COVID: OFF-LABEL

Medsafe approved Pfizer’s booster at six months. New Zealand moved to five, then to four, and finally, three. The more interesting story is how the law changed to allow it.

Photo by Fusion Medical Animation on Unsplash

The Government has now responded to Phase Two of the Royal Commission of Inquiry into COVID-19, released in March. It will not proceed with standalone pandemic legislation, preferring a review of the Health Act to balance public health, the economy and the rights of New Zealanders.

Before designing the next framework, it’s worth considering how some aspects of the last one worked. Notably, the story of the booster interval shows how quickly a medical decision can become a programme decision.

In late 2021 Medsafe assessed Pfizer’s evidence and approved its booster at least six months after the primary course. How boosters were deployed was then for the Government. The difficulty arose when the programme wanted to operate outside Medsafe’s approved terms.

New Zealand’s technical advisers, the COVID-19 Vaccine Technical Advisory Group (CV TAG), recommended shortening the booster interval to five months. Four days later, the Government announced four. Six weeks after that, it was three.

This is not an argument that three months was wrong. Australia and Britain were shortening intervals too, and there were real epidemiological reasons to boost people sooner. But as New Zealand moved from six months to five, then four, then three, Medsafe’s approved six month interval never changed.

What changed was the decision-making machinery. A judgment that began inside the medicines-regulatory process as an assessment of safety and efficacy became a programme judgment, balancing clinical risk against population benefit, operational simplicity and public communications.

On 8 November 2021, Medsafe updated the provisional approval for the Pfizer vaccine to state: ‘a booster dose of Comirnaty may be administered intramuscularly at least 6 months after completion of the primary course in individuals aged 18 years of age and older’.

On 10 November, CV TAG recommended that same six-month interval.

But as Omicron emerged, the COVID-19 Vaccine and Immunisation Programme asked CV TAG to reconsider its recommendation in order to shorten the booster interval.

On 17 December, despite noting that evidence about the advantages of early boosters was still accumulating, that there were no long-term safety data for early boosters, and that Medsafe had authorised boosters only from six months, CV TAG recommended five months.

Three days later, on 20 December, Cabinet agreed to reduce the booster interval to four months with effect from 5 January 2022. It was announced the following day.

Ministry papers reveal its source. An aide-mémoire for the 20 December Cabinet meeting described the proposed four month interval as “the Director-General of Health’s advice” and recorded that the Vaccine Ministers had agreed to put Dr Ashley Bloomfield’s proposal to Cabinet.

An implementation memo recorded that the programme had recommended five months after weighing “science advice, supply and operational impacts”, before Bloomfield proposed four. The shorter interval would lift the number eligible by 5 January from about 480,000 to 1.19 million, and allow Māori vaccinated later in the rollout to be boosted before winter.

Because four months fell outside the Medsafe-approved data sheet, the shorter interval could not simply be rolled out through the existing vaccination workforce. An authorised prescriber, such as a GP, could prescribe an off-label dose under section 25 of the Medicines Act, but most COVID-19 vaccinators were not authorised prescribers. Absent a Medsafe-approved change, a statutory mechanism was therefore required if the four month interval was to operate as a national programme rule without individual prescriptions.

The shorter interval was therefore implemented using an Immediate Modification Order made under section 14 of the Epidemic Preparedness Act 2006 - an exercise of what is commonly described as a Henry VIII power, because it enabled the Executive by Order in Council to modify the operation of primary legislation without first obtaining a further Act of Parliament.

This particular Order in Council modified section 25 so that a third dose could be administered after a four month interval. It also provided that the dose did not require a prescription and was not a new medicine for the purposes of its administration. The Director-General could specify by notice who could administer or procure it.

At the same meeting, Cabinet agreed booster requirements for specified MIQ, border, health and disability workers. For them, the interval helped to determine when another dose became a condition of remaining in work.

On 25 January 2022, CV TAG discussed calls for a three month interval but did not recommend a blanket reduction. It noted that immunity did not wane at the same pace for everyone and argued that any change “should not be applied to all” but should turn on age and risk: those over 50 or with high-risk conditions were more likely to benefit, with a lower threshold for Māori and Pacific peoples given their increased risk of hospitalisation, severe disease and death. The group also noted emerging evidence that a shorter interval might produce lower effectiveness against symptomatic infection than a longer one. Recommendations were to be drafted and brought back to CV TAG for approval.

A week later, on 1 February, CV TAG’s formal advice recorded that: “The Director-General has requested CV TAG’s advice about shortening the interval between the primary course and booster dose to 3 months.”

CV TAG again recorded that Medsafe had approved six months, and noted limited data on adverse events at intervals shorter than five or six months, the Pfizer and V-Safe evidence having involved longer ones. Its observations were careful: there was no evidence that a three-month booster produced a lower immune response than one at four or six months, and further follow-up was needed to determine whether the shorter interval had any negative effect on the secondary immune response. It nevertheless recommended three months.

Its reasons included earlier protection against hospitalisation from Omicron, particularly for older adults, who made up most hospitalisations, and that five Australian jurisdictions had already moved to three months amid “unprecedented strain on hospitals”.

The following day the Government made a new Immediate Modification Order, which came into force on 3 February, giving legal effect to a three month interval.

Cabinet later recorded that the Order permitted a booster at three months “rather than six months as is recommended in the approved data sheet.”

The Ministry’s Regulatory Impact Statement (RIS) described the earlier reductions as “off-label changes.”

Medicines are, of course, very commonly prescribed off-label every day in New Zealand. A GP or other medical practitioner can lawfully depart from a data sheet for an individual patient after discussing the risks and benefits, and obtaining informed consent. The emergency orders, however, moved that decision from a practitioner making a clinical decision relating to an individual patient to a centrally determined rule for classes of people.

That shift did not remove the clinical screening and informed-consent obligations that continued to apply at the point of vaccination. Medsafe guidance does, however, specifically provide that where an approved medicine is used off-label, the consumer should be told that the use is unapproved. The Health and Disability Commissioner has also treated disclosure of off-label status as capable of being required by the consumer’s right to be fully informed under Right 6 of the Code of Health and Disability Services Consumers’ Rights.

By early May 2022, as officials planned the rollout of fourth doses, the limits of the temporary arrangements had become unavoidable. Medsafe’s data sheet did not provide for a fourth dose, the groups under consideration numbered roughly 834,000, and Pfizer indicated that it did not intend to apply for consent for fourth doses in New Zealand or elsewhere, instead encouraging countries to find their own legal routes.

The programme considerations were unusually explicit. The Regulatory Impact Statement said that “alongside easy and equitable access to vaccinations, trust and confidence in the system and a good customer experience are also essential to encourage uptake”. People had “come to expect a certain experience”, and “adding additional steps into the customer journey could undermine the trust and confidence gained.”

The RIS added that the amendment would let the Ministry react quickly to future changes in dose or frequency. The object was therefore broader than the fourth dose: it was to preserve the mass-vaccination pathway whenever future changes would otherwise require additional clinical or prescribing steps.

Standing orders were rejected because they could not be used for what the Act treated as an unconsented ‘new medicine’. Officials put it bluntly: treating a fourth dose as other than a new medicine “undermines the scheme and purpose of the Medicines Act and the consent process under the Act.”

The solution was adding a new section 34A to the Medicines Act, placing the statutory power to make that programme-wide off-label judgment in the hands of the Director-General.

The Ministry described the solution as sitting “outside of the established Medsafe regulations of medicines process” while providing “an enduring and sound legal basis for the provision of any further doses of COVID-19 vaccines”.

Section 34A allowed the Director-General, by notice, to authorise administration of a consented COVID-19 vaccine otherwise than in accordance with its approved data sheet, including in relation to different populations, number of doses and intervals.

The amendment had a further consequence. Section 20 of the Medicines Act prohibits the sale, distribution or advertising of medicines outside the scope of the statutory consent regime. Medsafe’s position, as set out in its guidance, is that section 20 prevents the promotion of any off-label use of an approved medicine. Section 34A therefore provided that a COVID-19 vaccine would not be treated as a “new medicine” for the purposes of section 20 merely because the Director-General had authorised an off-label use.

In practical terms, the amendment removed the section 20 restriction on promoting an off-label use once the Director-General had authorised it under section 34A.

That was not an abstract problem. Vaccination promotion was already central to the programme. DPMC said its eight-day campaign featuring contestants from The Chase had been deliberately timed to coincide with the reduction of the booster interval from four months to three, which made another million New Zealanders eligible. The campaign pre-dated section 34A and ran under the temporary emergency regime. But it’s a reminder that the permanent legislation was passed at a time when the three-month off-label interval was being administered at scale and actively promoted as part of a national vaccination campaign.

The Ministry was candid about the regulatory departure. The key risk, it said, was that the amendment “would not be in keeping with the scheme of the rest of the Act”. It noted that the amendment “singles out COVID-19 vaccines from a regulatory regime that is intended to set requirements for the safety, quality and efficacy of all medicines.” Officials proposed monitoring the power to ensure that its ongoing use did not undermine ordinary consenting processes.

The Bill moved quickly. Treasury exempted the proposal from a full Regulatory Impact Statement on the grounds that it was needed urgently to mitigate the short-term impacts of the emergency, conditional on a streamlined RIS which “has not been formally quality assured as per the usual RIA [Regulatory Impact Analysis] requirements, given the time constraints”.

No public consultation was undertaken. The Bill was introduced on 7 June, referred to the Health Committee with a 20 June report-back, and passed its remaining stages on 21 June.

On 22 June 2022, the day before the new mechanism came into operation, CV TAG advised on second boosters. It recorded that Pfizer had not yet submitted an application and that “these recommendations require Medsafe approval.”

Medsafe approval for the fourth dose did not follow. The statutory solution did.

CV TAG found insufficient evidence for a broad recommendation of a second booster for healthy people under 30 and healthy pregnant women, and stated it would not support any further mandates.

It also deferred any specific recommendation for healthcare workers who did not otherwise fall within the recommended groups, noting that the data on benefits for healthcare workers remained marginal and that there was no evidence in the available New Zealand data that healthcare workers, particularly if young and without comorbidities, had a higher risk of acquiring or transmitting infection at work. That advice is reproduced in the Director-General’s section 34A decision memorandum.

Officials nevertheless proposed two additions to CV TAG’s list. One was healthcare workers aged 30 and over, relying on a different data point: higher observed infection rates among them compared with border workers. The other was everyone aged over 50.

The technical committee had provided its advice; officials brought the wider programme considerations to bear; and the Director-General applied the section 34A test to the specified groups, having regard to therapeutic value and risk of injury and being satisfied that vaccination was an appropriate measure for managing the risks associated with COVID-19. The surrounding programme advice also considered equity, accessibility, workforce and health-system pressure, simplicity, uptake and even vaccine stocks: the RIS noted that higher fourth-dose uptake would help use supplies “that otherwise could expire before they are used.”

By February 2023 the immediate Omicron emergency had passed and New Zealand had local data. A self-controlled case-series analysis found that the increase in pericarditis risk after the second Pfizer dose was greatest among children and youths aged 5 to 19.

CV TAG noted that these results suggested a longer minimum dosing interval, potentially twelve months rather than six, might be warranted for younger individuals. It also continued to note that myocarditis risk from mRNA vaccination is highest in young adults, particularly males.

The committee explicitly considered customising schedules by risk category:

“General guidelines could be customised for broad risk categories of the population to account for variations in immune responses to vaccines and infection among population groups. However, multiple schedules can be confusing and challenging to communicate clearly. Simple vaccination eligibility criteria and recommendations will enhance communication with the public.”

Yet the recommendation CV TAG made was a single interval - second boosters from six months after the previous dose - together with a further recommendation that dose intervals be “interchangeable and consistent across vaccine schedules to simplify vaccination regimes.” The differentiated model - shorter intervals for higher-risk groups, longer for lower-risk groups - was instead placed under a section called Future Considerations.

This was the second time the committee had contemplated a stratified schedule before ultimately supporting a single uniform interval. The same possibility had been canvassed in January 2022, a week before the three-month recommendation.

Today, the Comirnaty formulation used in New Zealand is not the 2021 product. Comirnaty LP.8.1 received Medsafe consent in October 2025. Its data sheet provides that, where someone has previously been vaccinated, the vaccine should be administered at least three months after the most recent dose.

Three months is therefore now within the regulator-approved terms. That does not retrospectively prove the 2022 judgment was optimal; it is a different formulation for a population with a different immune profile. It does, however, indicate that a three month minimum was not inherently extraordinary.

What is more interesting is what happened to the programme itself.

Current guidance is highly stratified. A healthy pregnant woman may have another dose from twelve months; one with a high-risk pregnancy or underlying condition, from six. Higher-risk children and adolescents aged 5 to 17 are generally on twelve months, severely immunocompromised people six-monthly, healthy 16 to 29-year-olds a single additional dose from six months. The guidance records that myocarditis incidence was highest in adolescent males after a second mRNA dose, and that longer intervals reduce adverse events.

Beneath the interval tables sits a small qualification: recommended spacing can be altered where there is clinical need, “or to facilitate mass vaccinations.”

There, in five words, is the tension that ran through New Zealand’s response to Covid: the dosing interval appropriate to the clinical characteristics of the person receiving the vaccine may not always align with the optimal dosing interval for a population-wide vaccination programme.

The Royal Commission examined Medsafe’s approval process, but not what happened when the programme moved outside it. The wider vaccine regulatory system was expressly out of scope of Phase Two, and neither phase appears to have considered the off-label authorisation or advertising issues.

In any event, New Zealand did not simply shorten its booster interval from six months to three. It changed the legal route by which programme-wide use outside Medsafe’s approved terms could be authorised and, ultimately, placed that off-label judgment in the hands of the Director-General under a broad statutory test.

It also changed the advertising rules. Once the Director-General authorised an off-label use under section 34A, section 20 no longer prevented that use from being promoted merely because it was outside Medsafe’s approved terms.

The advertising issue has not disappeared. In 2025, when considering new medicines legislation, Ministry officials proposed retaining the general prohibition on promoting both unapproved medicines and off-label uses, while allowing the Director-General to authorise such promotion in particular cases. For unapproved medicines, officials proposed that any authorised promotion should clearly disclose that the medicine had not been approved in New Zealand. They proposed that promotion of off-label uses should be treated in the same way, suggesting that the absence of Medsafe approval for the particular use would likewise need to be made clear to the public.

Officials also proposed a separate exemption for public health campaign statements approved by the Director-General, which would not be treated as advertising at all. They said this would permit campaigns “which are currently not permitted, such as immunisation campaigns involving unapproved medicines.” The papers do not make clear whether a public health campaign operating under that exemption and promoting an off-label use would itself be required to tell the public that the use had not been approved by Medsafe.

The underlying authorisation regime is also being redesigned. In April 2026, Cabinet agreed that the proposed Medical Products Bill should enable off-label medicines use through regulations and standing orders. Ministry advice specifically proposed that off-label vaccination without an individual prescription should be enabled through regulations, with the use based on clinical evidence and history of use. It is proposed that legal responsibility would rest with the person or entity in charge of the vaccination programme.

The current Government has promised that, in a future pandemic, key advice behind decisions affecting people’s rights will be released within five working days. That is an important safeguard. But transparency does not answer the harder question exposed by the booster rollout: who should have the power to authorise population-wide use outside Medsafe’s approved terms, and what safeguards should apply when they do?

Nor does it answer a more specific question: should an off-label use authorised by the Director-General for a population-wide vaccination programme then be capable of becoming a condition of employment?

Sir Ashley Bloomfield declined to comment, citing his current role as a chief executive in the wider public sector and noting that he had been interviewed at length during both phases of the Royal Commission.

The Ministry of Health and Associate Professor Helen Petousis-Harris, a member of CV TAG during this period, were also given the opportunity to comment but did not respond.

For the avoidance of doubt, this article does not express any view on the previous or current Government’s actions or policies in relation to Covid and/or future pandemic planning.