Medical Negligence

“But For” – It’s Just the Name of a Test

Introduction

For nearly a decade, the Supreme Court of Canada’s ("the SCC") decision in Clements has been regarded as the leading case on causation. Counsel and Judges alike have cited Clements as authority for the analytical framework outlining the test of causation—the “but for” test. When read carefully, however, Clements did not break new ground. Instead, the case merely summarized the law of causation that had long been established by the SCC. Chief Justice McLachlin’s carefully worded decision should only be attributed for clarifying when the “but for” test is used, not how.

The goal of this article is to outline various misconceptions about causation law, particularly as they relate to interpreting Clements. As the title suggests, we emphasize that “but for” is just the name of a test. The correct analytical framework of the “but for test” necessarily includes the language of “caused or contributed.” The origin of “caused or contributed” will be explored to support the phrasing of the causation question. Finally, we briefly discuss contemporary Canadian jurisprudence as it relates to causation and offer a potential solution to simplify the analysis going forward.

Pre-Clements

In 1990, the Supreme Court of Canada released its landmark decision on the law of causation, Snell v. Farrell. The Court considered the limitations of expert evidence to determine, with scientific certainty, the cause of a Plaintiff’s injury.1 Justice Sopinka reviewed the developments in causation law, and expressed dissatisfaction with the traditional approach and its overly rigid application by the Courts. His Honour held that the correct causation inquiry assesses whether the Defendant’s negligence caused or contributed to the Plaintiff’s injury. It was established that this analysis does not require scientific precision; instead, the trier of fact should conduct the causation inquiry in a robust and pragmatic manner.2

Six years later, the SCC released its decision in Athey v. Leonati. In Athey, the Plaintiff had a pre-existing back condition and was involved in two motor vehicle accidents. He engaged in physiotherapy and chiropractic treatment, and was encouraged by his treating physician to resume his regular exercise routine. While stretching at the gym, he suffered a herniated disc. The trial judge found that both the pre-existing back condition and the impact of both collisions were necessary to bring about the disc herniation.3

Writing for the majority, Justice Major stated that a strict application of the “but for” test would be unworkable in some situations. The Court held that causation may be established where the Defendant’s negligence materially contributed to the occurrence of the injury.4 Citing Snell, His Honour stated that the analysis is “essentially a practical question of fact which can be answered by ordinary common sense;” it need not be need not be applied too rigidly, and does not require scientific precision.5

To summarize, following the Court’s clarification of the law of causation in Snell and in Athey, the established principles can be stated as follows:

  1. Causation is established where the Plaintiff(s) prove on a balance of probabilities that the Defendant caused or materially contributed to the injury;
  2. This analysis does not require the Plaintiff to prove causation with scientific certainty; and
  3. The causation analysis requires a robust, pragmatic, and common sense application by the trier of fact.

Clements

In 2012, the SCC was faced with a factual matrix involving a single tortfeasor and a non-tortious contributing cause of the Plaintiff’s injury. The parties in Clements were “bike enthusiasts.” The Defendant was operating a motorcycle, which was overloaded by 100 pounds and was speeding by at least 20 km/h over the legal limit. The Plaintiff was a passenger. Unbeknownst to anyone, a nail had punctured the rear tire, which was slowly deflating. The bike wobbled, and the Defendant was unable to bring the motorcycle back under control. He crashed and threw the Plaintiff off the motorcycle.6

At trial, expert evidence adduced by the Defendant suggested that the probable cause of the accident was the deflated tire, and the accident would have occurred whether or not the Defendant was driving recklessly on an overloaded motorcycle. The trial judge found that the instability of the motorcycle was contributed to by the punctured tire, the speed, and the load on the motorcycle. After reviewing the law of causation, Justice Grauer concluded that the factual matrix gave rise to the “special circumstance” where it is appropriate to resolve causation by applying the “material contribution to risk” test, instead of the “but for” test of causation.7 On appeal, this approach was rejected, and the BC Court of Appeal held that the “but for” test of causation prevailed.8

Chief Justice McLachlin, writing for the majority, made it abundantly clear that the only circumstances in which the “material contribution to risk” test is to be applied is when there are two or more tortfeasors, and it is impossible for the Plaintiff to show that any one of the tortfeasors caused the injury because each one can point to the other(s) as a possible cause.9 Outside of this rare set of circumstances, the “but for” test is to be applied to resolve causation.

At paragraph 8 of the decision, the Court suggests:

The test for showing causation is the “but for” test. The Plaintiff must show on a balance of probabilities that “but for” the Defendant’s negligent act, the injury would not have occurred. Inherent in the phrase “but for” is the requirement that the Defendant’s negligence was necessary to bring about the injury – in other words that the injury would not have occurred without the Defendant’s negligence, This is a factual inquiry. If the Plaintiff does not establish this on a balance of probabilities, having regard to all the evidence, their action against the Defendant fails.10

This paragraph is often cited by those advocated for a strict application of the “but-for” test. However, it is important to read the decision in its entirety. Specifically, Chief Justice McLachlin clarified that the “but for” test is to be applied in a robust and common sense fashion. Her Honour also made clear that scientific evidence of the precise contribution to the Plaintiff’s injury by the Defendant’s negligence is not required.11

Returning to the initial proposition, it is clear that “but for” is just the name of the test. Clements merely provided guiding principles in applying the “but for” test, which had already been established:

  1. The “but for” test must be applied in a robust and common sense manner;
  2. The “but for” test does not require scientific certainty; and
  3. To establish causation, the Defendant’s negligence must be a necessary cause.

Lacking in Clements however, is a detailed explanation of the required robust and pragmatic application of the “but for” test. Put simply, Clements does not assist in outlining an analytical framework. What Chief Justice McLachlin did however—and this has often been overlooked—is list Snell and Athey as examples where causation was “resolved on a robust and common sense application of the “but for” test of causation.”12 Snell and Athey set out that causation is established where the Defendant’s negligence caused or materially contributed to the Plaintiff’s injury. Clements therefore can be seen to acknowledge that the analytical framework of “caused or contributed” is a robust and pragmatic application of the “but for” test demanded by Chief Justice McLachlin. In this phrasing, the use of the language “but for” is entirely absent—demonstrating that quite clearly, “but for” is simply the name of a test.

Caused or Contributed

It is perhaps one thing to tie a series of SCC decisions together and adduce that the well-established Canadian causation law requires a Plaintiff to show on a balance of probabilities that the Defendant’s negligence caused or materially contributed to their injury. However, it is important to explore the “caused or contributed” language, its origin, and why it is a necessary formulation of the “but for” test.

Contribution language actually pre-dates Snell and Athey. The use of such language is statutorily entrenched in the Negligence Act, RSO 1990. Section 1 states:

Where damages have been caused or contributed by the fault or neglect of two or more persons, the court shall determine the degree in which each of such persons is at fault or negligence, and, where two or more persons are found at fault or negligent, they are jointly and severally liable to the person suffering loss or damage for such fault or negligence, but as between themselves, in the absence of any contract express or implied, each is liable to make contribution and indemnify each other in the degree in which they are respectively found to be at fault or negligent.13

The analysis of whether one or more Defendants caused or materially contributed to a Plaintiff’s injury as set out in the Negligence Act simply recognizes that a Defendant’s negligence may not be the only cause.14 Contribution language is necessary because it is the only phrasing that properly encapsulates the burden the Plaintiff is required to prove on causation, while avoiding some of the pitfalls of the strict “but for” wording. These include:

  1. The recognition that the presence of other contributing tortious and non-tortious causes does not absolve a Defendant of liability;
  2. The onus on the Plaintiff to prove that their injury would have been prevented or reduced; and
  3. The risk of “circular causation” in situations of multiple tortfeasors.

Multiple Causes

Phrasing the causation inquiry with contribution language is necessary because it recognizes that even though there may be multiple tortious and/or non-tortious causes of injury, so long as the Defendant’s act is a contributing cause, then the Defendant is fully liable for that damage.

To contextualize, the factual circumstances that paved the way for the SCC in Athey are particularly helpful. Again, the trier of fact found that the Plaintiff’s herniated disc would not have occurred without the combined effect of the two motor vehicle accidents (two tortious causes) and a pre-existing back condition (non-tortious cause). While the tortious motor vehicle accidents were necessary to bring about the injuries of the Plaintiff, the underlying pre-existing back condition was also necessary. Each accident, on its own, could not possibly be any more than a contributing (albeit necessary) cause of the injury. A Defendant cannot point to the non-tortious cause and be absolved from liability. As carefully explained by Justice Major:

This position is entrenched in our law and there is no reason at present to depart from it. If the law permitted, apportionment between tortious causes and non-tortious causes, a Plaintiff could recover 100 percent of his or her loss only when the Defendant's negligence was the sole cause of the injuries. Since most events are the result of a complex set of causes, there will frequently be non-tortious causes contributing to the injury. Defendants could frequently and easily identify non-tortious contributing causes, so Plaintiffs would rarely receive full compensation even after proving that the Defendant caused the injury.15

This consideration of multiple necessary and contributing causes, specifically contributing non-tortious causes, is particularly prevalent in medical malpractice lawsuits arising out of the delay of diagnosis and treatment. In these cases, where there is an alleged delay in diagnosis, a Defendant’s conduct will never be the sole factual cause of the injury. Such cases are generally premised on the fact that the Plaintiff had an evolving ailment in the first place. A finding of material contribution to the injury by a Defendant’s negligence leading to delay does not excuse the Plaintiff from proving, on a balance of probabilities, that but for the Defendant’s negligence, the Plaintiff would not have suffered a loss: It simply recognizes that the Defendant’s negligence, albeit a necessary cause, cannot be anything more than a contributing cause.16

In these delay cases, the standard of care inquiry is not on what the Defendant did, but something they failed to do in; meanwhile, factually, the Plaintiff’s injury is caused by the disease process that has been left to evolve. The trier of fact is required to attend to the fact situation, as it existed in reality in the moment, and imagine that the Defendant took the action required by the standard of care. The trier of fact then must make a determination of whether the delay in appropriate treatment resulting from the Defendant’s omission(s) materially contributed to the Plaintiff’s injury, given that the underlying disease process prevents the omission(s) from being the sole cause of injury.17

It is important to clarify the purpose of the word “material” in describing the threshold which the contribution to the injury the Defendant’s negligence must rise to. A common pitfall and simply erroneous assertion often made is that the use of the phrase “material contribution to the Plaintiff’s injury” is an application of the “material contribution to risk” test. This is simply not the case. As expressly stated in Clements, a Defendant’s negligence must be necessary to bring about the injury. Justice Lauwers of the Court of Appeal grappled with this requirement, and stated that the word “necessary” was used in a purposive manner in order to emphasize the legal requirement for the Plaintiff to prove a real and substantial connection between the Defendant’s negligence and the Plaintiff’s injury.18 There is no magic to the word “necessary.” In fact, Chief Justice McLachlin in Clements, even used an alternate phrasing—“in other words the injury would not have occurred without the Defendant’s negligence.”19 With that in mind, the use of the phrase “caused or materially contributed” as opposed to simply “caused or contributed,” simply ensures that the causal connection between the Defendant’s negligence and the Plaintiff’s injury is beyond de minimis.20

There is a clear distinction between a material contribution to injury or harm (a robust and pragmatic application of the “but for” test) and the “material contribution to risk” test. The former is the correct analytical analysis to be applied in establishing causation, while the latter is a policy driven rule to be used sparingly if it is impossible to apply the “but for” test in rare multiple tortfeasor cases.21

Prevented or Reduced

Contribution language in the formulation of the causation inquiry is necessary for a related, but independent reason. In a negligence action, the burden placed on the Plaintiff in proving causation through the “but for” test is not to prove that they would have suffered no injury at all. The onus is to prove that, on a balance of probabilities, the specific injury would not have occurred if the Defendant had met the standard of care. Put succinctly, this requires the Plaintiff to prove that their unfavourable outcome would have been prevented or reduced.22

Proponents of an application of a strict “but for” test have scrutinized “prevented or reduced” language as lowering the onus the Plaintiff must meet to prove their losses on causation. However, this scrutiny is often accompanied with confusion on the distinction between the causation and damage inquiry. Chief Justice McLachlin provided helpful guidance in outlining this distinction in Blackwater v. Plint:

It is important to distinguish between causation as the source of the loss and the rules of damage assessment in tort. The rules of causation consider generally whether “but for” the Defendant’s acts, the Plaintiff’s damages would have been incurred on a balance of probabilities. Even though there may be several tortious and non-tortious causes of injury, so long as the Defendant’s act is a cause of the Plaintiff’s damage, the Defendant is fully liable for that damage. The rules of damages then consider what the original position of the Plaintiff would have been. The governing principle is that the Defendant need not put the Plaintiff in a better position than his original position and should not compensate the Plaintiff for any damages he would have suffered anyways.

Therefore, to the extent that the evidence shows that the effect of the sexual assaults would have been greater because of his pre-existing injury, that pre-existing condition can be taken into account in assessing damages.23

The thin skull rule simply recognizes that there was some pre-existing condition inherent in the plaintiff’s “original position.” If there is a measurable risk that the pre-existing condition would have detrimentally affected the plaintiff in the future or would have resulted in some of the same injury caused by the defendant, regardless of the negligence, then this can be taken into account in reducing the overall damages award. This does not prevent a Plaintiff from establishing their losses on causation.24

Blackwater was an action for damages for psychological suffering by a survivor of one of the residential schools that operated in British Columbia. The evidence demonstrated that the Plaintiff had pre-existing psychological distress and injuries due to trauma suffered throughout childhood, prior to attending the residential school. The abuse, specifically of the sexual nature, suffered at the residential school exacerbated the pre-existing psychological condition.25 With these factual circumstances in mind, one can practically understand why the Plaintiff must succeed by showing their injury would have been reduced, but not necessarily entirely prevented. To conclude otherwise would prevent victims of negligence from any entitlement for compensation where their pre-existing injuries are worsened by a Defendant’s wrongdoing.

In circumstances such as Blackwater, or in delay in diagnosis cases where some, but not all, of the Plaintiff’s injury would have occurred irrespective of the negligence of the Defendant, the negligence can be no more than a contributing cause. A strict application of the “but for” test, fails to account for these situations, because it is impossible for the Plaintiff to prove that “but for” the Defendant’s breach of the standard of care, their injury would not have occurred. It is, however, possible to show that their injury would have been prevented or reduced; a burden that is inherently captured by the phrase “caused or contributed”.

Circular Causation

In multiple tortfeasor cases, there may be circumstances where the Plaintiff’s injury was caused by the combination of multiple Defendants’ negligence, but, the conduct of a single Defendant alone cannot be proven to have brought about the injury. By way of example, the Plaintiff’s car is parked on the edge of a scenic overlook of a mountain. Three Defendants lean against the Plaintiff’s car and it is pushed over the edge. The combined force of two of three Defendants was required to push the car over the edge, but the force of one Defendant alone would have been insufficient. On a strict application of “but for”, the Plaintiff would be unable to hold any of the individual tortfeasors responsible. That is because “but for” the force of one individual Defendant, the force of the other two Defendants would have been sufficient to push the car down the cliff. In other words, “but for” the negligence of any one of the individual Defendant, the injury still would have occurred. This “problem” where the tortfeasors could sidestep liability on the basis that there was sufficient cumulative force by the remaining Defendants is known as circular causation.26

The pitfall of circular causation is a “striking outcome that would flow if the law refused to recognize as ‘causal’ a positive contribution.”27 These very specific but not uncommon circumstances again, are instances where a Defendant’s negligence can be no more than a contributing cause of the Plaintiff’s injury.

Conclusion

To reiterate, the “but for” test, has not been, nor ever will be anything more than simply the name of the test of causation. Recently, in 2019, the Ontario Court of Appeal released a well-written decision, which has provided further clarity to the law of causation. It outlined many of the concepts and misconceptions highlighted in this article, including the required use of contribution language in formulating the causation question, deciphering between causation and damages and the distinction between “material contribution to injury” and “material contribution to risk” of injury. In doing so, the following summary was provided on the proper causal inquiry:

Causation is made out under the "but for" test if the negligence of a defendant caused the whole of the plaintiff's injury, or contributed, in some not insubstantial or immaterial way, to the injury that the plaintiff sustained. Causation requires a "substantial connection between the injury and the defendant's conduct".

The conclusion that one or more defendants "materially contributed" to a plaintiff's injury or loss simply recognizes that the defendant's negligence was not the only cause.28

Unfortunately, however, misconceptions of causation law and the use of contribution language continue to prevail. Instead of recognizing that the “cause or contribute” phrasing is a simple way to clearly set out the “but for” test while ensuring the requirements of proof under the test are still met, arguments continue to be made in an effort to dismantle such phrasing.29 These include erroneous assertions that “cause or contribute” are two distinct concepts and provide two different pathways to establishing causation.30

What is the solution? Outside of simple cases involving a single Defendant who did something negligent and caused injury to the Plaintiff with no pre-existing injury, such as a hunting mishap or a rear-end motor vehicle accident, the causation analysis is typically complex, and involves multiple contributing factors. Moreover, in those simple cases, causation is not a live issue.31 In complex causation cases, where there are multiple tortfeasors, multiple tortious acts, non-tortious contributing causes, or the Plaintiff can only show their injury could have been reduced, but not prevented, a single Defendant’s negligence can only be a contributing cause of the Plaintiff’s injury—not the cause.

If we accept the above as true, and we also accept as true that in the simple single Defendant negligence cases, causation is not a live issue, then why not simplify the causation test where it is a live issue. Why not remove “caused” from the phrase “caused and contribute” and accept that the “but for” test of causation is made out where the Defendant’s negligence materially contributes to the Plaintiff’s injury? It is correct in law. As this article has emphasized, the Plaintiff discharges its burden in proving causation by showing that there was a material contribution to their injury by the Defendant’s breach in the standard of care. What value then, does the use of “caused” serve? Even in simple causation cases where it is not a live issue, if the Defendant’s negligence is found to be the sole cause of the Plaintiff’s injury, then it follows that it would also meet the onus of “material contribution.”

More notably, this “solution”, is not a new or novel idea. It simply proposes to rely on the simple yet sound causal analysis that the Supreme Court of Canada established over two decades ago. Returning to Athey, “causation is established where the defendant's negligence materially contributed to the occurrence of the injury.”32 This simple phrasing, which omits the phrase “caused” or “cause”, as conceded in Clements, is an example of a robust and common sense application of the “but for” test.


  1. Snell v Farrell, [1990] 2 SCR 311 at paras 2-7, [1990] SCJ No 73.
  2. Ibid, at paras 22, 29, 33 and 38.
  3. Athey v Leonati, [1996] 3 SCR 458 at paras 2-7 [1996] SCJ No 102.
  4. Ibid, at para 15.
  5. Ibid, at para 16.
  6. Clements (Litigation Guardian of) v Clements, 2012 SCC 32 at para 1.
  7. Clements (Litigation Guardian of) v Clements, 2009 BCSC 112 at paras 6-45 and 50-65.
  8. Clements (Litigation Guardian of) v Clements, 2010 BCCA 581 at para 64.
  9. Supra note 7, at para 46.
  10. Supra note 7, at para 8.
  11. Ibid, at para 9.
  12. Ibid, at para 28.
  13. Negligence Act, RSO 1990, c N1, s 1.
  14. Donleavy v Ultramar Ltd., 2019 ONCA 687 at para 72.
  15. Supra note 4 at para 20.
  16. Cottrelle v Gerrard, [2003] OJ No 4194 at para 20, 25, and 27, 67 OR (3d) 737 (Ont Sup Ct J); Beldycki Estate v Jaipargas, 2012 ONCA 537 at para 44; Vuong v Morton, [2009] OJ No 4661 at paras 153-162 and 164; Sacks v Ross, 2017 ONCA 773 at paras 122; and 130-132; White v St. Joseph’s Hospital (Hamilton), 2019 ONCA 312 at para 25.
  17. Sacks v Ross, 2017 ONCA 773 at paras 46 and 117.
  18. Supra note 7, at para 8.
  19. Supra note 18, at paras 118-122; Supra note 7, at para 8.
  20. Supra note 4, at para 15.
  21. Supra note 15, at para 65; MacDonald v Goertz, 2009 BCCA 358 at para 17.
  22. Supra note 18, at para 46; Cottrelle v Gerrard, [2003] OJ No 4194 at para 25, 67 OR (3d) 737 (Ont Sup Ct J).
  23. Blackwater v Plint, [2005] 3 SCR 3 at paras 78-79, [2005] SCJ No 59.
  24. Ibid; Supra note 4 at paras 34-36; Graham v Rourke, [1990] OJ No 2314 at paras 40-48 and 61, 75 OR (2d) 622 (Ont CA).
  25. Ibid, at paras 1-6 and 74-89.
  26. Supra note 18, at paras 113-117.
  27. Ibid, at para 115.
  28. Supra note 15.
  29. Cheeseman v Credit Valley Hospital, 2019 ONSC 4996 at para 69.
  30. Uribe v Tsandelis, 2019 ONSC 6242 at para 12.
  31. Supra note 18, at paras 45-46.
  32. Supra note 4, at para 15.