MLR review in Australia — the Code, the reviewers, the process.

Every piece of promotion an Australian pharmaceutical company puts in front of a doctor has been read, argued over and signed off before it ships, by medical, regulatory and compliance reviewers, one at a time. Global teams call that process MLR — medical, legal, regulatory. Australian affiliates call it the promotional review, medical sign-off, or simply "the Code review". The work is the same: hold the material to the approved Product Information and to the Medicines Australia Code of Conduct, find what fails, and send it back until it doesn't. This guide is about how that works here — what the Code actually asks, in its own words, and what review looks like from each desk it lands on.

What "MLR review" means here

In the United States, MLR review answers to the FDA's Office of Prescription Drug Promotion and a body of federal regulation. In Australia the authority behind the review is different in kind. Promotion of prescription medicines to healthcare professionals is governed by industry self-regulation — the Medicines Australia Code of Conduct — which describes itself as "a form of industry self-regulation" that "sets an accountable standard" § Introduction.

Two things stop that from being optional.

The Code binds companies that are not Medicines Australia members. The TGA makes compliance a condition of registration: "prescription medicines must be promoted in accordance with this Medicines Australia Code of Conduct, whether the Company is a member or non-member" § Scope. If you register a prescription medicine in Australia, you promote it under the Code.

The Code sits on top of statute. It "is underpinned by the Therapeutic Goods Act and Regulations" § Scope, and it is the Act — not the Code — that prohibits advertising prescription medicines to the general public. The Code refers to that prohibition rather than restating it, for instance where it warns that promotional content visible to patients during a consultation "may contravene the Therapeutic Goods Act, which prohibits advertising prescription medicines to the general public" § 2.3(a), and again for trade displays with a public audience § 4.3(d).

So an Australian reviewer is holding a piece to three things at once: the approved Product Information, the Code, and the law underneath both. The Code governs promotion to healthcare professionals; advertising of therapeutic goods to consumers is a separate regime under the TGA's advertising rules, and prescription medicines cannot be advertised to the public at all.

The edition that applies

Edition 20. It was adopted at Medicines Australia's annual general meeting in October 2024 and has been in effect since 30 March 2025. It is a restructure rather than an amendment — Parts A to F, ten Overarching Principles, seventeen sections, with the practical guidance moved out into the Code Tool Kit — and there are three changes a reviewer notices on the first piece they read against it.

The Minimum Product Information is no longer a mandatory inclusion. In its place, § 2.1(d) requires "a statement directing healthcare professionals to review Product Information (PI) before prescribing", including "the means for healthcare professionals to access the PI immediately in electronic or other form, or the telephone number for the Company medical information service". The Tool Kit's Prescribing Information Guidance treats a voluntary summary as permitted; the mandatory element is now the access statement.

The inclusions apply whether or not the piece makes a claim. § 2.1 opens with "all promotional material for a product, whether or not the material contains a promotional claim". A brand-and-logo placement in a journal carries the seven inclusions the same as a detail aid does.

Balance has its own clause. § 1.1 sets it out, and the Glossary defines it: balance "refers to information that provides proportionate weight to the benefits and risks of a product".

If your templates still carry a Min PI panel from Edition 19, it isn't wrong to keep one — but it isn't what the Code asks for, and if it's there it has to be accurate. The thing to check for is the access statement.

What a review actually checks

The Code's standard for promotion to healthcare professionals is one sentence. Companies may promote "so long as all information, claims and graphical representations are current, accurate, balanced, consistent with the approved product information, and do not mislead directly, by implication, or by omission" § 1. Everything a reviewer does is a test of one of those words. In practice a review runs five tests.

1. Consistent with the Product Information

Overarching Principle 8 puts it without qualification: "All promotional claims are consistent with the Australian Product Information document, including claims about competitor products, irrespective of the source on which the claim is based" § OP 8. Australian is the operative word — global masters are written to another market's label, and the reviewer's first job is to hold every indication, population, dose, endpoint and safety statement to the PI approved here.

"Consistent" is not "identical". A claim can paraphrase the PI and pass; it fails when it outruns the PI as a whole — a wider indication, a population the PI excludes, a dose the PI doesn't give, an endpoint the PI doesn't carry. That is what off-label promotion means in this system, and it is the most expensive finding a reviewer makes.

2. Balanced

§ 1.1(b): balanced promotional materials "provide proportionate weight to the benefits and risks of a product, and they recognise that product and risk information are as relevant as any therapeutic, promotional, or non-promotional claim". § 1.1(c) scales the requirement: the presentation of risks should be "appropriate to the complexity of the promotional material, the therapy area, stage of product lifecycle".

The working test is per claim: for each efficacy claim on the page, is risk information present with comparable prominence, placement and extent? A piece can be accurate line by line and still fail on the impression it leaves — which is why balance findings are the ones marketing argues hardest and loses most often.

3. Substantiated

Companies are "responsible for ensuring that all promotional claims are referenced and that cited reference/s provide the appropriate level of evidence for the claim being made, reflect the body of evidence, and allow healthcare professionals to independently evaluate the validity of the results" § 1(a). The Code then says what counts:

• Posters, abstracts, "personal communication" and unpublished data cannot be the sole substantiation for a claim § 1.2(a).
• "Data on file" cannot be the sole substantiation for a safety or efficacy claim § 1.2(b).
• Selecting consistent positive results while neglecting consistent negative ones from a systematic review or meta-analysis "is not appropriate" § 1.2(d).
• Animal or laboratory data "are insufficient to be used as the sole evidence", and where used need a qualifying statement that they "do not necessarily predict clinical effects" § 1.2(i).
• Comparative claims must show statistical significance; the accepted level is p < 0.05, and a result that isn't significant must say so "in full" § 1.2(j), § 1.2(k).
• "Hanging" comparatives — better, stronger, more widely prescribed, than nothing in particular — "should not be used" § 1(h).
• A healthcare professional's name, image or quotation needs their documented approval § 1(g).
• Substantiating data must be retrievable and supplied on request "within 10 working days" § 1(i).

The practical failure is rarely a claim with no reference. It is a reference that doesn't say what the claim says, a citation nobody can open, or a "data on file" that isn't.

4. Carrying the mandatory inclusions

§ 2.1 lists seven elements that "all types of promotional materials, in all media, must include". They're set out in the next section. Around them, § 2 adds that promotional material "must be clearly distinguishable as such" § 2(b), that PI statements, qualifying statements and references "must be clearly legible" § 2(d), and that visible information must be "accurate and consistent with the Code when read in isolation" § 2(e) — the fold, the first slide, the front of the leave-behind each have to stand on their own.

One thing that is not on the list: a job or approval code. Almost every Australian affiliate prints one, and it is good practice, but it is not a § 2.1 requirement — a missing job code is a workflow note, not a mandatory-inclusion failure.

5. Reaching only the audience it's for

Material covered by Section 2 must be "only able to be viewed or accessed by healthcare professionals" § 2(a); online, "a mechanism such as password protection" is what the Code has in mind § 2.2(a). On social media, content the public can see "should not advertise or include promotional claims for a prescription product", and content with claims "must be restricted to a verified healthcare professional audience" § 10.2(a), § 10.2(b). With the public, a company may make the CMI and PI available "in their entire form" but not "amended, abridged or displayed in a promotional manner" § 12(a), and product-specific programs and patient aids go "only to patients already prescribed the product and must not be promotional" § 12(d).

The seven inclusions, in Edition 20's words

Two of these move without anyone touching the material. The PBS status (e) changes when the schedule does; the boxed warning and the PI behind the access statement (c, d) change when the TGA approves a new PI. A piece that passed in March can be non-compliant in June with nothing on the page altered — which is why re-review triggers matter as much as the first review.

How review runs

The Code does not prescribe a review process — it prescribes the standard the material has to meet. What follows is how most Australian affiliates get there.

The reviewers. Review here is not a meeting. A piece leaves its originator, often as an agency draft, and goes to one reviewer at a time: typically Medical (the medical advisor or medical director), Regulatory Affairs, and Legal or Compliance. In smaller affiliates two of those desks are one person. The pieces come from every function, not only the brand team: medical education, patient support programs, events, social posts. The final signatory is usually the Medical Director, whose signature also appears on the Code's own paperwork — corrective letters under § 15.6 are signed by "the Subject Company Managing Director or Medical Director".

The stages. A brief; a draft, usually from the agency; the originator's own pre-review; the piece goes to each reviewer in turn with the references attached, with days at every step, and back to the originator whenever a reviewer returns it; sign-off; the date prepared and the job code go on; the piece goes to market. Then it lives — and every PI update, PBS change, new trial or competitor complaint is a reason to bring it back.

Where it stalls. Every step can bring a different opinion, and every returned piece is another round. The pattern is familiar. One reviewer finds the substantiation gaps — claims the reference doesn't quite carry. The next argues balance. Another finds a mandatory inclusion nobody thought to check, because the template "always had it". Somewhere along the way there is a reference that can't be opened, a "data on file" that isn't on file, or an agency version that drifted from the one the last reviewer read. Two reviewers can read the same claim differently, and the originator is left to reconcile them. None of this is judgment work; it is find-the-error work, done by the most senior medical people in the building, one desk at a time, at the end of the process instead of the start.

What shortens it. The affiliates that get to sign-off in days rather than rounds do a few unglamorous things: they run the seven inclusions and the PI check before the piece leaves the originator's desk, not after the first reviewer finds the gaps; they attach every reference as a file, not a citation; they agree a turnaround for every step, and send to reviewers in parallel where they can; and they keep a register of claims that have already been substantiated, so the same claim isn't re-argued on every piece that carries it.

What a complaint costs

Complaints are heard by the Code of Conduct Committee, and the numbers are in the Code.

A complainant "has the burden of proving their complaint on the balance of probabilities"; anonymous complaints are not accepted; and a complaint from one company against another is not accepted unless intercompany dialogue under the Tool Kit's Intercompany Dialogue Standards "has taken place and the complaint has not been resolved" § 15.1(a), § 15.1(b), § 15.1(c). A non-member company that complains lodges a bond of $20,000 § 15.1(d).

The clock is short. A complaint is acknowledged within five working days; the subject company has ten working days to respond; the committee's decision is notified within two working days of its meeting and its reasons within ten § 15.2(a), § 15.2(b), § 15.2(f). If the committee requires conduct to cease or material to be withdrawn, the company "shall at once comply … pending any appeal" § 15.2(h).

The sanctions in § 15.6 are cumulative: cessation and withdrawal, corrective action, and fines. Corrective letters and advertising are in a form the committee approves and must be completed within 30 calendar days. The schedule of fines is graded by the breach's effect on patients and prescribing:

Fines can be imposed "individually or cumulatively", up to "a maximum of $300,000 per complaint" § 15.6(b), and Medicines Australia may refer an unactioned finding to the TGA or the ACCC and publicise it § 15.6(d). The question worth asking of every piece before it ships is therefore not "would a reviewer pass this?" but "what would this cost if the other side read it?" — because a competitor's medical director will.

A pre-review checklist

The short version. A fuller, shareable one is in the works.

1. Every claim traced to the Australian PI — indication, population, dose, endpoints, safety — with nothing outrunning it § OP 8.
2. Every claim referenced, and the reference actually says what the claim says § 1(a).
3. No reference that is a poster, abstract, personal communication or unpublished data standing alone § 1.2(a); no "data on file" standing alone under a safety or efficacy claim § 1.2(b).
4. Comparative claims show significance; non-significant results say so in full § 1.2(j)–(k); no hanging comparatives § 1(h).
5. For each efficacy claim, risk information with comparable prominence, placement and extent § 1.1(b).
6. The seven inclusions present — brand, AAN beside the biggest brand mark, boxed warning or black triangle, PI-access statement with a means of access, PBS status, supplier name and location, date prepared § 2.1(a)–(g).
7. Everything visible on its own — first page, fold, first slide — accurate in isolation § 2(e); PI statements and references legible § 2(d).
8. The audience restricted to healthcare professionals, online included § 2(a), § 2.2(a).
9. Any quoted or pictured healthcare professional's documented approval on file § 1(g).
10. The substantiation file retrievable within ten working days § 1(i).

Where software fits

A review is judgment work sitting on top of a great deal of checking work, and the checking is what software should take. The bar for any tool Australian reviewers rely on is the Code's own bar: every finding it raises cites the clause and the PI section it rests on, so a reviewer can verify it in seconds; it never invents an inclusion the Code doesn't list; it holds the piece to the Australian PI and the current PBS position, not a global label; it keeps a register of claims and the sources that carried them, so the next piece starts already known; and it tells the people who signed off when a PI change has made a passed piece non-compliant. The signature stays human. That is what arcimedes' Reviewer is built to do, and the Faster approvals workflow is that review, told from the brand manager's desk.

Questions people ask

Does the Code apply if we aren't a Medicines Australia member? Yes. Compliance is a TGA condition of registration "whether the Company is a member or non-member" § Scope.

Do we still need a Minimum PI? Not as a mandatory inclusion. Edition 20 requires a statement directing the reader to the PI, with an immediate means of access § 2.1(d). A voluntary summary may still be used; if it is, it has to be accurate.

Does a reminder advertisement need the inclusions? Yes. § 2.1 applies to "all promotional material for a product, whether or not the material contains a promotional claim".

Can a claim rest on data on file? For prescribing frequency or cumulative exposure, it may; as the sole substantiation for a safety or efficacy claim, no § 1.2(b).

How quickly must we produce substantiation? Within ten working days of a request § 1(i).

What is the most a complaint can cost? Up to $300,000 in fines per complaint § 15.6(b), on top of withdrawal, corrective advertising, and referral to the TGA or the ACCC.