A Leading Canadian IP Case: Colour Confusion over Prescription Drugs

Eli Lilly and Company v Novopharm Limited, 1997 CanLII 5008 (FC), aff’d [2001] 2 FC 502, leave to appeal to SCC refused [2001] SCCA No 100 (QL). Do Canadians associate the colour of a prescription drug with the drug company that sells the drug? What evidence is useful for answering this question? These questions were tackled in Eli Lilly and Company v Novopharm Limited. Fluoxetine hydrochloride (fluoxetine) is the active ingredient in the anti-depressant drug sold by Eli Lilly under the brand name Prozac. The patent for fluoxetine expired in March, 1996. At that time, three generic drug companies began selling generic versions of Prozac. Eli Lilly sought to stop the three generic companies from selling capsules of fluoxetine that were green and cream or green and pale grey, the colour combinations that were used for the two dose sizes of Prozac, by suing them for "passing-off" under the Trade Marks Act. The defendant generic companies conceded that they had purposely copied Prozac's colours, but asserted that the reason for doing so was because in Canada the appearance of prescription medicines is associated with the type and dosage form of the drug. They denied that there was any intention to pass their capsules off as Eli Lilly's product. To succeed in proving passing-off, Eli Lilly had to establish: i) the existence of goodwill or reputation associated with the colours of Prozac in the relevant market; ii) a misrepresentation by the generic companies to the consumers by the use of the colours; and iii) actual or potential damage resulting from the likelihood of confusion of the consumers. In support of its arguments on each of these points, the parties produced an extensive amount of evidence, including testimony from physicians and patients. The most important evidence related to surveys that were conducted to determine the views of samples of the public, coupled with the testimony of the experts that designed the surveys. By her detailed analysis of the surveys presented in the case, Madam Justice Barbara Reed provided guidance on the kind of evidence that is useful to the Court in a passing-off case. In particular, it is clear from her decision that the Court will scrutinize the methodology and data of surveys; survey results gathered by methods that minimize any potential bias and that are supported by the data will be given the most weight. In this case, Justice Reed found that the survey evidence showed that, in general, people do not attach much meaning to the colour of prescription medicines; if meaning was attached, it was most likely that the content of the capsules was the same. She also found that there was a lack of any association between the Prozac name and the colour of the capsules, even among individuals who were taking the medication. Balancing the totality of the evidence, Justice Reed held that Eli Lilly failed to establish passing-off. The capsule colours did not play a role in misrepresentation because the capsule appearance did not function as an identifier of the Prozac brand. The colours had also not acquired any reputation in the marketplace as a distinguishing feature of Prozac compared to other brands of fluoxetine. Eli Lilly also failed to prove that a likelihood of confusion existed. Consumers who were prescribed fluoxetine for the first time would not be confused because they would be unlikely to have seen the capsule before. The consumers who had taken Prozac previously would not be confused because they would associate the colours with the character of the medicine, rather than the drug company that made the drug. Douglas Deeth, who represented Novopharm in this case, comments on the use of surveys as evidence:

The Prozac litigation, with seven different surveys, was perhaps the high water mark for the use of surveys via trade-mark litigation in Canada. In Masterpiece v Alavida (2011 SCC 27), the Supreme Court of Canada said that survey evidence should only be admitted where it was, inter alia necessary to provide evidence that was “outside the experience and knowledge of a judge”, and that judges were entitled to give effect to their own opinions, and use their own common sense, on the issue of likelihood of confusion. The difficulty with that approach, of course, is that the opinion of an expert witness, whether based on survey evidence or not, is known to both parties, and can be tested by cross-examination and challenged by other evidence. A judge’s opinion and common sense are not subject to the same scrutiny at trial.

Summary by: Cheryl Cheung