Promoting Cosmetic Procedures and the Duty to Warn of Risks

The Journal of Law and Medicine has just published a new article,¹ which I co-authored with Cameron Stewart. The article’s subject is the Therapeutic Goods Administration’s (TGA) clampdown on cosmetic injectables promotions. That clampdown was announced in an update of March 2024. As we write, the origins of the clampdown lie in the Australian Health Practitioner Regulation Agency’s (Ahpra) review of the cosmetic surgery sector. That review began as a review of only invasive surgeries but soon expanded into a review of all cosmetic treatments, such as botox injections. To explain this process of spillover, we introduce the conceptual notion of ‘scandal creep’ or ‘regulatory creep.’ Although imperfect, we think this works as a way to describe the way in which these regulatory changes can expand or metastasise into other related areas. As we say (footnotes omitted):

… Ahpra’s review of the cosmetic surgery sector, triggered initially by cosmetic surgery scandals in Victoria and New South Wales, ended up “widening to introduce stronger safeguards for the booming non-surgical cosmetic procedures sector?” These included “new guidelines for registered practitioners performing and advertising aesthetic treatments.” Such an expansion may be understood as a process of what we call “scandal creep” or, alternatively, “regulatory creep.” This, we propose, is a process whereby governance deficiencies revealed by one scandal in one area leads to reform, but then also leads to the exposure of regulatory deficiencies (sometimes less obvious than the first, and further reforms, in a related area.’ Although the word “creep” may have negative connotations, we do not mean to suggest that regulatory “creep” is undesirable.

Our newly published article covers a lot of terrain, and looks at the recent history of cosmetic injectables advertising regulation, as well as the history of what is known as DTCA-PM — ie, direct-to-consumer advertising of prescription medicines — in Australia and elsewhere. Of course, only the United States and New Zealand do not prohibit DTCA-PM. In those jurisdictions, prescription medicines can be lawfully advertised — although requirements as to what must be disclaimed or conveyed in those advertisements (eg, the mandatory inclusion of those sped-up warnings at the end of television advertisements) serve as a form of regulatory protection for vulnerable consumers. Indeed, the policy rationale for the liberal approach to prescription medicine advertising in the US and New Zealand is that this kind of ‘controlled advertising’ is preferable to the kinds of unofficial and unlawful advertising that one may receive in Australia (eg, on social media or online forums).

Speaking of ‘vulnerable’ consumers, the paper also deals with the changing understanding of consumers — as either ‘reasonable’ or ‘vulnerable’ — against the backdrop of the TGA’s changing policies. The purpose and history of Therapeutic Goods Act 1989 (Cth) and its related instruments, such as the Therapeutic Goods (Therapeutic Goods Advertising Code) Instrument 2021 (Cth) (‘Advertising Code’), is in some ways fixed, but in others ever-changing. For instance, before 2018, the TGA seemed to consider all consumers of therapeutic goods as ‘vulnerable’ for the purposes of the general advertising regulations. In the article, we set out that history as follows (footnotes omitted):

In 2018, the TGA introduced a range of advertising reforms, including a more expansive legal definition of the word “advertise.” Advertising now includes the making of any statement, picture or design that was intended, indirectly or directly, to promote the use or supply of therapeutic goods. Judicial dicta have confirmed the wide amplitude of the definition, as well as the basis of DTCA-PM under the TG Act to: “ensure that patients decide whether to use prescription-only medicines … based on information from trusted sources such as their medical practitioner”. The Advertising Code further provides that an advertisement is to be viewed objectively in its “total presentation and context” — not by reference to the advertiser’s intention, but by reference to its likely impact on a reasonable person to whom the advertisement is directed.

While current guidance indicates that a “reasonable person” under s 5(2) of the Advertising Code is “an average member of the public with sound judgement”, previous [TGA] guidance indicated the expression should be construed as a “consumer of therapeutic goods who may be more vulnerable than a prospective purchaser of ordinary consumer goods” The previous construction is partly supported by judicial dicta. In a case involving the unlawful advertising of unapproved hyperbaric oxygen therapy devices, Rofe J in the Federal Court of Australia (FCA) accepted that a reasonable person to whom an advertisement was directed might be a “vulnerable non-scientific reader” and that the TG Act “is intended to protect vulnerable consumers by regulating therapeutic goods and the manner in which they are advertised”.

The theoretical understanding of certain people as vulnerable in regulatory states has been developed, most famously, by Martha Fineman.² However, others, such as Terry Carney and Maria O’Sullivan, have also picked up the concept of vulnerability as it can be applied to certain ‘consumers’ (in their work, recipients of social security).³ Although this article does not deal with consumer vulnerability in detail, there are the beginnings of such an analysis. We consider, for instance, the particular vulnerabilities that may face those undergoing cosmetic treatments. Indeed, understandings of these vulnerabilities appear to be emerging from three separate sources: first, from Ahpra’s own investigation of cosmetic treatment complainants; second, from the general psychiatric and medical literature; and, third, from the courts’ recognition of duties to warn held by health practitioners administering certain cosmetic treatments.

Although we did not have space to deal with the third source in this article (and hope to write further on the NSW courts’ recognition of that duty to warn in future work), we do begin to develop analysis of patient vulnerability. As we write (footnotes omitted),

… Ahpra’s announcement provides four (deidentified) case studies based on examples of investigations it has undertaken, thus indicating the kinds of harms Ahpra has encountered. One example involved a patient with “a history of mental illness” and “body image issues” who had previously had 10 cosmetic injections. Ahpra found that the patient’s medical practitioner had failed to manage the patient’s mental health concerns and did not “properly gain consent.” Another patient was harmed by dermal fillers; they had reacted adversely to her pre-existing skin conditions and were required to be dissolved.

A growing body of research finds that health practitioners should be able to identify the psychological risks that some cosmetic procedures pose to some patients. These risks may be most acute for patients with body dysmorphic disorders, cosmetic procedure addiction, polysurgical addiction, dysmorphophobia, and other less recognisable forms of impaired judgment or psychological divergence. However, research also indicates that doctors should also be aware of the non-psychological complications that can arise from repeated cosmetic treatments, such as granuloma formation and filler migration. Although we do not, in this column, consider the duties of practitioners to warn of the inherent material risks of treatments involving cosmetic injectables, we acknowledge that, for some patients, the risk of increased psychological distress could amount an inherent material risk of a cosmetic procedure, which would thus create a duty on the practitioner to warn the patient of that inherent material risk.

As with all health practitioners, it is clear in Australia that a cosmetic practitioner holds a duty to determine which, if any, inherent material risks a patient would be unprepared to accept and, having identified those risks, to warn and advise the patient about those risks. So much is confirmed by the High Court in Rogers v Whitaker⁴ and the judgments in the Wallace v Kam cases.⁵ However, it is also true that the legal analysis of any breach of this duty may be complex, especially if it involves a failure to warn of multiple risks attaching to a single procedure.

That’s why in this post, I thought I’d work through a few of the issues related to the relationship between the promoting cosmetic procedures and the established duty of health practitioners to warn or caution patients about any inherent material risks involved in those same procedures. As it turns out, the relationship is far from clear.

One of the historical curiosities in this area of law — that is, in the area of duties to warn in cosmetic surgery — relates to the notion that cosmetic health practitioners may have a different or more expansive duty than other health practitioners. The idea is that those practitioners offering cosmetic and other so-called ‘elective’ or ‘non-therapeutic’ procedures (ie, health procedures that patients are not compelled by pressing health risks to undertake) may hold a more onerous duty to warn their patients about the inherent material risks of those procedures. More onerous than who? More onerous than the duty held by practitioners offering so-called therapeutic or ‘non-elective’ procedures. For instance, in 1999, when the NSW Minister for Health commissioned the Cosmetic Surgery Report, it included something like this point. It said as follows:

The courts regard the duty to warn as more onerous in cosmetic surgery because patients have a choice about whether to have the procedure. If there is a perceived level of inducement to have a procedure, coupled with failure to warn of risks, the doctor has a greater burden to demonstrate that the patient was properly informed.

Of course, this report was written before the Civil Liability Act 2002 (NSW) (and related legislation in other states) was enacted and before cases such as Wallace v Kam were heard. Another problem is that no citations are given for this principle in the Report and there there appears to me to be very few cases to support the distinction — if any. Any such cases (although I have not found one) would also contradict the ruling of the UK Court of Appeal in Gold v Haringey Health Authority.⁶ There, Lloyd LJ (Watkins and Stephen brown LJJ agreeing) rejected the distinction between therapeutic and non-therapeutic treatments (including ‘a face-lift’) as ‘elusive.’ His Lordship said that distinguishing between advice given in ‘a therapeutic context and advice given in a non-therapeutic context’ would be contrary to established principle.

In Tan v Benkovic (2000)⁷, the NSW Court of Appeal allowed an appeal by a cosmetic surgeon. However, the appeal was narrow: it concerned only the award of exemplary and aggravated damages. Before I address the damages appeal, however, it should be noted that Mason P (in a lead judgment with which Stein and Heydon JJA agreed) laid down a useful principle relating to the distinction between therapeutic and elective procedures. As his Honour the President stated, ‘the line between therapeutic and elective [procedures] — to the extent that it exists at all — is necessarily dependant in part upon the eye of the beholder.’⁸ That principle would be articulated in different words by Mason P again (as discussed below).

In any event, Tan involved a plaintiff who alleged (and had successfully claimed at first instance) that her cosmetic surgeon had failed to warn her of the risks of ‘facial asymmetry, nerve damage and dimpling of the lips.’ Mason P noted the finding of the trial judge that the doctor had negligently failed to inform the patient of the ‘full range of risks capable of attending the procedure.’⁹

The appellant, Dr Tan, submitted that all of the physical and psychological sequelae of the surgery had been compensated for with the award of compensatory damages that trial Judge had made ($30,000). In other words, Dr Tan did not challenge the finding of the trial Judge that there had been a negligent failure to warn. However, the trial Judge had also found, in the course of applying the Rogers test, that the surgeon, Dr Tan, had paid ‘inadequate attention’ to ‘the threshold matter of whether [the patient, Ms Benkovic] was a psychologically appropriate candidate for a cosmetic plastic surgery procedure.’ This was the point that Mason P picked up as the source of error. In this regard, his Honour said as follows:

… I feel bound to record a sense of unease with these aspects of the background findings. After all, the duty of care in Rogers is premised on the notion of the patient’s autonomy, albeit that there are patients and patients and that different proposed medical procedures call forth different matters that ought to be disclosed by the competent medical practitioner.¹⁰

In the above lines, Mason P grounds the duty to warn test articulated in Rogers in autonomy in order to distinguish it — importantly for this discussion of the advertising and promotion of cosmetic treatments — from the countervailing or separate act of inducement and promotion. Needless to say, medical practitioners can induce and promote particular kinds of treatments; however, how do these statements interact with the duty to warn? In the following paragraph, Mason P provides an in-principle answer to that question:

Blandishments of this nature may have been relevant to the now resolved issue whether the respondent would have undergone surgery had the full risks been disclosed, but their presence could not conceivably establish a case for aggravated or exemplary damages if otherwise unavailable in this particular case.¹¹

The answer, then, appears to be that promotional statements could indeed be material to the inquiry into whether a patient had been sufficiently cautioned, warned or informed for the purposes of the test in Rogers. However, that matter had been determined in this case: Dr Tan had breached the established duty to warn; that breach had caused the patient harm; and for that harm the patient had already been compensated via an order of compensatory damages, which the trial Judge had said was compensation for both her psychological and physical harms — including those harms arising by reason of the promotions. When it came to the real matter in the appeal, then — whether the promotions could sound in aggravated or exemplary damages — the Rogers test had no more work to do. As Mason P wrote with respect to the damages questions that comprised ‘the matters presently in issue’: ‘Nothing turns upon whether or not medical advice was accompanied by puffery or exaggerated bedside manner such as statements about making the respondent feel a different person or looking 20 years younger.’¹²

Thus, Mason P went on to find that the patient could not recover aggravated or exemplary damages for ‘disappointment [that] stemmed from her unfulfilled expectations for improvement,’ which was a distinct and separate form of ‘damage’ (if it could be recognised by the law) from that which arises from breaches of the duty to warn. As Mason P’s reasons indicate, the latter damage arises when patients, having not been warned of the material and inherent risks of a treatment, are deprived of the opportunity to make an autonomous decision about whether to hazard the risks of that treatment. Those damages had been proved in this matter and had also already been paid out. Thus, in order to determine whether aggravated or exemplary damages should be awarded, different tests would apply.

In respect of aggravated damages, the ordinary test would involve inquiring into whether the plaintiff has endured any special insult or humiliation by reason of the practitioner’s promotions.¹³ As Mason P found, however, while the plaintiff was ‘undoubtedly upset, vexed and depressed about her dashed expectations,’ this did not mean that Dr Tan’s ‘blandishments’ and ‘promises’ amounted to a ‘proper basis of any legal complaint.’¹⁴ In any event, the parties conceded that aggravated damages were not available in cases of ‘pure negligence’ such as this — a holding that Mason P had incidentally reached in an earlier decision¹⁵ — and so those damages, which were in the amount of $40,000, were set aside without further consideration. Of course, that position is reflected in the statutory bar that was later established by s 21 of the Civil Liability Act 2002 (NSW) and is operative today: in a personal injury cause of action framed in negligence, neither aggravated nor exemplary damages can be awarded.¹⁶

For exemplary damages, by contrast, the real test — and the test which had actually been applied by the trial Judge — was whether there had been a contumelious disregard of the plaintiff’s rights.¹⁷ Another test, seemingly not contemplated at trial (but deriving from the same source), was whether the court should express its opprobrium by way of imposing an order for damages on the defendant.¹⁸ In considering these matters, the trial Judge found that there had been such contumelious disregard and awarded the patient $60,000 in exemplary damages, as follows

… the tort involved was not one of mere inadvertence … it was significantly different from momentary inattention, and … the the blandishments and enticements amounted to contumelious disregard for the doctor-patient relationship obligations the defendant owed to the plaintiff…¹⁹

In essence, however, Mason P disagreed with the trial Judge’s reasons. His Honour said:

In my view the award of exemplary damages in this case must be set aside. Neither the facts as found nor the evidence in the case support his Honour’s conclusion that there was “an egregious error on the part of the defendant in talking the patient into assenting to a facelift operation”. Nor was this a case exposing a “need to stamp the defendant’s conduct with a mark of opprobrium”. Nor was there a “contumelious disregard for the doctor-patient relationship obligations” as adverted to in his Honour’s reasons for the award of aggravated damages.²⁰

As alluded to above, considerations as to whether aggravated or exemplary damages would be awarded in any case involving a failure to warn are likely only academic today, now that s 21 of the Civil Liability Act (NSW) bars the courts from awarding such damages. However, the key dicta that emerges from Mason P’s analysis in Tan is the broader point about the necessary/unnecessary distinction. At one point, Mason P states the law in this regard very clearly:

Whether surgery is ‘necessary’ or ‘unnecessary’ has no direct bearing upon whether it is performed competently or whether pre-surgical warnings satisfied the Rogers v Whitaker duty.²¹

Thus, if as Lloyd LJ’s and Mason P’s reasons indicate, established principle tells against the therapeutic/non-therapeutic or necessary/unnecessary distinction, it may mean that patients seeking unnecessary and non-therapeutic cosmetic treatments and cosmetic surgeries are to be treated by health practitioners in the same way as patients in need of necessary therapeutic medical treatment for a given pathology. Whether or not this state of affairs means that the specific vulnerabilities that may exist in some cosmetic treatment patients (indicated above) are more likely to be subject to a duty to warn of inherent materials risks will probably depend on the patient. Of course, this is consistent with Rogers. Paradoxically, however, this conclusion might have the effect of dispelling the apparent ‘myth’ that still seems to hold in some quarters: namely, that non-therapeutic patients require more exhaustive warnings than therapeutic patients.

Then again, I recognise that the quotation above from the 1999 Report seems to be directed not at the scope of a cosmetic practitioner’s duty per se but rather at the kinds of evidentiary expectations that might fall on a practitioner who seeks to defend a failure to warn claim framed in negligence. Specifically, where there is an accusation of negligence for failure to warn, and the plaintiff has already adduced evidence that discloses promotions and inducement, it is in that circumstance that the practitioner might face more difficulty in disproving any liability for the purposes of the Rogers test. Just how precisely this expanded evidentiary onus meshes with ex ante legal principle relating to the scope and content of the duty to warn about cosmetic treatments remains yet to be explored.

1. Christopher Rudge and Cameron Stewart, ‘Injecting Tighter Regulation: Implications of the TGA’s Clampdown on Cosmetic Injectables Advertising’ (2024) 31 Journal of Law and Medicine 464. ↩︎
2. See Martha Fineman, ‘The Vulnerable Subject and the Responsive State’ (2010) 60 Emory Law Journal 251; Martha Fineman, ‘Vulnerability and Social Justice’ (2019) 53 Valparaiso University Law Review 341. ↩︎
3. See Terry Carney, ‘Artificial Intelligence in Welfare: Striking The Vulnerability Balance? (2020) 46(2) Monash University Law Review 23–51; also see a 2024 themed issue of Public Law with essays on vulnerability in social security, titled ‘Public Law and Vulnerability’: eg, Maria O’Sullivan and Robert Thomas, ‘Public law and Vulnerability’ (2024) 2 Public Law 191. ↩︎
4. Rogers v Whitaker (1992) 175 CLR 479, 490 (Mason CJ, Brennan, Dawson, Toohey, McHugh JJ). ↩︎
5. See Wallace v Kam [2012] NSWCA 82; and Wallace v Kam (2013) 250 CLR 375. ↩︎
6. [1988] QB 481, CA. ↩︎
7. Tan v Benkovic (2000) 51 NSWLR 292 (‘Tan‘). ↩︎
8. Ibid 298 [32] (Mason P) ↩︎
9. Ibid 294 [10] (Mason P). ↩︎
10. Ibid 298 [29] (Mason P). ↩︎
11. Ibid 297 [30] (Mason P). ↩︎
12. Ibid. ↩︎
13. While Mason P gives no citation on the point in Tan, it was in Lamb v Cotogno (1987) 164 CLR 1 that the High Court (at 8) distinguished between aggravated damages and exemplary damages. As their Honours said: ‘Aggravated damages, in contrast to exemplary damages, are compensatory in nature, being awarded for injury to the plaintiff’s feelings caused by insult, humiliation and the like. Exemplary damages, on the other hand, go beyond compensation and are awarded as a punishment to the guilty, to deter from any such proceeding for the future, and as proof of the detestation of the jury to the action itself.’ ↩︎
14. Tan 299 [35]. ↩︎
15. Relying on Hunter Area Health Service v Marchlewski (2000) 51 NSWLR 268. ↩︎
16. Civil Liability Act 2002 (NSW) s 21. ↩︎
17. Tan 300 [42] (Mason P), citing Lamb v Cotogno (1987) 164 CLR 1. ↩︎
18. See my quotation from Lamb v Cotogno (1987) 164 CLR 1 in the note above. ↩︎
19. Tan 297 [24] (Mason P). ↩︎
20. Ibid 300 [43] (Mason P). ↩︎
21. Ibid 299 [33] (Mason P). ↩︎