FEDERAL COURT OF AUSTRALIA

Reckitt Benckiser (Australia) Pty Ltd v GlaxoSmithKline Australia Pty Ltd [2018] FCAFC 138

ORDERS

THE COURT ORDERS THAT:

1.    The appeal be dismissed.

2.    The parties have leave to file submissions on costs to be determined on the papers.

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

THE COURT:

Introduction and background

1    The primary judge found that, by publishing certain advertising material, the appellant, Reckitt Benckiser (Australia) Pty Ltd (Reckitt) engaged in conduct that was misleading or deceptive, or likely to mislead or deceive, in contravention of s 18 of the Australian Consumer Law (ACL) and made false representations in contravention of ss 29(1)(a) and (g) thereof.

2    The advertising material comprised print advertising (including in store advertising) and a television commercial used in a comparative advertising campaign in which Reckitt compared its product Nurofen and its active ingredient ibuprofen with a competing product, Panadol and its active ingredient paracetamol (also called acetaminophen). The respondents, GlaxoSmithKline Australia Pty Ltd and GlaxoSmithKline Consumer Healthcare Australia Pty Ltd (Glaxo), are the suppliers of the competing product.

3    The primary judge found that representations to the following effect were made:    

(a)    Nurofen, and thus ibuprofen, when taken in the recommended dose and as directed by the manufacturer, generally delivers faster and more effective relief from pain caused by common headaches including tension-type headaches (TTH) than Panadol or paracetamol.

(b)    For the reasons stated in (a), Nurofen is better than Panadol and paracetamol for the treatment of common headaches including TTH and performs in a superior manner to Panadol and paracetamol in delivering pain relief for such headaches.

(c)    At the time when Reckitt conducted its comparative advertising campaign (August to December 2015), there was a current adequate foundation in scientific knowledge to support the representations set out in (a) and (b).

4    These are not the precise representations formulated by Glaxo in its Amended Fast Track Statement dated 15 October 2015, and accepted by Reckitt in its Fast Track Response dated 22 October 2015. However, the parties take no issue with the primary judge’s summation.

5    The parties accepted before the primary judge that the question to be decided was whether the claims made in the comparative advertising campaign could be supported by an adequate foundation in scientific knowledge current throughout the period August to December 2015.

6    Reckitt argued that there was an adequate foundation. It relied on the results of a study published as Schachtel BP, Furey SA, and Thoden WR, Non prescription Ibuprofen and Acetaminophen in the Treatment of Tension­Type Headache (1996) 36 J Clin Pharmacol 1120–1125 (the Schachtel Study). The Schachtel Study is a report on a clinical trial conducted in 1996 which concluded:

The results of this study demonstrate that both ibuprofen at 400 mg and acetaminophen at 1000 mg are effective in the treatment of muscle contraction headache and that a single dose of ibuprofen at 400 mg is significantly more effective than acetaminophen at 1000 mg in the treatment of muscle contraction headache.

7    Glaxo accepted that the Schachtel Study was a properly designed and properly conducted clinical trial. The expert witness engaged by Glaxo—Professor Somogyi—and the expert witness engaged by Reckitt—Professor Dworkin —agreed that the Schachtel Study provided:

… convincing evidence for the superiority of ibuprofen 400 mg over paracetamol 1,000 mg in TTH, with differences that can be considered clinically meaningful.

8    Reckitt also relied on the fact that no other clinical trial proved the opposite of the results reported in the Schachtel Study or put those results in doubt. It further relied on its success before two panels of the Australian Self Medication Industry, which rejected the substance of Glaxo’s present complaints in 2014.

9    For its part, Glaxo relied on other scientific evidence, namely two clinical trials referred to by the primary judge as Study NL9701 and the NCT Study, and certain meta-analyses, namely Moore RA et al “Evidence for Efficacy of Acute Treatment of Episodic Tension-Type Headache: Methodological Critique of Randomised Trials for Oral Treatments” (2014, 155 Pain 2220) (the Moore Study) and two analyses undertaken by Cochrane et al in 2015 and 2016 (Cochrane 2015 and Cochrane 2016, respectively). A meta-analysis is a review which pools data across all available studies to report an outcome based on all those studies. Professor Somogyi gave evidence that this process minimises the weight placed on one particular study and gives appropriate weighting to all appropriate available data. Professor Dworkin agreed that, because they synthesise the results from all relevant studies, well-conducted meta-analyses are generally more informative than single clinical trials. He also agreed with Professor Somogyi that, in general, the Cochrane systematic reviews and meta-analyses are a gold standard for evaluating and comparing medical efficacy and safety.

10    Study NL9701 was commissioned by Boots Healthcare International (Boots) (which was later subsumed into Reckitt). The objectives of the study were stated as follows:

The primary objective of this study was to evaluate the analgesic effectiveness at 4 hours after administration of ibuprofen (Nurofen™) 400mg compared with commercially available paracetamol 1000mg and placebo for the treatment of tension-type headache. In addition, the study aimed to evaluate the analgesic effectiveness at 2 and 6 hours after administration, the recurrence rate of headache during the 24-hour period of dosing and the incidence and severity of adverse events associated with these three treatment regimens.

11    Professor Somogyi and Professor Dworkin agreed that Study NL9701 was well-designed, properly conducted and robust.

12    The primary judge recorded the study’s conclusion as follows:

There were no statistically significant differences between ibuprofen 400mg and paracetamol 1000mg, although a slight trend in favour of paracetamol was observed in the majority of efficacy parameters.

13    At [186], the primary judge concluded:

It was common ground and I find that Study NL9701 did not replicate the results of the Schachtel Study. On the other hand, Glaxo accepted and I also find that the results of Study NL9701 did not invalidate or impugn the results of the Schachtel Study. As Glaxo submitted, the results of Study NL9701 cannot be set aside or discounted. In any consideration or assessment of the relevant state of knowledge in science as at August 2015, the existence of Study NL9701 and the conclusion reached by the authors of that study need to be taken into account.

14    Study NL9701 was not published. The primary judge found that the decision not to publish was made by either Reckitt or Boots.

15    The NCT Study was conducted by Glaxo in relation to the treatment of episodic TTH comparing placebo, paracetamol plus caffeine, a marketed formulation of paracetamol, and a marketed formulation of ibuprofen. The study was terminated apparently due to “unforeseen difficulties with subject recruitment”. The primary judge found that the NCT Study did not support the results of the Schachtel Study. However, as was the case with Study NL9701, it did not support the opposite proposition either.

16    The Moore Study is a meta-analysis that was in existence and known to Reckitt when it commenced its comparative advertising campaign. The study included the Schachtel Study and Study NL9701.

17    The primary judge recorded the result of the Moore Study as follows:

It was possible to pool data from trials for some drugs (paracetamol 1000mg, ibuprofen 400mg, ketoprofen 25mg and aspirin 1000mg) for some of these three outcomes. Table 3 shows the amount of data available for pooling, the mean event rates, and risk ratio and NNT’s calculated by pooling. There was little difference among the drugs, and NNT’s were high, at about 9, for the IHS-preferred outcome of pain-free at two hours.

18    Professor Dworkin acknowledged that the results of the Moore Study did not support a difference between ibuprofen and paracetamol at the 2 hour pain-free endpoint.

19    The meta-analysis referred to by the primary judge as Cochrane 2015 (Derry S, Wiffen PJ, Moore RA, Bendtsen L Ibuprofen for acute treatment of episodic tension-type headache in adults The Cochrane Database of Systematic Reviews 2015, Issue 7. Art. No.: CD011474. DOI: 10.1002/14651858. CD011474.pub2.) was part of the body of scientific knowledge known as at August 2015. The Cochrane Collaboration is a not-for-profit organisation that carries out systematic reviews of healthcare interventions. The reviews are deposited in a database and are used as a reference source, including by health authorities such as, in Australia, the Pharmaceutical Benefits Advisory Committee.

20    Cochrane 2015 was based on 12 studies, including the Schachtel Study and Study NL9701. The authors concluded:

While ibuprofen 400 mg is one choice for treatment of episodic TTH, it may be that higher doses or different formulations might be better, but evidence on this is lacking. Ibuprofen 400 mg is probably not much different from any other treatment, based on what is known.

21    In comparing active treatments, the authors concluded:

No authoritative comparisons between active treatments is possible in the present state of knowledge.

22    Cochrane 2016 (Stephens G, Derry S, Moore RA, Paracetamol (acetaminophen) for acute treatment of episodic tension-type headache in adults Cochrane Database of Systematic Reviews 2016, Issue 6. No.: CD011889. DOI: 10.1002/1465858.CD011889.pub2) was based on 23 studies, including the Schachtel Study, Study NL9701 and the NCT Study. The primary judge referred to Cochrane 2016 even though it was not part of the body of scientific knowledge available as at August 2015. It seems that his Honour did so for completeness—Cochrane 2016 having been introduced into evidence via an affidavit made by Professor Dworkin.

23    With regard to the comparison between paracetamol and ibuprofen, the authors in Cochrane 2016 concluded:

… Based on very limited information, there was no difference between paracetamol and ibuprofen for pain free at two hours (low quality evidence), but a statistical difference was found for pain free at four hours (very low quality evidence).

24    As in Cochrane 2015, the authors concluded:

No authoritative comparisons between active treatments is possible in the present state of knowledge.

25    Before the primary judge, Glaxo submitted that, given the various studies forming part of the body of scientific knowledge known as at August 2015—which do not speak with one voice—Reckitt’s conduct in simply citing and reproducing the Schachtel Study in support of the representations it had made was misleading or deceptive or likely to mislead or deceive. It also submitted that the representations were false.

26    Reckitt submitted that it was not misleading or false to conduct its campaign in the way it did because the Schachtel Study provided support for its claims and no other study, whether a clinical trial or meta-analysis, suggested that the results of the Schachtel Study were invalid or unsupportable.

27    The primary judge accepted Glaxo’s submission:

In my view, it is misleading or deceptive or likely to mislead or deceive consumers in Australia for Reckitt to claim that ibuprofen (Nurofen) provides faster and more effective relief from pain caused by common headaches including TTH than does paracetamol (Panadol) when the only study which supports such a clear cut claim is the Schachtel Study and where the balance of the scientific knowledge is as I have explained it above. By making such representations, Reckitt also contravened s 29(1)(a) and s 29(1)(g) of the ACL.

28    The primary judge rejected Reckitt’s submission, saying:

… I think that that is too simplistic a view of the relevant body of scientific knowledge. Study NL9701 and the NCT Study did not replicate the Schachtel Study results and the meta-analyses, after appropriate weighting and application of systematic review principles, did not support the results of the Schachtel Study. The Cochrane meta-analyses concluded that no claim of superiority could justifiably be made by ibuprofen over paracetamol in the present state of scientific knowledge.

The appeal

29    Reckitt’s notice of appeal contains two grounds. First, the primary judge erred in failing to find that there was an adequate foundation in scientific knowledge for the representations that had been made in light of evidence, as to scientific knowledge, that was not addressed in his Honour’s reasons for judgment (Ground 1). Secondly, the primary judge erred in finding that the representations were misleading or deceptive or likely to mislead or deceive and/or false in contravention of ss 29(1)(a) and (g) of the ACL (Ground 2).

30    Ground 1 was developed, particularly in oral submissions, by reference to the following passage in Professor Somogyi’s cross-examination:

You agree that the Schachtel study was a well-designed and conducted study?---Yes.

You agree that the results of the Schachtel study remain valid, despite the studies NL9701 and NCT not coming to the same results?---Yes.

You agree that Schachtel remains part of the current body of scientific knowledge in assessing the efficacy of ibuprofen versus paracetamol for the relief of pain from tension-type headaches and common headaches?---Yes.

You agree that the Schachtel study provides convincing evidence of the superiority of ibuprofen 400 milligrams versus paracetamol 1000 milligrams for the relief of tension-type headache with differences that are clinically meaningful?---Yes.

Would you agree, therefore, that there is a foundation in the body of science concerning the efficacy of ibuprofen and paracetamol for the relief of tension-type headache or common headaches to make statements based upon the results of the Schachtel study?---Yes.

31    In substance, Reckitt submitted that the absence of any reference to this part of Professor Somogyi’s cross-examination in the reasons for judgment reveals an error in the primary judge’s process of fact-finding. Put simply, Reckitt submitted that the omission of any reference to this specific evidence shows a failure on the part of the primary judge to examine all the material relevant to the question in issue: Waterways Authority v Fitzgibbon [2005] HCA 57; (2005) 221 ALR 402 at [130] per Hayne J.

32    We do not accept that the primary judge’s reasons reveal any such error in fact-finding. It is clear beyond reasonable argument that the primary judge accepted that the Schachtel Study was a well-designed and conducted study that remained valid and part of the current body of scientific knowledge concerning the efficacy of ibuprofen and paracetamol for the relief of common headaches including TTH, notwithstanding the fact that Study NL9701 and the NCT Study did not report the same results: see at [167] – [168]; [174]; [177]; [186] and [188].

33    There is nothing in the primary judge’s findings or reasons which reasonably suggests that, in light of that acceptance, the primary judge did not also accept that the Schachtel Study, considered alone, provided a foundation in science for the representations that had been made. Indeed, the primary judge held as much at [148] – [150] where he reasoned that Reckitt’s reliance, inter alia, on the Schachtel Study and the circumstance that no clinical or scientific study had demonstrated that its conclusions were incorrect or unsupportable, demonstrated that it had satisfied the evidentiary requirements of s 4(2) of the ACL, concerning misleading representations with respect to future matters. Therefore, the fact that the primary judge did not specifically refer in his reasons to the quoted passage from Professor Somogyi’s cross-examination is without consequence.

34    We add the following observations. First, the question before the primary judge was whether, by making the representations, Reckitt had engaged in conduct that was misleading or deceptive, or likely to mislead or deceive, contrary to s 18 of the ACL, and whether the representations were false or misleading, contrary to ss 29(1)(a) and (g). As the primary judge correctly recognised at [153], the two experts could provide assistance to the Court in understanding the relevant scientific material, but they could not decide “the ultimate question”. Therefore, Professor Somogyi’s “concession”, if it truly be such (see below), could not rule the day. It was for the primary judge to evaluate whether, seen in the totality of the evidence, Reckitt’s conduct fell afoul of the prescriptive standards in s 18 and ss 29(1)(a) and (g) of the ACL.

35    Secondly, Professor Somogyi’s evidence in the quoted passage must be seen in its proper context. It does not reveal any real concession at all. In his evidence in chief, Professor Somogyi stated unequivocally that the Schachtel Study should be considered as part of the body of scientific evidence. However, his point was that using the Schachtel Study in isolation, and not taking into account other statistically valid and accurate results, was not scientifically acceptable. In the Joint Report prepared with Professor Dworkin, Professor Somogyi again emphasised that it was important to recognise that clinically meaningful results should be based on the totality of the best evidence and not simply on the results of individual studies. When Professor Somogyi’s evidence is seen in this context, it is clear that the cross-examiner was pushing at an open door when he obtained Professor Somogyi’s acknowledgement that the Schachtel Study provided a foundation in science for the representations that had been made. But this acknowledgement did not signify a departure from the evidence that Professor Somogyi had already given. What the cross-examiner did not ask was whether Professor Somogyi was prepared to depart from his evidence that the Schachtel Study should not be considered in isolation from the totality of all the available scientific evidence.

36    Ground 2 is closely related to Ground 1. Reckitt’s principal submission in relation to this ground is that, when considering the issue of whether there is an adequate foundation in scientific knowledge for the impugned representations, the Court is not concerned with determining— and ought not to determine— “the exact status and strength of the totality of the science on the superiority of ibuprofen over paracetamol…”. Reckitt submitted that the question whether there is an adequate foundation in scientific knowledge is answered by the Schachtel Study. According to Reckitt, the Schachtel Study alone provides that foundation.

37    In written submissions, Reckitt surveyed the findings of, and evidence given by the two experts in relation to, Study NL9701, the NCT Study, and Cochrane 2015 and 2016. The submissions omitted reference to the Moore Study. Once again, Reckitt drew attention to the fact that none of the subsequent studies (to which it referred) negated the findings of the Schachtel Study. Relatedly, Reckitt canvassed the possibility—for that is all it was in the evidence—that Study NL9701 and the NCT Study may have produced “falsely negative” or “false negative” results (meaning that, for some unknown reason, they failed to establish the claimed superiority of ibuprofen over paracetamol in the treatment of common headaches, including TTH). Further, Reckitt pointed to some parts of Professor Somogyi’s cross-examination where he accepted that he would not regard the results of the NCT Study with the same level of comfort as the results of the Schachtel Study. Reckitt also submitted that the reference in Cochrane 2016 to the evidence comparing paracetamol to ibuprofen being “low quality” at the 2 hour pain-free endpoint and “very low quality” at the 4 hour pain-free endpoint, did not mean that the actual efficacy of ibuprofen compared to paracetamol, as found in the Schachtel Study, was itself based on low quality or very low quality evidence.

38    These snippets of evidence must be seen in the overall context that, in carrying out the meta-analyses, Cochrane 2015 and Cochrane 2016 gave weightings to the quality of each study included in the analysis. For example, Study NL9701 was given a rating of 4/5 in the Cochrane reviews, whereas the Schachtel Study was given a rating of 5/5. However, the two experts agreed that this rating difference made no difference to the determination of the question in issue in the proceeding.

39    Reckitt’s submissions must also be seen in light of the acceptance by the two experts that the Cochrane reviews and meta-analyses are a gold standard for evaluating and comparing medication efficacy and safety.

40    Reckitt did not contend that any of the conclusions expressed in either Cochrane 2015 or Cochrane 2016 were inaccurate. As we have noted, it did not refer to the conclusions in the Moore Study, which Professor Dworkin accepted did not support a difference at the 2 hour pain-free endpoint between ibuprofen and paracetamol.

41    When evaluating whether there was an adequate foundation in the body of scientific knowledge to support the representations that were made, the primary judge did not err by taking into account the totality of the scientific evidence available at the relevant time. Further, the primary judge did not err by concluding that it would be misleading or deceptive, or likely to mislead or deceive, to make the impugned representations on the basis of the Schachtel Study alone, when the balance of the scientific evidence demonstrated no clear-cut superiority of ibuprofen over paracetamol in terms of faster and more effective relief from pain caused by common headaches including TTH. The body of scientific evidence, which took into account the findings of the Schachtel Study, but balanced them against the findings of other studies, did not support the making of simplistic comparisons of the kind found in Reckitt’s comparative advertising material. It was misleading for Reckitt to make the representations it did—which carried with them an unqualified and definitive statement of scientific fact—when the overall conclusion to be drawn from the scientific evidence was that no authoritative comparisons between active treatments were possible in the then state of scientific knowledge.

42    In this appeal, Reckitt did not argue that the primary judge’s conclusions with respect to contravention of ss 29(1)(a) and (g) of the ACL should be treated differently to his Honour’s conclusions with respect to contravention of s 18.

Disposition

43    The appeal fails. Reckitt must pay Glaxo’s costs.

Associate:

Dated:    24 August 2018