In St. Germain v. Benhaim, 2014 QCCA 2207, the Quebec Court of Appeal revived a seldom applied approach to causation in medical malpractice cases by applying an “unfavourable inference” of proof of causation against the defendant physicians who failed to diagnose a patient’s lung cancer.
The facts of the case are tragic. The patient, an active and seemingly healthy middle-aged non-smoker, received a chest x-ray in November 2005 as part of his annual physical. The radiologist assessing the scans noticed a lesion and recommended another scan. Another x‑ray was taken in January 2006 and the radiologist again recommended another scan. The third scan was not taken until December 2006 during the patient’s annual physical. By this time, the lesion had grown. At that point, the radiologist, suspecting cancer, sent the patient for a series of tests. The tests revealed, in January 2007, that the patient had stage IV cancer that could no longer be treated. The patient died in June 2008.
The trial judge found that the family doctor and the radiologist had committed a fault by failing to compare the patient’s 2005 scan with his previous records and by failing to take reasonable steps, in 2005, to determine if the lesions were cancerous. She found, however, that the plaintiffs had failed to prove on a balance of probabilities that the cancer would have been treatable had it been diagnosed in 2005, preferring the evidence of the defendants’ expert that the cancer would already have been at stage III or IV in November 2005. In other words, the trial judge found that the plaintiffs had failed to establish causation.
The Court of Appeal reversed in two sets of concurring reasons. Kasirer, J.A., writing for the majority, held that the trial judge erred in law in failing to apply an “unfavourable inference” of causation against the defendants. The “unfavourable inference”, first set out by the Supreme Court in Snell v. Farrell, [1990] 2 S.C.R. 311 and confirmed in a Quebec medical malpractice setting in St. Jean v. Mercier, [2002] 1 S.C.R. 491 provides that in certain cases, “very little affirmative evidence on the part of the plaintiff will justify the drawing of an inference of causation in the absence of evidence to the contrary” (Snell, p. 329).
The majority found that the conditions were met in this case.
First, because of the physicians’ negligence, it was impossible for the plaintiffs to prove, scientifically, by direct evidence, the stage of the cancer in November 2005. Staging, in order to be reliable, requires a biopsy. No biopsy was taken because of the physicians’ negligence.
Second, the plaintiffs did adduce authoritative medical statistics that 78% of cancers that are discovered fortuitously – as was the case here – are discovered at stage I. This statistical evidence was sufficient affirmative evidence to support an unfavourable inference of causation. In the majority’s view, the expert evidence adduced by the defendants was no more than a hypothesis and was insufficient to establish that the deceased was in the 22% of late-stage, fortuitously discovered cases.
Kasirer, J.A. was careful to explain that the unfavourable inference does not create a veritable presumption of causation. The burden is relaxed, but not shifted. The ultimate burden of proving causation on the balance of probabilities still rests with the plaintiff, in keeping with the curative function of extra-contractual liability. However, the unfavourable inference addresses the “potential unfairness of holding plaintiffs to an impossible standard and relieving doctors of liability by reason of their negligence” (para. 169). The need for such an evidentiary tool is heightened in an all-or-nothing legal system that does not recognize loss of chance.
Fournier, J.A. came to the same conclusion on causation in his concurring reasons, but via a different route. He found that the trial judge had erred in her appreciation of the evidence and that the appellants had succeeded in establishing fault on a balance of probabilities, without resort to any negative inference. In his view: “Il n’y a pas d’allègement du fardeau de preuve du fait que la faute des intimés empêche de faire la preuve directe de l’un des éléments essentiels de la responsabilité.” (para 63).
The majority of the Court of Appeal did not invent the unfavourable inference in St. Germain. However, as Kasirer, J.A. noted, since St-Jean – which did not actually apply the negative inference on the facts – its application in Quebec has been “rare”, and sometimes wrongly confused with a reversal of the burden of proof.

Whether the unfavourable inference will be more frequently or more widely applied in the future will undoubtedly depend on what the Supreme Court has to say in St. Germain. Stay tuned!

Image(s) and Emphasis added // Fair use – Education
Republished full commentary – Source –
Full case (French) – English reasons only follow at para 138
Référence : St-Germain c. Benhaim, 2014 QCCA 2207 (CanLII), <>, consulté le 2016-09-23