The Supreme Court of Canada Disagrees with Medical Experts in BC (Workers’ Compensation Appeal Tribunal) v Fraser Health Authority

 “Causation can be inferred⎯even in the face of inconclusive or contrary expert evidence⎯from other evidence, including merely circumstantial evidence.” British Columbia (Workers’ Compensation Appeal Tribunal) v Fraser Health Authority, 2016 SCC 25 at para 38.

When is it correct for a court to disagree with scientific experts? This is precisely the issue raised in the recent Supreme Court of Canada (“SCC”) case British Columbia (Workers’ Compensation Appeal Tribunal) v Fraser Health Authority, 2016 SCC 25 [Fraser Health Authority].

I believe that the approach in Fraser Health Authority sets a dangerous precedent for interpreting legislative standards of proof. These standards have been interpreted so progressively as to have the effect of removing the requirement of plaintiffs to put forth a tangible hypothesis or prove a direct causal link between the consequence they have suffered (in this case, their disease) and the alleged cause (in this case, employment).

The Litigants

While the SCC refrained from substantially commenting on the litigants before them, there is no doubt that the plaintiffs in this case invited serious sympathy. Fraser Health Authority was filed on behalf of three workers who were among seven technicians at a single hospital laboratory diagnosed with breast cancer. Each of them applied for compensation under the BC Workers Compensation Act, RSBC 1996, c 492, ss 6, 250(4) [the Act], which compensates workers for diseases arising out of and in the course of their employment.

Burden of Proof/Standard of Review

From the beginning, the SCC’s classification of the Act as a “comprehensive no-fault insurance scheme” with a legislative burden of proof below the civil standard of a balance of probabilities became determinative (Fraser Health Authority, para 4). The Act sets out an extremely relaxed burden of proof of “causative significance,” meaning that the conditions at stake must constitute “more than a trivial or insignificant aspect” in the unfortunate consequences in order to establish causation (Fraser Health Authority, para 6). This low threshold is even further compounded by section 250(4) of the Act, which states that where the evidence is evenly weighted with regard to causation, the issue must be resolved in the worker’s favour. This low standard is likely motivated by a political choice to allow claimants to more easily gain compensation under the Act.

Both the majority and minority applied the standard of “patent unreasonableness” as the appropriate standard of review. Simply put, “a reviewing court may not interfere with a finding of fact or law made by the Tribunal in respect of a matter over which it has exclusive jurisdiction unless it is patently unreasonable” (Fraser Health Authority, para 8).

The majority states that it should defer “where there is evidence capable of supporting (as opposed to conclusively demonstrating)” a finding of fact (Fraser Health Authority, para 30). While the dissent by Justice Coté frames her statement as a disagreement, I argue that she is effectively saying the same thing when she quotes the Board’s Rehabilitation Services & Claims Manual, vol. II [RSCM II]: “if the Board has no or insufficient positive evidence […] the Board’s only possible decision is to deny the claim.” Instead of using different standards of proof, where the majority and minority decisions differ is on whether the evidence before them can actually constitute “positive evidence,” given the expert’s explanation of it.

While the majority and minority seem to be in agreement that the legislative burden of proof in this case is quite low (it merely requires some positive evidence), they disagree over whether the scientists employed a high scientific standard of proof or the lower legislative burden of proof. The majority begins its critique of the lower court decisions by stating that scientific standards for findings of causation are more rigorous than the standards demanded by the law, especially under the Act. Ultimately, the majority assumed that the scientists came to the conclusion that there was no causation based on a higher scientific standard. However, as I will discuss later and as is addressed by the minority, the Report makes it clear that the scientists were not employing a higher scientific standard when investigating potential causes. I believe the Court would be in agreement if the experts presented some evidence of causation but dismissed it for not rising to the scientific threshold. However, that is not what occurred in this case.

The experts made it very clear that the information the SCC eventually relied upon to make their decision cannot be used to draw inferences of causation. This is different than saying evidence does not rise to a causal threshold. In the experts’ own words, they “did not find any scientific evidence for the plausibility of a laboratory work-related etiological hypothesis regarding breast cancer” [emphasis added]. Throughout their Report, the experts used words like “may be associated” and “might relate.” Only if they found potential evidence would they have balanced it on a scientific scale. Therefore, I argue that the SCC usurped the role of scientific professionals, essentially disregarding the multiple experts’ advice on how to use the information before them.


Three expert reports composed the only evidence in this case: an Occupational Health and Safety Agency for Healthcare in British Columbia (OHSAH) report, and two expert reports from specialists in occupational medicine.


The main evidence used by the SCC was the OHSAH report, which contained an epidemiological (statistical) analysis of the cancer cluster, along with a literature review and field investigation. It was the interpretation of the statistical information that was decisive in this dispute.

The majority used the statistics to prove “causative significance,” i.e. that the workplace was more than a trivial or insignificant aspect in the development of breast cancer. The OHSAH report confirmed that the number of current diagnoses (7 out of 63 workers) was a statistically significant cluster eight times the rate of the general population. Looking at this information, the majority felt there was a “strength of association” (a ratio of the incidence of disease among exposed workers when compared to the incidence within the general population) and a “temporal relationship” (a proximity in time between exposure and the disease onset), which were both evidence of causation.

The experts were clear to explain that statistical correlations do not equal causation. Clusters, such as the one observed, can occur naturally as a result of an uneven distribution of non-occupational risk factors. For example, the workers’ natural cluster of age, weight, family history, age of first menstruation, age at pregnancy and first birth, numerous lifestyle factors and even natural triggers currently unknown to science could have resulted in this statistical anomaly. The scientists made it clear that if there was a constant exposure, one would expect to see a cluster sustained over many years. However, this cluster is relatively recent (over a period of 34 years). Moreover, the field report could not find any current large exposures to carcinogens in the workplace. That would mean that past carcinogenic exposure, if from the workplace, would only be manifesting itself now. Based on the numerous explanations that could result in this anomaly, the experts cautioned against solely using this evidence, yet that is precisely what the majority did.

Potential Past Exposure to Carcinogens

Confusingly, the majority stated that their decision was not based on statistics alone but also “evidence of past carcinogenic exposure” (Fraser Health Authority, para 15). The Court based this conclusion on the fact that the experts did not rule out the possibility that past carcinogenic exposure could potentially explain the statistical anomaly. However, this finding misinterprets the onus needed for this case. The onus is on the party seeking compensation to establish some positive proof, not on the investigating experts to disprove all plausible explanations. Using a radical example, eating margarine could possibly be connected to breast cancer. The experts could have just as easily stated staff lunches could explain the breast cancer anomaly without providing additional proof for that hypothesis. I doubt that the Court would rule this bare statement, without more, as sufficient evidence demonstrating proof.

Not only are the majority’s inferences suspect, they misinterpreted the expert evidence that actually comments on why past carcinogenic exposure is unlikely to be the culprit. Of all the chemicals used in the laboratory, only one, ethylene oxide, has a known association with breast cancer and this association is weak. The current exposure to this chemical was minimal because of the small liquid volumes and handling procedures. The majority stated that “the experts acknowledged that present chemical exposures were minimal, [however] past exposures had likely been much higher and included at least one known carcinogen” (Fraser Health Authority, para 25). The statement that past exposures were higher is a misappropriation of the evidence. The experts’ comment that past chemical exposure was “likely much higher” was not in reference to the only known carcinogenic substance. One expert even testified that past exposure to this substance at high levels was “highly unlikely.” Therefore, not only is it wrong to use lack of expert conclusions as a form of evidence, the evidence on record actually points to experts being able to refute that speculative hypothesis.

As for the Court’s reliance on the expert’s inability to reach a scientific conclusion on causation, I am at a loss as to how a lack of conclusionary findings can become positive evidence. If there is no connection, then none should be found.


While my criticism of the majority has been harsh, I understand that my views would preclude several individuals from gaining compensation when they need it the most. In a way, I feel that the SCC has made a policy decision to change the use of evidence in order to allow seven women, who are already struggling, to gain resources for a terrible disease. While the length of the time it takes to reach the SCC is long, the studies needed to show stronger statistical correlations take longer. It would be heartbreaking to deny compensation only to have it be proved at a later time. One could interpret the objective of the Act as being to compensate workers quickly because things like diseases do not wait for scientific studies.

Moreover, I believe this case is significant in the way it tries to deal with the limits of scientific proof and innovation. Past exposure is extremely difficult to prove. The SCC is essentially saying that while no proof exists in the present, workers should be compensated for inferred past harms. I understand that science progresses slowly and is not infallible—only certain types of research are funded, further limiting the amount of proof—however, I worry that workers are being compensated simply on statistics. Is there any deterrence value in penalizing a workplace for statistical correlations, especially when whatever resulted in the harm has already been removed from the current workplace? On the flip side, because statistical anomalies, by their very nature are rare, this case will be unlikely to result in a fundamental change in overall compensation.

On that note, the Supreme Court has made the quasi-political decision to let these individuals be compensated anyways. I question whether that role is for the courts. Conversely, is there anything Parliament can really do? If statistical correlations are deemed to be enough for compensation, then what level of correlation is needed? This is a unique case whereby I have been happy with the result for the litigants involved but see the reasoning leaving open the potential for unpredictable future jurisprudence in the area.

I agree with the minority that there was no positive evidence showing that the seven workers’ breast cancer was due to their occupation. Although ultimately, I understand that this ruling is the only way these workers can get compensated for their disease. While I agree with erring on the side of caution, I disagree with how the SCC explicitly overruled expert explanations on how to use scientific information in the context of Fraser Health Authority. I agree that expert findings of causation can be overruled generally because legal causation is a different standard than scientific causation. That is not what happened here. However, what the SCC has implicitly said is that statistical correlations are enough to make a finding of causation without additional positive evidence. This case shows that claimants can be compensated for mere statistical correlations. On that note, I leave you with a question: how high do correlations need to be before other evidence is not necessary?

Michelle Cook

Michelle Cook is currently a 3L student at Osgoode Hall Law School in the Labour and Employment specialization stream. She started writing for the Court in 2L as a Contributor and now is an Editor for the blog. Her legal interests are diverse and she enjoys writing in the areas of labour and employment, administrative, corporate, commercial and resources law as well as on legal education.

You may also like...

Join the conversation

Loading Facebook Comments ...