In two recent cases, the Federal Court of Appeal provided "certain observations" on whether "the time has come to reconsider the view that appellate courts are to review patent construction on the basis of correctness."
In both decisions, Justice Stratas concedes that the law on the topic was settled by the Supreme Court of Canada some 15 years ago in Whirlpool Corp. v. Camco Inc. (2000 SCC 67). But he cites several reasons why a standard of review of correctness may not be the best approach.
First, he notes that the construction of a patent is conducted by the trial judge through the eyes of the skilled person at the relevant time, an area highly dependent on expert evidence. Since the weighing of expert evidence and the trial judge's conclusions on that evidence are questions of fact, those findings are to be given deference and are reviewable on the palpable and overriding error standard.
The upshot is that the evidentiary and factual context for the patent construction exercise is reviewable on a different standard than the construction itself. As Justice Stratas writes at para. 20 of Cobalt v. Bayer, "How are appellate judges supposed to cleave off those aspects of claim construction that flow from the trial judge's appreciation of expert evidence from the words of the claim per se?"
Second, he argues that judges will ensure that there is consistency and certainty in the interpretation of patents, even on a more deferential standard of review. In fact, it could be argued that given the central role of construction to most patent litigation, the correctness standard on appeal increases uncertainty. Counsel and their clients know that the landscape can shift dramatically if an appellate court substitutes its opinion on the meaning of the claims for that of the trial judge (something that will not be known until well after all arguments have been made).
Third, Justice Stratas noted that the U.S. Supreme Court recently decided to provide more deference to trial courts in the area of patent construction in Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., No. 13-854, 574 U.S. __ (2015) implementing a "clear error" test, not a de novo standard of review. His implication is that Canadian courts should consider doing the same.
Ultimately, Justice Stratas' musings don't alter the Supreme Court precedents, and the application of the correctness standard to patent construction remains the law. However, the "observations" of the Federal Court of Appeal should provide fertile ground for arguments in future cases, perhaps someday all the way back up to the highest court in the land.