IN Australia, an estimated 250 000 people per year are admitted to hospital because of harm from medicine use. In 2019, Australia identified medication safety as a national health priority. But are we doing enough?

The Therapeutic Goods Administration (TGA) plays a key part in medicine safety, both in pre-market evaluations to ensure that potential benefit outweighs harm, and through safety oversight of medicines once on the market. The TGA carries out post-market safety oversight (also known as “pharmacovigilance”) by collecting and analysing emergent evidence of harm, issuing safety warnings, and taking regulatory actions to promote safer prescribing, such as changes to product information, boxed warnings, or in the most serious cases, drug safety withdrawals. The TGA collects original data on patients’ experiences with medicines in Australia through its voluntary adverse drug reaction (ADR) reporting system.

The agency also communicates about harms with clinicians and the public, providing advice for safer use when new evidence emerges. In both roles, the TGA works closely with other regulators internationally.

Historic TGA leadership in post-market safety oversight

In the late 20th century and early 2000s, the TGA was considered a leader globally in safety oversight. Australia was a founding member of the World Health Organization Programme for International Drug Monitoring, which commenced in 1968. In a 2009 commentary on pharmacovigilance, Kathlyn Ronaldson described the Australian program of spontaneous reporting as “exceptionally effective”. This comment reinforced the statement in 2002 from Ian Boyd that the “Australian [ADR] reporting system is acknowledged as one of the best in the world”. At that time, Australia’s ADR reporting rate per capita placed it in the top few nations.

In 2001, the TGA received over 11 000 reports, of which 35% were from GPs and 27% from both hospitals and pharmaceutical companies. Reports from health professionals had uncovered a number of significant safety problems over the years. Of particular importance were flucloxacillin-induced hepatitis, amoxycillin/clavulanate-induced hepatitis, and the association of cystitis with tiaprofenic acid. The number and quality of the reports enabled an understanding of the characteristics of the reactions. Using reporters as a major source of cases, case–control studies were completed to identify risk factors.

More recent shift to fewer health professional reports

By 2009, the source of reports to the TGA had shifted considerably, with nearly half coming from pharmaceutical companies, around one-sixth from hospitals, one-third from other sources (states and territories, pharmacists, consumers, specialists) and only one-eighth from GPs. The Panel to Review the Transparency of the TGA in 2011 recognised this change in reporting and recommended that “the TGA more effectively facilitate the recognition and reporting of adverse events by health practitioners and consumers, and promote the adverse event reporting system”.

The percentages (and the absolute numbers) of both health professionals and GPs have decreased while the percentage of reports from companies have increased. This has resulted in a decrease in the quality of reports, as reports from health professionals represent an action arising from a patient–health professional interchange, whereas reports from companies are a second-hand report of such an interchange where detail will inevitably be lost. Follow-up is also straightforward with a report from a health professional, whereas follow-up from a company report is convoluted.

Fast forward to the 2019–2020 financial year (just before the effects of the coronavirus disease 2019 [COVID-19] pandemic), and the TGA received 23 476 reports. The TGA changed its descriptions of sources, making direct comparisons difficult, but over 61% were submitted by pharmaceutical companies and only 20% by health professionals.

In international comparisons, Australia is less likely to warn

Despite Australia’s well functioning ADR reporting system, an international comparison of product information for the 20 top-selling US medicines in 2011 found that 15 had boxed warnings in the United States, nine in Canada, and none in Australia. One example was US and Canadian boxed warnings of increased risks of death in older people with the antipsychotics aripiprazole, olanzapine and quetiapine. These drugs were also widely used in Australia, and Australians were at no lower risk.

The TGA also issues fewer safety advisories: over a 10-year period, from 2007 to 2016, the TGA issued 220, whereas the UK Healthcare Products Regulatory Agency issued 469, the US Food and Drug Administration issued 382 and Health Canada 370. An assessment of numbers of drug-specific safety issues covered in these warnings also found fewer from the TGA than the other regulators.

For example, although the TGA issued warnings about diabetic ketoacidosis and amputation risks with a newer class of diabetes drugs, sodium–glucose cotransporter 2 (SGLT2) inhibitors, it did not issue warnings about fracture risks, severe genitourinary infections or acute kidney injury. In 2019–2020, the UK Medicines and Healthcare products Regulatory Agency published drug safety updates on about 50 safety concerns, only three of which resulted in TGA warnings. The authors of this analysis describe the “near death” of medicine safety updates.

The reasons behind differences between the TGA and other regulators in the decision to warn remain largely elusive. There is one major difference, however, as compared with other regulators: the TGA does not publish “dear health professional communications” (DHPCs) sent by manufacturers to individual clinicians. Other regulators routinely post DHPCs on their websites.

DHPCs are the most common type of post-market safety warnings used in Europe. The TGA does not make DHPCs public, either routinely or on request. A query to manufacturers about whether DHPCs had been issued in Australia for 207 drugs subject to safety warnings in other countries resulted in no information for 170 (82%), with one-fourth of companies refusing to respond and several citing commercial confidentiality. The TGA refused a follow-up freedom of information request, citing too large an administrative burden.

Thus, although the TGA issues fewer safety warnings than other regulators, this partially reflects secrecy about some warnings. From a public health perspective, no safety warning should ever be considered a commercial secret. Doctors can easily miss an initial letter from a manufacturer or may throw it away before reading it, mistaking it for advertising.

No evidence Australians are better served with fewer warnings

Is the TGA simply avoiding “warning fatigue” by issuing fewer safety warnings?

A recent comparison of a systematic sample of safety advisories in five countries, including Australia, tested whether these warnings were effective in reducing harmful prescribing. On average, prescribing rates were reduced by around 6% post-warning in countries with a warning compared with countries without. This is similar in magnitude to other interventions for better quality prescribing. This study suggests that regulatory safety warnings do influence prescribing, and does not support a hypothesis of “warning fatigue” in countries with more warnings. A systematic review of UK safety warnings, similarly, found a substantial effect on prescribing.

Exemplary safety oversight during the pandemic

Whereas key performance indicators such as reports from health professionals, particularly GPs, and publication of safety advisories such as Medicines Safety Update articles have been in decline for several years, the performance of the TGA during the COVID-19 pandemic has been impressive.

Beginning on 3 March 2021, shortly after the roll-out of COVID-19 vaccines started, the TGA initiated a COVID-19 vaccine weekly update. This update was published on a weekly basis for the next 68 weeks and still continues on a fortnightly basis. The update includes a detailed analysis of all three widely used vaccines (AstraZeneca’s Vaxzevria, Pfizer’s Comirnaty and Moderna’s Spikevax), with a particular emphasis on the association of pericarditis with the two mRNA vaccines, Comirnaty and Spikevax, as well as the association of thrombosis with thrombocytopenia syndrome (TTS) with Vaxzevria.

The latest update, published on 3 November 2022, indicates that over 136 000 reports of suspected adverse reactions in association with COVID-19 vaccines have been received, including over 80 000 with Comirnaty, more than 48 000 with Vaxzevria and just over 7000 with Spikevax.

In total, in the most recent 6-month period with full details available, 1 July to 31 December 2021, the TGA received 78 544 ADR reports, over six times higher than in the same period in 2020, undoubtably due to interest and concern about COVID-19 vaccines, the subject of 67 000 of these reports. Consumers submitted almost 20 000 reports, up from 462 in the corresponding period in 2020.

Key barriers to safety oversight need to be addressed

The experience with COVID-19 vaccines highlights the TGA’s capacity to intensively monitor safety and report back to the public and the clinical community. This expertise needs to be extended more generally to post-market medicine safety.

To do so effectively, the TGA likely needs to be better resourced. However, more resources are not the only answer. There is an urgent need for improved transparency of post-market regulation. Not only should the TGA publish the rationale and evidence base for its decisions when issuing a warning or adding important new safety information to product information, it should also make public its safety reviews and decisions not to take action when other major regulators have issued warnings and/or shifted conditions for use of products also marketed in Australia. The decision not to warn when others have done so warrants public discussion.

Lastly, safety warnings should never be treated as confidential business information. The TGA should publish all DHPCs on its website. Secrecy about public safety is unacceptable.

Ian W Boyd, Ian Boyd Consulting, Shoalhaven, NSW, Australia.

Barbara Mintzes, PhD, School of Pharmacy and Charles Perkins Centre, Faculty of Medicine and Health, The University of Sydney.

Disclosures: Barbara Mintzes is an expert witness for Health Canada on a legal case concerning marketing of an unapproved drug product in Canada. Ian Boyd has no relevant disclosures.



The statements or opinions expressed in this article reflect the views of the authors and do not necessarily represent the official policy of the AMA, the MJA or InSight+ unless so stated.

Subscribe to the free InSight+ weekly newsletter here. It is available to all readers, not just registered medical practitioners.

If you would like to submit an article for consideration, send a Word version to

3 thoughts on “TGA’s oversight of medication safety: is it fit for purpose?

  1. Sandy Thomson GovernancePlus says:

    TGA reporting policies and procedures are a mandatory action for NSQHSS accreditation. While good policies and procedures are generally always in place and staff can describe the procedure for reporting it is not common to have evidence of any reports with the exception of larger tertiary facilities. A key point often raised is about the degree of the reaction and whether or not it should be reported. More education on adverse reactions and why reporting is important may increase reporting.

  2. Anonymous says:

    Impossible to provide financial incentive for clinicians to do ADR reporting. This would bias the reports. Post-market surveillance is for the public good. All clinicians have a responsibility, even those in private.

  3. Anonymous says:

    Thanks for an insightful article. Sadly the TGA is inadequately resourced – both the number of staff and their level of experience – to do this properly. Also in my opinion – far too easy to get drugs onto the market – big pharma pushing hard and difficult to say “no” if fda or ema already approved.

    Also as a clinician – not enough time and no financial benefit to reporting an adverse event – might work in public sphere – won’t happen in private

Leave a Reply

Your email address will not be published. Required fields are marked *