Antimicrobial Stewardship

You have said it at a bedside, at a discharge, on the phone to a worried daughter. Make sure you finish the whole course. Even if you feel better.

Australia's national ANTIMICROBIAL STEWARDSHIP CLINICAL CARE STANDARD runs to forty-eight pages and applies everywhere from an ambulance to a residential aged care home. It does not say that. It does not say it once.

The word 'course' appears TWICE in the whole document. Once in the glossary, defining the word 'assessment', where it means the course of an ILLNESS. And once in the list of INAPPROPRIATE USES, where the failure named is 'continuing treatment for longer than necessary by not time-limiting or cancelling courses'.

AND HERE IS THE SENTENCE THAT TRAVELS WITH THAT COUNT, EVERY TIME. This is NOT permission to stop an antimicrobial early, and it is not a nursing decision. The same list of eight inappropriate uses also names 'patients omitting doses' and 'patients not taking antimicrobials as prescribed by their clinician'. Both failures are on it, four lines apart. What the Standard removes is the RITUAL, not the duration.

Antimicrobial stewardship is 'the safe and appropriate use of antimicrobials TO REDUCE HARM, while curtailing the incidence of antimicrobial resistance'.

Read the order. Reduce harm comes first; resistance comes second. Stewardship is usually taught as an argument about a future generation, which asks somebody standing in front of an unwell person to weigh that person against a hypothetical population in twenty years. The Standard's own first reason is the patient in the bed.

The goal contains two more things nobody expects. 'Including assurance when an antimicrobial is NOT needed' - being able to tell somebody confidently that they do not need one is part of the goal, it takes longer than writing a prescription, and nobody is ever audited on it. And 'at the right time and for the right duration, based on accurate assessment and TIMELY REVIEW'.

The module works the Standard's own list of eight inappropriate uses - which runs in BOTH directions, because prescribing inadequate treatment and dosing too low are on it as well.

The whole Standard is eight quality statements. Most summaries reproduce page 5 and stop - which loses the part that tells you what is yours.

Each statement is written out three times over, in three registers: FOR PATIENTS, FOR CLINICIANS, and FOR HEALTH SERVICE ORGANISATIONS. Read all three and you can see exactly which sentence is addressed to you - and very often that what is missing where you work is a SYSTEM the organisation register has already described. 'We have no system under quality statement 4 for getting results back overnight' is a sentence a manager can act on.

Eleven indicators across the eight statements, and four of them measure documentation. Statements 1, 4 and 5 have no indicator at all - and one of those is the conversation with the patient.

One sentence the Commission says twice: 'NO BENCHMARKS ARE SET FOR THESE INDICATORS.' There is no national percentage to hit. If somebody tells you there is, ask where it comes from.

Statement 1: administer appropriate empiric antimicrobials as soon as possible, and DO NOT DELAY ADMINISTRATION AND DO NOT WAIT FOR RESULTS. Statement 4: samples taken as clinically indicated, PREFERABLY BEFORE starting therapy.

Side by side those sound like a contradiction, and a lot of confusion at three in the morning comes from having heard one and not the other. They are not in conflict, and if you take one sentence away from this programme take this one:

TAKE THE SAMPLE IF YOU CAN. GIVE THE DRUG REGARDLESS. REASSESS THE MOMENT THE RESULT LANDS.

The drug can be given at any time. The specimen can only be taken once, and statement 4 says why it matters - so treatment can be 'specific for the infecting organism' and the most appropriate NARROW-SPECTRUM agent used. The module also covers the paragraph nobody expects, inside the statement that says give it immediately: the patient's ADVANCE CARE PLAN should be considered, and it is important not to harm people approaching the end of life with burdensome treatments of no benefit.

Three words carry the whole statement: 'locally endorsed guidelines'. They are used every day to justify practices that differ from the national guideline - often perfectly properly, and sometimes not at all.

The Standard is unusually specific about what they mean, and the specification is checkable. A local adaptation must be BASED ON Therapeutic Guidelines; any deviation must carry a CLEAR RATIONALE based on published clinical evidence AND local epidemiology; that rationale must be DOCUMENTED; and it must be ENDORSED BY A GOVERNING BODY - a drug and therapeutics committee, an antimicrobial stewardship committee, a medicines advisory committee, or equivalent.

Which converts into three questions anybody can ask, none of which challenges anybody's clinical judgement: what is it based on, where is the written rationale, which committee endorsed it. They are questions about a document.

The module also covers the formulary, what 'restricted' actually means - three different reasons, and only one is about resistance - and the sentence that names the person giving the dose: 'PRESCRIBE, DISPENSE AND ADMINISTER'.

This is the module where the nursing work is, and it is the one that changes practice on Monday.

The 2020 revision of the Standard ADDED a whole quality statement - not a sentence, not an indicator, a statement with its own indicator. It is this one. Read that as the Commission looking at six years of implementation and concluding that what gets written in the allergy box was bad enough to need a statement of its own.

Box 2 lists six essential elements: the patient's description, the nature, the active ingredient, the assessment of likelihood, the severity, and the date and location of the care. And then the sentence that makes it usable on a busy admission: 'if any of the essential elements are UNKNOWN, this should be EXPLICITLY DOCUMENTED in the patient's healthcare record'.

You are not being asked to establish what happened in 1994. A word on its own in an allergy box is a rumour with a drug name attached, and it will follow the patient for thirty years. 'Penicillin - rash as a child, severity unknown, likelihood unable to assess' is a record, and it takes about forty seconds longer to write.

If not 'finish the course', then what? Quality statement 5 is the answer, and it is longer and better than the sentence it replaces.

Seven things the Standard says a patient needs to know, bookended by WHEN TO START and WHEN TO STOP - with the signs or symptoms that mean seek urgent care in the middle. And in the clinician register, three obligations collapsed into one sentence: the importance of using antimicrobials as prescribed, how long to take them, and when the treatment will be reviewed or ceased. The third is the one nobody says, and it is the one that tells a patient the plan can change and that stopping is a decision rather than a failure.

It also contains the hardest sentence in the Standard, and the shortest: 'IF ANTIMICROBIALS ARE NOT NEEDED, REASSURE THE PATIENT AND INFORM THEM OF OTHER TREATMENTS THAT CAN HELP WITH SYMPTOMS.'

This statement has no indicator. Nothing measured looks at what the patient was told, which is exactly why it is the one to hold onto.

Six things go in the record: the indication, the active ingredient, the dose, the frequency, the route, and the intended duration or review plan. Most charts have four.

The four that are always there are the ones needed to give the next dose. The two that go missing are the two the Standard built indicators for - and this is why:

No indication means the next clinician does not know what the antimicrobial is for. Not knowing what it is for, they cannot tell whether it is working. Not being able to tell whether it is working, they cannot justify stopping it. So it continues. THE UNDOCUMENTED ANTIMICROBIAL IS THE ONE THAT RUNS LONGEST - not through anybody's carelessness, but because the record has made continuing the only low-risk option.

And the commonest objection has an answer inside the statement. 'I don't know how long yet' is fine: the requirement is the intended duration OR A REVIEW PLAN, and indicator 6b accepts a review date.

Two clocks, two different starting points, and confusing them is the commonest error in this subject.

The 48 hours runs from THE FIRST PRESCRIPTION, and indicator 7a measures a review AND AN UPDATED TREATMENT DECISION documented within it. A drug started at ten on Friday night has to have one by ten on Sunday night - which in most services is nobody's shift.

The 24 hours runs from a microbiology result BECOMING AVAILABLE. Not from when somebody opened it, and not from when the round reached the bed. A result that sat unread over a weekend has already broken it.

Two switches: BROAD TO NARROW, and INTRAVENOUS TO ORAL - and the reason the Standard gives for the second is not cost or convenience but TO REDUCE HOSPITAL-ACQUIRED INFECTIONS. The cannula is the hazard, which puts the question in the hands of whoever looks at it every shift.

One more thing: the Standard never uses the word de-escalation. Zero occurrences. It says 'switched to a more narrow-spectrum agent'.

'Avoid prescribing antimicrobials POST-PROCEDURALLY, as prolonged antimicrobial use is not usually required.' And for patients, more plainly: 'after having a surgical procedure, antimicrobials are NOT USUALLY NEEDED UNLESS YOU HAVE AN INFECTION.'

This statement is not only about theatre. It names dental procedures, gastrointestinal endoscopy, implantable cardiac device insertions, cleaning and debridement of traumatic wounds, and cataract surgery.

It carries four indicators - more than any other statement - and one of them is unlike everything else in the Standard. Indicator 8d counts patients prescribed PROLONGED post-procedural antimicrobials DISCORDANT with guidelines: a count of something that should not have happened, rather than a proportion measured against a guideline.

And indicator 8c is the one that cannot be calculated without a nurse. Prophylaxis ADMINISTERED WITHIN THE RECOMMENDED TIME needs two times written down - the administration time and the incision time - and the prescription can be perfect while the indicator still fails.

A FACT ABOUT A DOCUMENT IS NOT AN INSTRUCTION ABOUT A PATIENT.

This programme opened with a count. A count is a powerful thing to hand somebody: checkable in a minute, contradicting something everybody says, and the kind of fact people repeat in a tearoom. Which is exactly why the last module is this one.

Nothing here authorises anybody to start, withhold, shorten, lengthen, change or cease an antimicrobial, and nothing here names a drug, a dose or a duration for any condition. What is not yours: the choice of agent, the duration, the decision that a patient should stop, the decision that an allergy is not real, and any verdict on whether anybody complies.

What IS yours is the longer list, and every item has a quality statement behind it: the specimen before the first dose, the six elements and the explicit unknowns, the missing indication and the missing stop date, the two clocks, the two switches, and the three things the patient is told.

The module closes on a habit. The Standard cites antimicrobial stewardship as NSQHS actions 3.15 and 3.16; those numbers now belong to workforce immunisation and workforce infections, and the stewardship actions are 3.18 and 3.19. Cite the instrument, not the document that quotes it.

Sixty items, 80 per cent to pass, two attempts - and twenty of the sixty are mandatory-correct.

A mandatory-correct item must be answered correctly whatever your total, because each one is a statement you will repeat to a manager, a colleague or a patient. They include the count and where both uses of the word sit; that it is NOT permission to stop early; that there are eight statements, eleven indicators and no benchmarks; the order in which statements 1 and 4 apply; the four-part test for a locally endorsed guideline; the six essential elements and the rule about unknowns; the three things a patient is told; the six things that go in the record; what indicator 7a measures; and that the duration belongs to a prescriber.

The five case studies come first and are completion-required rather than graded, with feedback on every option including the plausible wrong ones.

Five activities. Optional, and the time is not counted in the claimable CPD hours.

WHICH STATEMENT IS IT? matches fourteen situations to the quality statement that answers them. DOCUMENTED OR NOT? tests twelve record entries against what statements 3 and 6 require. WHICH CLOCK? sorts twelve prompts by which of statement 7's two clocks is running. YOURS OR NOT YOURS? sorts twelve actions by whose decision they are. And the scope self-check covers twenty-five actions across five roles.

None of them names an antimicrobial for a condition, states a dose or a duration, says what your local guideline contains, or decides whether anybody complies.

Four documents to take to work, three of them fillable.

The ANTIMICROBIAL RECORD CHECK takes one chart and tests it against the six things quality statement 6 asks for and the two clocks in statement 7 - and it produces no score and no verdict, deliberately, because it is not an audit tool. The ADVERSE REACTION RECORD walks the six essential elements with a column for marking which of them cannot be established, which is the whole point of it. The CITATION CARD asks which source settles a claim, and names the two this programme deliberately did not read.

And the WORKED EXAMPLE reads one constructed admission twice - once as it usually goes, once as the eight quality statements read it.

The documents you keep rather than read once.

The source and obligation mapping sets out every count, quality statement, indicator, definition and provision the programme relies on with the source named at each point, and ends with the sourced figures - every number the programme asserts and where it comes from. The scope matrix, the student workbook and the interactive register sit alongside it.

Built for anybody who has to defend a statement in a policy review, or find out quickly whether something they have been told is in the document at all.