Glaxo Wellcome UK Ltd (t/a Allen & Hanburys) & Anor v Sandoz Ltd & Ors [2017] EWHC 3196 (Ch)

[2017] EWHC 3196 (Ch)Case No HC-2015-005005
IN THE HIGH COURT OF JUSTICE
BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES
INTELLECTUAL PROPERTY LIST (ChD)
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 15/12/2017
Before
THE HON. MR JUSTICE BIRSS
Between
(1) GLAXO WELLCOME UK LIMITED (T/A ALLEN & HANBURYS)Claimants(2) GLAXO GROUP LIMITEDClaimant(1) SANDOZ LIMITEDDefendants(2) SANDOZ INTERNATIONAL GMBHDefendant(3) AEROPHARM GMBHDefendant(4) HEXAL AGDefendant
Simon Malynicz QC and Alaina Newnes (instructed by Stephenson Harwood LLP) for ClaimantsMartin Howe QC and Iona Berkeley (instructed by White & Case LLP) for DefendantsHearing Hearing dates: 29 November 2017
Judgment Approved
[1]This is an application by the claimants for permission to adduce survey evidence in passing off proceedings arising out of the colour and get-up of “Seretide” combination inhalers. Seretide is a product used by patients for the treatment of asthma and chronic obstructive pulmonary disease (“COPD”). It consists of a combination of two drugs, a long-acting β-agonist called salmeterol and an inhaled corticosteroid called fluticasone. The Seretide combination is delivered in two different forms of combination inhaler – an MDI (metered dose inhaler) called “Evohaler” and a DPI (dry powder inhaler) called “Accuhaler”. The distinction between these two forms of combination inhaler does not matter, nor does the fact that the name “Seretide” really refers to the drug combination rather than the inhaler device.[2]Seretide is said to be the first combination inhaler to be launched in the market. It is used twice daily by patients to control asthma or COPD and prevent an attack.[3]The claimants argue that the Seretide inhalers are sold in a get-up and packaging distinctive of the claimants and they are owners of significant goodwill not only associated with the name SERETIDE but with get-ups and packaging.[4]The claimants’ pleaded case refers to various features or “indicia” of the get-up they rely on. One of the features is the purple colour(s) used on the products themselves and on packaging. In fact two shades of purple appear on some of the claimants’ packs and in the marketing material (hence “colour(s)”) but it is clear that an important aspect of the claimants’ passing off case relates to one of the shades of purples itself. That is the darker shade. The claimants call it the “main purple colour”.[5]Members of the defendants’ group have launched a generic version of Seretide in various EU Member States called AirFluSal. It contains the same active ingredients as Seretide. AirFluSal is sold in a purple inhaler (called Forspiro) and has been sold by reference to a purple colour in the packaging. The claimants contend that the defendants are engaged in passing off. The first defendant is alleged to make a misrepresentation as to the trade origin of the product caused by the get-up, appearance, colour, livery etc. and the other defendants are alleged to be liable as joint-tortfeasors. The claimants also argue that the defendants’ acts have been deliberate. A further alleged misrepresentation relied on, not germane to this application, is as to the equivalence of AirFluSal to Seretide. That is a point about the scope of marketing authorisations.[6]Before me there was some debate about the scope of the allegation of passing off and in particular whether the claimants were relying on the use of the colour purple itself and on its own as a feature in which they have goodwill and as something the use of which amounts to passing off. The short answer is that the claimants do indeed advance that case. Moreover the defendants understand this, as can be seen from a letter to the UK IPO dated 13th December 2016 in parallel opposition proceedings concerning an application by the claimants to register a purple colour mark in the United Kingdom (Application No. 3108001, opposition now stayed pending this case). The defendants’ solicitors made the following point about the issues in this High Court action:
“Although other indicia are also relied upon, the claimants’ primary focus is on their use of purple (see section [in the Amended Particulars of Claim] headed “The use of purple for inhalers”, […]), and further it is clear that the claimants assert that the use of the colour purple as such (i.e. leaving aside the other alleged indicia) is sufficient to amount to passing off ([…]).”
[7]That is an accurate statement of what the claimants’ case is. There was some debate about the wording of the pleadings before me but a further amendment to the Particulars of Claim, not opposed, puts the matter beyond doubt.[8]The defendants deny passing off, including the allegation of intention. There is no need to summarise the entirety of the defendants’ case, but it includes a denial that the purple shade (or any colour) is distinctive of the trade origin of the goods as well as a denial that their use of a purple colour makes a misrepresentation.[9]A particular feature of the defendants’ case is a positive allegation that in the relevant market, colour has a particular meaning. It is said to be a reference to the type of medicine, so pale blue indicates a short acting β-2 agonist reliever product, brown/orange/burgundy represents a corticosteroid preventer product and purple (or more particularly two tone purple as used with Seretide) indicates a combination product. The claimants deny that these colours in general, and purple in particular, have this meaning.[10]Another particular feature of this case is that the market is a specialised one. Members of the general public do not select these products to buy because the products are sold on prescription. The relevant persons with whom the claimants would have goodwill and to whom the alleged misrepresentation is made are doctors and pharmacists. Given the nature of the prescribing the doctors will usually be GPs. The pharmacists may include pharmacy technical assistants (PTAs). Patients may have a role to play but it is clear that this case will involve a significant focus on doctors and pharmacists.[11]The survey evidence that the claimants seek to adduce was originally carried out to support acquired distinctiveness of the main purple shade on Seretide inhalers, Pantone 2587C. That is the subject of the pending UK trade mark application. This survey evidence was previously admitted into the UKIPO opposition proceedings by the hearing officer, Mr George Salthouse, following objections by defendants. Before they were stayed the opposition proceedings were well advanced. The evidence was complete and both sides had filed evidence about the surveys.[12]Once the opposition proceedings were stayed and it became clear that the passing off claim would come first, the claimants applied to be permitted to rely on the survey evidence in these proceedings. Normally on applications of this kind today, the main survey has not yet been carried out but in this case all the survey work has been done because of the history described.[13]The material is in two parts. The first survey exercise was carried out in 2015. In one survey 200 GPs were interviewed and in the other 200 pharmacists (including PTAs). The second survey exercise was carried out in 2016. Again one survey was with GPs (251) and the other with pharmacists (including PTAs), numbering 252. In all surveys the core of the survey involves putting a blank rectangle of the relevant purple colour shade to the interviewees and asking questions. The claimants say the results support their case on distinctiveness.[14]The 2015 survey used a method commonly used in a number of Continental courts (such as Germany, Austria, Swiss and Benelux) as well as the EUIPO. About 93%/92% of pharmacists/GPs respectively had seen the colour on an inhaler while the remaining 7%/8% said it looked familiar. Of these over 90% of both considered it indicated the origin of one pharmaceutical company and 90%/88.5% of pharmacists/GPs respectively indicated that was the claimants. In what the claimants say is a verbatim record of answers to the questions, the claimants rely on particular answers which indicate, the claimants contend, a high degree of conviction that this colour is a brand colour, exclusively associated with the claimants.[15]The 2016 survey used a method that closely follows the method designed for and approved by the court (Morgan J on permission and later Arnold J at trial) in Enterprise v Europcar ([2014] ETMR 50 and [2015] FSR 22). Once the purple colour was shown 7% of pharmacists and 5% of GPs mentioned GSK/Seretide before any questions were asked. In response to later questions 63.9%/75.7% of pharmacists/GPs respectively said they had seen the colour on Seretide, and many people, when asked if there was anything else they could tell the interviewer about the colour in relation to inhalers gave answers which the claimants contend support the idea that the colour is a “brand” (such as by saying the colour is a brand).[16]The conduct of all the surveys was overseen by Dr Pflüger, a well known German survey expert.[17]The claimants wish to use these surveys to support their case on distinctiveness. They do not seek to use the surveys as confusion surveys. Nor are the surveys being used as witness collection programmes, although the claimants do plan to call some GPs and pharmacists, those witnesses were not identified in these surveys.[18]The defendants oppose this application. The claimants contend that the survey evidence is of very limited value, based on an artificial premise and also is of minimal relevance to the actual passing off claim.[19]Although the parties had already exchanged expert evidence about the surveys in the opposition proceedings and the matter was ready to be decided, further evidence was served on this application. In particular the defendant served evidence against the surveys from a further expert, Mr Malivoire. He is a well known British survey expert. This was in turn responded to by the claimants’ survey expert witnesses.[20]A number of the points taken by the defendants have been answered by the production of further material from the claimants. The defendants criticised this approach and contended that the confusing manner in which the evidence had been presented was itself a reason to refuse permission. I sympathise with the defendants with the way the information came out but while it will be relevant for the court to take into account the impact of the evidence on the further conduct of the case, and confusing evidence is obviously undesirable, subject to a point about the number of witnesses which I will return to, I am not persuaded that the material relevant to the survey is now very difficult to follow or would be confusing at trial. Getting to this point has been painful but now we are here, if it was the right thing to do, going forward would be much less so.

The law

[21]In Imperial Group Plc v Phillip Morris[1984] RPC 293, Whitford J laid down guidelines concerning the admission of survey evidence in trade mark infringement and passing off cases, known as the “Whitford guidelines”. These were summarised by Lewison LJ in Interflora v Marks and Spencer [2012] EWCA Civ 1501 at [61] as follows:
“61 These (‘the Whitford guidelines’) can be summarised as follows: i) if a survey is to have any validity at all, the way in which the relevant interviewees are selected must be established as being done by a method such that a relevant cross-section of the public is interviewed; ii) any survey must be of a size which is sufficient to produce some relevant result viewed on a statistical basis; iii) the party relying on the survey must give the fullest possible disclosure of exactly how many surveys they have carried out, exactly how those surveys were conducted and the totality of the number of persons involved, because otherwise it is impossible to draw any reliable inference from answers given by a few respondents; iv) the questions asked must not be leading; and must not direct the person answering the question into a field of speculation upon which that person would never have embarked had the question not been put; v) exact answers and not some sort of abbreviation or digest of the exact answer must be recorded; vi) the totality of all answers given to all surveys should be disclosed; and vii) the instructions given to interviewers must also be disclosed.”
[22]In Interflora I, Lewison LJ also explained that to be admitted a survey must be of real value. He held at 150 as follows:
“150. In deciding whether to give permission, the court must evaluate the results of whatever material is placed before it. Only if the court is satisfied that the evidence is likely to be of real value should permission be given. The reliability of the survey is likely to play an important part in that evaluation, even then the court must be satisfied that the value justifies the cost. As Mr Hobbs said, this requires the court to conduct a cost/benefit analysis.”
[23]In Interflora 2 (Interflora v Marks and Spencer [2013] EWCA Civ 319) Lewison LJ was concerned that he had not made himself clear and explained that a judge should not let in survey evidence unless satisfied that it is of REAL value and that the likely value justifies the cost (paragraph 26).[24]The Court of Appeal in Zee Entertainment Enterprises v Zeebox [2014] EWCA Civ 82 applied this test in a passing off case. The survey in that case was flawed and not of real value.[25]Many of the surveys which are criticised are surveys designed to test confusion. In Enterprise v Europcar the survey that was permitted was not a confusion survey, it was aimed at supporting the case on distinctiveness.[26]Finally, if a survey can be seen to lack real value at this stage then it should not be permitted, but one must guard against turning the application into a mini-trial. It is not. The issue is one of case management.

The remaining objections

[27]Before me the defendants took two classes of objection. The first class was directed to their case that the surveys have an artificial nature and limited relevance. The second class was directed to the quality of the survey evidence.[28]The major objection in the first class was that the survey was artificial because it was based on a square of colour, rather than, for example, one of the claimants’ inhalers. It therefore took the colour out of context and ignored the other aspects of the claimants’ get-up. The surveys do not relate to the actual Seretide inhalers which the claimants contend generate the goodwill upon which it wishes to rely in this action. Therefore, the surveys are said to be based on an artificial premise that does not relate to the real world situation the court must assess.[29]If this was a good point it would be a powerful reason to refuse to admit the surveys but it is not. What I can say at this stage is that the claimants have a good argument that in this respect the surveys took an appropriate approach and did not take colour unduly out of context. The claimants’ case is (or at least involves) an allegation that the colour per se is distinctive. That colour is used by the claimants not only on the product itself but on packaging and in marketing material including leaflets and even the colour of trade stands at a trade show. Given that this is the claimants’ case, it is sensible to test it using a square of colour rather than a particular object like an inhaler or a leaflet or anything else. A survey based on a square of colour may well have real value in assessing such a case. The point will not be a costly one to decide, the experts can be cross-examined on it in fairly short order if need be.[30]The defendants took two other objections in the first class. The first was that the surveys did not test what the defendants called the “crucial issue” of deception or whether AirFluSal is actually deceptively confused with the Seretide products. It is true that the surveys are not aimed at misrepresentation or deception, they are aimed at distinctiveness. If distinctiveness was admitted then this objection would be well founded but it is not. The defendants deny that the colour is distinctive. They are entitled to do so and the matter will go to trial. However the fact the survey is aimed at distinctiveness and not deception is not an objection to it.[31]The second objection was that the survey did not relate to any pleaded indicia. I believe that objection was dropped. On any view as pleaded now, the survey does relate to the claimants’ pleaded case. The fact it does not relate to all the pleaded indicia is no objection. In another case if the cost and trouble of a survey was directed to what was obviously a minor feature of the get-up then a point like this might have force. But in this case the colour per se is not a minor aspect of this dispute.[32]So I reject the first class of objections. They do not establish that the surveys can be found now not to be of real value. I will consider the matters as a whole at the end.[33]The second class of objections related to the quality of the survey evidence itself. It was said to be poor and questionable. The first ground was that the evidence was confused, disorganised and convoluted. I have addressed that already. It is not a free standing reason to refuse to admit these surveys.[34]The other grounds are: i) The surveys are inadequately documented; ii) The surveys are so flawed that they offer little intrinsic value to the Court; iii) The significant costs that would be incurred by the inclusion of the claimants’ survey evidence in these proceedings are not justified by the limited value of the proposed evidence.[35]Taking each point in turn:

Inadequate documentation

[36]A major part of this objection is historic and overlaps with the defendants’ submission that the evidence has been confused and disorganised. The defendants had various questions which were not answered by the material they had but which have been subsequently answered. Nevertheless it is right to say that there are points which remain. In his reports Mr Malivoire makes a general point that the lack of proper documentation relating to a survey means that it is “difficult or indeed impossible to test the validity of any conclusions it makes”. At times in the written materials it seemed that the defendants’ case based on this evidence was that any problem with documentation of any significance justified the wholesale rejection of a survey. I do not think that is what Mr Malivoire meant to say but if it is then I would not agree that at this stage one can make such an assessment. Most of the outstanding points on documentation were answered. The issue turns on what is outstanding.[37]The major point which is outstanding is that it is said there is no adequate record of how the interviewers were trained and briefed, with the result that there are unanswered (and unanswerable) questions (c.f. Whitford Guideline (vii)). The main questions raised by the defendants are that it was not clear whether the interviewers were told to make sure no inhalers were visible, were told how the interviewers should be introduced, how questions the respondents might ask should be addressed and what prompts if any were permitted. The defendants are right that the way the surveys were documented means that one cannot look at a document and answer these questions. Nevertheless Dr Pflüger describes the training for interviewers in her fourth report. The interviewers were professionals. They did not know the client or the purpose of the interview but they knew it was about inhalers and part of the basic training of such interviewers is to position themselves so that the interviewee cannot see the product in question. Interviews with pharmacists were always conducted with the pharmacist facing away from the medicine.[38]The claimants’ main submission is that these are minor issues and the fact they cannot be answered does not fatally undermine the surveys or justify wholesale rejection. I agree. I am not convinced any of these points are so strong as to mean that at this stage the court can find that the survey can be ignored as fatally flawed. It is clear that Dr Pflüger delegated much of this work. The court at trial will be able to decide what impact these points have. It may (or may not) be a difficult decision to make but it will not be one which involves a lot of time and cost. The claimants have a properly arguable case that these points, even if they have any validity, might shift the numbers to a small degree but do not undermine the essential conclusion of the survey that the colour is readily recognised.[39]A different way of looking at this is the following. The claimants contemplated as an alternative to use these surveys as “pilot” surveys and then (with permission) embark on yet more surveys. It was not a course they wanted to take but it could be an approach to consider. That puts into perspective this argument about the records. Of course it would be better if a more direct record existed of all the matters raised by the defendants but to effectively put a party to a choice between calling no survey evidence or embarking on another whole survey exercise, just because of this point, would be disproportionate. The court may decide at trial that the absence of records is important but since that is the way the claimants wish to proceed, it seems to me that the court should not prevent them from doing so without stronger reasons.[40]As a separate way of answering the defendants’ points the claimants argued that they have disclosed all the information they have and sought to suggest that this discharged their duty under the Whitford Guideline (iii) (Disclosure). On that basis the fact that they could not answer some questions about how the survey was performed did not matter. I agree with the defendants that that is not a correct approach to the Whitford guidelines. This submission from the claimants would be a recipe for avoidance. The party conducting a survey should record exactly how the survey has been conducted in order to then disclose it. The reason I reject the defendants case on this is because the issues are, at least arguably, minor and can be explained by the manner in which the surveys were originally done.[41]Mr Malivoire is undoubtedly a very experienced expert in trade mark and passing off cases. Indeed as the claimants point out, he effectively designed the method used in their 2016 survey since that was based on Enterprise v Europcar in which Mr Malivoire did design the survey. In his evidence before me Mr Malivoire said that “in my considerable experience of such cases, I cannot recall a survey where the documentation has been so sparse”. Mr Malivoire referred to surveys which were actually conducted with English court proceedings in mind, however these surveys were conducted to support the trade mark application. The question is not determined by statements of that kind. What matters is the detail, which is addressed above.

So flawed as to be of little intrinsic value

[42]The defendants also contend that even with inadequate documentation provided about the surveys, the defendants’ experts, Mr Malivoire and Mr Rothman, both consider the surveys are so flawed that they offer little intrinsic value to the court. As with the other issues, dealing with this depends on the detail.[43]One point I can dismiss right away is the submission that the surveys are flawed because the interviewees did not include patients. However given the prescription nature of these products, the claimants approach of focussing on GPs and pharmacists is sensible.[44]A further point is that the defendants’ experts do not believe the answers are truly verbatim. This has two aspects. One is based on the characteristics of the answers actually given and the other is because the instruction to record verbatim in one survey only appears at the start and could be misunderstood as applying only to the first question. Neither point is very strong. Both could readily be dealt with at trial. They are not reasons to reject the survey at this stage as not having real value.[45]The defendants also criticise the sampling method used, the most significant point being that the approach used was based on sampling GP practices as the unit. So a medical practice would be sampled once, irrespective of its size, and therefore the results over-represent smaller practices relative to the number of GPs. That appears to be correct as an observation but it is not obvious to me why it matters unduly and in any case the trial judge will be in a good position to deal with this.[46]As mentioned above, the 2015 surveys were conducted using the standard three step methodology which is often used in some Continental European courts. The defendants’ experts say that all the questions after question 2 are leading, invite speculation and contravene Whitford guideline (iv). The essential structure of the three step approach is to start with the first step which asks two relatively open and non-leading questions. One of the first step questions is – have you seen this colour used on an inhaler? Then the survey goes to the second step only for interviewees who give an answer that they have or might have seen the colour used. The second step puts three options to the interviewee and the order in which the options are presented to different candidates is changed so that they do not all get presented with them in the same order. A record of the order is kept for each subject. The subjects are asked if the inhalers that use the colour “come from a particular pharmaceutical company/a number of different companies/tells me nothing at all”. The third step is only taken for those who answer “a particular pharmaceutical company” and they are asked for a name or brand.[47]The defendants criticise these questions and this structure as leading. I agree there is an element of speculation in the questions, but the essential structure is obviously intended to minimise any leading. Moreover it is relevant to consider the task which the court faces in deciding this issue. I do not need to add to the learning in this jurisdiction on the difficulty of drawing a distinction between recognition of a sign and whether the sign is used as an indication of origin. It is a fundamental difference but, in a case like this, not an easy one. It is made all the harder by factual circumstances in which a product has enjoyed a monopoly in the past (c.f. the Linoleum problem). So while asking subjects who do recognise the sign what they think it signifies in the terms of the questions does inevitably involve speculation, it seems to me a trial judge could find some assistance in the answers given to those questions. The judge would be in no way bound to do so by this decision but it will be better to allow this evidence to go to trial than preclude it. The judge will be able to see it for what it is.[48]The defendants also challenged the 2016 surveys on similar grounds but the defendants’ better case is against the 2015 surveys and so as I have rejected that, there is no need to deal with the 2016 surveys.[49]A further reason to permit the three step 2015 surveys to go to trial would be that I would be reluctant at an interim stage to rule out the possibility that a survey of a kind commonly used in other European Union Trade Mark courts is so flawed as to not be permitted at all.

Significant further cost not worth the limited value

[50]Individually I have rejected all the major submissions made by the defendants as reasons to refuse to admit this evidence. Nevertheless in case management terms the court should still stand back and look at the position overall. Although the costs of actually conducting the surveys are sunk costs and the claimants accept they are not recoverable in these proceedings, allowing the evidence in will still commit the parties to significant cost. The defendants say the extra cost from now on attributable to dealing with the surveys (adding both sides’ together) will be about £500,000. That is an enormous sum.[51]One costs driver is the number of experts who have already given evidence about the surveys, but that is caused by the fact that this material was originally set up to be used in the UKIPO proceedings. The claimants currently have three different survey experts giving evidence and suggested that they could cut down on the number of experts by having one or two of their experts cover the same ground as one or two of the existing experts. The defendants objected to that. I think part of the objective may have been tactical in that it may not suit the party trying to keep something out on proportionality grounds for the other party to find a way of streamlining the cost of it. Nevertheless the defendants also made a good point that they should be entitled to put questions to the relevant person. I suspect it would be a false economy to require the claimants to file yet another round of expert evidence simply designed to cut down the number of individuals giving evidence. However duplicative expert evidence should not be permitted, in other words at some stage each party must make clear, if it is not already, which witness is dealing with which issue.[52]Moreover I agree with the claimants submission that it is likely that the time required for cross-examination of the survey experts will be significantly less that the estimated time of four days of cross-examination on which the cost figure is based. The parties’ positions on the time required for this shifted in the course of this application. In the defendants’ evidence they suggested that four days of trial time (two days each) would be needed for this cross-examination. The claimants did not serve contrary evidence but in their skeleton argument they submitted that the issues would be likely to narrow and suggested the total time needed would be 1½ - 2 days. By the oral submissions this reduced to 1 day. In my judgment is it likely that a single day will be sufficient for cross-examination of the survey experts as a whole. Even if the time required is a bit more than that, it means the likely costs will be lower than the estimated sum.[53]Another important factor relating to costs is that the case itself is an important and high value dispute affecting two major pharmaceutical companies and the sales of a major medicinal product.[54]Finally the overall costs of the parties are estimated to be about £6 million. That is about £2 million for the claimants and £4 million for the defendants. The higher sum for the defendants apparently being attributable to the issue of intention to deceive and the disclosure obligation that entails. So even using the estimated costs of the surveys, they account for less than 10% of the costs of the proceedings.

Conclusion

[55]Standing back, there will be substantial costs caused by this evidence but not in a quantity which would be a disproportionate share of the overall resources to be devoted to the resolution of this high value commercial dispute. There are some potential difficulties with these surveys (not as many or as serious as the defendants contend) but they are not so significant as to establish now that the evidence as a whole will not be of real value to the court or that the cost is likely not to be justified by the value. I will allow the claimants’ application.[56]Finally, it seems to me that this case would benefit from a detailed discussion between the parties and a further case management conference if necessary to address exactly what sort of evidence is to be called, how many witnesses and of what kind and so on. As for the survey evidence, there is real scope for narrowing the issues. The survey experts should meet and draw up a statement of what is agreed and what is not. I must say I also believe that serious thought should be given to conducting a concurrent evidence session (“hot tub”) with some or all of the survey experts.

Cited in 1 later judgment