Wallace v Kam

The trial judge's reasons 44The trial judge identified the following issues for determination: (1) whether the respondent breached his duty of care to the appellant by failing to give a proper warning of the risks of the surgery he performed (the breach issue); (2) whether the appellant suffered harm in the circumstances (the injury issue); and (3) if harm was suffered, whether that harm was caused by the respondent's negligence (the causation issue). 45His Honour found that the appellant sustained bilateral femoral neurapraxia following the surgery on 22 November 2004 and that the further surgical procedure carried out on 23 November 2004 was a consequence of the surgery on 22 November 2004. There is no issue about these findings on the appeal and thus the injury issue can be put to one side. 46As to breach, the respondent conceded that he had not warned the appellant of the risk of bilateral femoral neurapraxia and that he should have done so. The trial judge found that this risk was an inherent risk of the surgery and was significant in this case because of the appellant's weight. In this regard, the appellant's weight increased both the likelihood of the occurrence of the condition as well as the extent and degree to which it might affect him if it occurred. His Honour held, at [49], that the risk of bilateral femoral neurapraxia was material. 47Flowing from the respondent's concession as to the need for and his failure to give a warning of bilateral femoral neurapraxia, and his Honour's finding that the risk was material, his Honour held, at [51], that the respondent breached his duty of care to the appellant in failing to warn him of that risk and/or its associated symptoms or consequences. Importantly for the issue on the appeal, his Honour held, at [96], that the failure to warn of bilateral femoral neurapraxia was the only potentially relevant failure to warn. 48His Honour concluded that if the respondent had warned of this risk, the appellant would nevertheless have undergone the surgery. Hence, the appellant had not satisfied the first limb of the test of causation in the Civil Liability Act 2002, s 5D(1)(a). 49His Honour made no finding as to whether the respondent had failed to warn the appellant of the 5 per cent risk of paralysis or whether such risk was material. His Honour found that the appellant's allegation that the respondent had failed to warn him of possible catastrophic consequences associated with known risks of the surgery was not relevant, because that risk had not come home. His Honour cited Rosenberg v Percival [2001] HCA 18; 205 CLR 434 in support of this conclusion. 50Apart from the conceded failure to warn of the risk of bilateral femoral neurapraxia, there was a dispute between the parties regarding what the respondent told the appellant as to the risks of the surgical procedure. 51The appellant's evidence in chief was that the respondent told him there was a 75 per cent chance of the surgery being successful and did not mention anything that could go wrong, leaving him with the impression that if the surgery was unsuccessful he would simply be the way he was before the surgery. In cross-examination, the appellant said the advice was of a 70-75 per cent chance of the operation being successful. He said he believed there were no negative aspects to the surgery. He reiterated his understanding from what the respondent told him, that the worst that could happen was that he would simply be the way he was before the operation. The appellant also said that the respondent told him that if the operation was successful he would be in no pain at all and would be able to stop taking medication which, prior to surgery, had quadrupled in quantity and strength from when it was first prescribed. 52The appellant agreed in cross-examination that the respondent specifically told him that any surgery had risks and complications. He also said he understood spinal surgery was "a serious business". He did not recall that the respondent specifically told him there was a risk of injury to his nerve roots and denied that he was told he could be left with weakness in either or both legs. He also denied he was told that he could have spinal cord injury, which could affect either or both of his legs, or that he could develop bladder and bowel incontinence as a result of the surgery. These denials were emphatic. In a later part of the cross-examination, the appellant said he could not remember the bladder and bowel being discussed. All he remembered was that the respondent said if he did not have surgery he would be incontinent. He also denied he was told the surgery could take longer because of his weight. 53The appellant's wife, who had accompanied the appellant on his consultations with the respondent, also gave evidence. She agreed in cross-examination that the respondent told the appellant that even though he had not lost enough weight, consideration needed to be given to proceeding with the surgery. She agreed the respondent said that any surgery was accompanied by risks and complications and there were no guarantees. She also agreed in cross-examination that the respondent told the appellant there was a risk of nerve root injury. However, she could not remember the respondent saying the appellant could be left with weakness in either or both of his legs after surgery. Nor did she remember if anything was said as to the risk of injury to the spinal cord. 54The respondent's evidence was that from the time of the first consultation he was concerned with the appellant's weight. His initial recommendation was for the appellant to attempt a significant weight loss program, with surgery as a last resort. The respondent said his advice to the appellant was that if there was no improvement in the appellant's level of pain after the weight loss, he may be a candidate for surgery. 55The last consultation prior to surgery was on 5 October 2004. At that time, the appellant described his condition as being "really bad. He said that he had not been able to leave the house or get out of bed to go to the kitchen. The respondent observed that the appellant had put on weight. The respondent said that at this time, given the appellant's condition, he again advised that consideration should be given to surgery notwithstanding that the appellant had not lost enough weight. 56The appellant's wife gave evidence that the respondent told the appellant he had no choice, he had to have surgery and if he did not, he would become incontinent. She said the respondent told the appellant he would be better off with the surgery and that otherwise he would be in a wheelchair, going to the toilet on the chair, and he would have no quality of life. 57The respondent's evidence was that, having regard to his usual practice when discussing the risks and providing warnings in relation to a proposed fusion and pedicle screw fixation on an obese patient, he would have informed the appellant of the following matters: any surgery was accompanied by potential risks and complications; there were no guarantees with the surgery; there was a risk the appellant could sustain nerve root injury, which could leave him with a weakness in either or both legs; there was a risk of spinal cord injury, which could affect both legs; and there was a risk that he could develop bladder and bowel incontinence after the surgery. 58The respondent also said he would have advised the appellant that due to his weight and size, the operation would take longer and his recovery would be prolonged. He said that he would have informed the appellant that even if the surgery was uneventful there was a chance that the back pain would not be resolved and would be ongoing. He said he would have informed the appellant that in his case he would estimate the probability of a successful outcome, being a significant decrease in pain, to be about 70 per cent. 59The respondent stated in his oral evidence that his usual practice was to give a warning that there was roughly a 5 per cent chance of a catastrophic outcome of nerve root injury. No mention had been made of this in his witness statement. The respondent explained this omission as an error. When challenged in cross-examination that he had never informed the appellant of the 5 per cent risk of a catastrophic outcome, the respondent disagreed. He also disagreed that he was telling an untruth in respect of this aspect of his evidence. The respondent agreed in cross-examination that his reference to a 70 per cent likelihood of a successful outcome was not a suggestion that there was a 30 per cent chance of a catastrophic outcome. Rather, he was suggesting there was a 30 per cent chance the appellant's condition would remain the same. 60The trial judge did not determine the factual dispute between the appellant and the respondent as to whether the respondent informed the appellant of the risk of a 5 per cent catastrophic outcome. He did, however, make observations as to the credit of the appellant, the appellant's wife and the respondent. As to the appellant, his Honour observed that he gave his evidence clearly and apparently truthfully and noted that, in any event, there was no attack upon his credit. His Honour also noted that there was no challenge to the truthfulness of the appellant's wife. 61However, there had been an attack on the respondent's credit, in that he was challenged as to his version of what he told or at least what, in accordance with his usual practice, he would have told the appellant in relation to the risk of a catastrophic outcome. His Honour commented that the respondent's concession that he had not given a warning about the risk of bilateral femoral neurapraxia or local nerve damage cast the balance of the respondent's evidence in a favourable light. His Honour said, however, that to the extent it was necessary to do so, he found the respondent did not warn the appellant of the risk he could be left with weakness in either or both legs.

Arguments on the appeal 62The appellant's principal contention on the appeal was that his Honour erred in confining the relevant breach of duty to the failure to warn of the risk of bilateral femoral neurapraxia. The appellant contended that the scope and content of the respondent's duty was to adequately warn the appellant of material risks of the operation, that duty being a "single, comprehensive duty": see Rogers v Whitaker [1992] HCA 58; 175 CLR 479. 63The appellant submitted that his Honour misapplied Rosenberg v Percival in finding that a failure to warn was not relevantly causative of harm unless the risk in respect of which there was inadequate warning, materialised. His Honour's consequent finding on causation was thereby erroneous. Alternatively, the appellant contended that Rosenberg v Percival ought to have been distinguished by his Honour on the basis that it involved a case where there was only one material risk, not two as was the case here. 64The appellant contended the respondent had breached his duty of care not only by failing to warn of the risk of bilateral femoral neurapraxia, but also in failing to warn of the 5 per cent risk of a catastrophic outcome. It followed that given the "single, comprehensive" nature of the duty to disclose the material risks inherent in the operation and, on the assumption that there had been a breach of that "single, comprehensive duty", the causation issue had to be determined having regard to the entire breach. 65The appellant submitted that the correct question for his Honour's determination on the causation issue was whether the appellant would have refused surgery on 22 November 2004 had he been warned, not only of the risk of bilateral femoral neurapraxia, but also of the 5 per cent risk of a catastrophic outcome. The appellant contended that the answer to that question should have been that he would have deferred or refused surgery at that time. 66The respondent submitted that his Honour correctly determined causation as a matter of principle and as a question of fact. He contended that even if it were to be found that he had not warned of the risk of a catastrophic outcome, it was not appropriate for the scope of his liability to extend to that failure, when that risk had not eventuated. 67The trial judge did not determine whether the risk of a 5 per cent chance of a catastrophic outcome was material or whether the appellant was warned of that risk. Both involve factual determinations of a different order. The question of materiality is determined objectively in the circumstances of the particular case: see Rogers v Whitaker at 490 (Mason CJ, Brennan, Dawson, Toohey and McHugh JJ); Rosenberg v Percival per Gummow J at 455. A reasonable person, contemplating operative treatment to resolve a condition that was not life threatening and potentially not urgent, may well attach significance to a 5 per cent risk of a catastrophic outcome in making a decision as to whether or not to undergo surgery and the respondent did not suggest to the contrary. It may be that there is no real issue as to the materiality of the risk. However, it is a question of fact for the determination of the trial judge. 68What the appellant would have done if warned of the risk of paralysis involves a subjective determination in accordance with s 5D(3). Both parties submitted that this Court could determine this question. In doing so, neither party directed attention to the necessity to first determine whether the risk was material. Assuming materiality of the risk, the appellant advanced a number of factual circumstances that he submitted would lead a fact finder to conclude that he would not have had the operation. These matters included: his weight being more than he believed and the respondent not having informed him of his correct weight; the scope for more weight loss before surgery was finally decided upon; the fact that the appellant was under a misapprehension that the back surgery was necessary at that time; that the surgery was not urgent and the respondent's reason for recommending it at that time was the appellant's loss of quality of life, given his failure to lose weight. The appellant also contended that the trial judge overstated the appellant's future mobility restrictions in finding that he would have been housebound and bedridden for many weeks had the surgery not been undertaken at that time. 69The respondent for his part contended that on the evidence, the Court would find that that the appellant would have had the operation in any event. The respondent referred to the appellant's condition in the weeks immediately preceding the last consultation when he was significantly incapacitated and effectively bedridden due to pain from his condition.

The statute 70As this case fell to be determined in accordance with the provisions of the Civil Liability Act, the question of causation had to be determined on the basis of the provisions of s 5D. That section provides, relevantly: "5D General principles (1) A determination that negligence caused particular harm comprises the following elements: (a) that the negligence was a necessary condition of the occurrence of the harm (factual causation), and (b) that it is appropriate for the scope of the negligent person's liability to extend to the harm so caused (scope of liability). (2) In determining in an exceptional case, in accordance with established principles, whether negligence that cannot be established as a necessary condition of the occurrence of harm should be accepted as establishing factual causation, the court is to consider (amongst other relevant things) whether or not and why responsibility for the harm should be imposed on the negligent party. (3) If it is relevant to the determination of factual causation to determine what the person who suffered harm would have done if the negligent person had not been negligent: (a) the matter is to be determined subjectively in the light of all relevant circumstances, subject to paragraph (b), and (b) any statement made by the person after suffering the harm about what he or she would have done is inadmissible except to the extent (if any) that the statement is against his or her interest. (4) For the purpose of determining the scope of liability, the court is to consider (amongst other relevant things) whether or not and why responsibility for the harm should be imposed on the negligent party." 71Reference should also be made to s 5E, which provides: "5E Onus of proof In determining liability for negligence, the plaintiff always bears the onus of proving, on the balance of probabilities, any fact relevant to the issue of causation." 72Division 6, s 5O, which makes provision for the standard of care for professionals has no application to cases involving failure to warn: see s 5P, and thus is not relevant to the problem with which the Court is presently concerned. 73The appellant submitted that in this case s 5D(1)(a), that is, the "but for" test, was satisfied: the respondent's breach of the duty to give him an adequate warning of the material risks of the operation was a necessary condition of the harm he suffered. He also submitted that s 5D(1)(b) was satisfied as it was appropriate for the scope of the respondent's liability to extend to the harm so caused. 74Section 5D is relatively recent legislation and there is at this point little High Court authority as to its scope and application and there is no authority on the application of s 5D in a case such as the present. There also appears to be scant consideration in the case law generally of the particular issue of causation that arises here.

The case law 75The Court was not referred to any Australian case law directly on point (even on the non-statutory test for causation) and was not referred to any overseas authority. The appellant's approach was to take existing High Court authority in failure to warn cases as building blocks to demonstrate, through the interrelationship between the elements of the cause of the action in tort: duty, breach, causation, harm and damage, why and how the trial judge erred in his causation finding. Cases in New South Wales, the United Kingdom and the United States, which from the Court's researches are relevant to the specific question of causation in issue in this case, are also discussed below. The High Court authorities upon which the appellant relied are discussed first.

Rogers v Whitaker 76As has already been indicated, the appellant's commencing point was with the scope and the content of the duty of care of a medical practitioner in a failure to warn case, which he submitted was a "single comprehensive duty" to take reasonable care: see Rogers v Whitaker. In that case, the plurality (Mason CJ, Brennan, Dawson, Toohey and McHugh JJ) stated, at 493, that the duty of care was not to be approached on the basis of an "artificial division or itemization of specific, individual duties, carved out of the overall duty of care". Rather: "The duty of a medical practitioner to exercise reasonable care and skill in the provision of professional advice and treatment is a single comprehensive duty." 77Their Honours also commented upon the emphasis in the United States authority to 'informed consent' as relevant to the action of trespass, noting that Australian law had not followed that path. Subject to the question of therapeutic privilege: see Canterbury v Spence 464 F 2d 772, 150 U.S.App.D.C. 263 (1972), which was not suggested in this case, the Australian position is that a failure to disclose risks inherent in medical procedures is founded in an action of negligence only. Their Honours stated that for the purposes of Australian law, the duty was to disclose material risks inherent in the procedure and that a risk was material if: "... in the circumstances of the particular case, a reasonable person in the patient's position, if warned of the risk, would be likely to attach significance to it or if the medical practitioner is or should reasonably be aware that the particular patient, if warned of the risk, would be likely to attach significance to it. This duty is subject to the therapeutic privilege." 78Gaudron J substantially agreed with the plurality judgment. 79Gaudron J stated the test for ascertaining whether a risk was material in slightly different terms. However, as the test stated by the plurality has been consistently applied in the authorities and is essentially the same as the test that has evolved in overseas jurisdictions, it is not necessary to separately consider her Honour's formulation of the test of materiality, other than to note that she referred to Canterbury v Spence on this issue. That case is considered further below. 80It should be noted that it was unnecessary for the Court in Rogers v Whitaker to deal with causation.

Chappel v Hart 81In Chappel v Hart [1998] HCA 55; 195 CLR 232, there was an inherent risk in the operation performed on the plaintiff of perforation of the oesophagus which, if infection set in, could cause damage to the laryngeal nerve and other damage to the plaintiff's voice. The plaintiff was not warned of these risks, although the risk of infection in connection with perforation of the oesophagus was very small. It was found that if warned, the appellant would have deferred the operation and would have had it performed by a more experienced surgeon. The randomness of the likelihood of infection meant that it was unlikely that, if the operation was performed at some other time, the inherent risk of perforation in combination with infection would have come home. 82The critical issue in Chappel v Hart was causation. The governing common law test required satisfaction of the "but for" test, but causation was otherwise determined in accordance with the "common sense" test of causation: see March v Stramare (E & M H) Pty Ltd [1991] HCA 12; 171 CLR 506. Gaudron, Gummow and Kirby JJ each held that the plaintiff's injury had been caused by the defendant's negligent failure to warn. The appellant relied upon the reasoning of Gummow J. However, reference should also be made to the reasoning of Gaudron J, to which Gummow J referred in the later decision of Rosenberg v Percival. 83Gaudron J had regard to the connectivity between the duty of care and causation. Her Honour stated, at [7] 238: "Questions of causation are not answered in a legal vacuum. Rather, they are answered in the legal framework in which they arise. For present purposes, that framework is the law of negligence. And in that framework, it is important to bear in mind that that body of law operates, if it operates at all, to assign a duty to take reasonable steps to prevent a foreseeable risk of harm of the kind in issue." (citation omitted) 84Gummow J, after referring to the test in March v Stramare, referred to the judgment of Lord Hoffman in Environment Agency v Empress Car Co (Abertillery) Ltd [1999] 2 AC 22, where his Lordship stated that: "... one cannot give a common sense answer to a question of causation for the purpose of attributing responsibility under some rule without knowing the purpose and scope of the rule ..." Gummow J then observed, at [65] 256-257: "The nature and purpose of a duty with the content established in Rogers v Whitaker concern the right of the patient to know of material risks which are involved in undergoing or forgoing certain treatment." 85In Mrs Hart's case, the very risk of which the plaintiff should have been warned, materialised. It was a risk about which the plaintiff had made a specific enquiry and the evidence established that the potential consequences of damage to her voice were more significant to her than the statistical risk. The respondent doctor conceded that had the operation been performed at a different time, then, in all likelihood, the damage would not have occurred. Gummow J considered that those additional factors, combined with satisfaction of the "but for" test, were sufficient to establish causation in the particular case. As his Honour explained, at [68] 257: "Here, the injury to Mrs Hart occurred within an area of foreseeable risk. In the absence of evidence that the breach had no effect or that the injury would have occurred even if Dr Chappel had warned her of the risk of injury to the laryngeal nerve and of the consequent risk of partial or total voice loss, the breach of duty will be taken to have caused the injury." (citations omitted) 86Gummow J also referred, at [70] 258, to the place of logic and policy in the determination of the causation issue. 87Although in dissent, the reasons of McHugh J are also relevant as his approach, on principle, was in accordance with the majority. His dissent was in his Honour's determination of the common sense and legal policy issues. His Honour accepted that the "but for " test had been satisfied. The failure to warn, in combination with other events, led to the perforation of the plaintiff's laryngeal nerve. However, as his Honour stated at [25] 243: "The defendant's failure to warn, therefore, must be regarded as a cause of the plaintiff's injury unless either common sense or legal policy requires the conclusion that, for the purposes of this action, the failure is not to be regarded as a cause of the plaintiff's injury." 88In his Honour's opinion, as a matter of logic, a defendant would only be held responsible where the defendant's conduct increased the risk of injury. His Honour then discussed the concept of material increase in risk, a matter not relevant to this case. However, it is useful to refer to his Honour's demonstration of that notion by reference to a different scenario in a negligent failure to warn of a landslide. That reference is relevant because of his Honour's acceptance that if a plaintiff would have abandoned a course of conduct, a failure to warn of a safer alternative route was an a fortiori example of a material increase in risk. 89The present case is different from Chappel v Hart, which involved a failure to warn of a single risk that in fact materialised. This case involves a failure to warn of two material risks of the surgery, where only one materialised. However, Gaudron, Gummow and McHugh JJ recognised that the question of causation in the tort of negligence must be resolved having regard to the duty that has been breached. In that regard, whilst Gaudron J referred to the "duty to take care", Gummow J expressly referred to the nature and purpose of the duty in a failure to warn case as being that established in Rogers v Whitaker. This was a reference to the "a duty to warn of material risks", that duty being a "single comprehensive duty" not the itemisation of specific individual duties carved out of the overall duty to take care. The example given by McHugh J recognises that if there was an alternative, safer option available to a plaintiff, including of not going on the journey, which in this case is the equivalent of not having treatment at all, a failure to warn of danger would be causative of harm suffered.

Rosenberg v Percival 90The appellant next referred to Rosenberg v Percival. That case involved a failure to warn of the risk of a particular complication of dental surgery, which caused severe and permanent pain. 91Relevantly, for present purposes, Gleeson CJ observed, at [3] 348, that the issues of breach of duty and causation were related. This is an important observation for the appellant's argument, which is based upon a breach of a "single comprehensive duty" to warn of material risks. His Honour further observed, at [14] 441: "Ordinary people live their lives surrounded by adverse contingencies that are foreseeable, in the sense that they are not far-fetched or fanciful. Transportation to a doctor's surgery may be accompanied by a foreseeable risk of serious injury in a motor vehicle accident, but such a risk is usually regarded as inconsequential. Even when surgical procedures are classified as elective, most people who undergo such procedures believe they have a serious reason for doing so; and doctors who recommend such procedures normally have the same belief. Thus, information about risk is being considered in the context of a communication between two people who have a common view that there is a serious reason in favour of the contemplated surgery. The more remote a contingency which a doctor is required to bring to the notice of a patient, the more difficult it may be for the patient to convince a court that the existence of the contingency would have caused the patient to decide against surgery." 92In that case, his Honour categorised the adverse contingency as remote. His Honour noted that the trial judge had found the plaintiff's evidence unconvincing that, if warned, she would not have had the operation. His Honour reiterated that, in determining causation in cases of failure to warn, regard was to be had to the context before or at the time of the event, in which a particular contingency was to be evaluated. The necessity of doing so was to guard against the hindsight importance that a plaintiff, understandably, was likely to give to the failure to warn if an adverse consequence eventuated. 93Gummow J, in noting the interrelationship of the issues of materiality and causation, referred to the observations of Gaudron J in Bennett v Minister of Community Welfare [1992] HCA 27; 176 CLR 408 at 422 where her Honour noted that "questions of the sufficiency of the precaution to avert the harm are inevitably subsumed in the finding that there was a duty". Gummow J continued, at [56] 452: "It should be emphasised that a determination of whether the failure to warn of a risk was causative depends to a large extent on the definition or identification of the risk in question. Central to the identification of the risk are considerations of degree and severity. Those same considerations are also central to the question of whether the risk was material. If the Court comes to one view of those considerations when dealing with the materiality issue, the same view will direct the resolution of the causation issue." 94In the context of the "identification of the risk", Gummow J, at [60] 453, stated that the duty that a medical practitioner owes to patient was "to warn ... of a material risk inherent in the proposed treatment". His Honour then turned to the question of the risk that was being spoken of in Rosenberg v Percival itself, noting that identification of the risk necessarily preceded the consideration of whether the risk was material. His Honour said, at [61] 453: "Where the action is brought in negligence and the plaintiff is seeking compensation for an injury suffered, the relevant risk is the possibility that the proposed treatment will result in the injury that in fact occurred" (emphasis added) 95In dealing with causation, his Honour explained the required connection between risk and injury, as involving two levels of inquiries. His Honour stated, at [83] 460: "The question of whether a failure to warn of a material risk was causative of the plaintiff's injury involves two distinct levels of inquiry. At the first level, the risk must be related in a physical sense to the injury that was suffered. Thus, a medical practitioner will not be held liable for the failure to warn a patient of a material risk of damage to 'her laryngeal nerve', if the injury that eventuated resulted from a misapplication of anaesthetic. This is so despite the fact that the patient would not have had the treatment and therefore would not have suffered the injury from the misapplication of anaesthetic if the patient had been warned of the risk to 'her laryngeal nerve' [see Chappel v Hart at 257]. This can be seen as an example of a situation where the application of the 'but for' test would lead to an unacceptable result [see March v Stramare at 516]." 96Gummow J had used the same example of the misapplication of the anaesthetic in Chappel v Hart at [66] 257. As had been explained by Mason CJ in March v Stramare at 516 (to which his Honour referred in the above passage), where an event: "... secures the presence of the plaintiff at the place where and at the time when he or she is injured [the event] is not causally connected with the injury, unless the risk of the accident occurring at that time was greater." 97Mason CJ accepted this as a recognised qualification to the application of the "but for" test: see Faulkner v Keffalinos [1970] 45 ALJR 80 at 86. See also the commentary by H L A Hart and Tony Honoré, Causation in the Law, 2nd ed (1985) at 122. In this context, Mason CJ commented, at 516-517, that the "but for" test "had to be tempered by the making of value judgments and the infusion of policy considerations": cf the contrary opinion that the "but for" test excluded such considerations: Weinrib, "A Step Forward in Factual Causation" (1975) 38 Modern Law Review 518, 530. 98The second level of inquiry involved finding "a causal connection in the legal sense, between the failure to warn of the material risk and the occurrence of the injury": see at [84] 460. In this regard, Gummow J, at [85] 460-461, noted that the legal concept of causation is primarily concerned with attributing responsibility. He again referred to Lord Hoffmann's analysis in Environment Agency v Empress Car Co, at 29, that the answer to the question whether a person is liable for conduct "will depend upon the rule by which responsibility is being attributed". In the case of a claim of negligence, the relevant rule is identified within the framework of "a duty to take reasonable steps to prevent a foreseeable risk of harm of the kind in issue". In the same part of his reasoning, Gummow J endorsed the statement of Gaudron J in Chappel v Hart at [7] 238, set out at [83] above. 99Earlier at 63, Gummow J had referred to the statement in Rogers v Whitaker, that "the law should recognise that a doctor has a duty to warn a patient of a material risk inherent in the proposed procedure" (original emphasis). Then, at [86] 461, in the particular passage upon which the trial judge in this case relied, Gummow J said. "In this case the purpose of the relevant rule is that stated in par (vi) of the extracts from the joint judgment in Rogers which have been set out above. This imposes an obligation upon medical practitioners adequately to warn their patients of the consequences of the treatment they are contemplating. If the medical practitioner should fail to warn the patient of a particular consequence and that consequence in fact eventuates then, subject to the question of materiality, the rule seeks to hold the medical practitioner liable for that consequence." (emphasis in trial judge's reasons) 100His Honour next referred to Bennett v Minister of Community Welfare where Gaudron J said, at 420: "[T]he issue is approached on the basis that 'when there is a duty to take a precaution against damage occurring to others through the default of third parties or through accident, breach of the duty may be regarded as materially causing or materially contributing to that damage, should it occur, subject of course to the question whether performance of the duty would have averted the harm." 101His Honour then (still within [86] 461) summarised the position as follows: "... causation, in the present kind of case, requires satisfaction of two criteria. The first criterion is a breach of the duty to warn of a material risk, that risk having eventuated and caused, in the physical sense, injury to the plaintiff. The second criterion is that, had the warning being given, the injury would have been averted, in the sense that the [plaintiff] would not have had the treatment in question." 102The appellant's response contended that the remarks in Rosenberg v Percival were to be understood in the context that there was only one material risk of harm. In essence, the appellant's position was that linguistic focus on "that risk" or "that consequence" eventuating or materialising, did not mean that a failure to warn of two or more material risks, only one of which eventuated as in this case, could not, as a matter of legal principle, be causative of harm. It was the appellant's submission that to the extent that the trial judge had found otherwise, he erred. 103The respondent's submissions emphasised first the subjective nature of the test in determining what the patient would have done if warned of the risk and secondly, the necessity for causation to be referrable to the risk that materialised. He relied, in particular, upon the approach of McHugh J in Rosenberg v Percival in the following passage: "[25] It follows from the test being subjective that the tribunal of fact must always make a finding as to what this patient would have done if warned of the risk. In some cases where there is no direct evidence as to what the patient would have done, the judge may infer from the objective facts that the patient would not have undergone the procedure ... (This passage must now be read subject to s 5D(3).) [45] In terms of causation theory, the critical fact is whether the patient would have taken action - refusing to have the operation - that would have avoided the harm suffered. But that fact can only be determined by making an anterior finding as to what the patient would have decided to do, if given the relevant warning. It is not possible to find what the patient would have done without deciding, expressly or by necessary implication, what decision the patient would have made, if the proper warning had been given. If the court finds that the patient would have decided not to have the operation, it concludes that he or she would not have had the operation. What the patient would have decided and what the patient would have done are hypothetical questions. But one relates to a hypothetical mental state and the other to a hypothetical course of action." (emphasis added) 104The respondent acknowledged that McHugh J was addressing his remarks to a case where there was "a risk of harm": see at [24] 443, but contended that his Honour's approach was of general application, that is, the question of causation was determined solely by reference to the risk that came home.

Ellis v Wallsend District Hospital 105Prior to the High Court authorities discussed above, the question of causation in a failure to warn case had been considered by this Court in Ellis v Wallsend District Hospital (1989) 17 NSWLR 553. 106In that case, the plaintiff brought proceedings against the defendant hospital in circumstances where, six days after she underwent surgery performed by a visiting honorary surgeon on her neck, she developed quadriplegia. The plaintiff brought her claim against the hospital only on the basis that it owed her a non-delegable duty of care which extended to the medical services provided by an honorary medical officer to a private patient. Alternatively, she alleged that the hospital was vicariously liable for the negligence of the honorary surgeon who performed the operation. 107The plaintiff was only warned of the risk of some slight numbness in her right hand consequent upon the operation. However, the surgery carried with it a risk of paralysis, because of the close proximity of the operation site to the spinal cord. That risk was small, but not fanciful. There was also substantial risk that the surgery might not relieve her neck pain. In fact, there was only a 30 per cent chance of success of relief. The surgeon was found to be negligent in failing to warn the plaintiff of those two risks. However, the trial judge concluded that the plaintiff would not have refused to have the operation even if warned of the slight risk of paralysis. The trial judge also held that there was no evidence that the plaintiff would have refused the operation had she been warned that there was only a 30 per cent chance of success in respect of long-term pain relief. He dismissed the plaintiff's claim. 108On appeal, the plaintiff was successful in establishing causation on the failure to warn case, although not ultimately successful in her claim against the defendant hospital, which was found not to owe her a non-delegable duty, nor to be vicariously liable for the honorary surgeon's negligence. In issue on the question of what course the plaintiff would have taken had she been adequately warned of the risks of the operation, was whether the trial judge had correctly assessed the evidence as to the plaintiff's subjective intention. 109Kirby P (who would have allowed the appeal) agreed with Samuels JA on the question of causation. In particular, Kirby P observed, at 561, that the plaintiff was entitled to be given a candid and balanced statement of the risks. As his Honour said, at 561, "the bigger the devastation of the possible risk, the greater is the obligation to lay it before the patient". His Honour continued: "An operation with a very high success rate and a very low risk of paralysis may, rationally, be accepted much more readily than one where the prospects of success were more circumscribed, though the risk of paralysis was still slight." 110Samuels JA, with whom Meagher JA agreed, stated, at 590, that the trial judge: "... did not give any attention to the possible subjective response which the [plaintiff] might have made had the information conveyed to her contained both a warning about possible paralysis and an accurate assessment of the likely prospects of success. This is curious ... because he found that [the surgeon's] negligence involved failure to warn about both these matters ... he took this course, I would judge, because of the way in which he eliminated from consideration the possible influence of the low rate of statistical success. But, in my respectful opinion, he was not only bound to consider that matter alone from a subjective standpoint, but bound also to consider it from the same standpoint as one of two combined elements in the warning which the appellant should have received." 111Ellis v Wallsend District Hospital is directly on point to the issue raised in the present case. However, it is not binding on this Court, as it was decided in accordance with common law principles and not under the statute. In addition, their Honour's comments were also strictly obiter. However, the Court's reasoning on the need to consider the combined risks of which a warning was required is persuasive as to the matters that may be relevant when determining whether it is "appropriate for the scope of a negligent person's liability to extend to the harm so caused" for the purposes of s 5D(1)(b). Their Honours' reasoning is also consonant with the interrelationship of duty and causation, established in the High Court authorities, and which has not been abrogated by the statutory provisions.

Chester v Afshar 112In Chester v Afshar [2004] UKHL 41; [2005] 1 AC 134, the plaintiff had not been warned of the risk of cauda equina syndrome, which had about 1-2 per cent chance of occurrence. That risk came home. 113The plaintiff's claim was upheld in circumstances where the majority of their Lordships concluded that the plaintiff had not established, on conventional principles, that her injury had been caused by the medical practitioner's failure to warn of the risk of the serious condition of cauda equina syndrome. The plaintiff had not adduced evidence as to what she would have done if warned of the risk. The headnote sufficiently summarises the reasoning of the majority for present purposes: "[The issue of causation was to be addressed] by reference to the scope of the doctor's duty, namely, to advise his patient of the disadvantages or dangers of the treatment he proposed; that such a duty was closely connected with the need for the patient's consent and was central to her right to exercise an informed choice as to whether and, if so, when and from whom to receive treatment; that, since the injury she sustained was within the scope of the defendant's duty to warn and was the result of the risk of which she was entitled to be warned when he obtained her consent to the operation in which it occurred, the injury was to be regarded as having been caused by the defendant's breach of that duty; and that, accordingly, justice required a narrow modification of traditional causation principles to vindicate the claimant's right of choice and to provide a remedy for the breach." (emphases added) 114The reasoning of the majority is relevant insofar as it refers to the interrelationship of the scope of the duty and causation. Lord Bingham and Lord Hoffmann dissented. Lord Bingham noted that if the plaintiff would not have undergone the surgery if warned of its inherent risks, she would, on conventional principles, have been entitled to recover damages. However, in the absence of a finding to that effect, the plaintiff had not established factual causation on the application of the "but for" test. In his Lordship's view, a defendant should not be liable when it had not been established that the defendant's conduct had worsened the physical condition of the plaintiff: see Chappel v Hart per McHugh J. 115Lord Hoffmann, at [31], considered that the relevant question on causation was whether the plaintiff would have taken the opportunity to avoid or reduce the risk, not whether the plaintiff would have changed the scenario. This was decisive of the outcome on his Lordship's view, because the trial judge had found that the risk would have been precisely the same whether the operation was performed then, or later, by the same or a different surgeon. His Lordship also agreed with the reasons of Lord Bingham, and thus with the judgment of McHugh J. 116Apart from Ellis v Wallsend District Hospital, the authorities considered thus far involved a failure to warn of a specific risk, which materialised. This case involves a failure to warn of material risks, only one of which materialised. That circumstance arose for determination in Moyes v Lothian Health Board 1990 S.L.T. 444 where the plaintiff suffered a stroke in the course of an angiogram. Her claim against the defendant included the alleged negligence of the neurological surgeon, who performed the angiography, in failing to warn her of the aggravated risk of a stroke occasioned by her hypersensitivity to the contrast medium used in the procedure and by her previous history of migraine. 117The plaintiff's action failed as she was unable to prove a previous hypersensitive reaction or that either hypersensitivity or allergy had caused her stroke. However, in commenting on the causation question in a failure to warn case, Lord Caplan stated, obiter, at 447: "The ordinary person who has to consider whether or not to have an operation is not interested in the exact pathological genesis of the various complications which can occur but rather in the nature and extent of the risk. The patient would want to know what chance there was of the operation going wrong and if it did what would happen ... It is perfectly conceivable that a patient might be prepared to accept the risk of one in 100 but not be prepared to face up to a risk of one in 20. If a doctor contrary to established practice failed to warn the patient of the four special risks but did warn the patient of the standard risk and then the patient suffered complication caused physiologically by the standard risk factor rather by one or other of the four special risks factors I do not think the doctor should escape the consequences of not having warned the patient of the added risks which that patient was exposed to. A patient might well with perfect reason consider that if there were five risk factors rather than one then the chance of one or other of these factors materialising was much greater. The coincidence that the damage which occurred was due to the particular factor in respect of which a warning was given does not alter the fact that the patient was not properly warned of the total risks inherent in the operation and thus could not make an informed decision as to whether or not to go through with it. In the example I give, by going through an operation with five risk factors rather than one the patient was exposed to a degree of risk materially in excess of what the patient had been warned about and was prepared to accept. If he had been given due warning he would have not risked suffering adverse complication from that particular operation and the fact that such complication occurred is causal connection enough to found a claim against the doctor."

The Canadian authorities 118There is also support for the appellant's argument in the Canadian authorities, although those authorities must be read with care, having regard to that jurisdiction's acceptance of the objective test in determining what a plaintiff would have done if warned. It is sufficient to consider the well-known authority, Reibl v Hughes [1980] 2 S.C.R. 880, 14 C.C.L.T.1, 114 D.L.R. (3d) 1, 33 N.R. 361, J.E. 80-894.

Reibl v Hughes 119In Reibl v Hughes, the Supreme Court identified the relevant duty of care of a surgeon to a patient "to make disclosure ... of ... all material risks attending the surgery which is recommended": see Hopp v Lepp [1980] 4 W.W.R. 645, 13 C.C.L.T. 66, 32 N.R. 145, 22 A.R. 361. The particular risks attending the surgery in that case were of stroke, paralysis and death. There were risks of stroke and death if the surgery was not performed. The Court was concerned with the question of how specific the information should be to enable the patient to make an "informed" choice between surgery and no surgery. The Court considered that the term "informed consent" should be abandoned when the question involved the duty of disclosure of material risks and that cases in trespass should be restricted to where there was no consent at all. This is in accord with the Australian authority that the relevant cause of action for failure to disclose is negligence, not trespass: see Rogers v Whitaker above at [77]. 120The Supreme Court held that the trial judge had correctly found that the surgeon had breached his duty of care to the patient. He had failed to give the patient an adequate disclosure of the risk of death or paralysis attendant upon the surgery itself and had also failed to inform the patient that the operation would not alleviate his headaches and hypertension, about which he had been concerned. On the question of causation, the Supreme Court held that a reasonable person in the plaintiff's position (that is, on the application of an objective test) would have refused the surgery, having regard to the combination of circumstances and, in particular, had he been informed of the particular serious risks of the surgery and of the fact it would not have cured his headaches and hypertension.

Canterbury v Spence 121The starting point for the American case law is Canterbury v Spence. That case is authority in the United States for the principle that the question of what a plaintiff would have done if adequately warned of the risks of treatment is to be determined on an objective basis. The approach in Canterbury v Spence on this question has been expressly rejected in Australia in favour of a subjective test. Section 5D(3), which governs this case also provides for a subjective test. Relevantly for the question in issue in this case, however, was the US Court of Appeals, District of Columbia Circuit approach to causation in cases of a failure to warn of material risks. 122The case, which was a single risk case, concerned a claim against a surgeon in trespass and negligence for a failure to warn of a 1 per cent risk of paralysis in respect of a laminectomy. Approximately one day after surgery, the plaintiff fell from his hospital bed. Some hours later, the plaintiff commenced to suffer from paralysis from the waist down. A second operation was performed to release pressure from the spinal cord, however, the plaintiff did not fully recover his mobility. 123There was a directed verdict at first instance in favour of the surgeon, on the basis that there was no evidence of negligence either in the diagnosis of the plaintiff's condition or in the performance of the surgery. The trial judge did not deal with the alleged negligence based on the surgeon's failure to warn of the risk of paralysis. In that regard, the Court of Appeals recognised the existence of a cause of action in negligence where a physician had failed to provide an adequate warning of the risks involved in any proposed treatment, independently of a claim in trespass based on an absence of informed consent. The matter was remitted for a new trial on all issues, other than the claim in trespass which was statute barred. 124Insofar as the claim was based in the failure to warn, the Court of Appeals defined the scope of the duty as being the disclosure of material risks. The Court explained "material risk", at [22], in the following terms: "[a] risk is thus material when a reasonable person, in what the physician knows or should know to be the patient's position, would be likely to attach significance to the risk or cluster of risks in deciding whether or not to forego the proposed therapy." (emphasis added) 125The Court of Appeals noted, at [24]-[25], that a medical practitioner was not required to disclose risks that had no apparent materiality to the patient's decision on therapy. However, the Court of Appeals noted it did not subscribe to the view that only risks that would cause a patient to forego treatment were required to be disclosed: "... for such a principle ignores the possibility that while a single risk might not have that effect, two or more might do so." 126The Court of Appeals then dealt with the issue of causation in the following terms: "[30] ... An unrevealed risk that should have been known must materialize, for otherwise the omission, however unpardonable, is legally without consequence. Occurrence of the risk must be harmful to the patient, for negligence unrelated to injury is nonactionable. And, as in malpractice actions generally, there must be a causal relationship between the physician's failure to adequately divulge and damage to the patient. [31] A causal connection exists when, but only when, disclosure of significant risks incidental to treatment would have resulted in a decision against it." (footnotes omitted) 127These passages, in their terms, support the proposition that it is only the risk that came home that is relevant to causation. However, it has been otherwise interpreted in the United States, given the requirement to establish "proximate cause".

Downer v Veilleux 128In Downer v Veilleux 322 A 2d 82 (Me. 1974) the Supreme Judicial Court of Maine was concerned, inter alia, with the question of causation where there had been an alleged failure to adequately disclose the various procedures available to treat the fractures the plaintiff sustained in a motor vehicle accident. The Court held that the proof of "proximate cause" required establishing that the non-disclosed risk, which should have been made known, had materialised. The Court stated, at [12], that absent occurrence of the undisclosed risk, the doctor's omission was legally inconsequential. This statement was firmly based on the United States test of "proximate cause". The Court had noted, at [11], that: "As in the case of any breach of a legal duty, the plaintiff must, as in malpractice actions generally, prove a proximate causal relationship between the physician's failure to adequately inform and injury to the plaintiff." 129The Court continued, at [12]: "Proof of proximate cause in such cases requires, initially, a showing that the unrevealed risk which should have been made known has materialized. Absent occurrence of the undisclosed risk, the doctor's omission is legally inconsequential." 130Canterbury v Spence and Downer v Veilleux are accepted authority in the United States for the proposition that an undisclosed risk must materialise in order for causation to be established: Cochran v Wyeth 3 A 3d 673 (Pa.Super, 2010). Cochran v Wyeth has a closer analogy to this case. The case involved a claim by a plaintiff against a drug company for inadequate disclosure of the side effects of a certain drug. The plaintiff suffered a side effect that had been disclosed. She asserted, however, that the drug company had failed to disclose another side effect, such that if disclosed, her medical practitioner would not have prescribed the drug for her. 131The Court observed, at 28, that in informed consent cases it appeared to be well settled that "the non-disclosed risk must manifest itself into actual injury in order for a plaintiff to establish proximate causation". The Court held that a plaintiff would fail to establish "proximate causation" where the non-disclosed risk failed to materialise in an injury. The Court stated, at 30, that: "... before a plaintiff can prove that a non-disclosed risk would have altered the physician's decision to prescribe a drug, the plaintiff must first demonstrate that he/she suffered from the precise injury that the manufacturer failed to disclose." 132In adopting this approach, the Court observed that Canterbury v Spence and Downer v Veilleux had been uniformly accepted by high courts of other jurisdictions in the United States. The Court further stated that it had been unable to locate any authority that had refused to adopt the "proximate causation" principle enunciated in those two cases. 133It should be noted that integral to the decision was that the tort of informed consent corresponded with the tort of failure to warn, so that the same test of "proximate causation" applied to both causes of action. 134The Court commented that a similar test had been imposed in claims in cases of fraudulent omissions or misrepresentations, stating, at 28 footnote 2, that in that area of jurisprudence: "These cases hold that in order to be compensable, the non-disclosed risk of misrepresented fact must materialize into an injury that is specifically related in scope to the peculiar nature of the omission or misrepresentation." 135Even making allowance for the objective test for determining what a plaintiff would have done if warned, these United States cases do not support the appellant's case. However, I do not consider them to be persuasive in the present context because of the different test of causation, that is, the test of "proximate causation" that applies in American law. That is not the test for the purposes of s 5D: see s 5D(3).

Should the appeal be allowed? 136This case involved a duty to give the appellant an adequate warning of the material risks of undergoing the surgery. The appellant contended there were two material risks: the neurapraxia, which the trial judge held was a material risk, and the 5 per cent risk of paralysis. The judge made no finding as to the materiality of this risk, although for the reasons I have given, there may be no real dispute about that. 137Only one of the risks, the neurapraxia, eventuated. The appeal directly raised the question whether, in determining causation in a case of failure to warn of material risks, the only relevant risk is the one that eventuated. Or is it sufficient that the harm suffered is within the area of foreseeable harm, and it is determined that the patient would not have had the operation if warned of a material risk that did not come home. 138That factual circumstance did not arise for consideration in the trilogy of High Court authority: Rogers v Whitaker, Chappel v Hart and Rosenberg v Percival, in which the parties respectively sought to find the principle which should govern such a case. In my opinion, the references in those cases to the undisclosed risk materialising does not preclude, as a matter of principle, the existence of a causative link to the harm suffered, when an undisclosed risk, being one of two or more material risks, does not materialise. 139In my opinion, the statements of Gummow J in Rosenberg v Percival, in particular at [86] 461, which led the trial judge to a contrary conclusion, have to be read in the context of the factual circumstances with which the Court was concerned. Paragraph [61] 453 of Gummow J's reasons is also relevant in understanding his Honour's reasons as a whole. 140At [61] 455, his Honour was dealing with the "identification of the risk". In that context, his Honour stated, at [64], that the "relevant risk" was: "... real and foreseeable ... [The] character of the injury or the precise sequence of events leading to the injury need not be foreseeable ... even if the extent of the injury was greater than expected." (at [64]) 141This principle derives from Hughes v Lord Advocate [1963] UKHL 1; [1963] AC 837 where liability was held to be established in circumstances where the risk of fire was foreseeable although an explosion (which occurred) was not. That is a different question from whether a relevant, material risk was causative of harm. That was the issue with which his Honour was concerned at [86] 461 and which, as I have explained, he dealt with by reference to the relevant rule of responsibility. In other words, that was his Honour's approach to the principle that governed the determination of the causation question. 142His Honour's statement that a "failure to warn of a consequence and that consequence [eventuating]" was, in my opinion, focussed upon the particular factual circumstances in that case, namely a failure to warn of a single material risk. Any different interpretation would undermine the two basal concepts upon which his reasoning proceeds, namely, the determination of a defendant's liability having regard to the relevant rule of responsibility and the interrelationship of causation with the nature and scope of the relevant duty. The relevance of those concepts to the application of s 5D(1) is dealt with below. 143Neither party referred to Ellis v Wallsend District Hospital, nor to any overseas authorities discussed. As I have already indicated, the reasoning in Ellis v Wallsend District Hospital is persuasive in relation to the causation issue that arises in this case. The English and Canadian case law discussed above is also supportive of the proposition that there may be a causative link to harm suffered, even though one of two or more material risks does not eventuate. None of those cases were decided under the statutory test of causation that governs this case. However, I do not consider that undermines their persuasiveness. They all recognise the interrelationship of the nature and scope of the duty of care to disclose material risks inherent in a medical procedure and causation. 144In my opinion, the statutory content of the causation question prescribed by s 5D(1) does not negate or modify this relationship, except to the extent that there may be policy considerations that impact upon the s 5(1)(b) question that compel a contrary conclusion. 145The Civil Liability Act, Pt 1B relates to claims in negligence: s 5A(1). Negligence is defined as "a failure to exercise reasonable care and skill": s 5. Apart from the definition, the statute does not, nor could it prescribe the nature, scope or content of the duty of care in a particular case. Accordingly, when a court is concerned to determine the nature, scope and content of the duty of care, it is concerned with the notion described by Lord Hoffmann in Environment Agency v Empress Car Co as the relevant rule of responsibility. The notion of the relevant rule of responsibility has found acceptance in express terms in Australia including, but not only, in the cases to which I have referred. In this regard, see also: Travel Compensation Fund v Tambree [2005] HCA 69, 224 CLR 627; Pledge v RTA [2004] HCA 13; 205 ALR 56; Cattanach v Melchior [2003] HCA 38; 215 CLR 1; Tame v NSW [2002] HCA 35; 211 CLR 317; Modbury Triangle Shopping Centre v Anzil [2000] HCA 61; 205 CLR 254. The same notion was expressed by Gaudron J in Chappel v Hart at [7] 238. 146The statute does specify what is involved in breach and the considerations relevant, non-exclusively, to that determination. Likewise, the statute specifies the elements of causation referred to as the "but for" test and the "scope of liability" rule. It is likely that the statutory test is conceptually different from the common law test in March v Stramare but, in any event, it is the statutory language that governs: see Adeels Palace Pty Ltd v Moubarak; Adeels Palace Pty Ltd v Bou Najem [2009] HCA 48; 239 CLR 420. 147However, the prescription of the statutory elements of causation has not abrogated the necessary interrelationship of and connection between the duty of care, its nature and scope and causation. On the contrary, I am of the opinion that the interrelationship plays an integral role in the determination of the "scope of liability" element of causation. Accordingly, in my opinion, the statements of Gaudron J at [7] 238 and Gummow J at [65] 256-257 in Chappel v Hart and Gummow J's further comments in Rosenberg v Percival at [86], continue to apply to questions of causation under the statute. 148As those statements apply to a case of a failure to warn, it means that the causation question has to be determined having regard to the content of the duty, namely a duty to warn of material risks. Accordingly, where there is more than one material risk, and a finding is made that a plaintiff would not have undergone surgery if warned of other material risks, it is difficult to see reasons in logic or policy why a negligent doctor should not be liable notwithstanding that a different risk eventuated. 149In every case however, the determination is a factual one. A court might readily conclude that where a risk, although material, has relatively minor consequences, it would not be appropriate for the scope of the defendant's negligent failure to warn of that risk to extend to the different harm that in fact eventuated. A more difficult factual question might arise where the failure is to warn of a different material risk with relatively serious consequences but with a low a chance of occurrence. 150In this case, the failure to warn was of a material risk with very serious consequences should it occur and with a high risk of occurrence. If in such a case, it were to be found that a plaintiff would not have undergone surgery if warned of the risk, then I consider that it is appropriate that the scope of the defendant's liability extend to the harm in fact caused. 151A case of failure to warn of more than one material risk is different from the example of the harm suffered whilst a patient was under an anaesthetic, to which Gummow J referred in Rosenberg v Percival. It is doubtful whether harm of that type would satisfy the "but for" test. It might also be arguable that such a circumstance constituted a novus actus. By contrast, the connection between the nature and scope of the duty, and what the plaintiff would have done if there had been no breach, are sufficiently and logically connected for it to be appropriate for the scope of the defendant's to extend to the harm in fact suffered. In my opinion, the trial judge erred in finding that the only potentially relevant risk was the neurapraxia.

Conclusion 152As I have stated, the trial judge did not resolve the factual dispute as to whether the respondent disclosed the risk of a 5 per cent catastrophic outcome. That is a necessary finding if, on the appellant's approach to the issue of causation, his case is to succeed. The Court was invited to make that finding if it came to the conclusion that his Honour erred in finding, at [88], that the only relevant warning was of the risk of bilateral femoral neurapraxia. However, whether or not the warning of paralysis was given involves questions of credit. Although his Honour referred to the credit of the witnesses, his comments are not of such a nature that the Court could conclude that the evidence of the appellant and his wife was to be preferred to the evidence of the respondent, or vice versa. 153The Court was also invited to make findings as to whether, if warned, the appellant would have had the operation. The parties' submissions on that question are to be found at [68]-[69] above. However, that finding also involves questions of credit and for the same reasons, it is not appropriate for the Court to engage in a fact finding task of this nature. 154It follows, in my opinion, that the appeal should be allowed. 155I propose the following orders:

1. Appeal allowed;
2. Set aside the verdict for the respondent and the costs order made by the trial judge;
3. Remit the matter for rehearing;
4. Order that the costs of the original trial be in the determination of the judge on the retrial (or abide the outcome of the retrial);
5. Order the respondent to pay the appellant's cost of the appeal. The respondent to have a certificate under the Suitors' Fund Act 1951 if qualified. 156BASTEN JA: It is curious that the circumstances of this case have not previously arisen, so far as the Court is aware. The appellant, Mr Ian Wallace, sought medical assistance in relation to a condition in his lumbar spine. He was offered surgery. Because he was significantly overweight, the surgery was likely to be lengthy and there was a risk that he would sustain transient local nerve damage in his thighs (described as bilateral femoral neurapraxia). The risk materialised, although the condition had entirely resolved by the time of the trial. At a consultation with the respondent, Dr Kam, before the operation, the appellant was given important information in respect of the procedure and its potential for success and the likely circumstances of rehabilitation, but was not advised of that particular risk. The primary judge (Harrison J) held that he should have been, but that, if he had been, he would nevertheless have proceeded with the operation. Accordingly, the judge held that the failure to warn did not constitute a relevant cause of the harm and dismissed the claim: Wallace v Ramsay Health Care Ltd [2010] NSWSC 518. 157The operation also carried with it a 5% risk of a catastrophic outcome. The appellant said he was not warned of that either. However, because it did not eventuate, the trial judge held that any failure to warn of a risk of a catastrophic outcome was causally irrelevant: at [88]. The appellant challenged that reasoning, contending that had he been so warned, he would not have had the operation, and would thus not have suffered the transient local nerve damage. Because the judge considered the failure to warn of a catastrophic outcome was causally irrelevant, he made no findings as to whether the practitioner did fail to provide adequate advice as to that risk and as to the likely response of the appellant, if warned. 158One reason why questions of causation give rise to difficult issues in this area is that there is no negligent mistreatment causing harm, the harm being the materialisation of a risk inherent in the procedure, even when carried out with all due skill and care. The failure to warn of attendant risks is not itself a direct cause of the harm suffered: it is not a breach of any duty on the part of the practitioner to carry out the procedure with due skill and care, but rather depends upon a legal requirement that it be the patient who ultimately determines (absent an emergency) whether the procedure is undertaken or not. 159The right to recover damages in the present case was put, in part, on the basis of upholding the autonomy of the patient to determine whether or not to have the procedure. That contention is, however, too broadly stated. Like the "somewhat amorphous phrase 'informed consent'", the concept is apt to mislead: see Rogers v Whitaker [1992] HCA 58; 175 CLR 479 at 490 (Mason CJ, Brennan, Dawson, Toohey and McHugh JJ). As the judgment continued: "Anglo-Australian law has rightly taken the view that an allegation that the risks inherent in a medical procedure have not been disclosed to the patient can only found an action in negligence and not in trespass; the consent necessary to negative the offence of battery is satisfied by the patient being advised in broad terms of the nature of the procedure to be performed." 160This passage has been identified as a warning against the uncritical use of concepts "which ought to be valuable currency, but which are susceptible to rhetorical inflation": Rosenberg v Percival [2001] HCA 18; 205 CLR 434 at [9] (Gleeson CJ). 161The appellant pressed upon the Court the importance of the statement in Rogers v Whitaker that "[t]he duty of a medical practitioner to exercise reasonable care and skill in the provision of professional advice and treatment is a single comprehensive duty": at 489. However, the adoption of that language was curious: the phrase "a single comprehensive duty" derived from the speech of Lord Diplock in Sidaway v Governors of Bethlem Royal Hospital [1985] AC 871, in which he (as one of the majority) held that the Bolam principle (requiring a standard fixed by reference to the views as to proper practice held by a responsible body of medical opinion) applied with respect to that element of the duty concerned with the provision of information and advice. The High Court rejected that approach, preferring the dissenting view of Lord Scarman in Sidaway, at 488-489; see also Gaudron J at 492. As the joint judgment explained, there is "a fundamental difference between, on the one hand, diagnosis and treatment and, on the other hand, the provision of advice or information to a patient": at 489. The consequence of the difference, identified in that case, was that the content of the duty of disclosure was not a matter to be determined by practice within the medical profession, but was to be objectively determined by a court, by reference to the state of relevant medical knowledge. The test was identified in the following terms (at 490): "The law should recognize that a doctor has a duty to warn a patient of a material risk inherent in the proposed treatment; a risk is material if, in the circumstances of the particular case, a reasonable person in the patient's position, if warned of the risk, would be likely to attach significance to it or if the medical practitioner is or should reasonably be aware that the particular patient, if warned of the risk, would be likely to attach significance to it." 162A differential approach to elements of professional negligence is now required by ss 5O and 5P of the Civil Liability Act 2002 (NSW). Courts are to approach the standard or content of the duty by reference to a national standard of peer professional opinion as to competent practice, if widely accepted and not "irrational": s 5O. However, that approach does not apply to the duty to give "a warning, advice or other information in respect of the risk of death or injury": s 5P. In that area the general law, as reflected in Rogers v Whitaker, remains applicable. Similarly, the provisions of the Civil Liability Act dealing with warnings of obvious risks of death or personal injury do not apply with respect to professional services: s 5H. 163The Civil Liability Act provides that a determination that negligence caused particular harm comprises two elements: s 5D(1), chapeau, set out at [70] above. It is to be read as requiring that unless the plaintiff proves the relevant facts (s 5E) and the court is satisfied as to the other elements, there can be no finding of causation. The apparent purpose of s 5D(1) is to ensure that physical and policy elements of causation are identified and addressed separately. That purpose is immediately muddied, however, by s 5D(2), which allows for an affirmative finding of causation, absent satisfaction of the "necessary condition" test of physical or factual causation, if "established principles" permit such a finding. Although sub-s (2) does not use the phrase "scope of liability" used in sub-s (1), it requires the court to consider "(amongst other relevant things) whether or not and why responsibility for the harm should be imposed on the negligent party", which is the language used in defining the "scope of liability" inquiry: s 5D(4). In short, the 'policy' considerations both limit the area within which the counterfactual inquiry results in liability and, "in an exceptional case", permit liability where the counterfactual inquiry is not satisfied. 164In truth, any attempt to treat elements of causation as entirely separate and distinct is doomed to fail in contestable cases, as is a rigid compartmentalisation of duty, breach and causation: see Rosenberg at [3] (Gleeson CJ); [56] (Gummow J), dealing with "materiality" of risk and causation. The scope and content of the duty to warn depends on the identification, with as much precision as the circumstances permit, of a foreseeable adverse consequence. It is only when that exercise is undertaken, on the basis of medical knowledge reasonably available to the practitioner at the time of the consultation, that the question of "materiality" can be addressed and, if the risk be found to be material, the likely consequence for the particular patient of being given the appropriate advice can be determined: Rosenberg at [60]-[61], [66]-[69] and [75]-[80] (Gummow J). 165The facts of Rosenberg have been explained by Beazley JA and need not be repeated here. Importantly, in dealing with the question of causation, Gummow J stated at [83]: "The question of whether a failure to warn of a material risk was causative of the plaintiff's injury involves two distinct levels of inquiry. At the first level, the risk must be related in a physical sense to the injury that was suffered. Thus, a medical practitioner will not be held liable for the failure to warn a patient of a material risk of damage to 'her laryngeal nerve', if the injury that eventuated resulted from a misapplication of anaesthetic. This is so despite the fact that the patient would not have had the treatment and therefore would not have suffered the injury from the misapplication of anaesthetic if the patient had been warned of the risk to 'her laryngeal nerve'. This can be seen as an example of a situation where the application of the 'but for' test would lead to an unacceptable result." 166Gummow J noted, in relation to "the second level", that a failure to warn of a risk cannot be said to "cause" an injury in the same way that one speaks of causation in respect of careless conduct in carrying out the operation. The second level requires identification of the purpose for which a causal relationship is being assessed, namely the attribution of responsibility for the harm. At [86], he continued: "[The relevant rule] imposes an obligation upon medical practitioners adequately to warn their patients of the consequences of the treatment they are contemplating. If the medical practitioner should fail to warn the patient of a particular consequence and that consequence in fact eventuates then, subject to the question of materiality, the rule seeks to hold the medical practitioner liable for that consequence." 167The reference to warning of "a particular consequence", followed by the condition that "that consequence in fact eventuates" is not ambiguous. Nor is it to be dismissed as an incautious or inaccurate use of language. 168The US cases referred to by Beazley JA are to similar effect. The results in those cases are not to be ignored because the judges used the language of "proximate cause" not presently in vogue in Australia. That phrase may be traced to Sir Francis Bacon: Michael S Moore, Causation and Responsibility: An Essay in Law, Morals and Metaphysics (OUP, 2010), p 101. It is, like "scope of liability", a label to describe the boundaries of liability derived from the purposes or policies of the law. It does not dictate one conclusion rather than another. 169The arguments in favour of permitting recovery in this case, subject to relevant facts being found, are broadly twofold. One is that the duty to warn should not be subdivided because the decision to undergo treatment is a single decision made by assessing all the appropriate information: it cannot sensibly be partitioned. A second, though related, argument is that the risks were not entirely distinct, as were the risk of severing a nerve and a problem with anaesthesia, in the example used by Gummow J in Chappel v Hart [1998] HCA 55; 195 CLR 232 at [66] and in Rosenberg at [83]. 170These arguments clearly turn on matters of degree and do not rely on watertight legal categorisation. Thus, in respect of the first argument, there is a close correlation between the tasks of assessing the materiality of a risk and the causal relationship with the harm. Neither can be resolved without careful attention to the particular risk as identified in the medical literature, the manner in which it should be described to the patient and the way in which it will be understood and assessed by the patient. Although the patient has but one decision to make, the factors to be placed into the balance in making that decision (in the somewhat idealised paradigm adopted by the law) are separate and each has its own weight. 171The correlation between the tasks of assessing the materiality of a risk and the causal relationship with the harm is also close in a practical sense. Absent some subjective element known to (or which should have been apprehended by) the practitioner, "materiality" involves an objective test of assessing the likely significance of a particular risk, but by reference to a reasonable person in the plaintiff's position. The test of causation is said to be subjective, namely, asking whether this patient would have declined to undergo the treatment. However, given the inadmissibility of any statement by the patient after suffering the harm, and favourable to his or her interests, as to what he or she would have had done had the appropriate warning been given (s 5D(3)(b)), the difference between the two tasks is diminished. At least that is so in a case where the plaintiff has exhibited no specific concern before the operation as to the risks involved. Yet the materiality analysis, relevant to the identification of breach, does require separate assessment of each risk as to which it is said a warning should have been given. A trial judge could accept that, although the warning with respect to a specific risk would not have been significant to this patient, objectively it should have been given. Such a conclusion must carry with it a strong implication in relation to the possible causal connection with the harm in fact suffered. 172In respect of the second argument, it is not self-evident that the two risks in question were closely associated. The catastrophic risk of paralysis was inherent in the site and nature of the procedure, close to the spinal column. It could have a close connection with the skill and care exercised by the surgeon in carrying out the operation. By contrast, the bilateral femoral neurapraxia, which materialised, arose from the fact that the patient was required to lie prone for a significant period. The distinction between the two is not far removed from the examples relied on by Gummow J, being the possible injury to the laryngeal nerve and the risk of misapplication of the anaesthetic. 173As a matter of principle, if the purpose of requiring payment of compensation is to require the defendant to bear the cost of the harm eventuating from his or her negligent conduct, it is far from clear why that burden should be imposed in circumstances where the harm could only have arisen from breach of another element of the duty. Once the trial judge was not persuaded that a warning as to bilateral femoral neurapraxia would have led to a postponement of the operation, why should the practitioner bear responsibility for breach of a separate duty to warn of a risk which did not materialise? Recovery in such circumstances appears opportunistic. 174If recovery of compensation were to be accepted in this case, it would be on the basis of the negligent failure to warn of the catastrophic risk, while the injury suffered was the bilateral femoral neurapraxia. The causal relationship would be established by a finding (yet to be made) that, if warned of the potentially catastrophic risk, the appellant would not have undergone the procedure. On that analysis, any harm which resulted from the operation would have been sufficient to attract compensation. Whether the harm which in fact resulted was a risk as to which a warning should have been given would appear not to be a relevant consideration, once it is found that, if a warning as to that risk had been given, the appellant would nevertheless have proceeded with the operation. The same result would follow if the warning had in fact been given and, consistently with the finding of the trial judge, the appellant had proceeded to undergo the operation. However, it is hard to identify a principled basis upon which a practitioner should be made legally responsible to pay compensation to a person who suffered an adverse outcome as to which he or she had been properly advised and had accepted. By contrast, there is a clear justification for imposing liability on a medical practitioner in respect of the harm suffered where an adverse outcome of which a warning should reasonably have been given, but was not, has come to pass. The principle which supports recovery in the latter case cannot, without some significant variation or expansion, justify recovery in the present case. 175Section 5D does not clearly compel one answer to this question, rather than another. The language of the section involves (inevitably) a level of generality, which does not easily cover the circumstance where the negligence is a failure to warn of risks which inhere in a procedure carried out with due care and skill. Although it may be said that a number of physical conditions were "necessary" for the occurrence of the harm in the present case, the failure to warn of the risk which came to pass was not one. Accordingly, that aspect of the negligence of the respondent did not bear that causal relationship with the outcome sufficient to warrant imposing on him responsibility for that harm. That conclusion is at least consistent with the language of s 5D(1)(a). Alternatively, if the exclusion of something which otherwise qualifies as a necessary condition of the injury depends on a normative judgment, then it is appropriately excluded in the present case, pursuant to s 5D(1)(b) and (4). 176The general law authority (limited as it is) weighs in favour of the conclusion reached by the trial judge, namely that relevant negligence had not been established. 177If the weight of general law authority, to be applied pursuant to the statutory scheme, were not against a finding of liability in the present case, it would be necessary to consider whether legal principle warranted an extension of existing authority to cover the case. There are significant reasons for this Court declining such an expansion of the area of liability. 178As explained by Kirby J in Rosenberg at [143], there is an extensive literature suggesting that the law in respect of this aspect of the patient-practitioner relationship has tended to part company with reality, in the sense that the assumptions upon which the "duty to warn" is based are fictitious or at least exaggerated. There is empirical evidence that many patients do not want and do not know how to cope with the uncertainty created by such warnings. Further, many patients go to a practitioner for skilled advice, not to be given a smorgasbord of considerations they are ill-equipped to assess. (These and related issues are exposed and analysed with clarity by Gerd Gigerenzer in Reckoning with Risk (Penguin, 2002) at pp 100-101.) The law rarely asks whether a warning should be given in a particular form and how it would be understood by the patient. For example, if a practitioner advising a particular procedure says, "the medical literature reveals that this procedure is associated with adverse consequence 'x' at the rate of 1%, but I have done the procedure 1,000 times, with no case of adverse consequences", the patient might infer that this practitioner is particularly expert, or that she is facing a string of adverse outcomes in the near future. Without attention to the precise form, content and context in relation to particular risks, the imposition of legal liability will not further the purposes underpinning the law. 179In Rosenberg, Kirby J supported the obligation mandated by Rogers on a number of bases at [145]. Whether those considerations would favour an expansion of the current limit of liability is not a matter which can properly be resolved by this Court, but should properly be addressed by the High Court, or the Parliament. 180The trial judge undertook a careful analysis of the evidence and the law and dismissed as irrelevant the possible failure to warn of the risk of a catastrophic outcome (as to which there was conflicting evidence). He was correct to do so. The appeal should be dismissed with costs. 181Since writing the above, I have had the opportunity to read the reasons of the President. To the extent to which our reasoning is consistent, I agree with his more expansive analysis of the issues. To the extent we differ, primarily as to whether s 5D demands a rigid separation of physical and policy considerations in assessing causation, I would accept his alternative analysis, if thought to be preferable. It does not affect the result in the present case.

DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated. Decision last updated: 13 April 2012