Generic Health Pty Ltd v Bayer Pharma Aktiengesellschaft

Disposition of the appeal 4 For the following reasons, the appeal should be allowed in part with respect to appeal Grounds 5, 8, 9 and 12. We would dismiss the appeal in relation to s 115 of the Act, and the interest and costs questions. But we would allow the appeal on some aspects concerned with the question of quantifying damages on a one-for-one basis and an aspect of the Petibelle issue.

Background facts 5 It is useful to commence with a brief description of the oral contraceptive (OC) market and Yasmin's place therein. As discussed by the primary judge at [198] of the Damages Judgment, there are at least two aspects to the OC market: the Pharmaceutical Benefits Scheme (PBS) market where OCs are sold at government-subsidised prices, and the private prescription market comprising of OCs not on the PBS. All require registration on the Australian Register of Therapeutic Goods (ARTG) and all require a doctor's prescription. There are three generations of OCs, each with different molecular make-ups. All first and most second generation OCs come under the PBS scheme. The cost of PBS OC products to a consumer was, at the relevant time, between $19 and $24 for a packet of four cycles or four months. Some second and all third generation OCs are prescribed with no subsidy. The primary judge found that the cost of these to a consumer was between $35 and $40 for second generation and between $71 and $84 for third generation OCs, for a three month supply. (The relative costs of these products are important, and were to a degree, contentious on the appeal.) 6 Yasmin is a third generation OC. It was registered on the ARTG on 6 July 2001. It contains EE with the progestogen DRSP. Putting Isabelle and Petibelle aside, Yasmin was the first and only OC with DRSP as the progestogen component in the Australian market. There was no bioequivalent product for Yasmin until the appearance of Isabelle on the market. Even though Bayer launched new OCs containing a combination of EE and DRSP in 2008 and 2012 called Yaz and Yaz Flex respectively, these were not Yasmin substitutes. They had a different molecular make-up, and so affected users differently to Yasmin. 7 All OCs require a prescription from a doctor and can only be purchased at a pharmacy. Generally, prescriptions are for the originator brand names, not their bioequivalent generics, unless specifically requested by a patient. The primary judge noted that no evidence indicated that any doctor had prescribed Isabelle instead of Yasmin. This does not mean patients must request a prescription for a generic in order to obtain a generic. Unless the prescribing doctor ticks a box indicating that "brand substitution [is] not permitted", pharmacists can supply the generic on a prescription for the bioequivalent originator. Thus, a pharmacist could supply Isabelle in response to a prescription for Yasmin, its bioequivalent. 8 Prescriptions also tend to last patients around a year, as repeat prescriptions are usually given in addition to an initial prescription for some months. The use of a single OC over such periods of time is not uncommon. Women tend not to switch between OCs regularly. Each OC has a different set of side effects for different women. Once a person is satisfied with a particular brand (and the effect of its particular molecular compounds), she is unlikely to change to another unless she is experiencing a material change in circumstances. The situation may differ in highly genericised markets. Where the active ingredients of products are identical, and hence their impact on a woman is identical, brand loyalty is likely to be weaker. However, the market for Yasmin was not highly genericised. Indeed, it featured no generics and no other substitute for Yasmin until the introduction of Isabelle. 9 The above characteristics of the OC market make it unlike many other markets for goods. It is a market within which the consumer is highly constrained. There is low substitutability of products. In Yasmin's case, there was no substitutable product until the introduction of Isabelle, and then Petibelle. Patients needing a prescription for a particular OC can only purchase it (or its generic, if one is available) at a pharmacy, may not be permitted to substitute the brand for a generic depending on their prescription, and are further restricted by how they individually respond to the different formulations in different OCs. 10 This very personal response to OCs makes their side-effects profile another important characteristic. Third generation OCs are marketed not only for their contraceptive function, but also for associated benefits and minimised side effects. In Yasmin's case, this included reduced weight gain and fluid retention, positive effects on mood and the avoidance of symptoms such as breast tension, headaches, bloating and, in some cases, increased blood pressure. Yasmin was the only OC on the market to provide this specific side-effect profile. The side-effect profile of OCs tends to be relevant in the prescribing process. As found by the primary judge at [209], doctors consider the following when prescribing OCs: (a) Whether the patient has a contra-indication that would make it unsafe to take any COCP [combined oral contraceptive] such as migraine with aura, history of thromboembolic disease or significantly elevated blood pressure; (b) The effectiveness of that formulation in that woman. Generally all COCPs are all equally effective as contraceptives in all women with the exception of women taking other medications that alter the metabolism of the COCP e.g. women taking medications for epilepsy. In these women higher dose COCPs or alternative forms of contraception may be required. (c) Other problems that may be effected (helped or hindered) by the COCP including acne, weight gain, fluid retention, menstrual cycle control; and (d) Patient preference - side effect profile, cost, their friends' recommendations. 11 In January 2012, Generic Health introduced the generic Isabelle. Isabelle was the only bioequivalent of Yasmin, with the same relevant molecule of 3 mg DRSP and 0.03 mg EE, and was registered on the ARTG on this basis. 12 Whereas Bayer only marketed Yasmin to doctors and mainly sold to wholesalers, Generic Health marketed and sold Isabelle to pharmacies. As a generic, it was sold at a cheaper price than Yasmin. For a three month supply, Yasmin cost around $71 to $73 and Isabelle cost around $60 to $61. 13 Generic Health ceased to sell Isabelle as of 19 June 2014 in accordance with Court orders. One week later, on 26 June 2014, Bayer began to sell Petibelle. Petibelle was priced at a mid-way point between Isabelle and Yasmin, at around $66 for a three month supply. Bayer advised doctors and pharmacists of Petibelle's release. 14 Bayer had already obtained an ARTG listing for Petibelle on 8 April 2011, prior to the release of Isabelle. In primary proceedings, Bayer submitted that "[t]his was in keeping with Bayer Australia's usual strategy where if a third party started selling a generic, it may launch its own generic brand of that product": [220]. 15 On 15 February 2012, Bayer commenced patent infringement proceedings against Generic Health. On the same day, Bayer applied to amend the Yasmin patent. On 14 December 2012, Yates J allowed the amendment under s 105 of the Act in the Amendment Judgment. 16 There were a number of amendments made to the Yasmin patent. These were described by Yates J at [2012] FCA 1510 at [24]: The amendments, as they relate to claims 3, 11, 27 and 32: (a) limit the dosage of drospirenone to 3 mg (the 3 mg feature); (b) confine the composition or preparation to a tablet form (the tablet feature); and (c) specify that the stated dissolution test is performed in 900 ml of water (the 900 ml feature). 17 For the purposes of this appeal, the relevant amendment to the patent related to the quantity of DRSP. The specification claimed a range of 2 mg to 4 mg DRSP. This was amended in some claims to be 3 mg. This was the amount of DRSP in both Yasmin and Isabelle.

Appellate review 18 Some debate took place about the role of the appeal court in an appeal under s 28 of the Federal Court of Australia Act 1976 (Cth). The debate was caused by reliance on certain passages in Robinson Helicopter Company Inc v McDermott [2016] HCA 22; 331 ALR 550. The question of appellate review and Robinson Helicopter was dealt with by Allsop CJ in Aldi Foods Pty Ltd v Moroccanoil Israel Ltd [2018] FCAFC 93; 133 IPR 375 at [2] to [10]. We refer to and adopt those reasons.

section 115(1): Damages prior to amendment (grounds 1-4) 19 We commence our consideration of this aspect of the appeal by reviewing the legal requirements of s 115(1) of the Act, before noting relevant aspects of the specification as filed and Bayer's successful application to amend the specification in 2012, both of which provide important matters of context. We then summarise the primary judge's findings on the application of s 115(1) at the damages hearing and Generic Health's submissions on appeal. Finally, we consider whether Generic Health's grounds of appeal in relation to the s 115(1) issue are made out. 20 We should state at the outset that we have recorded the primary judge's findings in some detail, particularly her Honour's analysis of the documentary record before her at trial. A shorter exposition may have sufficed but, in fairness to the careful arguments of Generic Health, we have approached the matter in the way we have in order to provide the appropriate setting in which to understand the relevant grounds of appeal and Generic Health's submissions which, in large measure, seek to demonstrate error in the primary judge's findings of fact, particularly the inferences her Honour drew. Nonetheless, as we come to explain (see [131] to [144] below), we are of the view that Generic Health's challenge in this regard is misplaced because it proceeds on a fundamental misunderstanding of the obligations placed on a patent applicant and a patentee with respect to drafting a complete specification, as well as on a fundamental misunderstanding of the scope and purpose of s 115(1)(a).

Legislation 21 Section 115(1) of the Act provides: (1) Where a complete specification is amended after becoming open to public inspection, damages shall not be awarded, and an order shall not be made for an account of profits, in respect of any infringement of the patent before the date of the decision or order allowing or directing the amendment: (a) unless the court is satisfied that the specification without the amendment was framed in good faith and with reasonable skill and knowledge; or (b) if the claim of the specification that was infringed is a claim mentioned under subsection 114(1). 22 The object of attention in this appeal is s 115(1)(a). 23 Section 115(1)(a) has a long lineage. In the United Kingdom, at a time when specifications were filed or enrolled in the Court of Chancery, the Master of the Rolls, as Keeper of the Records in Chancery, exercised the power to correct what Lord Langdale, MR described in In re Sharp's Patent; ex parte Wordsworth (1840) 49 ER 96 at 99 as "mere slips or clerical errors" to make accurate an enrolment which, by inadvertence, was not what it was intended to be. Speaking as at 1840, his Lordship said at 98: … Patents for inventions are granted on condition of a specification of each invention being enrolled in a limited time, and except for the purpose of correcting mere verbal or clerical errors, proved to have arisen from mistake or inadvertence, I am of opinion that I have no authority to make any alteration in the enrolment of the patent or of the specification. The party enrolling his specification does it at his own peril; and if in his specification he expresses something by which his patent is rendered invalid, he must submit to all the legal consequences; and those who have a right to take advantage of any error of his, must do so in a legal course … 24 Apart from the power of the Master of the Rolls to make corrections for mere slips and clerical errors, there was, at that time, no facility for substantive amendments to be made to a patent specification that might be considered necessary to secure the validity of a patent granted on that specification. 25 By an Act of 1835 (5 & 6 Will 4 c 83) (the 1835 Act (UK)), it became possible, for the first time, for a patentee (whether as grantee or assignee) to file with the Clerk of the Patents a disclaimer or a memorandum of alteration in respect of both a specification and the letters patent themselves. However, in Perry v Skinner (1837) 2 M & W 471; 150 ER 843, s 1 of the 1835 Act (UK), which introduced this provision, was construed as not having retrospective effect but only effect from the filing of the disclaimer or alteration. Thus, even if the defendant's acts would have been infringements if carried out after the filing of the disclaimer or alteration, the defendant might have a good defence (for example, on the ground that the patent granted on the original specification was invalid) where the acts occurred before the filing. At 845, Parke B said: … The construction Mr. Rotch contends for is, that it shall be deemed and taken for part of the letters patent as originally enrolled. The rule by which we are to be guided in construing acts of Parliament is to look at the precise words, and to construe them in their ordinary sense, unless it would lead to any absurdity or manifest injustice; and if it should, so to vary and modify them, as to avoid that which it certainly could not have been the intention of the legislature should be done. Now, if the construction contended for by Mr. Rotch was to be considered as the right construction, it would lead to the manifest injustice of a party who might have put himself to great expense in the making of machines or engines, the subject of the grant of a patent, on the faith of that patent being void, being made a wrong-doer by relation. That is an effect the law will not give to any act of Parliament, unless the words are manifest and plain. We must engraft, therefore, a modification upon the words of the act in this case for the purposes of its construction, and read it as though it had been, "shall be deemed and taken as part of the said letters patent &c. from thenceforth," so as not to make the defendant a wrong-doer. The only doubt arising in this case is from the words of the proviso; but we cannot think the legislature meant to do so unjust a thing as to restrict a party from doing that which he has a lawful right to do; and therefore, though there is some obscurity in the words of the act, we are bound to put a reasonable construction upon them; and undoubtedly the effect of it is to make the patent good for the future. 26 The Patents, Designs, and Trade Marks Act 1883 (UK) (46 and 47 Vict c 57) (the 1883 Act (UK)) introduced a new regime for amendments. Under s 18, a patentee or patent applicant could, by request left at the Patent Office (itself established by the 1883 Act (UK) to succeed the Great Seal Patent Office established under the Patent Law Amendment Act 1852 (UK)), seek leave to amend a specification by way of disclaimer, correction or explanation. Section 18 also provided the procedure for dealing with a request. First, it provided for the request to be advertised. Secondly, it provided for any opposition to the amendment to be heard before the comptroller, with an appeal to the law officer. However, an amendment under s 18 could not be sought where infringement or other proceedings in relation to the patent were pending. Where an action for infringement or a proceeding for revocation had been commenced, s 19 of the 1883 Act (UK) provided that the court or a judge may at any time order that the patentee be at liberty to apply at the Patent Office for leave to amend their specification by way of disclaimer. The court or judge could also direct that, in the meantime, the trial or hearing of the action be postponed. The facility provided by s 19 marked a significant departure from the position under the 1835 Act (UK), which, under s 1, also provided that no disclaimer or alteration could be received in evidence in any action or suit (except a proceeding by way of scire facias) pending at the time when the disclaimer or alteration was enrolled. Notwithstanding the new regime provided under ss 18 and 19 of the 1883 Act (UK), the Master of the Rolls continued to exercise a particular jurisdiction to correct mere slips and clerical errors in specifications of granted patents, but not in specifications for patents lying in the Patent Office before sealing, where s 18 was the only avenue for amendment: In re Gare's Patent (1884) 26 Ch D 105. 27 Another significant departure from the 1835 Act (UK) was the enactment of s 20, which provided: Where an amendment by way of disclaimer, correction, or explanation, has been allowed under this Act, no damages shall be given in any action in respect of the use of the invention before the disclaimer, correction, or explanation, unless the patentee establishes to the satisfaction of the Court that his original claim is framed in good faith and with reasonable skill and knowledge. 28 The parallel between s 115(1)(a) of the Act and s 20 of the 1883 Act (UK) is immediate and obvious. When introduced, s 20 was intended to balance the interests of the patentee and the public in the way explained in the Second Reading of the Bill on 16 April 1883, recorded in Hansard as follows (UK, Parliamentary Debates, House of Commons, 16 April 1883, 359 (Joseph Chamberlain, President of the Board of Trade)): … There was also a new provision as to disclaimer. Where no action was pending, and the patentee desired either to amend his specification or disclaim any part of it, he could do so by leave of the Law Officer. Where, however-and this was the new portion to which he called attention-an action was pending, they provided that the patentee could by leave of the Court disclaim, and that the action might still be continued; but it was provided, in that case, that no damages should be obtained for any previous infringement of the patent, unless the inventor showed his original claim was made in good faith and with reasonable skill and knowledge. The principle of this provision was that the inventor who knowingly or carelessly claimed bad matter was not an object of sympathy. On the other hand, an infringer who knowingly infringed the good and valid part of a patent because he had discovered some portion, perhaps unimportant, which was bad or invalid, was not entitled to sympathy or protection; and, accordingly, he had provided that while, in the first case, the inventor would lose the advantage of his patent until he had amended his claim, in the second case the infringer might be made to suffer and pay damages for his infringement without benefiting by the technical defects or misdescription of the patentee. … (Emphasis added) 29 Writing in 1884, Professor Crawford Munro observed in J E Crawford Munro, The Patents, Designs, and Trade Marks Act, 1883 (46 & 47 Vict. c. 57) with the rules and instructions, together with pleadings, orders, and precedents (Stevens & Sons, 1884), 32: Under this Act an amendment may be allowed in two cases: (1) by the comptroller, under sect. 18; (2) by the Court or judge, under sect. 19. This section does not say that the amendment may not be given in evidence where the alleged infringement occurred previous to the amendment, but only that in such a case damages are not to be given, except as provided by the last proviso. A patentee may be able to establish the validity of the patent as amended - a matter of some importance in view of the provisions of sect. 31 [dealing with the certification of claims]. Unless the patentee can satisfy the Court that his original claim was framed in good faith and with reasonable care and knowledge, it may be better not to amend until the action is at an end - inasmuch as the result of the action will show the patentee where his specification is wanting. It is quite possible for the Court to give leave to amend, for the amendment to take place, and yet for the patent to be upset in the action. 30 In the same work, at xlix, Professor Crawford Munro predicted: As to amendment during action, it is not likely that patentees will often avail themselves of the provision of the Act, inasmuch as they cannot recover damages in the action, save where they satisfy the court that the claim was framed in good faith, and with reasonable skill and knowledge. 31 Section 20 of the 1883 Act (UK) was substantially reproduced in s 23 of the Patents and Designs Act 1907 (UK) (the 1907 Act (UK)), which remained unaffected by amendments subsequently introduced by the Patents and Designs Act 1932 (UK). Section 23 of the 1907 Act (UK) was then replaced by s 59(3) of the Patents Act 1949 (UK) (the 1949 Act (UK)), which provided, in slightly modified form: Where an amendment of a specification by way of disclaimer, correction or explanation has been allowed under this Act after the publication of the specification, no damages shall be awarded in any proceeding in respect of the use of the invention before the date of the decision allowing the amendment, unless the court is satisfied that the specification as originally published was framed in good faith and with reasonable skill and knowledge. 32 Section 59(3) of the 1949 Act (UK) was replaced with s 62(3) of the Patents Act 1977 (UK) (the 1977 Act (UK)), in substantially similar terms: Where an amendment of the specification of a patent has been allowed under any of the provisions of this Act, no damages shall be awarded in proceedings for an infringement of the patent committed before the decision to allow the amendment unless the court or the comptroller is satisfied that the specification of the patent as published was framed in good faith and with reasonable skill and knowledge. 33 The 1977 Act (UK) also introduced a provision concerning the relief to be granted for infringement where the patent was found to be only partially valid. In that regard, s 63(2) provides: Where in any such proceedings it is found that a patent is only partially valid, the court or the comptroller shall not grant relief by way of damages, costs or expenses, except where the plaintiff or pursuer proves that the specification for the patent was framed in good faith and with reasonable skill and knowledge, and in that event the court or the comptroller may grant relief in respect of that part of the patent which is valid and infringed, subject to the discretion of the court or the comptroller as to costs or expenses and as to the date from which damages should be reckoned. 34 A similar evolution can be seen in the precursory Australian legislation. Before Federation, provisions identical or substantially identical to s 20 of the 1883 Act (UK) were enacted in State legislation: see s 26 of the Patents Act 1889 (Vic) and s 18 of the Patents Act 1899 (NSW). Following Federation, a provision identical to s 20 of the 1883 Act (UK) was enacted in s 82 of the Patents Act 1903 (Cth) (the 1903 Act). 35 Section 82 of the 1903 Act was replaced by s 87 of the Patents Act 1952 (Cth) (the 1952 Act). Section 87 was expressed in substantially similar terms to s 59(3) of the 1949 Act (UK): Where an amendment is made under this Act to a complete specification after publication, damages shall not be given, and an order shall not be made for an account of profits, in an action for an infringement of the patent occurring before the date of the decision or order allowing or directing the amendment, unless the patentee establishes to the satisfaction of the court that the specification without the amendment was framed in good faith and with reasonable skill and knowledge. 36 This section was repealed by s 17 of the Patents Act 1962 (Cth) and replaced with the following provision as s 159B: Where an amendment is made under this Act to a complete specification after the specification became open to public inspection, damages shall not be awarded, and an order shall not be made for an account of profits, in an action for an infringement of the patent occurring before the date of the decision or order allowing or directing the amendment - (a) unless the court is satisfied that the specification without the amendment was framed in good faith and with reasonable skill and knowledge; or (b) if the claim of the specification in respect of which the infringement is found is a claim in respect of which the court is required, by virtue of sub-section (1.) or sub-section (3.) of the last preceding section, to treat a certain date as the priority date. 37 Section 159B underwent amendments enacted by s 35 of the Patents Act 1969 (Cth) and s 66 of the Patents Amendment Act 1979 (Cth). These amendments are without consequence to the present discussion. Section 159B was repealed by the Patents Act 1990 (Cth), which brought in the present s 115(1). 38 From this brief summary, it can be seen that, stemming from s 20 of the 1883 Act (UK), Australian patent law and United Kingdom patent law have developed along similar lines so far as concerns the availability of pecuniary relief for pre-amendment acts of infringement. United Kingdom law has taken the further step - via s 63(2) of the 1977 Act (UK) - of limiting the pecuniary relief for infringement of a patent found to be only partially valid at trial. Even so, the availability of pecuniary relief in those circumstances is conditioned on the same requirements of good faith and reasonable skill and knowledge in framing the specification in suit.

Case law 39 Despite the lineage summarised above, there is remarkably little case law on the application of provisions such as s 115(1)(a) until the latter part of the twentieth century: see, in this connection, the observations in Thomas A Blanco White, Patents for Inventions and the Registration of Industrial Designs (Stevens & Sons, 1st ed, 1950) at 251, repeated in subsequent editions of that work up to and including the 4th edition in 1974. The sparseness of the case law on this topic perhaps gives life to Professor Crawford Munro's prediction in 1884 that it would be unlikely that patentees would avail themselves of the facility provided by the provision. Such references as there are in the cases up to 1974 yield little commentary on the meaning and scope of the provision beyond reflecting on the fact that, where an inventor frames a specification knowing that it includes subject matter that the inventor has not invented, or frames a claim so as to include within it much that is old, then it cannot be said that the specification has been framed in good faith or with reasonable skill and knowledge: see, for example, the observations in Wenham and Co v Carpenter, Todd and Co (1888) 5 RPC 68 at 69; Meyer v Sherwood (1890) 7 RPC 283 at 286; Hopkinson v The St. James' and Pall Mall Electric Light Company, Limited [1893] 10 RPC 46 at 62; Kane & Pattison v J. Boyle & Co [1901] RPC 325 at 338; British United Shoe Machinery Company Ld v A. Fussell & Sons Ld (1908) 25 RPC 368 at 386; and British United Shoe Machinery Co, Ld v Gimson Shoe Machinery Co, Ld [No 2] (1929) 46 RPC 137 at 164. 40 For our purposes, it is convenient to commence with the observations of the Court of Appeal in 1975 in General Tire & Rubber Company v Firestone Tyre & Rubber Company Limited (1975) 92 RPC 203 (General Tire (1975)) concerning the application of s 59(3) of the 1949 Act (UK). 41 As to the good faith requirement, the Court of Appeal said at 269: On this question of good faith in framing a specification one is, we apprehend, basically in a field of inquiry whether the patentee or his agent knew something detrimental to the patent, as applied for in the form in which the specification was framed, which escaped the eagle eye of the examining officer in the Patent Office. If a patent agent puts forward something of which he has no knowledge, which suffers from some fatal imperfection in the patent field we do not consider that, when the Patent Office accepts it without demur, it can be said that it was framed otherwise than in good faith. It is, after all, the function of a patent agent to argue in honesty for the width of the application. 42 Later, the Court of Appeal said at 270: … On the question of good faith it is not to be expected that the patentee should in evidence first raise up and then exorcise the ghost of every possible defect in the unamended claim: if he destroys the defects that are suggested, that suffices. … 43 As to reasonable skill and knowledge, the Court of Appeal distinguished this requirement from the notion of reasonable care. The Court of Appeal saw the reasonable skill and knowledge requirement as directed, and limited, to "the field of expertise peculiar to those concerned with framing specifications" and not to "mere commonplace slips" that are "remote from patent expertise": see General Tire (1975) at 270. 44 The requirements of s 63(2) of the 1977 Act (UK) have been discussed in a number of cases. 45 In Hallen Co v Brabantia (UK) Ltd [1990] FSR 134, Aldous J referred to General Tire (1975) and said at 143: Taking guidance from those statements, it is my view that section 63(2) imposes upon a plaintiff a duty to prove on the balance of probabilities two things: First that the specification was framed in good faith. That requires a plaintiff to prove that the specification was framed honestly with a view to obtaining a monopoly to which, on the material known to him, he believed he was entitled. Secondly, that the specification was framed with reasonable skill and knowledge. The words "skill and knowledge" are a composite phrase relating to the competence employed in framing the specification and require the specification as framed to be in the form in which a person, with reasonable skill in drafting patent specifications and a knowledge of the law and practice relating thereto, would produce. 46 In Chiron Corporation v Organon Teknika Limited [No 7] [1994] FSR 458, Aldous J at 467 to 468 made clear that the reasonable skill and knowledge requirement assumes that the person drafting the specification has been properly instructed on the details of the invention: … the words in the Act require the specification as framed to be in a form in which a person, with reasonable skill in drafting patent specifications and knowledge of the law and practice relating thereto, would produce with the patentee's knowledge of the invention. (Emphasis added) 47 In that case, Aldous J also addressed a submission that claims directed to vaccines for Hepatitis C Virus (HCV) and to certain tissue cultures should not have been included in the patent based on certain knowledge at the time the specification was drafted - namely, that HCV was a relative of the pestiviruses and also related to the flaviviruses. By way of background, Dr Houghton was an employee of the patentee and the project leader of the team which identified and sequenced HCV. Aldous J recorded the following evidence at 464: At the time that Dr Houghton recognised the similarity between HCV and flaviviruses, he believed that it was possible to predict, extrapolating from flaviviruses, that vaccines could be prepared and that tissue culture could be carried out by straightforward means. As further sequencing of HCV took place, Dr Houghton realised that HCV was a relative of the pestiviruses, but was also within the flaviviridae family. He subsequently concluded that HCV was related to the flaviviruses, but was more closely related to another genus within the flaviviridae family, namely the pestivirus genus. It was Dr Houghton's evidence that the realisation that HCV was pesti-like did not alter his view as to the likelihood that a vaccine could be produced using routine techniques. That was because his prediction, based on the similarity of HCV to flaviviruses, would have been the same if he had realised that it was like a pestivirus. As he said (Evidence 1, p. 16D) whether HCV was truly like a flavivirus or a pestivirus did not matter. 48 The defendants argued that the specification was not framed with reasonable skill and knowledge in that a properly instructed and competent draftsman would have deleted the vaccine and tissue culture claims because of knowledge then available that HCV was "pesti-like". 49 In considering that submission, Aldous J said at 468 to 469: The basis of the defendants' submission was the knowledge gained during 1988 and 1990 that HCV was pesti-like. No doubt it would have been better if that information had been in the specification. However I do not believe that that knowledge led Dr Houghton to alter his views as to the validity of the prediction made in the patent. Nor do I believe that a competent draftsman, turning his mind to that fact, would have deleted the claims from the patent. As the submissions and evidence showed, the claims were proper to place before the court for adjudication. It is also not without significance that Professor Almond did not refer to the fact that HCV was closer related to pestiviruses than flaviviruses in his evidence. His reasons for doubting the validity of the plaintiffs' predictions were based on other grounds. 50 Aldous J held at 469: Upon the basis of the conclusions of insufficiency in my judgment, it is apparent that the inclusion of claims relating to vaccines and tissue culture was an error. However I am satisfied that the specification was framed in good faith and with reasonable skill and knowledge. The proof of the pudding is in the eating. The specification was the subject of substantial scrutiny during these proceedings and at no time was it shown to be a document which a competent properly instructed patent agent would not have produced. 51 In Hoechst Celanese Corporation v BP Chemicals Limited [1997] EWHC 370 (Pat); FSR 547, Laddie J said at 575, with respect to s 63(2) of the 1977 Act (UK): As Aldous J. said in Hallen, this imposes on the patentee a duty to prove on the balance of probabilities two things: first that the specification was framed in good faith and secondly that it was framed with reasonable skill. The first of these is a test of honesty. It is assessed by reference to what the patentee or his agent actually knew. The fact that there is something detrimental to the patent which a competent patentee would have found out about, but this patentee did not, does not mean that the latter lacks honesty or that the specification is not framed in good faith. Furthermore, taking into account what the patentee knows, the court has to decide whether its framing was carried out with reasonable skill and knowledge. Carelessness is not, per se, enough. The court must ask whether, in the circumstances existing at the time, including what was known about the prior art, a competent patentee or, more usually, patent agent would or could have been expected to frame the specification in the way it was in fact framed. (Emphasis added) 52 In Nutrinova Nutrition Specialities & Food Ingredients GmbH v Scanchem UK Limited [No 2] [2000] EWHC 98 (Pat); [2001] FSR 43, Pumfrey J at [6] described the requirements of good faith and reasonable skill and knowledge in the following terms: It is well settled that to demonstrate good faith and reasonable skill and knowledge a patentee must prove two things: (a) that the specification was framed honestly with a view to obtaining a monopoly to which on the material known to him he believed he was entitled; (b) the specification was framed with reasonable skill and knowledge, a phrase which relates to the care exercised by the person who drafted the specification. 53 Pumfrey J then remarked at [7] on the role of the patent agent in drafting a specification: Often, a specification is a collaborative effort between (for example) a patent agent and the inventor. It must always be remembered that it is the task of the patent agent to exercise his skill to apply to obtain as wide a coverage for the invention which is disclosed to him by the inventor as the state of the art known to him at the time of the application would reasonably warrant. So too, if he detects a wider potential invention in what has been described by the inventor (since inventors often approach the patent agent with what will eventually be the preferred embodiment of the patent in suit) it is his duty to exercise his skill in generalising the features of the invention disclosed to him. It is neither possible nor desirable to give a list of examples of want of good faith or reasonable skill and knowledge; but an example of the former might well be knowingly drawing a claim which covers the prior art, and of the latter failure to appreciate that a claim as drawn does cover prior art known to the draftsman. 54 In Kirin-Amgen Inc's Patent [2002] EWHC 471 (Pat); 119 RPC 851 (Kirin-Amgen), Neuberger J considered a submission to the effect that, when applying s 63(2) of the 1977 Act (UK), the drafter of the specification should be taken as having been equipped with the actual knowledge of the reader of the specification. He rejected that submission, saying at [145]: There is no warrant, in my view, for equating all the components of skill and knowledge to be expected of the notional reader of a Patent to the actual writer of the Patent, even accepting that the writer under section 63(2) is treated as extending to the patentee, its employees and its patent agent. There is nothing to suggest a departure from reality when asking whether the patentee was guilty of want of reasonable skill and knowledge under section 63(2). The Statutory language and the authorities appear to require the court to take the patentee and his patent agent as it finds them, albeit that one obviously is entitled to expect them to be honest and to expect the patent agent to be "properly instructed" and professionally competent (see the guidance given in Chiron (No. 7) at [1994] FSR 458 at 467 to 468). (Emphasis added) 55 On appeal (Kirin-Amgen Inc v Transkaryotic Therapies Inc [2002] EWCA Civ 1096; [2003] RPC 31) (Kirin-Amgen/Transkaryotic), the Court of Appeal explained at [152] that the requirement that the specification be framed with reasonable skill and knowledge is one to be determined objectively. Further, the relevant knowledge "must encompass the knowledge of the person which formed the basis of the information in the specification". This observation is not to be understood (as Generic Health's oral submissions suggested) as limiting the relevant knowledge to the actual knowledge of the drafter or those attending upon him or her for the purpose of providing instructions for framing the specification. The word "encompass" was obviously used advisedly by their Lordships as an expression of amplitude, not of limitation or restriction. The Court of Appeal illustrated its proposition by referring to the case where the relevant knowledge might be sourced in a contractor who had been engaged by the patentee to carry out work for it. 56 Finally, in Unilin Beheer BV v Berry Floor NV [No 2] [2005] EWCA Civ 1292; [2006] FSR 26, Jacob LJ (with whom Mummery and Neuberger LJJ agreed) stated at [29] that any rational view of policy suggested that the provision should be construed so as to deprive the patentee of damages only where the patentee has done something that might mislead. This statement is consistent with the objective stated in the Second Reading speech when s 20 of the 1883 Act (UK) was enacted: see [28] above. 57 His Lordship also considered and rejected the following submission at [30]: [Counsel for the appellants] submitted that a patentee who has amended his patent should be regarded as the grantee of an indulgence. If, prior to amendment, he committed the past sin of framing the specification without reasonable skill and knowledge, he must be punished for that sin by depriving him of damages. And this punishment must be visited on him even though the fact that the patent needed amendment was not his fault. I reject this medieval view of the section. 58 There is little authority in the Australian context. Section 115(1)(a) arose for consideration in Pracdes Pty Ltd v Stanilite Electronics Pty Ltd (1995) 35 IPR 277 but, on the facts of the case, it was not necessary for the trial judge to engage in any analysis of the scope and operation of the provision. Given the provision's antecedents, we proceed on the basis that the United Kingdom cases provide persuasive guidance on how s 115(1)(a) should be construed and applied. In doing so, we bear in mind the caution expressed in Lockwood Security Products Pty Ltd v Doric Products Pty Ltd [2004] HCA 58; 217 CLR 274 at [63] to [67] concerning the relevance of post-1977 United Kingdom cases. However, there is nothing to suggest that the jurisprudence to which we have referred has itself been guided by considerations of European patent law.

The specification as filed 59 The application for the patent is based on International Patent Application No. PCT/IB00/01213 filed on 31 August 2000 (the PCT application). The patent was sealed on 30 June 2005, and claims a priority date of 31 August 1999 based on the filing of European Patent Application No. 99202826.6 and US Patent Application No. 09/386,274. The complete specification accompanying the application (the specification) was amended pre-grant in 2005 to ensure consistency with a European patent application, which was also based on the PCT application, that was being prosecuted at the time. The specification describes the invention as relating to: … a pharmaceutical composition comprising drospirenone and ethinylestradiol, a method of providing dissolution of drospirenone, methods of inhibiting ovulation by administration of drospirenone and the use of drospirenone and ethinylestradiol for inhibiting ovulation. 60 As background to the invention, the specification states that oral contraceptives containing a combination of a gestagen component and an estrogen component have been in use since the 1960s. The specification explains that contraceptive reliability is mainly provided by the gestagen component. The estrogen component acts to increase the ovulation inhibitory effect of the gestagen component and to ensure cycle stability. The daily dosage of the oral contraceptive should contain, at least, the minimum dosage needed for the gestagen component to inhibit ovulation effectively. The specification discloses that, since the introduction of oral contraceptives, the daily dosage of gestagen has been reduced through the development of new and more efficient gestagens. It has also been possible to reduce the daily dosage of estrogen. 61 DRSP is a gestagenic compound. The specification teaches, by reference to various German and European patents, that: (a) DRSP can be used as a diuretic. (b) DRSP can be used as a contraceptive at dosage levels of 0.5 mg to 50 mg per day. (c) DRSP does not give rise to increased blood pressure and, therefore, can be administered to women who have, or are at risk of, developing increased blood pressure. (d) DRSP can be administered together with EE in an amount of 0.03 mg to 0.05 mg per day. (e) DRSP can be used for the treatment of gynaecological irregularities and for contraception at a dosage level of 0.5 mg to 50 mg per day. (f) DRSP can be used for the treatment of androgen-induced disorders, aldosterone-induced disorders, and hormonal irregularities as well as for contraception at dosage levels of 0.5 mg to 50 mg per day, preferably 1 mg to 10 mg per day (including with EE, administered at 0.02 mg to 0.04 mg per day). 62 When summarising the invention, the specification states: In the course of research leading to the present invention, it has surprisingly been found that a hitherto undisclosed minimum dosage level of drospirenone is required for reliable contraceptive activity. Similarly, a preferred maximum dosage has been identified at which unpleasant side effects, in particular excessive diuresis, may substantially be avoided. 63 The specification then describes an aspect of the invention as follows: Accordingly, in a first aspect, the present invention relates to a pharmaceutical composition comprising, as a first active agent … (drospirenone) in an amount corresponding to a daily dosage, on administration of the composition, of from about 2 mg to about 4 mg, and, as a second active agent, … (ethinylestradiol) in an amount corresponding to a daily dosage of from about 0.01 mg to about 0.05 mg, together with one or more pharmaceutically acceptable carriers or excipients. 64 When providing a more detailed description of the invention, the specification teaches that DRSP is a sparingly soluble substance in water and aqueous buffers at various pH values. Further, DRSP is rearranged to an inactive isomer under acid conditions and hydrolysed under alkaline conditions. The specification discloses that it has been found that, when DRSP is provided in micronized form in a pharmaceutical composition, rapid dissolution of the active compound from the composition occurs in vitro. The specification also discloses that, instead of providing DRSP in micronized form, it is possible to dissolve it in a suitable solvent and spray it onto the surface of inert carrier particles that can be incorporated in the composition. 65 Prior to the amendments discussed below, claim 3 in the specification was directed to a pharmaceutical composition, in part defined by reference to a dissolution test: A pharmaceutical composition in oral dosage form comprising; from 2 mg to 4 mg of drospirenone and 0.01 mg to 0.05 mg of ethinylestradiol, together with one or more pharmaceutically acceptable carriers or excipients, wherein at least 70% of said drospirenone is dissolved from said composition within 30 minutes, as determined by USP XXIII Paddle Method II using water at 37°C as the dissolution media and 50 rpm as the stirring rate. (Emphasis added.) 66 Claims 11, 27 and 32 also included a dissolution test in substantially the same terms (together with claim 3, the dissolution test claims).

Amendment of the specification 67 The specification was amended by orders made on 14 December 2012 in the Amendment Judgment. The effect of the amendments as ordered was to replace the existing claim set with a new claim set. Relevantly for present purposes, claim 3 (a pharmaceutical composition) and claim 11 (a pharmaceutical preparation in the form of separately packaged and individually removable daily dosage units) were amended to limit the composition/preparation to one containing 3 mg DRSP as opposed to 2 mg to 4 mg DRSP. In the Amendment Judgment, this was referred to as the 3 mg feature. Importantly, the body of the specification was not amended. Thus, the description of the invention in the form of a pharmaceutical composition comprising 2 mg to 4 mg DRSP as a first active agent remains. 68 The new claim set also included other amendments (including with respect to claim 3 and claim 11) concerning (what the Amendment Judgment termed) the tablet feature and the 900 ml feature. Only the amendments relating to the 3 mg feature are of significance to the present appeal. 69 It is important to note that, neither at the time of the amendment application nor at the time of the liability hearing before the primary judge, was it alleged by Generic Health that a composition/preparation containing 2 mg to 4 mg DRSP is inutile or that the patent had been granted on a false suggestion or misrepresentation in that regard. Significantly for present purposes, there was no challenge to the description of the first aspect of the invention quoted at [63] above as an accurate and proper description of the invention. Further, there was no challenge to the accuracy of the disclosure that 2 mg DRSP was the minimum dosage level for reliable contraceptive activity and that 4 mg DRSP was a preferred maximum dosage at which unpleasant side effects, in particular excessive diuresis, may substantially be avoided: see [62] and [63] above. 70 As the Amendment Judgment recorded at [26] and [27]: (a) The amendments sought by reference to the 3 mg feature, the tablet feature and the 900 ml feature were sought to forestall any argument, particularly in the principal proceeding (then to come before the primary judge), that the dissolution test claims of the patent were not fairly based on the matter described in the specification: s 40(3) of the Act. (b) There was no evidence to suggest that the amendments had been sought to overcome any prior art or to address any other asserted, apparent or possible ground of invalidity. (c) In applying for the amendments, Bayer had made clear its position that it did not concede that, without the amendments, the dissolution test claims would not be fairly based on the matter described in the specification. 71 As to the matter noted in para (c) above, page 4 of the specification states: Drospirenone, which may be prepared substantially as described in, e.g., US 4,129,564 or WO 98/06738, is a sparingly soluble substance in water and aqueous buffers at various pH values. Furthermore, drospirenone is rearranged to an inactive isomer under acid conditions and hydrolysed under alkaline conditions. To ensure good bioavailability of the compound, it is therefore advantageously provided in a form that promotes rapid dissolution thereof. It has surprisingly been found that when drospirenone is provided in micronized form (so that particles of the active substance have a surface area of more than 10,000 cm2/g, and the following particle size distribution as determined under the microscope: not more than 2 particles in a given batch with a diameter of more than 30 µm, and preferably ≤ 20 particles with a diameter of ≥ 10 µm and ≤ 30 µm) in a pharmaceutical composition, rapid dissolution of the active compound from the composition occurs in vitro ("rapid dissolution" is defined as the dissolution of at least 70% over about 30 minutes, in particular at least 80% over about 20 minutes, of drospirenone from a tablet preparation containing 3 mg of drospirenone in 900 ml of water at 37°C determined by the USP XXIII Paddle Method using a USP dissolution test apparatus 2 at 50 rpm). Instead of providing the drospirenone in micronized form, it is possible to dissolve it in a suitable solvent, e.g. methanol or ethyl acetate, and spray it onto the surface of inert carrier particles followed by incorporation of the particles containing drospirenone on their surface in the composition. (Emphasis added.) 72 At [28] of the Amendment Judgment, Yates J reflected on Bayer's reservation of position concerning the question whether the dissolution test claims, without amendment, would be fairly based on the matter described in the specification: Without seeking to decide that question in the present application, I should record what can be no more than a provisional view that there is support for Bayer Pharma's position in this regard. As I have recounted, the specification refers, in a number of places, to dosages of drospirenone "from about 2 mg to about 4 mg" and to the fact that the composition of the invention can be formulated in a number of dosage forms, including as a tablet. The dissolution test in each of the relevant claims, although not stated in precisely the same terms as those used on page 4 of the body of the specification, nevertheless has an apparent close connection with the dissolution test so stated. In deciding whether the relevant claims, in their present form, are fairly based on the matter described in the specification, one issue to be determined (there may well be others) would be how those claims should be constructed through the eyes of the person skilled in the art and whether the dissolution test stated in, for example, claim 3 would, as a matter of substance, be read and understood by such a person materially differently from the test stated on page 4 of the specification. 73 As to the reason(s) to be attributed to making a claim amendment, the Court of Appeal in General Tire (1975) stated at 270: … It is to be observed that the withdrawal of a claim does not necessarily involve admission that that claim cannot be or could not ever have been sustained as valid. It is common practice to include in good faith a wider claim so as to have the benefit of a wider search in the Patent Office. It is also common practice for an applicant to decide that in terms of commerce a narrower claim suffices, and therefore to drop a wider claim which will not serve him commercially and may land him in opposition litigation. … 74 Following this line of authority, Yates J stated in the Amendment Judgment at [218]: I do not find it to be exceptionable that a patentee or patent applicant would seek to keep the scope of its patent claims as broad as possible in order to fully protect its invention as it perceives it to be. Similarly, I do not find it to be exceptionable that, for pragmatic reasons, a patentee or patent applicant might be prepared to compromise on, and resile from, its desired position and not seek to vindicate to the fullest its patent rights or entitlements, as it perceives them to be. The question is whether, in so doing, the patentee or patent applicant, improperly, has sought to prosecute or maintain claims of unjustified width. … I do not regard Bayer Pharma's conduct, as revealed by the evidence, to involve it seeking to prosecute or maintain claims which it knows to be either impermissibly or indefensibly wide. In my view, none of the material relied upon … in this regard shows any untoward or inappropriate conduct on the part of Bayer Pharma which, on the present state of the evidence, would warrant criticism of Bayer Pharma by the Court. 75 In the following paragraphs it is now convenient to refer to the form of the specification before the introduction of the new claim set as the pre-amendment specification.

The primary judge's findings 76 As we have said, in order to place Generic Health's appeal in relation to the s 115(1)(a) issue in context, it is necessary to record the primary judge's findings in some detail.

The role of the patent attorney 77 The primary judge at [53] to [57] accepted the following propositions, which were not in contest in this appeal: (a) Patent drafting is a collaborative process between the inventor/client and the patent attorney, who each bring different skills and information to the process. A "balancing act" between the commercial, legal and technical position of the patent applicant is involved. (b) Claim drafting involves an exercise in judgment as to how broadly an invention can be defined. (c) It is common for patentees to seek to claim a dosage range for an active pharmaceutical ingredient. (d) It is typical and reasonable for a patent attorney to rely on information provided by the inventor/client on technical matters, such as the appropriate dosage range for an active pharmaceutical ingredient. Although these instructions should be reviewed with a critical eye by the patent attorney, it should nevertheless be recognised that the inventor/client has the relevant expertise on technical matters. (e) When a range is claimed, the patent attorney expects that the inventor/client will have a sound basis for an honest belief in the range claimed. This belief may be based on experimental evidence or a plausible explanation. It is not up to the patent attorney, when drafting a patent specification, to make "sound predictions" of a technical nature unless the patent attorney has particular technical expertise in a specific area. (f) The standard set by s 115(1)(a) of the Act for framing a specification is one of reasonable skill and knowledge, not the highest possible level of skill and knowledge on the part of either the patent attorney or the inventor/client. (g) It follows that it is not necessarily the case that the standards applied by an expert witness to his or her own practice as a patent attorney set the boundaries of competent practice.

The selection of 2 mg to 4 mg as the dosage range of DRSP in the pre-amendment specification

Dr Broesamle's evidence 78 The PCT application was drafted by Plougmann, Vingtoft & Partners (P&V) - a firm of European patent attorneys. At that time, Dr Broesamle was the Senior Patent Counsel at Bayer Intellectual Property GmbH with responsibility for "the managing, filing and prosecution" of a significant number of patents within Bayer's patent portfolio around the world. One of these was the PCT application. Dr Broesamle was a highly qualified chemist and a patent attorney. He commenced working as a patent attorney at Schering AG in November 1987. His particular responsibilities included work on fertility control and hormone replacement therapy. His work with DRSP commenced in 1989, 10 years before the PCT application was prepared. Schering AG had been working with DRSP since 1976. Some of its work is reflected in the patents briefly referred to at [59] above, which are cited in the PCT application. Dr Broesamle became responsible for this patent portfolio in 1989. 79 We observe that, in 2006, Schering AG merged with the Bayer Group of companies. The name of the merged entity was changed to Bayer Schering Pharma AG. On 1 July 2011, the name was again changed to Bayer Pharma AG, the first respondent's current name (which we have abbreviated to Bayer AG). 80 Dr Broesamle was assisted by Dr Ulrich in providing instructions to P&V. Dr Broesamle was senior to Dr Ulrich. The final decision concerning any issue in relation to the patent family for which he was responsible was made by Dr Broesamle, in consultation with Bayer's Head of Patents and Licensing. 81 Dr Broesamle gave evidence in support of the amendments covered by the Amendment Judgment. He also gave evidence before the primary judge at the damages hearing. Dr Ulrich did not give evidence. 82 Before the primary judge, Generic Health did not challenge Dr Broesamle's status as a highly qualified chemist and patent attorney who, since 1987, was specifically responsible within Schering AG (and then Bayer AG) for fertility control and hormone replacement therapy and, since 1989, had been responsible for Schering AG's DRSP patents. Generic Health did not suggest that Dr Broesamle, as Senior Patent Counsel, had failed to act as a competent patent attorney in respect of the framing of the PCT application. 83 Further, before the primary judge, Generic Health did not challenge Dr Broesamle's evidence that P&V was a highly reputable patent attorney firm which completed all work with a high level of skill and care and in accordance with instructions and expectations. Generic Health did not suggest that Dr Broesamle had failed to hold P&V to the high standards and competence which he expected from them in respect of the PCT application. 84 However, Dr Broesamle did not give evidence about the selection of the 2 mg to 4 mg dosage range. Therefore, before the primary judge, Generic Health submitted that there was an "evidentiary vacuum" about the selection of this range. 85 In this connection, Generic Health submitted that: (a) The documents produced by Bayer on discovery were ambiguous and incomplete. (b) There was no evidence that Schering AG had relied on any testing for dosages above 3 mg for the purposes of formulating the PCT application. (c) There was no evidence that P&V were advised by Schering AG personnel that the therapeutically effective range of DRSP was 2 mg to 4 mg or the relevant Schering AG personnel formed a state of mind that there was a reasonable or sound basis for this dosage range. (d) There was nothing to suggest that a report known as the 9274 report was considered in the process of framing the PCT application. (e) The documentary record indicated a concern within Schering AG as to whether side effects could be avoided at 4 mg DRSP. (f) The evidence indicated that Schering AG had doubts about the efficacy of a dosage of 2 mg DRSP. 86 We record the primary judge's findings in relation to these submissions in later parts of these reasons. Of present significance is the fact that the primary judge found that, given his role and expertise, it was difficult to imagine that Dr Broesamle did not familiarise himself with the research and patents obtained by Schering AG for DRSP when he became involved with work on DRSP in 1989, or at least before giving instructions to P&V in 1999. By the same token, the primary judge accepted that, as Dr Broesamle must have been involved in giving instructions on numerous pharmaceutical compositions over many years, and as dosage ranges are commonplace in patents for pharmaceutical compositions, the notion that Dr Broesamle, or any other person, could reliably recall instructions to a patent attorney about aspects of one patent application from 16 years ago "would stretch credulity". 87 Thus, the primary judge was not persuaded that, in giving his evidence, Dr Broesamle deliberately remained silent on matters that would not have assisted Bayer's case. Her Honour considered such a suggestion to be "far-fetched and fanciful". The primary judge concluded that Dr Broesamle did not give evidence about the instructions he gave to P&V, or about what particular documents meant, because he had no independent recollection of the framing of one patent application amongst many some 16 years ago, and because he could add nothing useful to the documents themselves. The primary judge further concluded that the fact that a document was created at least 16 or more years ago provided a rational basis for inferring that nothing Dr Broesamle or anyone else could have said about the document would be other than potentially self-serving reconstruction. 88 Similarly, the primary judge concluded that no significance could be given to the fact that Dr Ulrich was not called to give evidence. The primary judge accepted that Dr Broesamle was the person who made decisions about the PCT application, not Dr Ulrich.

The documentary record 89 The primary judge then turned to consider the documentary evidence by reference to three categories: (a) documents relating to the preparation of the PCT application (and related priority documents); (b) other documents created within Schering AG, which pre-dated the PCT application; and (c) the PCT application itself (including documents referred to in the PCT application).

First category 90 The documents in the first category (documents relating to the preparation of the PCT application and related priority documents) comprise: (a) A table identifying five reports in respect of different formulations of DRSP/EE including 0.5, 1, 2 and 3 mg DRSP. This table is dated August 1998. Based on the fact that the table identifies the status and availability of each report, the primary judge found that Schering AG was collating information it considered of obvious relevance to the patents for DRSP for which it subsequently applied. The primary judge found that Schering AG was at least aware of all the information in the five reports when it gave instructions to P&V for framing the specification for the PCT application. These reports are conveniently referred to as the 9693 report, the 9692 report, the 9970 report, the A187 report and the A892 report. (b) A document headed "Open questions for patent application regarding DRSP". On the basis of this document, the primary judge found that Schering AG was considering the ratio of DRSP to EE to achieve the optimal therapeutic effect and minimal side effects for contraceptive purposes and, separately, for hormonal disorders. The primary judge found that Schering AG was doing so for the purposes of seeking patent protection for the Yasmin formulation. (c) A typed document dated 13 November 1998, which was apparently prepared by two of the named inventors. The document refers to DRSP in the range of 0.5 mg to 7.0 mg, with a handwritten note indicating 2 mg as the lower limit for ovulation inhibition. On the basis of this document, the primary judge found that Schering AG was considering dosage range and held the view that 2 mg was the lower limit for ovulation inhibition. Further, the primary judge found that, even though side effects are not referred to in the document, it did not mean that the dosage range of 0.5 mg to 7.0 mg was recorded without regard at all to possible side effects. (d) A handwritten document, which identifies the structure and characteristics of DRSP and contains other information which resembles the information in the PCT application. The primary judge found that this document had been prepared for the purpose of instructing P&V and that the substance of the information had been communicated to P&V. The document refers to DRSP in the amount of 2 mg to 4 mg, and also 2.5 mg to 3.5 mg. The document includes a notation, which queries the upper limit of the range and the presence of side effects. On the basis of this document, the primary judge found that 4 mg DRSP was being considered as the upper limit at which side effects might be largely avoided. The primary judge reasoned that this was consistent with Schering AG's rejection of a suggestion by P&V that the upper limit of the range be increased to 5 mg: see (i) and (j) below. (e) An undated two-page handwritten document with notes arranged in two columns, one headed "Anti-Contraceptive" and one headed "Hormonal Replacement". Both columns referred to DRSP and EE. The first column refers to a dosage of DRSP in the range 2 mg to 4 mg, and 3 mg. We observe that there was evidence before the primary judge that suggested that these notes were prepared at a meeting between Schering AG personnel (Dr Broesamle and Dr Ulrich) and Marianne Johansen of P&V on 14 January 1999. On the basis of this document, the primary judge found that either Schering AG or P&V was considering the optimum dosage of DRSP and EE for two compositions (the first being a contraceptive preparation and the second being a hormonal preparation), in which a dosage of DRSP of 2 mg to 4 mg, or 3 mg, was being considered in order to achieve "the optimum effect and the optimum side effect level". (f) Handwritten notes prepared by Dr Broesamle that he found behind a letter from P&V dated 19 January 1999. The primary judge found that these were notes of the meeting conducted with P&V on 14 January 1999 and that Dr Broesamle was considering 2 mg to 4 mg, or 3 mg, as the dosage range of DRSP for contraceptive purposes in a formulation which provided an optimum effect with minimised side-effects. The primary judge found that, at this meeting, Dr Broesamle informed P&V of his consideration of the possible dosage of DRSP for contraceptive purposes for optimum effect and side-effect level to be 2 mg to 4 mg, or 3 mg. The primary judge also found that Dr Broesamle was aware of some, or perhaps all, of Schering AG's research about DRSP because he could not have considered any range in a vacuum and Schering AG had already carried out substantial research about, and obtained other patents involving, DRSP. The primary judge further found that Dr Broesamle's handwritten notes contained other details, which were inconsistent with any inference other than that, as at 14 January 1999, he was well aware of DRSP, its solubility, operation and potential dosage levels for optimum contraceptive effect and side-effect levels. Given his responsibility for fertility and hormone patents, the primary judge concluded that Dr Broesamle would have familiarised himself with the information available within Schering AG about DRSP as part of P&V's briefing. (g) A facsimile from Schering AG to P&V dated 18 January 1999 enclosing "open questions" dated 13 November 1998, and the material needed to answer those questions in order to prepare a first draft of the priority documents on which the PCT application is based. (h) The first draft of a patent specification sent to Dr Broesamle and Dr Ulrich by letter from P&V dated 17 March 1999. The draft specification included the following description and questions: The claimed pharmaceutical composition is novel because it cites a narrower dosage range than that indicated in the prior art. The lower limit has been set to 2 mg per dosage unit/day because it is indicated to be the lowest dosage required to obtain ovulation inhibition. The upper limit has arbitrarily been set at 5 mg per dosage unit. Is that realistic, or could the limit be higher without causing adverse effects? What are the adverse effects at higher doses? This dosage level has been found to be particularly advantageous for the purposes of contraception as well as for other suggested therapeutic uses. Is this correct, or are the indicated dosage levels inappropriate for the therapeutic uses of the compound? (Emphasis in original.) The draft also included the following claim and question: A pharmaceutical composition comprising, as an active agent, 6β, 7β; 15β, 16β-dimethylene-3-oxo-17α-pregn-4-ene-21, 17-carbolactone (drospirenone) together with one or more pharmaceutically acceptable carriers or excipients, the amount of drospirenone corresponding to a daily dosage, on administration of the composition, of from about 2 mg to about 5 mg (what is a suitable upper limit of the dosage - that is, one at which the severity of adverse effects is not prohibitive?). (Emphasis in original.) From this, the primary judge found that P&V had considered the material about DRSP which had been provided to them and that they were querying whether the dosage range of DRSP could be more than 4 mg (as Schering AG had indicated at the 14 January 1999 meeting) and up to 5 mg or more without causing adverse side effects. The primary judge noted that no question had been raised about the 2 mg lower limit of the range. The primary judge concluded that P&V had accepted that 2 mg was the lowest dose of DRSP for optimum therapeutic effect for contraceptive purposes, but had queried whether the maximum dosage range which could be claimed was 4 mg (as had been discussed) or could be 5 mg or more, without adverse side effects. (i) A letter to Dr Ulrich sent by P&V on 26 August 1999, which asked for comments on the draft claims. Draft claim 1 referred to a dosage of DRSP from about 2 mg to about 4 mg. The primary judge found that, either in discussions or in a fax communication (known to be missing from the documentary record), Schering AG had instructed P&V that the dosage range of DRSP to be claimed was about 2 mg to about 4 mg. The primary judge further found that these instructions would have been conveyed by Dr Ulrich who, in turn, would have been acting on instructions from Dr Broesamle, given in consultation with the Head of Patents and Licensing at Schering AG. The primary judge found further that the instructions were given with the objective of achieving a formulation with optimal therapeutic effect with minimised side effects, consistently with the objective discussed at the meeting on 14 January 1999. The primary judge held: 124 … Given the contemporaneous references to optimal therapeutic effect and optimal (minimal) side effects, and the question raised by P&V about a possible 5mg dosage for the upper end of the range in the context of side effects, I consider that there is no reason to infer other than that Schering AG gave this instruction to P&V honestly believing that the instructed range would achieve these objectives. To infer otherwise, would suggest that Schering AG and P&V were in the business of making inaccurate and misleading contemporary notes of their discussions, in effect, identifying objectives of the dosage range which neither of them believed they were seeking to achieve in the patent specification and claims. This is nonsensical. (j) The draft specification with comments. This was sent to P&V by Dr Ulrich on 30 August 1999. The comments (referred to as "our comments") included details about Schering AG's patents, a dissolution test, and other matters. The primary judge noted that there was no comment about the 2 mg to 4 mg dosage range of DRSP. The primary judge found, however, that Schering AG was satisfied that this range was "appropriate and fulfilled the initially identified objectives of optimum therapeutic effect and optimum (meaning minimal) side effects". 91 The primary judge found that Schering AG had provided P&V with information on at least two occasions - before 17 March 1999 and before 26 August 1999 - and that the communications between Schering AG and P&V showed active consideration of a dosage range to achieve the objectives of optimum therapeutic effect and minimal side effects. The primary judge reasoned that P&V must have had access to substantial information about DRSP to prepare even the first draft of the specification. 92 The primary judge further noted that the appellant's overall approach to these documents was to assume that Schering AG did not intend to identify a dosage range of DRSP that would meet the objective of optimum therapeutic effect with minimal side effects. In light of the evidence before her Honour, the primary judge rejected that approach. 93 In February 2002, P&V instructed FB Rice & Co to prepare the Australian patent application, based on the priority documents (which had been filed on 31 August 1999) and the PCT application. The Australian patent application was filed on 14 March 2002. The prosecution of that application was thereafter undertaken by Davies Collison Cave. The primary judge noted that, so far as the 2 mg to 4 mg dosage range is concerned, there was no suggestion that the preparation, filing or prosecution of the Australian patent application was carried out other than in good faith and on the basis of reasonable skill and knowledge.

Second category 94 The documents in the second category (other documents created within Schering AG, which pre-dated the PCT application) comprise: (a) a report entitled "Controlled study on pharmacodynamics and pharmacokinetics of the combination drospirenone/ethinylestradiol over 3 months with Microgynon® as a reference" dated 1 July 1993 (the 9274 report); and (b) each of the five reports identified in the table dated August 1998 referred to at [90(a)] above. 95 The 9274 report concerns a study to investigate various pharmacodynamic effects and the pharmacokinetics of multiple doses of DRSP when administered orally in combination with EE, compared with a marketed standard preparation containing levonorgestrel and EE (Microgynon). The study was designed as a single-centre, double-blind, randomised and controlled interindividual group comparison. A total of 27 healthy fertile women aged 19 to 42 years participated. Each was assigned to one of three treatment groups: Microgynon (0.15 mg levonorgestrel with 0.03 mg EE per day); 2 mg DRSP with 0.03 mg EE per day; and 4 mg DRSP with 0.03 mg EE per day. All treatments were administered for three cycles (3 x 21 days followed by a treatment-free interval of seven days). 96 The study concluded that, from a safety perspective, all treatments were well-tolerated. The study provided evidence of the antiovulatory and antimineralocorticoid properties of 2 mg and 4 mg DRSP in combination with 0.03 mg EE. 97 The 9274 report was referred to in an affidavit made by Dr Broesamle on 19 August 2016. Whilst describing the study referred to in the report, Dr Broesamle did not explain where or how he located the report. 98 Before the primary judge, Generic Health submitted that the 9274 report could not shed any light on the s 115(1)(a) issue because there was no proof that it was relied on, or even known, by those framing the specification (in particular, those involved in the selection of the 4 mg upper limit). 99 The primary judge rejected this submission. Her Honour found that, on its face, the 9274 report describes a sophisticated system for information control and retention. Further, the study must have involved more than a trivial cost to Schering AG. The primary judge noted that the 9274 report deals with the effect of DRSP as an aldosterone antagonist, an effect referred to in the specification: see [61(f)] above. 100 The primary judge concluded at [141]: Weighing up all of the circumstances, I am unable to accept that in instructing P&V, Schering AG had forgotten the information in the 9274 report and overlooked the existence of that report. Information control must be vital to an innovator pharmaceutical company. Schering AG was working on DRSP for years, carrying out various studies, and filing various patents along the way. Yet Generic Health would have it that a clinical trial directed at the specific quality of the aldosterone antagonist effect of DRSP compared to another contraceptive which concluded that the combination of DRSP and EE tested had a clearly demonstrable aldosterone antagonist effect, was not considered by Schering AG when it instructed P&V to prepare a patent application for the same compounds (DRSP and EE) claiming the very same aldosterone antagonist effect. The notion that Schering AG would overlook one of its own clinical trials, let alone one that was concerned with comparing DRSP with a commercially available contraceptive to test the aldosterone antagonist effect of DRSP when it was arranging for a patent to be prepared for DRSP as a contraceptive with an aldosterone antagonist effect, is one I am unable to accept. 101 On the basis that Schering AG did consider the 9274 report when framing the pre-amendment specification, the primary judge did not accept Generic Health's submission that the report did not provide proper scientific support for the 4 mg upper limit. Here, the primary judge reasoned that the ratio of DRSP to EE was important to Schering AG's "objective of minimal side effects". EE acts to increase water retention, and thus possibly blood pressure. As an aldosterone antagonist, DRSP acts as a diuretic to reduce the water retention which EE promotes. 102 The primary judge found at [144]: Thus, to submit as Generic Health does that the 9274 report says nothing about diuresis, when the report is concerned with the aldosterone antagonist effect of DRSP, is to miss the point of the report. The 9274 report reported that both dosages (2mg DRSP + 30µg EE; 4mg DRSP +30µg EE compared to Microgynon which has 0.15mg levonorgestrel + 30µg EE) were "well tolerated, subjectively and objectively", with "no adverse drug reactions" and the aldosterone antagonistic effect was clearly demonstrated for DRSP in both doses with initial increased sodium excretion being later neutralised by the body's counter regulation. No one with the expertise within Schering AG or P&V would have failed to appreciate that the report was dealing with water retention and removal by diuresis. The dosages being well tolerated must have meant that the water retention effects of EE were effectively countered by the DRSP without excessive diuresis at either dose. 103 The primary judge noted Generic Health's submission that the 9274 report does not involve testing doses higher than 4 mg DRSP. Her Honour reasoned, however, that Generic Health's reliance on this submission was misconceived because it proceeded on the basis that the only way that a pharmaceutical innovator, in good faith and with reasonable skill and knowledge, could decide upon a maximum dosage range was to conduct a test which showed adverse side effects and then, in effect, "conduct a precise titration exercise, lowering the dose by degrees to find the highest absolute dose at which the effect is avoided". The primary judge said that this approach "has nothing to commend it"; it ruled out the role of sensible prediction by the person skilled in the art. 104 The primary judge rejected other criticisms advanced by Generic Health as to the weight that should be given to the 9274 report. For example, even though the 9274 report says that the ovulation inhibitory effect of the two tested formulations had not been conclusively proved (although the results indicated that ovulation was probably inhibited), the inhibition of ovulation was dealt with in the reports identified in the August 1998 table. 105 The primary judge then turned to consider these reports, by noting: (a) The 9692 report concerns a study of two formulations - 2 mg and 3 mg DRSP, each with 0.03 mg EE. No differences (worth mentioning) were observed between the two formulations. A distinct aldosterone-antagonistic effect was demonstrated for each, and each was equally well-tolerated. (c) The 9693 report concerns a study to test the ovulation inhibition of two formulations - once again, 2 mg and 3 mg DRSP, each with 0.03 mg EE. The report concludes that, although no differences were demonstrable between the two preparations, a 3 mg formulation should be developed because an ovulation-inhibiting effect was demonstrable in all cases examined with this preparation. The report states that it was to be expected that the contraceptive reliability of a 3 mg formulation will be greater than that of a 2 mg formulation on wide-spread use. The report concludes that the 2 mg formulation was in the threshold region of ovulation inhibition. We note that this report provides information used in Example 5 in the specification. (d) The 9970 report concerns a study of formulations containing 3 mg DRSP with three varying dosages of EE, compared with Microgynon. We observe that, based on the clinical results, the report recommends that the formulation containing 0.03 mg EE should be selected for further development, although the differences between the three formulations were small and not statistically significant. We also observe that the report notes that a distinct aldosterone-antagonistic effect was demonstrable for all formulations. Blood pressure and body weight remained constant or showed a slight tendency to fall. (e) The A187 report concerns a further study of the three formulations (referred to immediately above) using 3 mg DRSP. We observe that, once again, the report concludes that the differences between the three formulations were small and not statistically significant, but that the combination of 3 mg DRSP and 0.03 mg EE was regarded as the trial formulation with the most prominent trend towards favourable clinical efficacy and tolerance. (f) The A892 report concerns a comparison between formulations containing 0.5 mg, 1 mg, 2 mg, and 3 mg DRSP. The report states that adequate ovulation inhibition was found only with the 2 mg and 3 mg formulations. The report states that the results of the study confirm what had already been shown in earlier studies - with 2 mg DRSP, the formulation lies in the threshold range of ovulation inhibition. The report concludes that a formulation with 3 mg DRSP provides a safety margin and could be recommended for use as an oral contraceptive in combination with estrogen.

Third category 106 The primary judge noted the patents referred to in the PCT and Australian specifications and found that, given his role, Dr Broesamle must have been involved in the preparation of all the applications for these other patents.

The primary judge's conclusions 107 The primary judge did not accept Generic Health's submission that the first respondent had not discharged its onus of proof to establish that the pre-amendment specification was framed in good faith and with reasonable skill and knowledge so far as concerns the selection of the dosage range of 2 mg to 4 mg DRSP. 108 A feature of Generic Health's case before the primary judge was that the pre-amendment specification contains a promise that the side effects of the formulation would be avoided at the maximum dosage of 4 mg DRSP. The primary judge noted, however, that the pre-amendment specification made no such unqualified promise. Rather, the promise was qualified by words such as "may", "preferred" and "substantially". 109 Further, the primary judge accepted Bayer's submissions that: (a) Schering AG was in possession of the 9274 report, which demonstrates the effectiveness and safety of DRSP within the range of 2 mg to 4 mg. This information would have been known to the inventors. (b) Schering AG and P&V were in possession of numerous experimental reports demonstrating that there was no statistically significant difference between DRSP dosages of 2 mg and 3 mg. (c) Dr Broesamle met with P&V and discussed the dosage range of 2 mg to 4 mg DRSP. (d) P&V initially proposed a dosage range of 2 mg to 5 mg DRSP, and queried whether the upper limit could be increased. (e) Schering AG reviewed P&V's draft specification and decided to reduce the claimed dosage range of DRSP to 2 mg to 4 mg, as had been discussed at the meeting on 14 January 1999: see [90(f)] above. 110 The primary judge concluded at [164]: Even if Bayer had somehow forgotten one of its own clinical trials as reported in the 9274 report (which seems highly improbable), it had available to it a wealth of information enabling it to make a sound prediction as to the upper limit of the dosage range of 4mg. The fact that P&V tried to put the range up to 5mg, which Schering AG must have rejected, is irreconcilable with any suggestion of lack of good faith or reasonable skill and knowledge in the framing of the unamended specification. As Bayer said, the evidence demonstrates "that the bounds of the range were determined very conservatively within the extensive knowledge accumulated by Bayer over the period of time its invention had been developed".

The grounds of appeal 111 Generic Health's appeal on this issue (Grounds 1 to 4) is directed to the selection of 4 mg as the upper limit of the dosage range of DRSP in the formulation described and claimed in the pre-amendment specification. The grounds are: 1. The primary judge erred in finding that the specification without the amendment was framed with reasonable skill and knowledge in that her Honour ought to have found that the Respondents (Bayer) did not discharge their onus of proving that the reference to the amount of "4 mg" of drospirenone in the claims was framed with reasonable skill and knowledge (Main Judgment at [161]-[164]). 2. The primary judge erred in finding or inferring that Bayer had considered the "9274 Report" in framing the specification (Main Judgment at [141]). 3. The primary judge erred in finding (if her Honour did so find) that Bayer made "a sound prediction", or relied upon information in making "a sound prediction", as to the upper limit of the claimed dosage range of drospirenone (Main Judgment at [164]). 4. The primary judge erred in finding (if her Honour did so find) that the fact that the upper limit to be claimed in respect of the dosage range for drospirenone had been considered by those responsible for framing the specification was sufficient to establish that the specification was framed with reasonable skill and knowledge (Main Judgment at [122], [124], [163(c)]).

Generic Health's submissions 112 In summary, Generic Health argued that Bayer's case before the primary judge relied on her Honour drawing inferences. No contemporaneous witness gave evidence concerning the selection of 4 mg DRSP as the upper limit of the range and, according to Generic Health, the documentary record before the primary judge was both incomplete and inconclusive on that matter. Generic Health submitted that the evidence did not enable the primary judge to draw the inferences her Honour did, and was insufficient to discharge the onus cast on Bayer to show that, in this particular respect, the pre-amendment specification was framed with reasonable skill and knowledge. 113 Generic Health also submitted that the validity of a patent granted on an unamended specification is not the touchstone of whether reasonable skill and knowledge has been deployed. A patent granted on an unamended specification may have been valid, but nonetheless drafted without reasonable skill and knowledge. In this connection, Generic Health submitted that an amendment to a patent specification is an indulgence and that s 115(1)(a) is "one of the mechanisms which incentivises patentees to frame their granted patents carefully and properly from the outset, rather than falling back on the possibility of amendment at a later stage". Thus, according to Generic Health, if, for whatever reason, a patentee has insufficient evidence of the framing process to establish that it was done with reasonable skill and knowledge, "this is the patentee's cross to bear". The infringer's conduct does not bear upon how s 115(1)(a) is to be applied. 114 More specifically, Generic Health submitted that claiming a suitable compositional range in pharmaceutical patents is important, including because the utility requirement for validity applies to "everything within the scope of a claimed invention". It submitted that the primary judge's findings, which we have recorded at [77(b)-(e)] above, indicate the type of evidence that is pertinent to the s 115(1)(a) requirement in this case. However, according to Generic Health, the primary judge appears not to have taken this evidence into account. 115 Generic Health noted that, even though Dr Broesamle had no independent recollection of the framing process in relation to the pre-amendment specification - a matter about which it raises no complaint - he was silent on how he and Bayer chose the claimed dosage ranges and on whether (and, if so, how) he and Bayer, in general, made predictions in claiming dosage ranges. Generic Health further submitted that Dr Broesamle was also silent on Bayer's general record-keeping practices, including with respect to the retention and accessibility of previous studies that had been conducted. It submitted that evidence of these matters, if adduced, would have been admissible as bearing on the probabilities that a general practice was followed in respect of the framing of the pre-amendment specification: Connor v Blacktown District Hospital [1971] 1 NSWLR 713 at 721. It submitted that the failure to adduce such evidence in the present case bears upon the inferences that could be drawn from the documentary record before the primary judge. 116 In this connection, Generic Health submitted that the documentary record itself does not support a finding that reasonable skill and knowledge was exercised. Its argument commenced with the 9274 report. Generic Health accepted that the 9274 report provides proper scientific support for the 4 mg upper limit and that, had Bayer relied on the report in framing the specification, its (Bayer's) case in relation to s 115(1)(a) would have been made out. Generic Health further submitted that, if Bayer had considered the 9274 report, it would have appreciated that it was dealing with water retention and its removal by diuresis. 117 However, Generic Health submitted that the primary judge erred in finding or inferring that Bayer had considered the 9274 report in framing the specification. Generic Health submitted that there was no fact which positively suggests that the report was relied on: see, for example, the observations of Gordon J in Re Day [2017] HCA 2; 340 ALR 368 at [18]. Further, the material available to the primary judge strongly indicated that the report was not relied on. Generic Health advanced four reasons in support of this submission: (a) First, the 9274 report was merely annexed to a late-filed affidavit made by Dr Broesamle without Dr Broesamle stating that the 9274 report was relied on in framing the specification. Perhaps more importantly, Generic Health submitted that Dr Broesamle did not explain who found the report, or how or where it was found, in circumstances where no evidence was given as to how Bayer stores its reports. Generic Health submitted that, given the absence of evidence of these last-mentioned matters, the primary judge ought to have been less likely to draw favourable inferences from other evidence (Kuhl v Zurich Financial Services Australia Ltd [2011] HCA 11; 243 CLR 361 at [63]) or, indeed, should not have drawn favourable inferences at all, when Bayer made no attempt to prove these matters by direct evidence: Commercial Union Assurance Company of Australia Limited v Ferrcom Pty Ltd (1991) 22 NSWLR 389 at 418. Generic Health pointed to other evidence (which it described as showing Bayer's "haphazard discovery" of documents), and the absence of evidence (certain documents discussed below at [122] to [129]), as showing that Bayer's record-keeping was far from being as organised as the primary judge found it to be. (b) Secondly, contemporaneous documents - such as the handwritten document described at [90(d)] above - make no reference to the 9274 report, even though they referred to other reports. (c) Thirdly, notwithstanding discovery given by reference to categories (including a category covering drafting instructions for the selection of the 2 mg to 4 mg range), the 9274 report was not discovered. Generic Health submitted that, when the 9274 report was later adduced in evidence through Dr Broesamle, it was not suggested that this was as a result of Bayer's ongoing discovery obligations. (d) Fourthly, Generic Health contrasted the use of the word "found" in relation to the lower limit of 2 mg with the use of the word "identified" in relation to the upper limit of 4 mg, in the passage of the pre-amended specification quoted at [62] above. Generic Health submitted that, if the 9274 report had been relied upon, it is reasonable to assume that the word "found" would also have been used in relation to the upper limit of 4 mg. 118 Generic Health submitted that it is not enough that Schering AG was in possession of the 9274 Report or that the information in it was known to the inventors. The question is whether the 9274 Report was relied on when framing the pre-amendment specification, and there is no fact which shows that it was. 119 Generic Health submitted that the other documents referred to by the primary judge provide no contemporaneous evidence, based on knowledge and skill, that the 4 mg upper limit meets the promises of the specification concerning efficacy and the substantial avoidance of unwanted side effects. Generic Health argued that, on the evidence before the primary judge, the basis on which the upper limit was chosen remains unexplained. Apart from the 9274 report, there is no evidence to suggest that an investigation in respect of 4 mg or any upper limit was even undertaken. 120 In this connection, Generic Health submitted that, taken together, the other documents only show that, having considered the dosage range to claim, Bayer merely selected 4 mg as the upper limit at which, in its view, side effects would be substantially avoided. According to Generic Health, this is not enough, even if it does involve the rejection of the upper limit of 5 mg DRSP suggested by P&V. This is because mere discussion and selection of an upper limit does not establish reasonable skill and knowledge in framing that limit. As Generic Health put it, the fact that Schering AG "decided not to push the limit any higher does not mean that the limit it did settle upon was based on reasonable skill and knowledge." 121 Further, Generic Health submitted that, to the extent that the primary judge relied on experimental reports testing 2 mg and 3 mg DRSP (which demonstrated that there was no statistically significant difference between those two formulations) to conclude that Schering AG and P&V had information which enabled a sound prediction to be made as to the upper limit of 4 mg, her Honour erred. Generic Health submitted that this conclusion was not open on the evidence. First, there is no evidence to suggest that such a "sound prediction" was made, let alone that these reports were relied on for that purpose. Secondly, even if a prediction was made, there is no evidence to suggest that it would have been indicative of reasonable skill and knowledge. Thirdly, the reports on the 2 mg and 3 mg DRSP formulations do not support or justify the selection of the 4 mg upper limit or demonstrate how the invention could be framed as involving the identification of a maximum dosage of 4 mg DRSP at which unpleasant side effects, in particular excessive diuresis, may substantially be avoided: see the passage in the pre-amendment specification quoted at [62] above. Generic Health noted, in this connection, that these reports did not, on their face, test for diuresis. 122 Generic Health submitted that these difficulties are compounded by the fact that two "potentially critical" documents are missing from the evidence. 123 The first was described by Generic Health as "material sent by Bayer to P&V in response to P&V's letter of 14 January 1999 regarding the specification". This description does not appear to be correct. So far as we can see, there is no letter from P&V dated 14 January 1999. There is a letter from P&V dated 19 January 1999, which thanks Dr Broesamle and Dr Ulrich "for a very pleasant and interesting meeting in Berlin on 14 January 1998" (obviously meaning 14 January 1999). This is the meeting at which Dr Broesamle made the notes described at [90(f)] above. It is also the meeting at which, it seems, the handwritten notes described at [90(e)] above, were made. The letter continues: We look forward to cooperating with your [sic] regarding the drawing up of a new patent application and expect to receive further material from you by the end of January. 124 Generic Health referenced its submission to [119] of the Damages Judgment where, after referring in [118] to the letter from P&V dated 17 March 1999 enclosing the first draft of the specification (see [90(h)] above), her Honour said: Given the detail already within the first draft of the specification and the questions posed by P&V, I infer that Schering AG must have sent to P&V a considerable amount of material relevant to the proposed patent application, consistent with the terms of the 14 January 1999 letter. 125 It seems to us that the primary judge's reference in this passage to the 14 January 1999 letter was mistaken and that, most likely, her Honour intended to refer to the letter dated 19 January 1999. So understood, the point made by the primary judge at [119] is that Schering AG must have sent, presumably after the meeting on 14 January 1999 (as referred to in the letter dated 19 January 1999), a considerable amount of material to P&V, which enabled them to produce the first draft of the specification and to make the following comments in the 17 March 1999 letter: Please find enclosed a first draft of the claims of the proposed new patent application relating to formulations of drospirenone. In addition, we have briefly outlined the proposed structure of the description, including the elements we believe are essential to provide novelty and inventive step of the claimed invention. We have also made suggestions as to the experimental work to be included in the examples. Please let us know if you consider the suggested experimental reports to be appropriate for this purpose, or if you would rather have us include other data in the application. We are looking forward to further discussions with you on this matter. In particular, your comments on the claims as well as our questions will be greatly appreciated. 126 Generic Health's point is that the material sent by Schering AG after the 14 January 1999 meeting is missing from the evidence. 127 The second document referred to by Generic Health is a fax sent to P&V by Dr Ulrich between 17 March and 26 August 1999. In their letter to Schering AG (for the attention of Dr Ulrich) dated 26 August 1999, P&V (Ms Johansen) said: Thank you for your fax with details regarding DRSP. We now forward a draft set of claims and ask you kindly to let us have you [sic] comments to these as soon as possible. We look forward to hearing from you. 128 Generic Health submitted that there is no evidence as to why these documents are missing. There was no direct evidence as to Bayer's record-keeping and, according to Generic Health, there is a general lack of clarity as to the nature and scope of the searches undertaken into Bayer's records. 129 Generic Health contrasted the documentary record with respect to the selection of the 4 mg upper limit with the documentary record supporting the selection of the 2 mg lower limit which, it submitted, does represent knowledge from which it was reasonable to frame the lower limit. 130 Generic Health submitted that principles governing the onus and standard of proof must be applied faithfully: Australian Securities and Investments Commission v Hellicar [2012] HCA 17; 247 CLR 345 at [165] per French CJ, Gummow, Hayne, Crennan, Kiefel and Bell JJ. It submitted that reasonable satisfaction should not be produced by inexact proofs, indefinite testimony or indirect inferences: Briginshaw v Briginshaw [1938] HCA 34; 60 CLR 336 at 362. The implication of this submission is that the primary judge did reach her Honour's conclusions based on inexact proofs, indefinite testimony or indirect inferences.

Analysis 131 The thrust of Generic Health's case on appeal is that, in order to discharge the burden cast on it by s 115(1)(a) of the Act, Bayer must establish that, at the time that the pre-amendment specification was drafted, the drafter had before him or her the very documents or information that would be necessary to justify or substantiate scientifically all features of the invention described and claimed. Thus, the scientific justification for or substantiation of the upper limit of 4 mg DRSP (as the dosage at which side effects can be substantially avoided) - as shown in test reports and the like - were required to be before Dr Broesamle and P&V, or actually known by them, at the time that Dr Broesamle gave instructions to P&V to draft the specification with reference to the 2 mg to 4 mg dosage range of DRSP. Absent that proof, Generic Health contended that it could not be the case that the pre-amendment specification was framed with reasonable skill and knowledge. It is for this reason that Generic Health placed significance on the inferences of fact made by the primary judge as to the documents and information before Dr Broesamle and P&V at the time that the priority documents and the PCT application were drafted, and whether those facts were properly found. 132 This challenge is misplaced. As we recorded at the outset, it proceeds on a fundamental misunderstanding of the obligations placed on a patent applicant and a patentee with respect to the drafting of a complete specification, as well as on a fundamental misunderstanding of the scope and purpose of s 115(1)(a) of the Act. 133 Take first the legal obligation with respect to a complete specification. The obligation of an applicant for a standard patent is to provide a complete specification that complies with s 40 of the Act. Thus, the patent applicant must provide a complete specification that describes the invention fully, including the best method known to the applicant of performing the invention. The specification must end with a claim or claims that define the invention. The claim or claims must be clear and succinct, and fairly based on the matter described in the specification. (We speak here, relevantly, of patents granted before the amendments introduced by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (Cth)). It is not, however, an obligation of the patent applicant to explain how the invention was made or the theoretical basis underlying any stipulated integer: NV Philips Gloeilampenfabrieken v Mirabella International Pty Ltd [1992] FCA 493; 24 IPR 1 at 27. It follows, necessarily, that the patent applicant does not need to supply the proof that the invention works. And if the patent applicant does not need to supply that proof, it does not need to have that proof for the purposes of drafting the specification. As Fletcher Moulton LJ explained in British Ore Concentration Syndicate Ld v Minerals Separation Ld (1909) 26 RPC 124 at 144: An inventor patents a process and not its scientific explanation. He may not, and in many cases does not, know the modus operandi of Nature in bringing about the results that he obtains, and I know inventions in which, to this day, it is a matter of controversy as to how they act to produce their results. 134 The test of whether a complete specification fully describes the invention, as required by s 40(2)(a) of the Act, was stated authoritatively in Kimberly-Clark Australia Pty Ltd v Arico Trading International Pty Ltd [2001] HCA 8; 207 CLR 1 at [25]: The question is, will the disclosure enable the addressee of the specification to produce something within each claim without new inventions or additions or prolonged study of matters presenting initial difficulty? 135 Plainly, this statement of the legal requirement is not concerned with either the scientific justification for, or the scientific proof of, the invention, but with the patent applicant's obligation to enable the public to work the invention, whatever its scientific or technical underpinnings, by a sufficient description of what the invention is. This legal requirement is not transformed into a requirement to be in possession of scientific proof or substantiation of the invention simply because the patent applicant or, in the case of a granted patent, the patentee, seeks to amend a complete specification that has been filed. 136 To what end is s 115(1)(a) directed? The evident purpose of s 115(1)(a) is to be found in the reason expressed for introducing s 20 of the 1883 Act (UK). As we have noted, this was to balance the interests of the patentee and the public. As Jacob LJ remarked in Unilin, the provision should not be approached as a measure of punishment to be inflicted on the patentee. However, the patentee should not take advantage of its own conduct that is found to be wanting when measured against the standards of good faith and reasonable skill and knowledge. By the same token, the infringer should not be permitted to seek refuge from the consequences of its own infringing conduct when the patentee establishes that it has met the relevant standard in drafting the specification prior to amendment. 137 The requirement for "reasonable" skill and knowledge bespeaks the objective standard of reasonableness familiar in English law. The matter is put beyond doubt by the Court of Appeal in Kirin-Amgen/Transkaryotic. Such a standard must be measured against all the circumstances of the case. This, no doubt, is the reason why, in Chiron, Aldous J went on to emphasise that the specification must be one that would be produced with "the patentee's knowledge of the invention". The same notion of reasonableness is evident in Laddie J's reasons for judgment in Hoechst, where his Honour remarked that the standard of reasonable skill and knowledge must be one that takes into account "what the patentee knows". Of course, "what the patentee knows" must also encompass what reasonableness requires the patentee to know in the circumstances. Once again, the matter is put beyond doubt in Kirin-Amgen/Transkaryotic. 138 The burden of these observations is that the requirement of reasonable skill and knowledge - and no less the requirement of good faith - focuses on knowledge that the patentee had, or reasonably should be taken to have had, at the time the specification (before amendment) was framed, not simply information that the patentee might choose to place before the drafter. As Neuberger J remarked in Kirin-Amgen, the drafter of the specification must be taken to have been "properly instructed", as well as professionally competent and honest. If this were not so, the standard of reasonableness against which the patentee's conduct is to be judged would be imperfect and would not meet the evident purpose of s 115(1)(a). By the same token, the patentee is not required to actually put before the drafter the scientific proof of, or the scientific justification for, its invention in order for the specification to be framed. For the purposes of s 115(1)(a), the drafter is taken to have the patentee's knowledge of the invention, or knowledge that the patentee reasonably should be taken to have had given its circumstances. The exercise of reasonable skill is to be judged against that body of knowledge, as well as reasonable knowledge of the law and practices relating to the drafting of patent specifications. 139 In the present case, Generic Health did not dispute the sufficiency of the specification for the purposes of s 40(2)(a) of the Act - it being remembered that the amendments made by the orders on 14 December 2012 were to replace the (then) existing claim set with a new claim set. The body of the specification, including its description of a composition/preparation containing 2 mg to 4 mg DRSP, remained, post-amendment, in its pre-amendment form. As we have noted, Generic Health did not allege, either at the time of the amendment application or at the time of the liability hearing before the primary judge, that such a composition/preparation was inutile or that, based on the pre-amendment specification, the patent was obtained by fraud, false suggestion or a misrepresentation: see s 138(3)(e) of the Act. As recorded by Yates J in the Amendment Judgment, the amendments were sought to forestall an argument in the case, then to come before the primary judge, that the dissolution test claims of the patent were not fairly based on the matter described in the specification (an argument which Bayer did not accept would be soundly based in any event). There was no evidence to suggest that the amendments had been sought to overcome any prior art or to address any other asserted, apparent or possible ground of invalidity. 140 Further, Generic Health does not now contend that there is no proper scientific support for the selection of a dosage range of 2 mg to 4 mg DRSP. It accepts that there is scientific support for both the lower limit and the upper limit. As to the upper limit, it accepts that the 9274 report provides that support. This report was plainly within Schering AG's knowledge. Generic Health's only point is that Bayer has not proved that the 9274 report was actually before Dr Broesamle and P&V when the pre-amendment specification was drafted or that, at that time, Dr Broesamle and P&V were actually cognisant of that report. Further, it submits that Bayer has not proved that it made a prediction as to the 4 mg upper limit based on the other reports that were actually before Dr Broesamle and P&V, or that such reports were sufficient to make such a prediction in any event. 141 Generic Health's challenge misses the mark. When assessing whether reasonable skill was exercised, s 115(1)(a) requires the assessment to proceed on reasonable knowledge and, as we have explained, reasonable knowledge is not limited to documents actually before the drafter of the specification. It is knowledge that the patentee had or reasonably should be taken to have had - whether or not the drafter, or those providing instructions to the drafter, was or were cognisant of that knowledge or its source at the time when the specification was framed. Where, of course, the patent applicant or the patentee is a body corporate, the relevant corpus of knowledge is corporate knowledge. Further, the drafter is taken to have been "properly instructed". There is no doubt that, at the time when the pre-amendment specification was framed, the 9274 report was knowledge which Schering AG had as to the appropriateness of the dosage range of 2 mg to 4 mg DRSP. In those circumstances, it cannot be said that the selection of that range, at that time, was not based on reasonable knowledge. 142 Thus, it is a diversion for Generic Health to challenge the primary judge's conclusion that the requirements of s 115(1)(a) have been met by arguing that her Honour erred in finding or inferring that Bayer (Schering AG) had considered the 9274 report in framing the specification. The primary judge correctly found that Schering AG was in possession of the 9274 report, which demonstrated the effectiveness and safety of DRSP within the range of 2 mg to 4 mg, and that this information would have been known to the inventors. Once that fact is accepted, Generic Health's challenge to the primary judge's ultimate conclusion cannot succeed. In fact, the primary judge went further and found that three of the named inventors were involved in the process of preparing the priority documents on which the PCT application was based and on which, in turn, the pre-amendment specification was based: see at [102] and [117]. In the circumstances, it is not necessary for us to decide whether the primary judge was correct to conclude that Schering AG had available other information that would have enabled it to make a sound prediction as to the upper limit of 4 mg. We do observe, however, that the evidence before the primary judge justified her conclusion at [110] and [132] that, at the time of, and in the context of, framing the pre-amendment specification, Schering AG was actively considering the upper limit of the dosage range of DRSP in a way that was consistent with the objective of minimising side effects. The evidence shows, persuasively, that the upper limit considered to be appropriate to achieve that objective was 4 mg. As we have explained, if that conclusion was arrived at by prediction, then whether the prediction was sound (in the sense of justified as a matter of scientific proof) is not to the point. No such obligation of justification is cast upon a patent applicant or patentee in drafting a complete specification. 143 Before leaving this aspect of the appeal, we record that, in the course of oral argument, Generic Health raised the question whether some aspects of Bayer's submissions required it to file a notice of contention. This included Bayer's reliance on its (and Schering AG's) accumulated corporate knowledge. While it is true that, in considering the application of s 115(1)(a), the primary judge directed attention to whether the 9274 report was consciously considered in the course of drafting the pre-amendment specification, it seems to us that her Honour's ultimate conclusion (that Bayer had discharged the onus cast on it by s 115(1)(a)) was founded, at least in part, on what Bayer corporately knew. This is made clear by [163] and [164] of her Honour's reasons and, in particular, her finding at [163(a)] that Bayer was in possession of the 9274 report and that the information in that report would have been known to the inventors. This finding was plainly based on a passage in Bayer's written submissions at trial. Further, at trial, Bayer and Generic Health each referred to a number of the United Kingdom authorities we have discussed. Indeed, Generic Health urged on her Honour the correctness of Hallen, Chiron, Kirin-Amgen and Kirin-Amgen/Transkaryotic: see, for example, at [36] to [37] of the primary judgment. In the circumstances, no notice of contention was required. 144 For these reasons, this aspect of Generic Health's appeal fails.

Lost sales of Yasmin (Grounds 5-9) 145 We now turn to Grounds 5, 6, 8 and 9 of the notice of appeal concerning the Damages Judgment, Ground 7 not being pressed. Grounds 5, 6, 8 and 9 are expressed in the following terms: 5. The primary judge erred in holding that Bayer's damages were to be assessed on the basis that every sale of Isabelle was a lost sale of Yasmin (Main Judgment at [345]) (the Lost Sales of Yasmin Conclusion). 6. The primary judge erred in failing to apply the five-step approach to discounting that is used where the infringer and the intellectual property rights holder compete in the same market and the intellectual property rights holder's claim is for lost profits (Main Judgment at [227]-[229]). … 8. The primary judge erred in finding, and relying upon the finding, that third generation oral contraceptives (OCs) other than Yasmin and Isabelle had prices "ranging from around $71 to $84 for a packet of three cycles" (Main Judgment at [206]) in circumstances where there was evidence that cyproterone acetate brands such as Brenda, Carolyn, Estelle, Juliet and Laila, being other third generation OCs, had prices ranging from $18.90 to $39.10 for a packet of three cycles. 9. Further or in the alternative to grounds 5 and 6, the primary judge erred in holding that it was unnecessary for the Court to attempt to quantify the size of any discount to be applied in respect of the Lost Sales of Yasmin Conclusion (Main Judgment at [299], [301], [302]). 146 Her Honour held that Bayer's damages were to be assessed on the basis that every sale of Isabelle was to be treated as a lost sale of Yasmin, with no discount to be made to reflect any uncertainty in the relevant hypothetical counterfactual, namely, that if a particular sale of Isabelle had not been made, then an equivalent sale of Yasmin would have been so made. But Generic Health submitted that her Honour erred in that conclusion by not making a discount for such uncertainty. It submitted that we ought to apply a discount rate in the order of 10% to 15%, which would result in a reduction in the damages award of $2.4 to $3.6 million (excluding interest), not including the portion of the award made in respect of the Petibelle sales. 147 The purpose of an award of damages for patent infringement is compensatory, rather than punitive: Bailey v Namol Pty Ltd [1994] FCA 733; 53 FCR 102 at 110-111 and the other cases collected in Seafolly Pty Ltd v Fewstone Pty Ltd [2014] FCA 321; 313 ALR 41 at [506]. The appropriate quantum of damages is that which will restore the patentee to the position it would have been in had the infringement(s) not occurred: General Tire and Rubber Company v Firestone Tyre and Rubber Company Limited (1976) 93 RPC 197 (General Tire (1976)) and the other cases cited in Pacific Enterprises (Aust) Pty Ltd v Bernen Pty Ltd [2014] FCA 1372; 321 ALR 715 at [5]. But when one comes to assessing damages based upon a hypothetical counterfactual, a question that arises is whether any discount should be made, and if so what, to reflect any uncertainty. We will return to that question later. 148 It is appropriate to begin by noting the following facts, most of which were the subject of the primary judge's findings at [197] et seq and not disputed before us unless we indicate otherwise. 149 First, women could only purchase OCs at a pharmacy, with all OCs requiring a prescription from a doctor. Doctors would prescribe originator brand names, and were not generally aware of generic versions of brands unless a patient specifically requested them. 150 Secondly, and as her Honour found at [208], whether before or after Isabelle was introduced, doctors would have prescribed Yasmin if they intended the patient to take the desired formulation of EE and DRSP. 151 Thirdly, Yasmin consisted of 3 mg DRSP and 0.03 mg EE. It offered a unique compound. When first introduced in Australia, it was the only OC with DRSP as the progestogen component. This offered non-contraceptive benefits which were not otherwise available in the one OC, including by reason of DRSP's aldosterone-antagonist effect, thereby avoiding increased plasma volume and body weight and symptoms such as breast tension, headache, bloating and increased blood pressure. From 23 January 2012, Isabelle and Yasmin were the only products on the market with 3 mg DRSP and 0.03 mg EE. 152 Fourthly, Isabelle was cheaper than Yasmin. Isabelle was sold by pharmacists to customers for around $60 to $61 for a three month supply and around $25 to $26 for a one month supply. Contrastingly, Yasmin was sold by pharmacists to customers for around $71 to $73 for a three month supply and around $28 to nearly $30 for a one month supply. 153 Fifthly, there was high compound loyalty by women to their OC, or, as the parties described it before us, "molecule" loyalty. Women did not like to 'chop and change' OCs, but loyalty was not absolute. 154 Sixthly, the OC market was crowded and split into two segments. One segment was third generation OCs, which were marketed for their side-effect profiles and benefits. These were not PBS-listed and were for the most part priced at a premium. The other segment was first and second generation OCs, which were PBS-listed (other than some second generation OCs) and which were not priced at a premium. Each of Yasmin and Isabelle was a third generation OC. Yasmin had a side-effect profile based on reduced weight gain, fluid retention and effect on mood. Other third generation OCs had different profiles, including lighter bleeding (Zoely and Qlaira) and better management of androgenisation, such as facial hair and acne. These included the cyproterone acetate combined OC brands (CPA brands). We will return to discuss the pricing of CPA brands later. 155 Generic Health submitted that the primary judge erred in various respects in not making an appropriate discount. Generally speaking, Generic Health's appeal grounds fall into two categories. The first category relates to what was said to be the price sensitivity of women concerning their purchase of OCs. Generic Health submitted that her Honour ought to have made a discount on the basis of such sensitivity to reflect the fact that a woman who had purchased Isabelle may not have necessarily purchased Yasmin if Isabelle had not been available. The second category relates to the appropriateness of the methodology that her Honour used. It is appropriate to begin with the first category of complaint. 156 Generic Health submitted that the primary judge misunderstood evidence, which showed that Bayer had engaged in conduct recognising the price sensitivity of patients to Yasmin, being: (a) the "Win/Lose" Scenarios document (the Win/Lose document), being an internal document dated 14 March 2013 prepared by Mr Peace, Bayer Australia's former Business Unit Manager, Women's Health Care Unit; and (b) the strategy adopted by Bayer of releasing Petibelle, the generic version of Yasmin. 157 Further, Generic Health pointed out that, under the heading "Outcome/behaviour", the Win/Lose document stated that, if Bayer won at trial and Isabelle was removed from the market, then "Pharmacy offer Yasmin at normal price or more likely suggest a switch to a CPA brand", thereby indicating a price sensitivity that might manifest itself by switching to a CPA brand. Contrastingly, the primary judge found that the existence of the Win/Lose document containing such a possibility was not unexpected, and gave it little weight. 158 Further, Generic Health submitted that the primary judge's conclusion of an absence of specific price sensitivity appears to have been infected by her Honour's overlooking of critical evidence, which showed that CPA brands featured many brands (for example Brenda, Carolyn, Estelle, Juliet and Laila) that had prices ranging from $18.90 to $41.45 for three cycles. It submitted that those prices were much lower than Yasmin's $71 to $73. We note that table 1.1 of Mr Houston's 8 July 2016 report relied upon by Generic Health just lists the CPA brands, not the prices; Mr Houston was an expert economist called by Generic Health. Further, in annexure FM-2 to Mr McCann's first affidavit, the position is unclear concerning Carolyn and Juliet. 159 Generic Health submitted that her Honour failed to make factual findings as to the prices of the third generation CPA brands and that they were far cheaper than Yasmin. In this respect, Generic Health submitted that the primary judge erred in respect of her findings at [206]: Third generation OCs are priced at a premium with the prices (leaving aside Isabelle and Petibelle) ranging from around $71 to $84 for a packet of three cycles (i.e. three months). 160 Generic Health submitted that the primary judge overlooked evidence concerning the price of CPA brands and that this is confirmed by what were said to be the erroneous findings that: (a) third generation OCs, which included CPA brands, had prices "ranging from around $71 to $84 for a packet of three cycles" at [206]; and (b) Isabelle was priced "below the price of any non-PBS third generation OC" at [238]; in relation to the reference to "non-PBS", her Honour found that at the relevant time all third generation OCs were non-PBS. 161 We agree that her Honour erred in this respect and that, accordingly, Ground 8 is made out. But where that takes Generic Health is another matter. We will discuss the significance of making out Ground 8 to the question of whether the damages award should have been discounted later. 162 Further, Generic Health submitted that the Win/Lose document was important internal evidence of Bayer's perception of the price sensitivity of consumers, with its significance elevated by the fact that it was prepared on 14 March 2013, before the judgment on liability and, accordingly, before the quantum proceedings were on foot. It is said that unguarded, internal and contemporaneous communications often provide the best evidence of the relevant state of mind or perception. Further, Generic Health submitted that the reference to the cheaper third generation CPA brands is significant. Generic Health submitted that it confirmed Bayer's appreciation of the competition it faced from the CPA brands, including on price. Generic Health contended that Bayer's concern as to the CPA brands can only have related to price, otherwise Mr Peace would have been concerned about pharmacists suggesting a switch to other third generation OCs, including those more expensive than Yasmin. Generic Health submitted that the price concern was a very real one to Bayer. Moreover, Mr Peace considered that it was "more likely" that a pharmacist would suggest switching to a CPA brand rather than purchasing Yasmin "at normal price". 163 Further, Generic Health submitted that Bayer's strategy in relation to Petibelle was also instructive competitive conduct concerning Bayer's perception of the competitive consequences of its own pricing and, in essence, the price sensitivity of women to Yasmin. 164 Bayer started selling Petibelle on 26 June 2014, a week after Generic Health was required to cease selling Isabelle. It was sold by pharmacists for about $66 for a three-month supply and nearly $22 for a one month supply, being approximately halfway between the Yasmin and Isabelle prices. Generic Health submitted that Bayer's expert economist, Dr Williams, accepted that, if Bayer believed that consumers were indifferent to a price difference of $5, being the rough price difference between Yasmin and Petibelle, then they would not have introduced Petibelle. According to Dr Williams, Petibelle's introduction suggested that there was "potentially a class of customers of Isabelle who would not be happy paying the higher Yasmin price and might shop elsewhere in the market". Generic Health submitted that the primary judge did not refer to this evidence. Instead, according to Generic Health, her Honour found at [300] that, in introducing Petibelle, "… Bayer's concern was that Isabelle had disrupted the market possibly introducing a price sensitivity that would not have existed but for Isabelle". 165 Further, and in any event, Generic Health submitted that her Honour's finding at [300] was not open on the evidence. According to Generic Health, Dr Williams readily accepted that, if consumers buy at a high price, and then have a lower price for a period, that does not necessarily prevent them from going back to the original price. Rather, so it was said, the people who were likely to be affected by the removal of the low price were people who were new customers at the low price, i.e. people who were never prepared to pay the high price, which would include some Isabelle customers. Accordingly, so it was submitted, the introduction of Petibelle demonstrated Bayer's acceptance that some customers were price-sensitive to Yasmin and, based on its price, would have preferred other OC brands instead of Yasmin. 166 Generic Health also submitted that Dr Williams accepted that there was a price-value trade-off for consumers between different contraceptive products. As applied to Yasmin, Generic Health submitted that Dr Williams accepted in cross-examination: (a) the possibility that some customers might desire the characteristics of Yasmin but not be willing to pay its retail price; and (b) it was likely that there was a class of customers who might have been prepared to pay the Isabelle price, but not the Yasmin price, although Dr Williams was not asked at that point about its size. 167 Dr Williams added that "… it has always seemed to me that this issue is about quantum. That is is it de minimis or is it something that's material[?]". According to Generic Health, this was a concession that a one-for-one approach, without any discount, was inapposite. Yet, so Generic Health complains, her Honour accepted and then relied on Dr Williams' earlier suggestion that "it is appropriate to assume that every cycle sold of Isabelle was one less cycle sold of Yasmin" (see at [295], [296] and [301]). 168 Further, Generic Health submitted that the primary judge incorrectly discounted the evidence of its witness, Mr McCann, who owned a pharmacy in New South Wales. Mr McCann's evidence was the only evidence from a pharmacist, or evidence at the 'point of sale'. Generic Health submitted that Mr McCann's evidence provided examples of women purchasing Isabelle in circumstances where they would not have purchased Yasmin, contrary to the last line of [247] of her Honour's reasons. 169 Further, Generic Health submitted that her Honour found at [272(1)] that, if there was no material change in a patient's financial circumstances, then the desire of a patient taking Yasmin or Isabelle "would not be based on price but continued therapeutic effectiveness". Generic Health submitted that this entailed a finding that Yasmin and Isabelle were highly price inelastic. But, Generic Health contended such a proposition was not supported by any evidence. Further, the finding was contrary to the evidence of the expert economists and what had in fact occurred in the market. Generic Health argued that if Yasmin was price inelastic, customers would not have purchased Isabelle in its place. 170 It is appropriate to now turn to the other category of challenge relating to the methodological approach used by the primary judge. 171 More generally, Generic Health submitted that where, as here, the parties have competed in the same market and the claim is for lost profits, it is appropriate to apply the following "five-step approach" (see, for example, Norm Engineering Pty Ltd v Digga Australia Pty Ltd [2007] FCA 761; 162 FCR 1 at [266] to [271] per Greenwood J), being to: (a) examine the number of the infringer's sales; (b) assume the infringer was trying to capture sales from the patentee; (c) assume the hypothetical counterfactual that the number of sales made by the infringer was equal to the number of sales lost by the patentee on the assumption that the patentee was the market leader; (d) discount the number of lost sales in (c) to reflect the fact that not all sales made by the infringer could be considered sales lost by the patentee; and (e) apply any further necessary discounts for other factors and contingencies. 172 For present purposes concerning the debate before us, the relevant integer to focus on is integer (d). 173 Generic Health submitted that the five-step approach was especially apposite in the present case because the evidence showed that Yasmin sales were declining, which made it difficult to ascertain mathematically Isabelle's impact on Yasmin's sales. We would note, however, that at [259] of her reasons, the primary judge referred to Mr Peace's evidence, which said that only a "very, very small" part of the Isabelle share of the market was new business and otherwise affirmed his evidence that an Isabelle sale was a lost sale of Yasmin. More generally, whatever the sales trends of Yasmin before and after the introduction of Isabelle, from our assessment of the evidence before her Honour, there was little, if anything, in the way of probative evidence before her that demonstrated that the introduction of Isabelle significantly grew or slowed the rate of decline of the market for the relevant compound or "molecule" from what it would have been absent the introduction of Isabelle, such that one could readily conclude that not all sales of Isabelle translated into lost sales of Yasmin. 174 Further, Generic Health contended that at [227] to [229], the primary judge appears to have held that the five-step approach was inapposite because it related to markets that "give a different context", being one of "consumer choice and substantial substitutability", which her Honour considered, at [229], was unlike the context in the OC market: …in which the product can be obtained only on prescription, the prescription will be for a specific brand, the consumer will take advice from a doctor as to the brand which best meets the consumer's requirements, brand substitution may be permitted or prohibited, the fact that consumers do not change from one OC to another without good reason and, in the case of Isabelle, the only brand for which it could be substituted was Yasmin. 175 Generic Health submitted that her Honour's justification for the rejection of the "five-step approach" was in error. 176 On the question of the difference in markets, Generic Health submitted that integer (b) (as we have described above) assumes that it was trying to capture sales that would otherwise have been made of Yasmin. It submitted that this was consistent with her Honour's findings as to its marketing and, more generally, the purpose of a generic composition. Accordingly, Generic Health submitted that this supported the correct approach, being to consider the nature of the OC market as part of the determination of the size of the discount to be applied under integer (d) (as we have described above), rather than to reject the "five-step approach" because of the different market context. 177 More generally, Generic Health submitted that, as the burden of proof regarding damages for patent infringement lies with the patentee, the burden of proving the paucity of any discount lay with Bayer. Generic Health made reference to Gerber Garment Technology Inc v Lectra Systems Ltd [1995] RPC 383 at 393 per Jacob J, although we note that Jacob J did not refer in terms to the burden of proving the paucity of any discount. We also note that Jacob J's decision was partly set aside on appeal (Gerber Garment Technology Inc v Lectra Systems Ltd [1996] EWCA Civ 1245; [1997] RPC 443), although not in relevant respects that need trouble us for the moment. 178 Further, Generic Health submitted that the primary judge was obliged to assess the size of the discount in accordance with the principles set out earlier and to do the best her Honour could in calculating damages: Commonwealth of Australia v Amann Aviation Pty Ltd [1991] HCA 54; 174 CLR 64 at 83 per Mason CJ and Dawson J. Further, it was submitted that the inevitable degree of imprecision in quantifying a discount, or the fact that Generic Health was the "wrongdoer", did not entitle her Honour to fall back on a one-for-one assumption and to draw the conclusions that her Honour did at [301] and [302]. We set out these paragraphs of her Honour's reasons later, to which we will return. 179 Generally, Generic Health submitted that, in light of the errors contended for, we should set aside the primary judge's damages award and determine afresh the quantum of damages, applying a discount in the range of 10% to 15% to reflect the following matters: (a) the significant price difference between Yasmin and Isabelle; (b) Bayer's own recognition that it faced price competition in the OC market, especially from CPA brands; (c) the crowded nature of the OC market; (d) the evidence that at least some women were price-sensitive; and (e) that "the hypothetical circumstances in which a woman might have come to purchase Isabelle", but would not have purchased Yasmin if Isabelle had not been available, were "capable of infinite variation" ([232]). 180 We reject most of the attacks made by Generic Health, save that (as we have recorded above) we agree that her Honour made an error concerning [206]. We also respectfully differ with her Honour in terms of the appropriate methodological approach, which in our view warrants some discount being made, even if only a very modest one. 181 We are dealing with a hypothetical situation of the past, namely, whether if a sale of Isabelle had not been made, then an equivalent sale of Yasmin would have been made. Accordingly, we must form an estimate of the likelihood that the hypothetical situation would have occurred: see Malec v J. C. Hutton Proprietary Limited [1990] HCA 20; 169 CLR 638 at 639 per Brennan and Dawson JJ. In that respect, we must assess the degree of probability that such an event would have occurred and adjust any award of damages to reflect that degree of probability: Malec at 643 per Deane, Gaudron and McHugh JJ. 182 In another sense, the question can be looked at through the lens of the value of a lost opportunity. The lost opportunity was to make a sale of Yasmin if the sale of Isabelle had not been made. In our view, it is not in doubt that it was established before the primary judge that, on the balance of probabilities, for all sales of Isabelle there was a loss of an opportunity to make a sale of Yasmin. The question then is: what is the value of the lost opportunity? On that question, the value is to be "ascertained by reference to the degree of probabilities or possibilities" involved in the hypothetical counterfactual: Sellars v Adelaide Petroleum NL [1994] HCA 4; 179 CLR 332 at 355 per the plurality. 183 Another way to express the same point is as Brennan J described it in Sellars at 368: Although the issue of a loss caused by the defendant's conduct must be established on the balance of probabilities, hypotheses and possibilities the fulfilment of which cannot be proved must be evaluated to determine the amount or value of the loss suffered. Proof on the balance of probabilities has no part to play in the evaluation of such hypotheses or possibilities: evaluation is a matter of informed estimation. 184 This theme of informed estimation resonates with some observations made by Hayne J in Placer (Granny Smith) Pty Ltd v Thiess Contractors Pty Ltd [2003] HCA 10; 196 ALR 257 at [37] and [38]. Any estimation must be done "with as much precision as the subject matter reasonably [permits]". Mere difficulty in estimating damages does not relieve a court "from the responsibility of estimating them as best it can". 185 Interestingly, Staughton LJ in Gerber Garment Technology [1997] RPC 443 at 459 cited Sellars with approval and discussed the matter in terms of evaluating the loss of a chance. Gerber Garment Technology both at first instance ([1995] RPC 383 per Jacob J at 407 and 408) and on appeal ([1997] RPC 443 per Staughton LJ at 459 and 460) evaluated and applied a loss of a chance approach based on the possibilities and probabilities. Moreover, Jacob J at [1995] RPC 383, 395 also invoked the themes of Lord Diplock in Mallett v McMonagle [1970] AC 166; [1969] 2 WLR 767 at 176, which resonate with the themes discussed in Malec at 639 and 643 that we have referred to above. 186 Whichever way one expresses it, in assessing the possibilities or probabilities of a hypothetical counterfactual, one is engaged in the task of estimation, even if the estimation involves an assessment of the counterfactual as being close to a certainty. But being close to a certainty is not the same thing as a certainty. If one is estimating, one still needs to apply a discount, albeit a very modest one, to reflect the assessment that one is not at a certainty. If one is looking at the value of a lost opportunity which is not certain to occur, then the valuation must involve some discount, even if a very modest one. 187 Given that Generic Health is a wrongdoer, it may be accepted that damages should be liberally assessed, although not, of course, to punish it: General Tire (1976) at 212 per Lord Wilberforce. The object is to compensate Bayer. To say that damages should be liberally assessed in no way cuts across what we have just said: that, in estimating or valuing a lost opportunity or in assessing a hypothetical counterfactual for any scenario short of certainty, some discount must be made to reflect that less than certain position, even if the discount is very modest indeed. 188 Also, whilst the "five-stage approach" in Norm Engineering may be helpful in some circumstances to order analysis, it should not be seen as a definitive test for what is an evaluative assessment on all the available evidence. 189 We raise one other matter before we proceed further. In assessing the relevant hypothetical counterfactual, we are in the realm of inferential assessment. We are in as good a position as the primary judge to consider that question. 190 It is appropriate now to consider the primary judge's reasoning on this aspect of the case. 191 In our opinion, the primary judge correctly reasoned that one is "dealing with wrongdoers" and that, in one sense, her Honour was entitled to fix damages on a liberal scale (at [193]). But as will be apparent from what we have just said, one needs to be careful of statements in the authorities to the effect that estimating damages is not an exercise that requires mathematical precision, that it may need to be dealt with in the rough, or that it should be done liberally. Whilst all such statements are apposite in one sense, in another sense when one is valuing a loss of an opportunity, there still needs to be a discount if the relevant probabilities or possibilities reflect less than certainty concerning the relevant counterfactual. 192 Generic Health submitted that the primary judge ought to have applied a discount to the sum awarded to Bayer to account for the "completely speculative" ([291]) "mere possibility" ([301]) that "a very, very small" ([259] and [291]) or "immaterial" number of women who purchased Isabelle would not have purchased Yasmin if Isabelle had not been available. We agree. Although the primary judge found that there was a "very real" doubt as to whether any such customers actually existed ([301]), we respectfully reject the idea that this doubt was to be resolved in Bayer's, but not the wrongdoer's, favour (cf TS & B Retail Systems Pty Ltd v 3Fold Resources Pty Ltd [No 3] [2007] FCA 151; 158 FCR 444 at [207]). In our view, some very small discount should have been made to reflect the uncertainty. 193 The Win/Lose document has some importance at this point. It was a recognition by the business person at the time that there was a risk of price sensitivity or elasticity. That in part could have been brought about by the disruption to the market by Generic Health. Putting a percentage on this is not straightforward. It is more helpful to think in terms of customers. A 10% discount attempts to reflect that one in ten customers who bought Isabelle would not have bought Yasmin. Five percent is one in twenty. Two percent is one in fifty. One percent is one in a hundred. 194 We agree with Bayer's submission that there is no evidence to justify a discount of the order of 10% to 15% for which Generic Health contended. The only basis propounded for such a large discount is as to a "very small possibility" ([302]), which the primary judge found on the evidence before her was unlikely to exist. Some discount was necessary given that we are not dealing with a certainty, but not a discount of 10% to 15%. It is appropriate to elaborate further on some of the relevant facts. But we would say now that we have little difficulty with her Honour's primary and secondary factual conclusions and reasoning, save for what we have said earlier concerning the pricing of CPA brands. Rather, based on those conclusions and reasoning, and for the reasons that follow, we would put the range as between 1% and 3%. We do not think greater precision is warranted, or possible. Given that is our view of the range, and given the principle of liberality, we think that, for the reasons that follow, the choice of 2% is appropriate. 195 First, as her Honour correctly found, Genetic Health's contentions and the views of Mr Houston, its economic expert, "lack[ed] sufficient foundation in the reality of the OC market in Australia" ([270]). 196 Secondly, Dr Williams in his report (at [36]) stated that "[a]lthough it is possible that some of the sales of Isabelle were substituting Isabelle for an OC other than Yasmin, the extent (if any) of such substitution is difficult to determine". He concluded at [48] and [49]: The detection of the magnitude of substitution between products can only be undertaken by means of empirical analysis. However, in this case, even a full-blown empirical analysis of demand is unlikely to reveal any significant substitution of Isabelle for oral contraceptives (OCs) other than Yasmin because of other effects that were operating at the same time. These other effects so dominated any possible substitution in favour of lower-priced products, that the net effect was a reduction in sales coinciding with the period in which Isabelle was in the market. The available evidence (including the evidence of Peace, Walker and McCann) suggests that the best available assessment of Bayer's claim for damages would be by treating each sale of Generic Health's Isabelle product as a lost sale of Bayer's Yasmin product. 197 While Dr Williams' views supported her Honour's findings, we do not accept that he was expressing the matter as one of certainty. 198 Indeed, one could not exclude the possibilities adverted to by Mr Houston in the following terms: My reasoning relies on some patients having made one of two forms of substitution, from one product to another, ie: • the availability of Isabelle meant that some patients that would otherwise have switched from Yasmin to another COC purchased Isabelle instead of making that switch; and • similarly, the availability of Isabelle meant that some patients with a new prescription for Yasmin purchased Isabelle instead whereas, had Isabelle not been available, they would not have proceeded to fill their Yasmin prescription, because it was too expensive. 199 His reasoning rejected the idea that each sale of Isabelle was a lost sale of Yasmin. He also concluded that "since patients were able to fulfil their prescription for Yasmin more cheaply after Isabelle was introduced, this would have led to more sales for Yasmin and Isabelle combined than would have been the case for Yasmin alone". 200 While many of Mr Houston's opinions were speculative (and his last statement was rightly rejected given its ambitious breadth), his evidence did support the notion that one could not rule out the possibility, in a very small number of cases, that a sale of Isabelle did not necessarily transpose into a lost sale of Yasmin. 201 Thirdly, as her Honour correctly observed ([226] to [229]), her estimation of damages depended upon a factual context which was quite different to the context discussed in the cases referred to by Generic Health. Whilst the principles referred to in those cases may have been appropriate in their context ([226]), including the application of the posited "five-step approach", we agree with her Honour that such authorities do not set any universal step-wise approach or rules for every damages estimation. 202 In our view, her Honour correctly distinguished other markets from the context before her. As her Honour described at [227] to [229] in pellucid terms: The context of the present case is different. A person wishing to buy swimwear, a T-Shirt or an industrial bucket is not constrained in any relevant way. The person is free to buy whichever swimwear, T-Shirt or industrial bucket best meets their current needs and circumstances. A woman who wishes to use an OC is not in that position. Every woman who wishes to use an OC must do so under prescription from a doctor. As Yasmin was the only formulation of 3mg DRSP and 30 µg EE on the market for some 10 years before Isabelle entered the market, all sales of Yasmin represent a prescription given by a doctor to a woman for Yasmin. Having received that prescription, a woman could obtain only Yasmin from a pharmacist. If the woman did not fill that prescription (because of price or for some other reason), her choice was to not use any OC or to return to a doctor to obtain a prescription for a different OC. If she returned to her doctor, she would be informed that Yasmin was the only OC containing 3mg DRSP and 30 µg EE available and that this formulation had (or at least was said to have had, based on clinical trials) a unique suite of benefits in terms of not causing weight gain, bloating, breast tenderness and perhaps increased blood pressure. Further, given that women generally do not change their OC other than by reason of some major change in circumstances, it may be inferred that women who had been prescribed Yasmin in the past would continue to be prescribed and use Yasmin in the future unless subject to a major life event which may cause them to interrupt their use of an OC, to cease using an OC altogether, or to change their OC. When Isabelle entered this market, it did so on the basis that it was bioequivalent to Yasmin. On the evidence, doctors would not prescribe by reference to the brand "Isabelle". Doctors would continue to prescribe Yasmin but, provided the doctor did not tick the "brand substitution not permitted" box on the prescription, a pharmacist could supply Isabelle instead of Yasmin. Accordingly, a woman could obtain Isabelle only if she held a prescription for Yasmin. A woman with a prescription for any OC other than Yasmin could not, on that prescription, obtain Isabelle. As a result, it is accurate to say that the only OC for which Isabelle could be substituted was Yasmin. These are different circumstances from the markets for items of clothing or industrial equipment. They give a different context. The context for clothing and equipment (and many other consumables) is one of consumer choice and substantial substitutability. This is not the context for the OC market in which the product can be obtained only on prescription, the prescription will be for a specific brand, the consumer will take advice from a doctor as to the brand which best meets the consumer's requirements, brand substitution may be permitted or prohibited, the fact that consumers do not change from one OC to another without good reason and, in the case of Isabelle, the only brand for which it could be substituted was Yasmin. 203 Moreover, at [230] her Honour went on to say: Accordingly, it is at least clear that every sale of Isabelle represents a prescription for Yasmin which resulted in Isabelle rather than Yasmin being prescribed. This is because the evidence was that doctors generally prescribe by reference to the originator brand, there was evidence of pharmacists dispensing Isabelle in respect of a prescription for Yasmin, and there was no evidence of any prescription for Isabelle. This proposition is not the same as Bayer's proposition that every sale of Isabelle (and Petibelle) must be a lost sale of Yasmin. It is, however, the necessary starting point for the consideration of the issue of Bayer's lost profits. Against this starting point, Generic Health's contentions can be assessed. 204 The primary judge did not err in not applying the "five-step approach" that we described earlier. That said, even on that approach, the application of integer (d) (as we have described above) would justify a discount of 2%. This is the discount that also ought to be applied under the Malec, Sellars or Placer approaches. Thus, any perceived methodological error in not using the "five-step approach", if there be one (which there is not), goes nowhere. 205 Fourthly, the complex and unique factual circumstances found by the primary judge were amply supported by the evidence of a pharmacist, a doctor, a pharmaceutical marketing expert, a project manager and economic evidence. Taken with other important contextual matters that we will discuss in a moment, this cumulatively led the primary judge to conclude that Bayer's loss was best assessed on the basis that every sale of Isabelle was a lost sale of Yasmin. But, as we have said, if the position was less than certainty, as it was on the evidence before her Honour, then some discount ought to have been made. The primary judge emphasised that the "one for one proposition is not a mere assumption in any event". She noted that it was "an expert opinion of Dr Williams and Mr Peace (from different perspectives) by reference to non-quantitative but detailed data about the OC market in Australia and the place of Yasmin (and Isabelle) within that market" (see at [299]). All of this is true. But no expert was certain and could not be certain that each sale of Isabelle was at the expense of Yasmin. Accordingly, some discount had to be made. 206 It is now appropriate to discuss some of the contextual matters the primary judge found, which were amply supported by the evidence. 207 First, a woman who wished to use an OC obtained a prescription, after a consultation with a doctor regarding the benefits (including non-contraceptive benefits) and side effects of various available OCs. Further, the benefits of originator OCs such as Yasmin were primarily promoted to doctors, rather than to pharmacists, to whom generics were promoted. Moreover, doctors were not generally aware of the cost of OCs and did not generally discuss the cost of OCs with patients. 208 Secondly, when Isabelle entered the market, it did so on the basis that it was bioequivalent to, and directly substitutable with, Yasmin, which was the only such product then on the market. The Isabelle ARTG registration was based on the ARTG registration for Yasmin and the Isabelle Product Information (PI) sheet contained substantially the same information as the Yasmin PI sheet. It is not in doubt that Generic Health promoted Isabelle to pharmacists as being a substitute for Yasmin. Indeed, this could not be seriously denied. The various forms of "Exclusive introductory offer" in evidence before her Honour, marketing Isabelle to pharmacists, contained the prominent statement "Originator Brand-Yasmin" and explicitly compared the Isabelle net in-store price to the Yasmin net in-store price. 209 Thirdly, doctors were not generally aware of generic versions of OCs such as Isabelle and would prescribe the branded product Yasmin. But a pharmacist could supply only Isabelle (i.e. no other OC) instead of Yasmin if the doctor did not tick the "brand substitution not permitted" box on the prescription. A woman with a prescription for any OC other than Yasmin could not obtain Isabelle on such a prescription. 210 Fourthly, as we have touched upon earlier, the OC market included premium and standard products. The premium products generally were newer, so-called "third generation" products offering non-contraceptive benefits. The standard products generally were cheaper OCs listed on the PBS. Generally speaking, price was not a motivating factor within the private prescription segment of the OC market, as opposed to between private and PBS segments ([210]) and ([273]). Private market OCs were highly differentiated ([279] and [280]). Choices between these private market OCs were based upon side-effect profiles and non-contraceptive benefits, rather than being based on cost ([240], [243] and [273]). 211 Fifthly, the relevant CPA products were launched in 1992 as acne treatments but prescribed off-label as third generation OCs. Prior to, and after, the entry of Isabelle, they were highly genericised. Generally speaking, Mr McCann considered the CPA originator and generic OCs to be effectively within their own market. We agree with Bayer's submission that the notion that pharmacists would suggest switching to CPA brands rather than purchasing Yasmin was properly found by her Honour at [255] to be generally inconsistent with the evidence of Mr McCann as to pharmacist behaviour. 212 Sixthly, as Bayer pointed out, prior to the entry of Isabelle in January 2012, Yasmin had been on the Australian market for 10 years and was the only OC comprising 3 mg DRSP and 0.03 mg EE. Yasmin offered a unique suite of benefits over other OCs (including CPA OCs), including reduced weight gain, fluid retention and positive effect on mood. Yasmin was not, and has never been, listed on the PBS and was priced at a premium over other OCs. Further, women with a prescription for Yasmin could obtain only Yasmin from a pharmacist. If the woman did not fill that prescription, her choice was either to not use any OC or to return to her doctor to obtain a prescription for a different OC without the unique suite of benefits offered by Yasmin. 213 Seventhly, it was open to the primary judge to find that, generally speaking, women do not change their OC when they are happy in terms of side effects, other than by reason of some major change in circumstances which may cause them to interrupt or cease altogether their use of an OC or to change their OC. Such a conclusion was well-supported by the evidence, including that of Mr Peace and Mr McCann. 214 Mr McCann gave the following evidence concerning price in his first affidavit at [27] to [31]: In my experience, price does influence whether a patient will fill a prescription. In this regard, private scripts (i.e. a script that is not covered by the PBS) can be out of reach for some patients, such as those who hold concession cards. In my experience, patients do regularly ask whether there is a generic brand available particularly in circumstances where they hold a private script. In this regard, in my practice as a pharmacist, I am frequently asked by patients questions in words to the effect of "can you tell me how much this is?" followed by "Is there a cheaper brand?" In my experience, the product Yasmin has produced many comments from customers as to the high price of the product. In my experience my customers are not made aware by the prescribing doctor of the price of filling such a script as Yasmin and they are surprised at the pharmacy when the script is filled or to be filled. Upon a patient's own enquiry or upon offer of a generic by the pharmacist, my experience is that many patients will chose [sic] to take a generic product because of the cost saving. In my experience if the cost saving is $10 or more (as in the case of Isabelle and Yasmin), the consumer will usually choose the generic product. If the saving is $5 or less some will take the generic option but not as many in the case of a higher saving. From reviewing the sales figures for Yasmin and Isabelle products in annexure FM-2, in my view the costs savings on the Isabelle script were significant enough to cause patients to opt for the generic product. In this regard, I note that I dispensed 30 3x28 boxes of Isabelle during January 2012-June 2014 versus 16 3x28 boxes of Yasmin. 215 These passages do not provide direct evidence that, if Isabelle was not available, then a woman would have gone for something other than Yasmin. Further, they do not stand as evidence of any woman switching from another brand to Isabelle, who would not have also switched to Yasmin if Isabelle had not been available. 216 Mr McCann gave evidence in cross-examination on [30] of his affidavit in terms that, when he was talking about a $10 difference, he was talking about people purchasing Isabelle instead of Yasmin. This was because their prescription was for Yasmin. He was not talking of a person purchasing Isabelle instead of another OC such as a CPA. 217 In his second affidavit, Mr McCann sought to explain at [6]: I also refer to paragraph 29 of my First Affidavit. On numerous occasions, I received a comment from a patient to the effect that Yasmin was too expensive and they would not be able to fill their script. On these occasions (and provided the doctor has not specified no substitution), in order to assist the patient and possibly enable them to have the therapeutic benefit that their doctor had prescribed, I informed the patient about the possibility of purchasing a cheaper brand, namely Isabelle. On some of those occasions, the patient then purchased Isabelle as a cheaper alternative, either immediately or after having received information about Isabelle and subsequently returning to the pharmacy to fill their script with Isabelle rather than Yasmin. 218 But again this did not establish that, if Isabelle was not available, a woman would not have purchased Yasmin. Put another way, this was not evidence that, if Isabelle was not available, a woman would have gone to a CPA, for example. 219 Eighthly, Yasmin was targeted to women with higher disposable incomes who wanted the unique overall non-contraceptive benefits it gave them. As her Honour found, women buying Yasmin were likely to be the least price-sensitive consumers in the OC market ([270] and [273(b)]). Moreover, a woman taking Yasmin or Isabelle was paying a premium price. Contrastingly, women for whom price was a relevant factor were unlikely to afford either product: [273(a)-(b)]. 220 Ninthly, the price differential between Yasmin and Isabelle was only approximately $10 over three months. Isabelle's price ($60 to $61 for three months) was well above the price of OCs on the PBS ($19 to $24 for four months), second generation OCs not on the PBS ($35 to $40 for three months) and the alleged price of the CPA OCs ($18.90 to $41.45 for three months). The retail price of CPA OCs ranged between $18.90 and $77.55 for three months (Laila-35 lowest and Diane-35 highest) according to annexure FM-2 to Mr McCann's first affidavit. According to his second affidavit, which clarified aspects of FM-2, the retail price of CPA OCs ranged between $14.90 and $59.77 for three months (NIP + retail mark-up of $6.50 to $8.00). The $10 price differential between Yasmin and Isabelle caused, as Generic Health intended, some women to purchase Isabelle instead of Yasmin, but not any other OC. This places the Win/Lose document into more context. Though the document reveals some risk of price elasticity, it is difficult to see how that price differential of $10 over three months would have been important to deter many women who did buy Isabelle from buying Yasmin on the hypothesis that Isabelle was not on the market. Put another way, for those who did buy Isabelle for its price, if it had not been on the market, it is difficult to see why $10 over three months would have been a deterrent to many women. 221 Tenthly, there was no probative evidence that the availability of Isabelle increased the sales of the DRSP and EE compound OCs, that is Yasmin + Isabelle or Yasmin + Petibelle. 222 Eleventhly, there was also no direct evidence of any woman who had purchased Isabelle who would not have purchased Yasmin. In other words, there was no evidence of a woman being supplied Isabelle as a substitute for an OC other than Yasmin. Mr McCann gave evidence that when Brenda, a CPA brand, "took all the market", he stated that all the sales came from Diane, the same compound. Mr McCann gave no evidence of any woman with a prescription for another OC returning with a prescription for Isabelle. Moreover, there was no evidence that any doctor issued a prescription for Isabelle. 223 Generally, there was no evidence that a pharmacist would encourage a woman to switch from one OC to a different OC involving a different molecule ([265]). For example, there was no evidence that a single consumer who switched from a CPA OC to Isabelle and paid, for example, an extra $7 to $14 per month, and who would not have switched to Yasmin and paid an extra $11 to $18 per month. The existence of any such consumers was "the merest possibility" ([267]). 224 Contrastingly, and as is pointed out by Bayer, switching from Yasmin to its "substitute" Isabelle was specifically promoted and intended by Generic Health, as illustrated by Generic Health's own marketing to pharmacists, as the primary judge rightly described and found on the evidence ([219], [236] to [239], [246] and [247]). 225 Generally, in terms of price sensitivity, it is also to be recalled that the primary judge made the finding that "a very price sensitive woman would have a far greater incentive to obtain a PBS prescription than to save $10 by purchasing Isabelle rather than Yasmin" [245] and also went on to say at [246] and [247]: If Isabelle sales had been dependent on women described by Mr McCann (that is, shocked at the price of Yasmin but willing to pay for Isabelle rather than getting a PBS prescription) then the market for Isabelle, I infer, would have been absurdly narrow. Isabelle's market, it seems obvious to me, depended on women who were willing to pay the kind of price that third generation OCs commanded. This, of course, is the very market on which Generic Health's marketing to pharmacists focused - converting women on Yasmin to Isabelle. In other words, Generic Health's marketing seems to me to have been inherently rational and to reflect the proven circumstances of the OC market in Australia. Its submissions in this case, suggesting that there was a substantial segment of the market consisting of women who would not pay the price for Yasmin but would pay the price for Isabelle (at a difference of about $10 over three months), when PBS subsidised prescriptions could result in a saving of around $40 per month compared to Isabelle, does not correspond well with the evidence. It is possible that there were some women, as Mr McCann said, who wanted the benefits of Yasmin but could not afford to pay for it but could afford to pay for Isabelle. If such women existed they were more likely than not to be in an area such as that serviced by Mr McCann's pharmacy. Yet Mr McCann's evidence did not identify any actual woman who purchased Isabelle who would not have purchased Yasmin. 226 Further, at [252] her Honour said: In other words, while I accept there is price sensitivity for women purchasing OCs it is a sensitivity mediated through a number of other considerations. There is the doctor prescribing the product who, it seems, should consider long-term financial capacity to pay for an OC, but the evidence suggests does not in fact do so routinely. There is the fact that the market involves radically different price points - a low price point for PBS prescriptions and a high price point for private prescriptions of third generations [sic] OCs. There is the fact that women do not usually change from one OC to another OC involving a different molecule unless there is good reason to do so. There is the fact that Yasmin had been on the market for 10 years at a premium price so only newcomers to Yasmin (or private prescription OCs) could have been surprised at the price of Yasmin. There is the fact that OCs could not readily be seen as optional by women dependent on an OC, in contrast to some other prescription medicines. All of these considerations tend to support the inference that Isabelle's most likely source of sales was women who had been taking Yasmin and would have continued to do so until a good reason arose to change to another molecule. 227 No error has been demonstrated in these observations. 228 Generally, it was the primary judge's consideration of the context and nature of the OC market which led to her Honour's decision not to apply any discount. So her Honour found that Bayer's damages for each and every infringing sale of Isabelle were best assessed on the basis that each sale of Isabelle would have been a sale of Yasmin, absent Isabelle's unlawful presence in the OC market. 229 We do not deny the reasonableness of her Honour's conclusion in one sense. But given that one is involved in estimation or valuing the loss of an opportunity, if the position ultimately found is one which is less than certainty, then, as we have emphasised, some discount is required to be made, even if it is very modest. Her Honour, with respect, was in error in not doing so. 230 First, it is true that the one-for-one proposition was not a mere assumption but based upon expert opinion from Dr Williams and also the evidence of Mr Peace: [299]. But that did not require her Honour to accept it without applying a discount if the probabilities relating to evaluating the counterfactual were less than certainty. 231 Secondly, her Honour's qualified statement at [300] referring to Bayer's concern that "Isabelle had disrupted the market possibly introducing a price sensitivity that would not have existed but for Isabelle" (our emphasis), still left open the possibility of some other additional price sensitivity outside this in any event for a very small number of women. 232 Thirdly, at [301] her Honour said the following: It is not the case that given "its onus and quantification burden, Bayer's approach is not of assistance to the Court". The onus of proof can be discharged in many ways. I am satisfied that no discount should be made in the particular circumstances of this case in order to allow for the possibility that some women bought Isabelle but would not have bought Yasmin. Apart from the fact that I agree with Mr Peace and Dr Williams that this possible class of women is likely to be very small, to the point of immateriality, it is relevant that Generic Health was the wrongdoer so that the assessment of damages ought to be liberal, not miserly. There is doubt whether such a class actually exists; it is a mere possibility. There is doubt that such a class, if it exists, in fact acted on price signals in the way Generic Health's case would have it. All of these very real doubts are to be resolved in favour of Bayer and not the wrongdoer. (Emphasis in original.) 233 Given that these observations indicate less than certainty, some discount should have been applied, even if the possible class of relevant women was likely to be very small. Her Honour's statement of "to the point of immateriality" in context seems to equate immateriality with very small, and something to be disregarded if one "ought to be liberal, not miserly". But, as we have said, even a very small possibility (indeed, her Honour's acceptance of a "mere possibility") warranted some discount given that one is evaluating a hypothetical counterfactual. 234 The primary judge also went on to say at [302]: This approach reflects the reality that every assessment of damages is an estimate which is imprecise in a multitude of respects. In a case such as the present, where the evidence supports the conclusion that sales of Isabelle which were not lost sales of Yasmin is likely to represent nothing more than a very small possibility, there is no reason to make an adjustment of 1% (as Bayer suggested as a possible adjustment in the alternative to its principal position). While 1% may sound trivial, it represents not only one woman in every 100 who purchased Isabelle but, as Generic Health posited, one woman in every 100 who purchased Isabelle and had the particular form of price sensitivity which meant she would not have purchased Yasmin if Isabelle had not been available. It also represents not insubstantial monetary sums given the overall sales and period of infringement. Given that Generic Health is the wrongdoer I see no reason to make such an adjustment given the evidence which I have accepted in the circumstances of this case (which, as I have said, bears no resemblance to that of other consumables) and that, on any view, Bayer is entitled to the benefit of doubt, not Generic Health. (Emphasis in original.) 235 Again, the recognition of "a very small possibility" warranted a discount. The fact that a 1% discount reflected a substantial monetary sum did not justify not discounting, but rather the converse conclusion. Further, the fact that Generic Health was a wrongdoer did not justify not discounting. In terms of the "benefit of doubt", there was no doubt that there was a very small possibility that a woman who purchased Isabelle could have had a form of price sensitivity, which meant she would not have purchased Yasmin if Isabelle had not been available. 236 Given that her Honour erred in not applying any discount, it is necessary for us to quantify the relevant discount. While her Honour's reasons refer to a 1% discount, this was not in terms of quantifying any discount at that level; the 1% was Bayer's suggestion as a possible adjustment. Accordingly, we are free to set the discount for ourselves, but bearing in mind her Honour's factual findings. This is not a case of needing to show why her Honour was in error at the 1% level, for she did not quantify any discount. And this is not a case of needing to justify moving from a 1% level to a 2% level, for again there was no such 1% quantification of the discount by her Honour. As we have said, we are free to set the discount for ourselves. As we have earlier said, it is more helpful to deal with this in terms of numbers of women (who would not have bought the higher priced Yasmin but who did buy a lower priced Isabelle) rather than percentages. On the preponderance of the evidence before her Honour, it is apparent that there would have been only a small number of women who, if Isabelle had not been available, would have purchased another OC other than Yasmin or would not have purchased an OC at all. The Win/Lose document reveals an appreciation by Bayer of some risk of price elasticity, though with the qualification that we have already discussed. On the evidence, that number was, we think, likely within a range of one in one hundred to, at the very most, one in thirty-three (1% to 3%). Assessing damages liberally in circumstances where Generic Health was the wrongdoer, we consider that a fair percentage is 2% and we will apply 2% as the discount rate. There is no precision or science involved. The best estimation is the start and finish of the task. We think a percentage at the higher end of the previously mentioned range of 1% to 3% would risk infringement of the principle of liberality in the task. We consider 2% is appropriate in the circumstances. 237 Accordingly, we uphold appeal Grounds 5, 8 and 9, but not Ground 6. 238 First, and importantly as we have explained, it was necessary for the primary judge to quantify the discount as her Honour was engaged in the exercise of evaluating the probabilities and possibilities of the relevant hypothetical counterfactual or, in essence, valuing the loss of an opportunity. And her Honour ought to have quantified the discount, even if it were very small. Accordingly, Grounds 5 and 9 must be upheld. 239 Secondly, the primary judge erred in her finding at [206], and accordingly Ground 8 must be upheld. But, at the end of the day, that error adds little to our ultimate finding that her Honour ought to have quantified the discount at 2% as previously discussed. The error at [206] merely reinforces the point that, for a very small percentage of women who were sufficiently price-sensitive, instead of purchasing Isabelle they may not have purchased Yasmin if Isabelle had not been available. That is, they may have purchased another cheaper third generation OC instead or purchased none at all. In other words, to accept the error does not justify any additional discount over the 2% that we have already found. The 2% assessment factors in and assumes Generic Health's success on Ground 8. 240 Thirdly, we do not consider that the primary judge erred in not applying Generic Health's "five-step approach" to discounting that we have discussed earlier. While such an approach may have been justifiable in different markets involving different product substitutability questions, the task at hand is essentially one of judicial estimation unconfined by the application of rigid formulae. The primary judge was entitled to consider the "five-step approach" to be inapposite to the context before her. In any event, even if it had been appropriate to apply that approach, no different result would follow. Integer (d) would warrant, as we have said, a discount of 2%.

Petibelle (Grounds 10 - 12) 241 Grounds 10 to 12 of the notice of appeal concerning the Damages Judgment are expressed in the following terms: 10. The primary judge erred in holding that Generic Health was liable to pay any damages in respect of the sales of Bayer's Generic Product (Petibelle) made by Bayer in the period from the introduction of Petibelle until 30 June 2016 (the Petibelle Sales Period) (Main Judgment at [315], [345]), as that head of damages was too remote to result in any damages award. 11. In the alternative to ground 10, the primary judge erred in holding that Generic Health was liable to pay damages to Bayer in respect of Bayer's sales of Petibelle for the whole of the Petibelle Sales Period (Main Judgment at [316], [345]), as Bayer's decision to continue to supply Petibelle for the whole of the Petibelle Sales Period was not causally connected to, and/or was too remote from, Generic Health's supply of Isabelle. 12. In the alternative to ground 10 and further or in the alternative to ground 11, the primary judge erred in holding that Bayer's damages were to be assessed on the basis that every sale of Petibelle would have been a sale of Yasmin, had Isabelle not been on the market (Main Judgment at [318], [345]). 242 After Isabelle was removed from sale on 19 June 2014, Bayer released Petibelle, a generic version of Yasmin, on 26 June 2014. It was priced at $66 for a three month supply, which was approximately halfway between the retail prices for Yasmin and Isabelle. Bayer claimed for losses in margin by way of sales of Petibelle, rather than Yasmin, from approximately June 2014 through to June 2016. Bayer contended before the primary judge that every sale of Petibelle was in effect a lost sale of Yasmin. It may be accepted that Petibelle was only introduced after Isabelle had been withdrawn, and therefore never competed against Isabelle. 243 The primary judge allowed damages in respect of Petibelle and for the time period of approximately two years. 244 As patent infringement is a statutory tort, Bayer was required to show that damages in respect of Petibelle were not too remote to be recoverable. Whether the relevant loss is a foreseeable consequence of infringement is to be assessed on the evidence: Krueger Transport Equipment Pty Ltd v Glen Cameron Storage and Distribution Pty Ltd [No 2] [2008] FCA 1493; 79 IPR 81 at [38] per Gordon J. In our opinion, the primary judge correctly identified the relevant principle at [313]. But Generic Health contends that her Honour erred in finding that Bayer had established that any losses it suffered in respect of Petibelle met the lack of remoteness requirement ([315]). 245 At [314] and [315], the primary judge held: Generic Health's submission that Bayer intended to capitalise on Isabelle's removal from the market misses the point. The point is that it was reasonable of Bayer to try to respond to Isabelle's unlawful disruption of the market. It was foreseeable that Bayer would respond to the infringement in various ways including the introduction of its own generic version at a cheaper price than Yasmin. This was a natural and direct consequence of the introduction of Isabelle. The fact that Bayer waited until it had succeeded in having Isabelle removed from the market does not undermine either causation or foreseeability. The manifest temporal connection between the removal of Isabelle and the introduction of Petibelle (within the week of Isabelle's removal Petibelle was on sale) confirms the causal connection. If, for example, there had been substantial time between the removal of Isabelle and the introduction of Petibelle, the result might well have been different. As it is, however, it cannot be said that the infringement was merely a background circumstance of, or set the scene for, the introduction of Petibelle. The infringement was the direct and obvious cause of the introduction of Petibelle. The fact that Dr Williams agreed that "it would only be profitable if the loss in margin on the units that switched from Yasmin to Petibelle would be made up for in extra volume on the sales of the Petibelle" goes nowhere given that Bayer's reasons for introducing Pettibelle were not to increase its profits (and Dr Williams concluded that there was no financial incentive for Bayer to introduce Petibelle as it would not make more money by so doing). As such, the submissions that "Bayer clearly understood the risks to the volume of Yasmin sales which were associated with the introduction of Petibelle. Its reasons for doing so cannot be dissociated from its commercial motivations. Its introduction is Bayer's cross to bear, not that of Generic Health" are unpersuasive. It is the infringement that created the risks which Bayer reasonably addressed. As such, it is Generic Health's cross to bear. Generic Health's submission that the loss is too remote because originator companies usually introduce their own generic while the competing generic is on the market does not mean that Bayer's decision, to put Petibelle on the market immediately Isabelle had been removed [sic], involves loss that is other than reasonably foreseeable and thus is too remote. 246 The primary judge's principal conclusion (at [314]) was that it was foreseeable that Bayer would respond to the infringement in various ways, including by introducing its own generic version of Yasmin at a cheaper price. The primary judge reasoned that this was a natural and direct consequence of the introduction of Isabelle. Generic Health submitted that there was no evidence to support such a proposition, insofar as it concerned the introduction of a generic (Petibelle) after the infringing generic (Isabelle) was taken off the market. Generic Health also contended that there was no evidence to suggest that it foresaw that Bayer would release, let alone make a claim in respect of, its own generic after Generic Health had been restrained. 247 Generic Health also submitted that the primary judge overlooked evidence to the contrary, being Mr Peace's evidence to the effect that originators usually introduce their own generic whilst the competing generic is on the market, rather than after the competing generic has been removed and the originator is free to supply its higher-priced specific brand. Generic Health also submitted that the primary judge overlooked the Win/Lose document, which evinced a strategy to release Petibelle if Bayer "lost" its injunction application (i.e. if Isabelle remained on the market), rather than if Bayer "won" (i.e. Generic Health was restrained from selling Isabelle). Generic Health contended that such evidence militated against a finding of lack of remoteness. 248 In summary, Generic Health submitted that Bayer, which had re-established its monopoly by restraining Generic Health, unexpectedly and unforeseeably declined to take advantage of that monopoly. Rather, it introduced to the market its own rival product (Petibelle) at a cheaper price rather than re-instating its higher priced product (Yasmin). 249 We reject appeal Ground 10. 250 Claimable loss for patent infringement can include lost profits arising from a patentee, because of the infringement of its rights, having to reduce the price, and accordingly the profitability, of the patentee's goods in order to retain its market or regain its profitability due to "price" depression caused by the infringer's conduct; see Gerber Garment Technology [1995] RPC 383 at 394 and 417 and also Ultraframe (UK) Limited v Eurocell Building Plastics Limited [2006] EWHC 1344 (Pat) at [47], [154] and [155] per Kitchin J. Moreover, such "price depression" loss may be continuing and may be awarded for a reasonable period following the withdrawal of an infringing product from the market (Ultraframe at [184] and [185]). 251 Mr Peace gave evidence that the "strategy was to have Petibelle available to those women who had become accustomed to paying less for Isabelle (compared to Yasmin) and who would not be willing to pay the higher price for Yasmin once Isabelle was no longer available". Further, Bayer wrote to doctors on 20 June 2014 (after the injunction was granted against Generic Health) announcing the immediate availability of Petibelle and stated that it considered "Petibelle as a logical replacement for your Isabelle patients…" Moreover, in a letter to pharmacists, Bayer wrote: To ensure minimal inconvenience to patients Bayer Australia is pleased to advise the immediate availability of its latest brand of oral contraceptive - Petibelle (ethinylestradiol 30 mcg/drospirenone 3mg). Petibelle is the second brand of Yasmin meaning the products are interchangeable based on doctor or pharmacist preference and would also be a suitable alternative for returning Isabelle patients. 252 At an earlier point in time, the decision-making process concerning Petibelle contemplated the potential for generic on generic competition i.e. if Generic Health was not enjoined. But that did not exclude Bayer using Petibelle to deal with the "price depression" or reputational damage caused by Isabelle and its consequences even after Generic Health had been restrained. If Isabelle had not been supplied, then Petibelle would not have been sold. 253 Mr Peace gave evidence in his third affidavit at [67] that: After Generic Health began marketing and selling a generic version of Yasmin (under the brand name Isabelle), I together with my colleagues in management at Bayer Australia, considered whether to start selling Petibelle as Bayer Australia's own "generic" version of Yasmin to try to minimise the damage that Isabelle was causing to the reputation, pricing structure and market share of Yasmin. In mid 2012, Bayer manufactured stocks of Petibelle and delivered them to Bayer Australia for storage. However the strategy adopted by Bayer Australia was as follows: (i) Bayer Australia did not start marketing and selling Petibelle in competition with Isabelle because it was ultimately considered that this could further undermine the established price premium for Yasmin, especially in circumstances where the outcome of this proceeding was uncertain; (ii) Bayer Australia would start marketing and selling Petibelle if the Bayer parties were successful in the Court proceedings and the Court ordered Generic Health to stop supplying and selling Isabelle in Australia. The strategy was to have Petibelle available to those women who had become accustomed to paying less for Isabelle (compared to Yasmin) and who would not be willing to pay the higher price for Yasmin once Isabelle was no longer available; (iii) If Bayer Australia started marketing and selling Petibelle: (a) Bayer Australia would supply Petibelle with a recommended retail price (RRP) that was around halfway between the retail prices offered by the large discount pharmacies for Yasmin and Isabelle; (b) Bayer Australia would supply Petibelle to its wholesalers at a price which was likely to generate the same margin for the wholesalers and for pharmacists as the margin which the wholesalers and pharmacists would make on sales of Yasmin, so as to minimise any erosion of Yasmin sales caused by the availability of Petibelle. 254 The primary judge held that the launch of Isabelle and its later court-ordered withdrawal "fractured the price/value perception of Yasmin" ([307]) and caused Bayer to start selling a less profitable version of Yasmin, namely, Petibelle ([304], [308], [310], [311], [312], [314] and [316]). Of course, these are findings of causation rather than reasonable foreseeability. Relying on the evidence of Mr Peace, who was cross-examined on the subject, the primary judge found that Bayer was concerned that some women who had obtained a discount by buying Isabelle instead of Yasmin would have felt aggrieved if there was no option softening the blow of being forced back to the fully-priced Yasmin. Accordingly, Bayer launched Petibelle immediately following the exit of Isabelle from the market in June 2014 in order to mitigate reputational damage, as her Honour found ([266] and [267]). The idea was to return Bayer to the position it enjoyed before Generic Health's infringement. As her Honour found, and as we have already noted, absent infringement, Bayer would not have launched Petibelle. Importantly, none of these causal findings were challenged before us, except for the two-year question which we will discuss later when dealing with Ground 11. 255 It was well open to the primary judge to find that it was reasonably foreseeable that Bayer would respond to infringement in various ways, including by introducing a generic version of Yasmin whilst the infringing generic product was on the market. But Generic Health submitted that the introduction of Petibelle immediately after withdrawal of the infringing generic product was not foreseeable by it. In our view, the matters referred to by Generic Health do not support the assertion that the launch of Petibelle immediately after the withdrawal of Isabelle was unexpected and unforeseeable. As the primary judge said, it was reasonably foreseeable that Bayer would launch Petibelle in response to the infringement ([306] to [308], [310], [312], [314]). As the primary judge correctly found, "[t]he introduction of Petibelle to rectify the damage, to the extent it could be rectified, was a reasonable response to the difficult commercial position in which Generic Health's infringing conduct had placed Bayer" (see [308], [311] and [312]). 256 Generic Health submitted that there is no direct evidence of reasonable foreseeability as opposed to evidence of causation. This is too narrow a focus. It may be accepted that causation and reasonable foreseeability are separate concepts. But facts and evidence concerning the former can inform the evaluative judgment to be made concerning the latter, as it clearly did in relation to her Honour's assessment. Moreover, reasonable foreseeability and remoteness involve normative questions, and for the latter policy questions. For our own part, Bayer's actions concerning Petibelle were reasonable, predictable and a proportionate response to the harm caused by Generic Health's conduct and ought well to have been within the reasonable contemplation of Generic Health as a natural commercial response to its infringing behaviour. 257 We therefore reject Ground 10. 258 As to Ground 11, the primary judge found that Generic Health was liable to pay Bayer damages in respect of Bayer's sales of Petibelle for the whole of the two-year post-Isabelle period for which damages in respect of Petibelle were claimed. At [316] her Honour said: Nor do I accept that the claim period, of two years, would involve compensating Bayer for loss that is too remote to be recovered. Bayer's reasons for introducing Pettibelle [sic] were multiple but all involved trying to ameliorate the harm done to the market which the infringement caused. It is true that Bayer cannot recover the loss resulting from the cheaper price of Petibelle in perpetuity but the period of two years is reasonable. There is no basis in the evidence for Generic Health's submission that "a loss (if any) resulting from Petibelle would be a short-term loss or no more than a month or two" or would be rectified on the first pharmacy visit to obtain Yasmin or Petibelle. Again, any doubt in this regard should be to the account of Generic Health, not Bayer. I can see no reason why the claim over two years offends the requirement that loss be compensated only if it satisfies the test of causation, foreseeability and is not otherwise speculative. 259 Generic Health submitted that the primary judge erred in this respect both in terms of causation and remoteness. Generic Health submitted that the evidence of Mr Peace was that Petibelle addressed a short-term issue. In those circumstances, any "losses" caused by the continued presence of Petibelle on the market for two years (as opposed to three months, being the maximum multicycle purchase) were the result of an independent commercial decision by Bayer, for which Generic Health cannot be liable because of the causation and lack of remoteness requirements. 260 The primary judge was entitled to award Bayer the amount they would have earned in the two-year period immediately following Isabelle's withdrawal as if every sale of Petibelle during that period would otherwise have been a sale of Yasmin ([294], [295] and [345]) subject to a 2% discount, by parity of reasoning with our earlier analysis. Her Honour considered whether the claim was too remote, concluding that it was not ([310] to [316]). In our opinion this conclusion was open on the facts as found and the inferences drawn by her Honour. There is also no substance to the complaint of lack of factual causation. 261 As to Ground 12, this ground only applies if Ground 10 fails. It stands or falls on Grounds 5 to 9. If a 'discount' is applied under those grounds, in our view it follows that a discount ought to be applied in respect of Bayer's sales of Petibelle. In other words, a sale of Petibelle is not in all cases necessarily a lost sale of Yasmin. 262 As we have already said, Ground 10 fails. Moreover, and consistently with our earlier ruling concerning Grounds 5 to 9, we would uphold Ground 12 in part and allow a discount of 2%.

Interest (Ground 13) 263 The primary judge calculated pre-judgment interest on pre-tax loss (being the profit from lost sales). Error is said to be displayed in not assessing the tax that would have been payable on the profits that were lost, and only awarding interest on that post-tax sum. 264 It can be accepted that pre-judgment interest is compensatory in nature: Haines v Bendall [1991] HCA 15; 172 CLR 60 at 66 to 67. 265 The proper approach to that compensatory task in the context of a statutory remedy involves a choice or discretion. In Cullen v Trappell [1980] HCA 10; 146 CLR 1 at 22, Gibbs J said that the award of interest should always be approached in a broad and practical way and should not be allowed to assume disproportionate importance in the resolution of the dispute. 266 In Daniels v Anderson (1995) 37 NSWLR 438; 118 FLR 248 at 586, Clarke and Sheller JJA expressed the view that interest should generally be on a pre-tax basis. See also Hodgson v Amcor Ltd [No 9] [2012] VSC 205. 267 One of the reasons for this approach is the limited capacity of the Court, in a case about lost sales and profits, to estimate the tax liability, especially in the absence of the revenue authorities. This may, perhaps, be especially so in circumstances where one may have to attribute intra-group payments for the value of patent rights and assess legitimate intra-group costs and charges. 268 The primary judge's approach at [338] to [342] saw no call to undertake a taxation exercise. We see no error in her Honour's approach, in particular in the light of the above considerations.

Indemnity Costs (Ground 14) 269 In the Costs Judgment, the primary judge awarded indemnity costs under r 25.14 of the Federal Court Rules 2011 (Cth). 270 On 6 May 2015, Bayer offered to settle the proceedings for $19,891,858. 271 Rule 25.14(3) is in the following terms: If an offer is made by an applicant and not accepted by a respondent, and the applicant obtains a judgment that is more favourable than the terms of the offer, the applicant is entitled to an order that the respondent pay the applicant's costs: (a) before 11.00 am on the second business day after the offer was served - on a party and party basis; and (b) after the time mentioned in paragraph (a) - on an indemnity basis. 272 The primary judge considered whether to make an order under r 1.35 inconsistent with r 25.14(3). That consideration was detailed and careful and was expressed as follows in [3]: Bayer seeks orders for costs in its favour on a party-party basis up to 11 May 2015 and an indemnity basis thereafter under r 25.14(3) of the Federal Court Rules 2011 (Cth). The rule, and the various statements of principle about its operation are clear, but that does not mean that the application of the rule and the principles to any particular set of factual circumstances is necessarily straightforward. In the present case, a number of factors have persuaded me that I should not make an order under r 1.35 of the Rules inconsistent with r 25.14(3), with the consequence that orders as sought by Bayer for indemnity costs ought to be made. Those factors are as follows: (1) It is for the party seeking to rebut the presumption in r 25.14(3) to persuade the Court to make an inconsistent order under r 1.35 (Futuretronics.com.au Pty Ltd v Graphix Labels Pty Ltd [2009] FCAFC 40 at [10]). (2) The objective circumstances indicate that Generic Health, if it had wished to do so, could have sufficiently informed itself about its overall exposure to liability for infringement of Bayer's patent when the offer was made for the purpose of deciding whether or not the offer represented a genuine compromise (which it did). This is so despite the fact that, when the offer was made, Bayer had not quantified its claim for damages or filed its evidence in support. (3) Generic Health's evidence on the issue of costs consisted of an affidavit from its solicitor who, while experienced in patents cases, did not have the experience and expertise available to Generic Health (a large generic manufacturer of pharmaceuticals) which, I infer, before or at least by the time the offer was made: (a) knew the price of Yasmin (because it set its own price for Isabelle materially below that of Yasmin); (b) knew or had a good idea about Yasmin's total market share and thus the potential total revenues available for the sale of the Yasmin molecule (matters it must have considered before launching Isabelle); (c) knew the costs of manufacturing, supply and sale of Isabelle (a generic version of Yasmin) and its own margin on the sale of Isabelle; (d) knew the profit it had made on Isabelle; (e) knew that its solicitors held information about Bayer's EBIT on the sale of Yasmin for the period 2006 to 2011 to which access could have been requested on the giving of a confidentiality undertaking; (f) contrary to the submissions for Generic Health, did not have a reasonable basis to assume it would be rebuffed if it had sought access to this information for the purpose of assessing the offer (because, in short, Generic Health's case based on s 115 of the Patents Act 1990 (Cth), in my view, was merely speculative, being based on nothing more than the hope that Bayer might not discharge its onus of proof); (g) knew that Bayer was considering launching Pettibelle as a result of the sale of Isabelle, as the introduction of a generic version by an originator company is a known strategic response to a generic competitor entering the market; (h) knew that Bayer, from 26 June 2014, had in fact launched Pettibelle at a substantially discounted price compared to Yasmin and continued to sell Pettibelle at that discounted price; (i) knew that Bayer intended to claim damages for the period of infringement before the amendment of the patent on 14 December 2012; (j) knew that Yates J had rejected the contention that Bayer had engaged in covetous claiming in Bayer Pharma Aktiengesellschaft v Generic Health Pty Ltd [2012] FCA 1510; (2012) 99 IPR 59 (the amendment decision); (k) knew that the patent in its unamended form had been based on overseas applications prepared under the supervision of Dr Broesamle, an experienced European patent attorney and highly qualified chemist, and by Plougmann and Vingtoft, which could readily have been identified as a firm of highly regarded patent attorneys; (l) knew that Bayer's position was that the amendment was not proposed "to overcome any prior art or, indeed, to address any other asserted, apparent or possible ground of invalidity" (at [26] of the amendment decision); and (m) had had the opportunity to cross-examine Dr Broesamle at length and knew that Yates J had concluded that nothing which emerged suggested "any untoward or inappropriate conduct on the part of Bayer Pharma which, on the present state of the evidence, would warrant criticism of Bayer Pharma by the Court" in respect of the amendments (the amendment decision at [218]). (4) As explained further below, these inferences I draw about Generic Health's knowledge (or capacity to become aware if it had wished to do so when the offer was made) compel me to conclude that, had it wished to do so, Generic Health would have had no real difficulty in assessing its maximum total liability for the infringement and had no reason to suppose that Bayer, if forced to litigate, would claim anything materially less than that maximum amount. The fact that Generic Health sought no explanation from Bayer as to how it had calculated the amount of the offer, when the precision of the amount indicates that it resulted from a calculation, supports my conclusion that Generic Health was not interested in whether the offer represented a genuine compromise. This is also confirmed by the fact that, before the hearing and when Bayer's claim was the subject of evidence, Generic Health made its own offer for a lesser amount than Bayer's offer. The difference between the amount Bayer claimed and Generic Health's offer represented the damages for the infringements before the amendment of the patent based on s 115 of the Patents Act. In other words, Generic Health was prepared to chance its hand to save itself some $6 million by seeing if Bayer could discharge the onus of proof on it under s 115 to establish that the patent specification in its unamended form had been framed in good faith and with reasonable skill and knowledge. (5) Nothing in the evidence from Generic Health's solicitor suggested that she had sought or obtained instructions from Generic Health to the effect that, despite the matters set out in (3) above, it was unable to assess whether or not the offer represented a genuine compromise because it did not know or could not estimate the quantum of Bayer's claim when the offer was made. 273 Her Honour elaborated upon these reasons at [5] to [9] of the Costs Judgment. Her Honour expressed the view that Generic Health had an intimate familiarity with the pharmaceutical market in Australia. From this familiarity, Generic Health would have been able to work out its maximum liability to Bayer, that is, to make an estimate. At [9], her Honour dealt directly with the submission that the claim had not been sufficiently identified: This is not a case in which Bayer made an offer and then changed its case. The sole reason for Generic Health's contention that an order inconsistent with r 25.14(3) should be made is that, when it made its offer, Bayer had not informed Generic Health of the quantum it claimed so that, as I understand it, Generic Health was not able to ascertain if the offer represented a genuine compromise or not. While I do not doubt that an offer made before notice of the quantum of a claim may well mean that a party is unable to assess whether an offer represents a genuine compromise, and that such a circumstance also may well constitute proper reason to make an order inconsistent with r 25.14(3), the bare statement - that Bayer had not yet given notice of the quantum of its claim - does not reflect the full spectrum of circumstances relevant in the present case. The objective facts, which I have identified above, support an inference of sufficient knowledge on Generic Health's part at the time the offer was made to make an assessment of the genuineness of the offer as a compromise of Bayer's claim. In the circumstances of this case, Generic Health has not adduced sufficient evidence to support a contrary inference when it is the party which bears the persuasive burden for any departure from r 25.14(3). It has not adduced any evidence at all from an officer of Generic Health who might be taken to have familiarity with the factors and circumstances identified above, such as Mr de Alwis, General Manager and Director of Generic Health, who swore two affidavits in the proceeding including one relating to Generic Health's profits from Isabelle for the purpose of Bayer electing between an account of profits and damages (an election made before Bayer made its offer). 274 It was submitted that Generic Health did not know the quantum of the claim and the basis on which it would be calculated. The offer was made some months before Bayer served its evidence in support of the quantum claims and before discovery. 275 The difficulty with the submission is that there was no attempt made by Generic Health to prove that it could not have made a reliable estimate. The primary judge said the following at [8]: The affidavit from Generic Health's solicitor, to the extent relevant to this issue, says only that when she received the offer she did not know the quantum of the claim or the basis upon which the claim would be calculated and nor, to her knowledge, did Generic Health. So much may be accepted. Ms Owen does not say that Generic Health had made no estimate of the claim against it at the time the offer was made. She does not say that it was not obvious to her or to Generic Health that Bayer's claim was for loss it had suffered in circumstances where the only product for which Isabelle could be substituted was Yasmin and Bayer, to Generic Health's knowledge, had introduced Pettibelle [sic] in response to Isabelle. She does not say that Generic Health was unable to estimate Bayer's claim. She does not say that she or anyone else within her firm discussed the offer with Generic Health or decided that it could not be assessed because of lack of information. 276 The primary judge considered that Generic Health had failed to persuade her that there were sufficient circumstances to depart from r 25.14(3). That was founded substantially on her Honour's view as to a lack of evidential basis to conclude that a sophisticated competitor could not have estimated their liability on the claim. 277 We do not see any appealable error in her Honour's approach.

CONCLUSION 278 We have partly upheld the appeal concerning the lost sales of Yasmin (Grounds 5, 8 and 9) and on the Petibelle question (Ground 12), but would otherwise dismiss the appeal. I certify that the preceding two hundred and seventy-eight (278) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Chief Justice Allsop, and Justices Yates and Beach.