told you cannot do something here that you have done elsewhere, the sentence you are actually being told is usually about the building, not about you.

Where the money comes from, honestly

Australian public hospitals are funded substantially through activity based funding. Activity is classified, counted, and converted into National Weighted Activity Units. A national body sets a price per unit — for 2026–27 it is $7,418 per NWAU. The Commonwealth pays a proportion, the state pays the larger share, and money reaches a Local Health District through a service agreement that also carries its activity targets and performance measures. The district then allocates to its facilities.

Now the honest version of the thing people say about our nine-year-old. You will hear that transfers are funded badly, or that the hospital is penalised for stabilising and sending on. There is no transfer penalty in the pricing model. That claim will not survive contact with anyone who works in finance, and using it will cost you credibility you need for better arguments.

What is true is more interesting and harder to fix:

• Averaged prices have tailsActivity based funding pays an average price for a class of patient. It does not pay for this patient. Six hours of one-to-one nursing, a consultant off the floor, a registrar unavailable and a resuscitation bay closed are real costs that an averaged price was never designed to reflect. That is a property of prospective payment, not a conspiracy.
• The unit of count matters more than the priceIf the child is retrieved without ever being formally admitted, the episode is counted as an emergency department presentation, which carries far less weight than any admitted episode. Whether an admission happened can matter more to the recorded activity than anything about the diagnosis.
• The paediatric loading goes elsewhereThere is a pricing adjustment recognising that children cost more to treat — but it attaches to establishments identified as delivering specialised paediatric services. The loading follows the children's hospital, not the regional hospital doing the resuscitation. That asymmetry is real, and it is a much better argument than the one about transfers.
• Some things never touch your budget at allThe retrieval service is funded centrally and is free at the point of use to you. It appears nowhere in your hospital's activity or expenditure. A great deal of what makes the system work is like this, which is one reason it is invisible from the bedside.

What happened to the form you filled in at 4 am

Most trainees file an incident report believing it disappears into a void, or worse, that it is a mechanism for allocating blame. Neither is true, and the actual process is worth knowing.

A Harm Score 1 incident sets a formal clock running, and it runs fast: notification the same day or as soon as practicable, a brief to the Ministry within 24 hours and a fuller one within 72, a preliminary risk assessment within 72 hours, and a Serious Adverse Event Review completed within 60 days.

That review is the statutory framework — and here is another detail people get wrong: it did not abolish root cause analysis. RCA is now one of four approved methodologies that can be used inside a SAER, alongside the London Protocol and two NSW analysis methods. The chief executive decides which is used. The whole process is conducted under statutory privilege, which is what allows people to speak freely inside it.

Alongside it sits open disclosure — the obligation to tell a patient and family openly when something has gone wrong. The national framework was revised in June 2026 and now runs on five principles, including a restorative process and cultural safety. On the point trainees most often agonise over: saying I am sorry is described as essential, and every Australian jurisdiction has apology legislation protecting such statements in defined circumstances. Apologising is not an admission of liability. It is the expected standard.

Why the observation chart looks like that

You have almost certainly complained about the observation chart. It is worth knowing where it came from, because it is the clearest example in Australian practice of administration doing exactly what it is supposed to do.

In November 2005 a sixteen-year-old named Vanessa Anderson was admitted to Royal North Shore Hospital after being struck on the head by a golf ball. She died two days later. The coroner found she died of respiratory arrest caused by the medication she had been given, and identified poor communication between doctors, inadequate staffing, absent or inadequate notes, poor clinical decisions and ignorance of protocols. His finding was that "almost every conceivable error or omission occurred and continue to build on top of one another". He recommended a full and open inquiry into the delivery of health services in New South Wales.

That inquiry — the Garling Special Commission — was established within weeks and reported in November 2008 with 139 recommendations. On deteriorating patients it said this:

Detection and management of deterioration in a patient's condition by hospital staff can be problematic, particularly where the problems occur overnight when patients are under the care of junior clinicians. The Clinical Excellence Commission has suggested a plan to NSW Health which will assist to improve the care being provided for patients who may fall into this description. My recommendation endorses that plan and commends it to NSW Health for its consideration.

Special Commission of Inquiry into Acute Care Services in NSW Public Hospitals, 2008

Note the sequence carefully, because it is how change actually happens. The Clinical Excellence Commission already had the plan. It had identified unrecognised deterioration as a problem years earlier. What the inquest and the inquiry supplied was not the idea but the political conditions in which the idea could be funded and mandated across an entire state. Good proposals frequently sit waiting for a moment. Part of administration is having the proposal ready when the moment arrives.

That plan became Between the Flags, which commenced in 2010 and now runs across more than two hundred facilities. The colour-coded chart in front of you is its visible part: a yellow zone triggering a clinical review within thirty minutes, a red zone triggering an immediate rapid response. Underneath it sits National Safety and Quality Health Service Standard 8, Recognising and Responding to Acute Deterioration, one of the eight standards your hospital is accredited against.

Accreditation is not decorative. Assessment happens on a three-year cycle with mandatory short notice assessments, and a Not Met rating against an action starts a remediation clock with real consequences attached. This matters to you for a practical reason we will come to.

The job itself

Something should be said plainly here, because the default attitude of clinicians towards administration ranges rather widely. Ambivalence, mostly. Confusion, frequently. Bemusement, suspicion, and at the far end something close to contempt. All of which is, in the end, kind of lazy.

Consider what you do. You allocate a scarce resource — a bed, your attention, a ventilator, your registrar — under uncertainty, with incomplete information, knowing that the allocation will determine outcomes, and knowing you will be judged on results you do not fully control. You do it at three in the morning with the data you happen to have.

That is also the administrator's job. The differences are the timescale and the unit of analysis. You are optimising for the patient in front of you, over hours. They are optimising for every patient who will arrive over the next year, including all the ones nobody can see yet, over a budget cycle. Both of those are legitimate, and they pull against each other, and that tension is not a malfunction of the system. It is the system working. Somebody has to argue without compromise for this child. Somebody else has to hold the whole thing. A hospital where the clinicians stopped pushing would be badly run; so would one where the executive simply said yes to everything until the money ran out in March.

• They mostly cannot win visiblyWhen an administrator gets it right, the thing that happens is that nothing happens. No incident, no gap, no crisis. Nobody is thanked for an event that did not occur, and the counterfactual is invisible. You at least get to see your patient improve.
• Many of them were youA considerable number of hospital executives are clinicians who took a harder and less rewarding job because somebody had to. The ones who irritate you in meetings have frequently been in the room you are describing.
• The refusal is often a relayThe person declining your request is regularly explaining a decision taken three levels above them, years ago, in a document neither of you wrote. Arguing with them about it is like arguing with the registrar about the roster.
• They carry risk you never seeEvery gap in the service sits on somebody's risk register with their name against it. They are as aware of what the hospital cannot do as you are — usually more, because they can see all of it at once rather than the part that came through your door.

None of which means every decision is right, or that every administrator is good, any more than every clinician is. It means the caricature is not a useful tool, and that a trainee who arrives believing it will be unable to achieve anything.

Communication, which is the whole skill

Clinicians speak in cases. Administrators speak in populations, recurrent cost, risk registers and standards. Neither dialect is wrong and neither is superior — but they are genuinely different languages, and almost every failure between the two groups is a translation failure rather than a disagreement about what matters.

A case is the most powerful thing you have and the least persuasive way to deploy it. "I had a terrible night with a sick child and we did not have what we needed" is true, and it will be heard as an anecdote, because an executive's week is full of people with true anecdotes. The same material, properly translated, is formidable.

And the underlying point, which took me a long time to learn: role delineation, credentialing, the standards and the service agreement are not obstacles arranged against clinicians. They are the vocabulary in which resources are allocated. A registrar who can say we are a Level 4 unit, our networked tertiary service expects this of us, Standard 8 requires that, our last three incidents show this pattern, and here is the recurrent cost of each option is not fighting administration. They are speaking to it in its own language, which is very nearly the whole of the skill.

Learn it early. You will be doing this for thirty years.