'But for' and Complications. Clements v Clements
In Clements v Clements (SCC, 2012), after an extensive (and worth reading) review of the state of tort causation law in the major common law jurisdictions, the court stated as follows on the present state of this law in Canada:
 The foregoing discussion leads me to the following conclusions as to the present state of the law in Canada:. Fowlow v. Southlake Regional Health Centre
(1) As a general rule, a plaintiff cannot succeed unless she shows as a matter of fact that she would not have suffered the loss “but for” the negligent act or acts of the defendant. A trial judge is to take a robust and pragmatic approach to determining if a plaintiff has established that the defendant’s negligence caused her loss. Scientific proof of causation is not required.
(2) Exceptionally, a plaintiff may succeed by showing that the defendant’s conduct materially contributed to risk of the plaintiff’s injury, where (a) the plaintiff has established that her loss would not have occurred “but for” the negligence of two or more tortfeasors, each possibly in fact responsible for the loss; and (b) the plaintiff, through no fault of her own, is unable to show that any one of the possible tortfeasors in fact was the necessary or “but for” cause of her injury, because each can point to one another as the possible “but for” cause of the injury, defeating a finding of causation on a balance of probabilities against anyone.
In Fowlow v. Southlake Regional Health Centre (Ont CA, 2014) the Court of Appeal engaged in this consideration of the treatment of causation in a medical malpractice case:
 On appeal, the appellants submit the trial judge erred by unduly focusing on the specific mechanism that led to the detachment. In doing so, they say he required them to establish causation with scientific precision. They submit he failed to apply the robust and pragmatic approach to causation set out in Snell v. Farrell, 1990 CanLII 70 (SCC),  2 S.C.R. 311, and that he failed to apply the “modified objective test” for causation in Reibl v. Hughes, 1980 CanLII 23 (SCC),  2 S.C.R. 880 in relation to the issue of informed consent. The appellants submit it is impossible for them to prove the specific mechanism of the detachment, and that these cases indicate that it is sufficient that they established that the respondent’s use of a thin wall graft materially contributed to the risk of detachment and that there was a relationship between the increase in risk and the death of the patient. They also submit that the trial judge failed to consider whether a reasonable person in the patient’s position would have declined treatment, had he been informed of the use of a “not recommended” graft.. Resurfice Corp. v. Hanke
 We do not agree.
 Second, this is not a case for the application of the “material contribution” approach. As the Supreme Court of Canada recently pointed out in Clements v. Clements, 2012 SCC 32,  2 S.C.R. 181, the “material contribution” approach may be employed where it is truly impossible to establish “but for” causation, and is particularly apt for cases involving multiple tortfeasors. Not only was there only a single tortfeasor in this case, but the appellants had not established it was truly impossible for them to satisfy the “but for” test. We agree that without evidence from the manufacturer or from a pathologist who performed or witnessed the post-mortem examination, or at least evidence of why these witnesses would not be helpful, the “impossibility” criterion was not satisfied.
 Third, the robust and pragmatic approach does not dispense with the requirement that plaintiffs lead some evidence of causal connection, whether the claim is based on negligence or on breach of the duty of disclosure. Moreover, the approach is most robustly applied in cases in which the facts lie particularly within the knowledge of the defendant. Counsel for the appellant recognizes that the appellants in this case were equally able to tender evidence from the manufacturer or from a pathologist who performed or witnessed the post-mortem examination.
 We do not doubt that the trial judge engaged in a robust and pragmatic analysis of the evidence. After his careful review, he declined to infer that the respondent’s use of a graft contrary to the manufacturer’s recommendation and failure to disclose were sufficient, in itself, to establish a causal link to the patient’s death. We are not persuaded that he committed a reversible error.
In Resurfice Corp. v. Hanke (SCC, 2007) the Supreme Court of Canada commented on the law of causation in tort that the general rule is the 'but for' standard (ie. 'but for' the actions of the defendant there would have been no damage caused), but that an allowable exception in some circumstances (explained below) is the 'material contribution' standard:
21 First, the basic test for determining causation remains the “but for” test. This applies to multi-cause injuries. The plaintiff bears the burden of showing that “but for” the negligent act or omission of each defendant, the injury would not have occurred. Having done this, contributory negligence may be apportioned, as permitted by statute. . Donleavy v. Ultramar Ltd.
22 This fundamental rule has never been displaced and remains the primary test for causation in negligence actions. As stated in Athey v. Leonati, at para. 14, per Major J., “[t]he general, but not conclusive, test for causation is the ‘but for’ test, which requires the plaintiff to show that the injury would not have occurred but for the negligence of the defendant”. Similarly, as I noted in Blackwater v. Plint, at para. 78, “[t]he 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.”
23 The “but for” test recognizes that compensation for negligent conduct should only be made “where a substantial connection between the injury and the defendant’s conduct” is present. It ensures that a defendant will not be held liable for the plaintiff’s injuries where they “may very well be due to factors unconnected to the defendant and not the fault of anyone”: Snell v. Farrell, at p. 327, per Sopinka J.
24 However, in special circumstances, the law has recognized exceptions to the basic “but for” test, and applied a “material contribution” test. Broadly speaking, the cases in which the “material contribution” test is properly applied involve two requirements.
25 First, it must be impossible for the plaintiff to prove that the defendant’s negligence caused the plaintiff’s injury using the “but for” test. The impossibility must be due to factors that are outside of the plaintiff’s control; for example, current limits of scientific knowledge. Second, it must be clear that the defendant breached a duty of care owed to the plaintiff, thereby exposing the plaintiff to an unreasonable risk of injury, and the plaintiff must have suffered that form of injury. In other words, the plaintiff’s injury must fall within the ambit of the risk created by the defendant’s breach. In those exceptional cases where these two requirements are satisfied, liability may be imposed, even though the “but for” test is not satisfied, because it would offend basic notions of fairness and justice to deny liability by applying a “but for” approach.
26 These two requirements are helpful in defining the situations in which an exception to the “but for” approach ought to be permitted. Without dealing exhaustively with the jurisprudence, a few examples may assist in demonstrating the twin principles just asserted.
27 One situation requiring an exception to the “but for” test is the situation where it is impossible to say which of two tortious sources caused the injury, as where two shots are carelessly fired at the victim, but it is impossible to say which shot injured him: Cook v. Lewis, 1951 CanLII 26 (SCC),  S.C.R. 830. Provided that it is established that each of the defendants carelessly or negligently created an unreasonable risk of that type of injury that the plaintiff in fact suffered (i.e. carelessly or negligently fired a shot that could have caused the injury), a material contribution test may be appropriately applied.
28 A second situation requiring an exception to the “but for” test may be where it is impossible to prove what a particular person in the causal chain would have done had the defendant not committed a negligent act or omission, thus breaking the “but for” chain of causation. For example, although there was no need to rely on the “material contribution” test in Walker Estate v. York Finch General Hospital, this Court indicated that it could be used where it was impossible to prove that the donor whose tainted blood infected the plaintiff would not have given blood if the defendant had properly warned him against donating blood. Once again, the impossibility of establishing causation and the element of injury-related risk created by the defendant are central.
In Donleavy v. Ultramar Ltd. (Ont CA, 2019) the Court of Appeal considers that the standard causation test is the 'but for' test, and when other causation measures might apply:
(i) “But for” is the correct test for causation in this case. Sack v Ross
 The “but for” test is generally applied in establishing causation in the tort of negligence. It requires a plaintiff to prove, on a balance of probabilities, that without the negligence of one or more of the defendants, the injury would not have occurred. A defendant’s negligence is thus a necessary factor to bring about the injury: Clements, at para. 8.
 The “but for” test applies even where a defendant’s negligence is not the sole cause of the plaintiff’s injury. A defendant will be liable for all injuries caused or contributed to by his or her negligence, even if other non-tortious causes are present: Athey v. Leonati, 1996 CanLII 183 (SCC),  3 S.C.R. 458, at paras. 12, 17. Indeed, there are usually a number of background factors that cause an injury and not only a single cause. A defendant only needs to be “a” cause of “some” harm to be found liable in tort: Erik S. Knutsen, “Clarifying Causation in Tort” (2010) 33:1 Dal. L.J. 153 at p. 159.
 The alternative, and exceptional, basis on which legal causation can be established is where the defendant’s act or omission “materially contributed” to the plaintiff’s risk of injury. As McLachlin C.J.C. pointed out in Clements, at para. 46, “material contribution” eliminates the need to prove factual causation, and it is the appropriate test only in exceptional cases:
Exceptionally, a plaintiff may succeed by showing that the defendant’s conduct materially contributed to risk of the plaintiff's injury, where (a) the plaintiff has established that her loss would not have occurred “but for” the negligence of two or more tortfeasors, each possibly in fact responsible for the loss; and (b) the plaintiff, through no fault of her own, is unable to show that any one of the possible tortfeasors in fact was the necessary or “but for” cause of her injury, because each can point to one another as the possible “but for” cause of the injury, defeating a finding of causation on a balance of probabilities against anyone. I emphasize here that the alternative basis for legal causation is the defendant’s material contribution to risk of injury (and not, as I point out below, material contribution to the plaintiff’s loss or damage). Material contribution is not a test of causation but a policy-oriented rule that imposes liability on the basis of tortious risk creation. The departure from the general “but for” rule is justified in these specific cases, because it satisfies the underlying goals of negligence law: Clements, at para. 41; Lewis N. Klar & Cameron S.G. Jefferies, Tort Law, 6th ed. (Toronto: Thomson Reuters, 2017), at p. 545.
 In this case the trial judge began her discussion of causation at paras. 153 and 154, where she correctly observed that, in the usual case, causation is determined on a “but for” standard, and that in exceptional cases, a defendant might be held liable on the basis of “material contribution” to the risk of injury or loss. The trial judge stated:
To succeed in an action in negligence, a plaintiff must demonstrate on a balance of probabilities that the defendant’s breach of the standard of care is causally connected to the injuries or losses alleged. In all but exceptional cases, the test for causation is the “but for” test: a plaintiff has the onus of establishing that their injuries or losses would not have occurred without the defendant’s negligence (see Clements v. Clements, 2012 SCC 32 (CanLII), at para. 8). The trial judge then went on to conclude that this case fell within the meaning of “exceptional cases” and she stated that, in determining causation, she would rely on the material contribution test. Her explanation for resorting to the “material contribution” test was that there were three named defendants, and more than three actors whose conduct was considered, including the plaintiffs: at para. 155.
Exceptional cases include where there are negligent acts by multiple actors and it is established that one or more of them in fact caused the injuries or losses, but the plaintiff is unable to determine which of a number of negligent acts caused the injuries or losses. In such cases, a defendant may be held liable on the basis that his or her acts and omissions “materially contributed” to the risk of injury or loss (Clements, at para. 13).
 The fact that there are multiple defendants, or more than one potential cause of an injury, is not a reason to depart from the “but for” test for causation: Resurfice Corp. v. Hanke, 2007 SCC 7 (CanLII),  1 S.C.R. 333, at para. 19. McLachlin C.J.C. emphasized in Clements that “but for” causation continues to apply as the default test in cases where there are multiple tortfeasors, and that something more is required before the material contribution to risk approach is warranted. She stated, at para. 43:
It is important to reaffirm that in the usual case of multiple agents or actors, the traditional “but for” test still applies. The question… is whether the plaintiff has shown that the negligence of one or more of the defendants was a necessary cause of the injury. Degrees of fault are reflected in calculations made under contributory negligence legislation. By contrast, the material contribution to risk approach applies where “but for” causation cannot be proven against any of multiple defendants, all negligent in a manner that might have in fact caused the plaintiff’s injury, because each can use a “point the finger” strategy to preclude a finding of causation on a balance of probabilities. The critical threshold, then, for the application of the material contribution to risk approach is the impossibility of proving which of two or more possible tortious causes is in fact a cause of the injury. Before resorting to material contribution, it is necessary to establish that fault or negligence caused the event and that there is more than one negligent actor, but that it is impossible for the plaintiff to tease out whose negligence was a “but for” cause.
 Professor Knutsen, in “Coping with Complex Causation Information in Personal Injury Cases” (2013), 41: 2 and 3 Adv. Q. 149, notes that the fact that there are multiple tortfeasors and an indivisible injury does not mandate the use of the “material contribution” test. Rather, he states, at pp. 153-154:
Nothing in the Clements case’s approach to the material contribution test alters traditional damages and contribution doctrinal analyses whereby multiple tortfeasor defendants may be partially responsible for some of a plaintiff’s injury, but not all of it (i.e., divisible and indivisible injuries) after they have been proven to be a cause of some of the plaintiff’s harm. The Clements analysis concentrates entirely on the causation threshold question – did the defendants cause the plaintiff’s injury, yes or no. It does not target the “how much” question. That is a damages question, not causation, and left for later on in the negligence analysis (along with joint and several liability issues about responsibility to pay for the harm). Indeed, this point was made by McLachlin C.J.C. in Blackwater v. Plint, 2005 SCC 58 (CanLII),  3 S.C.R. 3, a case involving civil liability for sexual assault. The Chief Justice stated, at para. 78:
In nearly all instances [of an indivisible injury involving multiple tortfeasors], it will be possible for the plaintiff to apply the “but for” test in proving that the target tortfeasors’ negligent conduct was a cause of some of the global total harm. If the question is about “how much” of that harm each target tortfeasor is responsible for, that is not a causation question but one of extent and damages, dealt with later in the negligence analysis. The fact that the injury is indivisible does not affect the workability of the causation test.
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. [Emphasis added.] Even in the “but for” context, there are situations involving multiple defendants where courts have used “contribution” language to recognize that a defendant’s negligence is “a” cause of the plaintiff’s injury. This is a potential source of confusion. The phrase “caused or contributed” is used in the Negligence Act, R.S.O. 1990, c. N.1, s. 1. As Lauwers J.A. observed in Sacks v. Ross, 2017 ONCA 773 (CanLII), 417 D.L.R. (4th) 387, leave to appeal refused  S.C.C.A. No. 491, at para. 117, this is the way the causation test has been described in several cases in the Supreme Court (including as I noted earlier in Athey), and in this court; see also White v. St. Joseph’s Hospital (Hamilton), 2019 ONCA 312 (CanLII), at para. 25. 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”: Resurfice, at para. 23, Clements, at paras. 20-21, 28. As Paciocco J.A. noted in Smith v. Safranyos, 2018 ONCA 760 (CanLII), at para. 128, McLachlin C.J.C.’s reference to “a robust and common sense application of the ‘but for’ test of causation” is a “manifestation of this limit”. Used in this way, 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.
 Accordingly, even in a case where there are multiple tortfeasors, the “but for” test presumptively will apply. In the case of a negligent omission, as in the present case, the application of the “but for” test requires the trier of fact “to attend to the fact situation as it existed in reality the moment before the defendant’s breach of the standard of care, and then to imagine that the defendant took the action the standard of care obliged [it] to take, in order to determine whether [its] doing so would have prevented or reduced the injury”: Sacks, at para. 46. The “but for” causation question in respect of each appellant, is whether, if that party had discharged its duty to the respondents, and had not been negligence, the oil spill would have been prevented.
In Sack v Ross (Ont CA, 2017), a complex medical malpractice case, which bears reading in full, the Court of Appeal engages in an extensive consideration of principles applicable to causation in negligence, and the sometimes difficulty of applying the standard 'but for' causation test when dealing with some situations. These situations are: an omission rather than an action, several alleged tortfeasors, and delayed diagnosis and treatment:
E. Issue One: Did the trial proceed on a correct understanding of causation in negligence cases?. White v. St. Joseph's Hospital (Hamilton)
 The common law feels its way on a case by case basis recognizing that judges are not omniscient. A carefully formulated legal proposition that works justice in a specific case might later be found, in the light of different facts, to be over-inclusive or under-inclusive, or as otherwise flawed or inadequate. The paradoxical genius of the common law is that, on principle, the process of refinement and adaptation is interminable.
 The doctrine of causation in negligence is a signal example. As McLachlin C.J. observed in Resurfice Corp. v. Hanke, 2007 SCC 7 (CanLII), at para. 20: “Much judicial and academic ink has been spilled over the proper test for causation in cases of negligence.” The Supreme Court’s decision in Clements v. Clements, 2012 SCC 32 (CanLII),  2 S.C.R. 181 ushered in another wave of ink.
 In picking through the law relating to causation in negligence actions, there is some conceptual clarity to be gained by distinguishing between simple negligence cases involving a plaintiff and a single defendant, and complex cases involving multiple tortfeasors.
(1) Simple Negligence Actions
 The paradigm negligence case is one in which a single defendant is alleged to have negligently done something that caused the plaintiff’s injury. To succeed the plaintiff must prove both the defendant’s breach of the standard of care and causation.
 The current state of the law regarding causation in negligence cases was set out by McLachlin C.J. in Clements at para 8:
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, her action against the defendant fails. [Emphasis by McLachlin C.J.] The determination of causation is understood as a factual inquiry. In all instances of determining causation, Professor Russell Brown (as he then was) observed that the trier of fact proceeds by inference and not as a direct witness of the events: “The Possibility of “Inference Causation”: Inferring Cause-in-Fact and The Nature of Legal Fact-Finding” (2010) 55 McGill L.J. 1; “Known Unknowns in Cause-in-Fact” (2010-2011), 39. Adv. Q. 37; “Cause-in-Fact at the Supreme Court of Canada: Developments in Tort Law in 2012-2013”, (2014), 64 S.C.L.R. (2d) 327.
 Very few fact situations demonstrate the causal clarity of a game of billiards in which the combinations of balls striking balls can be easily replicated. The closest equivalent would be a simple case involving a single defendant who did something in breach of the standard of care that physically hurt the defendant, where the “but for” test is relatively easy to apply. Consider the cases in which the defendant accidently shoots the plaintiff in a hunting mishap, or where the defendant runs a red light and collides with the plaintiff who suffers physical trauma. Causation is not usually a live issue in such cases because the causal inferences are so easy to draw.
 Things are more complicated where the complaint is not about something the defendant did, but about something the defendant failed to do in breach of the standard of care. When what is in issue is not the defendant’s act, but an omission, the trier of fact is required to attend to the fact situation as it existed in reality the moment before the defendant’s breach of the standard of care, and then to imagine that the defendant took the action the standard of care obliged her to take, in order to determine whether her doing so would have prevented or reduced the injury. Even though this exercise is bounded significantly by the actual facts, it counts as “factual” because the task is to consider how the events would actually have unfolded had the defendant taken the action she was obliged to take.
(2) The Causal Reasoning Process
 Regardless of whether the defendant’s breach of the standard of care is an act or an omission, the trier of fact’s cognitive process in determining causation has three basic steps. The first is to determine what likely happened in actuality. The second is to consider what would likely have happened had the defendant not breached the standard of care. The third step is to allocate fault among the negligent defendants.
 There are two possible outcomes to the trier of fact’s imaginative reconstruction of reality at the second step. On the one hand, if the trier of fact draws the inference from the evidence that the plaintiff would likely have been injured in any event, regardless of what the defendant did or failed to do in breach of the standard of care, then the defendant did not cause the injury. On the other hand, if the trier of fact infers from the evidence that the plaintiff would not likely have been injured without the defendant’s act or failure to act, then the “but for” test for causation is satisfied: but for the defendant’s act or omission, the plaintiff would not have been injured. The defendant’s fault, which justifies liability, has been established.
 I move now to consider some more factually complicated situations.
(3) Negligence Actions Involving Multiple Defendants
 Many factual combinations and permutations are possible in negligence cases, including the complexities arising from multiple negligent parties, multiple potential causes, successive and cumulative injuries, divisible and indivisible injuries, injury aggravation, and the thin and crumbling skull scenarios: see Erik Knutsen, “Clarifying Causation in Tort”, (2010) 33 Dal. L.J. 153, and “Coping with Complex Causation Information in Personal Injury Cases”, (2013) 41 Adv. Q. 149.
 Delayed diagnosis medical negligence cases involving multiple tortfeasors typically raise several such complexities, and are among the most complex to assess from the perspective of causation.
 Chief Justice McLachlin took the view in Clements that the basic “but for” test also applies in these more complex cases, noting at paras. 12 and 43:
In some cases, an injury -- the loss for which the plaintiff claims compensation -- may flow from a number of different negligent acts committed by different actors, each of which is a necessary or "but for" cause of the injury. In such cases, the defendants are said to be jointly and severally liable. The judge or jury then apportions liability according to the degree of fault of each defendant pursuant to contributory negligence legislation. In his 2010 article, at p. 170, Professor Knutsen made the following observations, basically agreeing with McLachlin C.J.:
It is important to reaffirm that in the usual case of multiple agents or actors, the traditional "but for" test still applies. The question, as discussed earlier, is whether the plaintiff has shown that one or more of the defendants' negligence was a necessary cause of the injury. Degrees of fault are reflected in calculations made under contributory negligence legislation.
“But for” causation is therefore not complicated in successive or cumulative injury context when one focusses solely on each actor and each actor’s role in a causal story. The confusion results when one forgets two things: first, that it is a defendant’s breach of the standard of care that is the locus of the causal inquiry, and second, that a defendant’s negligence need only be proven to be “a” cause of “some” injury to the plaintiff. However, it is worth observing that the Supreme Court has never considered cases beyond the simple. Even in Athey v. Leonati, 1996 CanLII 183 (SCC),  3 S.C.R. 458, which involved successive accidents with different defendants, the plaintiff’s injuries were treated as a single tort by agreement of the parties, which greatly simplified the analysis. I do not consider Cook v. Lewis 1951 CanLII 26 (SCC),  S.C.R. 830, which was a very strange case about a hunting accident, to be an example applicable to conventional delay of diagnosis and treatment cases.
 As I explain below, this case shows that the trier of fact’s cognitive process in reasoning through causation in a delayed diagnosis and treatment case does not unfold as simply as these statements of the doctrine might suggest.
In White v. St. Joseph's Hospital (Hamilton) (Ont CA, 2019) the Court of Appeal considers medical delay in the context of the but-for test for causation:
 In an action for delayed medical diagnosis and treatment, a plaintiff must establish that the delay caused or contributed to the unfavourable outcome: Sacks v. Ross, 2017 ONCA 773 (CanLII), 417 D.L.R. (4th) 387, leave to appeal to SCC refused, 2018 CarswellOnt 10678-10679, at para. 117; Beldycki Estate v. Jaipargas, 2012 ONCA 537 (CanLII), 295 O.A.C. 100, at para. 44. The phrase “caused or contributed” originates in the Negligence Act, R.S.O. 1990, c. N.1, at s. 1, and is the normative test applied by this court, as set out in Sacks v. Ross, at para. 117, and embodied in the “but for” test prescribed by the Supreme Court in Clements v. Clements, 2012 SCC 32 (CanLII),  2 S.C.R. 181, at para. 8. In other words, “but for” the alleged delay would the plaintiff have suffered the unfavourable outcome? (Nothing in Sacks v Ross revived the “material contribution to injury” test.) The trial judge found that Mr. White failed to prove that the delay in treatment caused or contributed to his injuries. She noted: “Dr. Fong was not able to articulate what injury he says was avoidable.” The trial judge stated: "Dr. Fong agreed that most of Mr. White's outcome was unavoidable. The court finds this evidence is devastating to the plaintiffs' case."