IpsofactoJ.com: International Casses [2001] Part 5 Case 5 [HCA]


HIGH COURT OF AUSTRALIA

Coram

Rosenberg

- vs -

Percival

GLEESON CJ

McHUGH J

GUMMOW J

KIRBY J

CALLINAN J

5 APRIL 2001


Judgment

Gleeson CJ

  1. The facts of the case are set out in the reasons for judgment of Callinan J.

  2. Two main issues were argued in this Court. The first is whether the appellant was in breach of his duty of care to his patient, the respondent, when he failed to bring to her notice the risk of a certain kind of harm she might suffer in consequence of surgery he undertook to perform. The second is whether, if there had been such a breach of duty, it was causally related to the respondent's injuries. That in turn involved the question whether, if she had been made aware of the risk, the respondent would have decided not to undergo the surgery. The trial judge found against the respondent on both issues. The Full Court of the Supreme Court of Western Australia reversed the first finding, and ordered a re-trial on the second ([1999] WASCA 31).

  3. I agree with Callinan J that the appellant is entitled to succeed on the second issue, and that the Full Court was in error in over-ruling the trial judge's findings on causation. It is therefore unnecessary to decide the first issue. However, since the two issues are related, it is convenient to make some comments about the first matter, by way of introduction to the second.

  4. The case was conducted at trial, in the Full Court, and in this Court, upon the basis that the decision of this Court in Rogers v Whitaker (1992) 175 CLR 479 established the nature of the appellant's duty of care, as distinct from its practical content in the circumstances of the particular case. It is important to note what Rogers v Whitaker decided; and what it did not decide.

  5. The case concerned a patient who undertook elective ophthalmic surgery to the right eye. When considering whether to have the surgery, she questioned her doctor closely about possible complications, including possible damage to her left eye. There was a remote risk, of which she was not told, that the operation to the right eye could affect her left eye. The risk eventuated. She was left totally blind. She brought an action for negligence on the basis of a failure to warn. She succeeded at trial, and in the Court of Appeal of New South Wales, and in this Court. None of the judges who considered the matter found in favour of the doctor. The principal issue was whether the doctor should have informed the patient of the risk. The surgery was elective. The outcome was catastrophic. The patient had questioned the doctor incessantly about risks. The doctor had not told her of the risk which eventuated. A conclusion that he should have done so was hardly a startling result. By the time the case came to this Court, the doctor's case was argued on a narrow basis. The trial judge had found that there were two responsible but different bodies of medical opinion as to whether the patient should have been warned. The doctor's argument was that, once it was found that there was a responsible body of professional opinion which supported the view that a warning was not required, that was the end of the matter (See the argument of counsel (1992) 175 CLR 479 at 480). It was not open to the trial judge to choose between two bodies of professional opinion. The existence of a body of professional opinion supporting the defendant was conclusive in his favour. That is the argument that was rejected.

  6. Rejection of the doctor's argument involved deciding not to follow the English decisions of Bolam v Friern Hospital Management Committee [1957] 1 WLR 582; [1957] 2 All ER 118, and Sidaway v Governors of Bethlem Royal Hospital [1985] AC 871. This Court preferred the approach of Lord Scarman, who dissented in Sidaway, and who said (at 876):

    In my view the question whether or not the omission to warn constitutes a breach of the doctor's duty of care towards his patient is to be determined not exclusively by reference to the current state of responsible and competent professional opinion and practice at the time, though both are, of course, relevant considerations, but by the court's view as to whether the doctor in advising his patient gave the consideration which the law requires him to give to the right of the patient to make up her own mind in the light of the relevant information whether or not she will accept the treatment which he proposes.

    (emphasis added)

  7. As the above passage, which was quoted with approval in Rogers v Whitaker, makes clear, the relevance of professional practice and opinion was not denied; what was denied was its conclusiveness. In many cases, professional practice and opinion will be the primary, and in some cases it may be the only, basis upon which a court may reasonably act. But, in an action brought by a patient, the responsibility for deciding the content of the doctor's duty of care rests with the court, not with his or her professional colleagues.

  8. In Rogers v Whitaker, once the issues of duty, and breach of duty, had been resolved, there was no serious question of causation. The trial judge had accepted that, if the patient had been warned of the risk of what ultimately occurred, she would not have undergone the surgery. The patient had shown anxious concern about the possible risks. The trial judge's finding on that factual issue was affirmed in the Court of Appeal (Rogers v Whitaker (1991) 23 NSWLR 600 at 608, 618-619). Causation was not in dispute by the time the matter reached this Court.

  9. Although this Court came to the same conclusion about the duty of care as had been reached in North American cases, the Court warned against uncritical use of certain concepts, which ought to be valuable currency, but which are susceptible to rhetorical inflation. Mason CJ, Brennan, Dawson, Toohey and McHugh JJ said ((1992) 175 CLR 479 at 490):

    In this context, nothing is to be gained by reiterating the expressions used in American authorities, such as 'the patient's right of self-determination' or even the oft-used and somewhat amorphous phrase 'informed consent'. The right of self-determination is an expression which is, perhaps, suitable to cases where the issue is whether a person has agreed to the general surgical procedure or treatment, but is of little assistance in the balancing process that is involved in the determination of whether there has been a breach of the duty of disclosure. Likewise, the phrase 'informed consent' is apt to mislead as it suggests a test of the validity of a patient's consent. Moreover, consent is relevant to actions framed in trespass, not in negligence. 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. In Reibl v Hughes the Supreme Court of Canada was cautious in its use of the term 'informed consent'.

  10. Having rejected the Bolam principle, the Court went on to formulate the test to be applied in determining whether there was a breach of duty. However, in the light of the way the case had been argued, and in the light of the facts of the case, it was sufficient to do so in general terms.

  11. The joint judgment continued ((1992) 175 CLR 479 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.

  12. The trial judge had found that the case fell within the second of the two alternative tests of materiality, but not the first. Having regard to the patient's conduct before surgery, there was no challenge in this Court to that finding.

  13. In the circumstances of that case, it was unnecessary for the Court to explore some of the potential factual difficulties in the concept of a duty to warn about a material risk inherent in the proposed treatment. What was meant by materiality was explained; but the practical application, in a given case, of the concepts of warning, and risk, may raise difficult issues for judgment.

  14. 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.

  15. That is the difficulty which confronted the respondent in the present case. The adverse contingency which resulted in her post-operative problems was remote. The respondent failed to prove that her problems resulted from incompetence in the manner in which the surgery was performed. In the alternative, she sought to make out a case based on a failure to warn of the contingency. Expert medical opinion at the trial was divided on the degree of remoteness of the contingency, and the need to warn about it. However, as the evidence emerged, it became clear that there was a serious question whether, even if the contingency had been brought to the attention of the respondent, it might have been expected to make any difference to her decision. This question was then addressed by the respondent in evidence, belatedly, and, in the opinion of the trial judge, unconvincingly.

  16. There is an aspect of such a question which may form an important part of the context in which a trial judge considers the issue of causation. In the way in which litigation proceeds, the conduct of the parties is seen through the prism of hindsight. A foreseeable risk has eventuated, and harm has resulted. The particular risk becomes the focus of attention. But at the time of the allegedly tortious conduct, there may have been no reason to single it out from a number of adverse contingencies, or to attach to it the significance it later assumed. Recent judgments in this Court have drawn attention to the danger of a failure, after the event, to take account of the context, before or at the time of the event, in which a contingency was to be evaluated[1]. This danger may be of particular significance where the alleged breach of duty of care is a failure to warn about the possible risks associated with a course of action, where there were, at the time, strong reasons in favour of pursuing the course of action.

  17. The trial judge's findings on the issue of causation did not depend solely upon the adverse opinion he formed as to the respondent's credibility, although that was important. He also took into account the seriousness of her need for corrective surgery, her evident willingness to undergo the risks of a general anaesthetic, with which she was familiar by reason of her professional background, her failure to ask specific questions about risk, and the fact that the possibility of which, on her case, she should have been warned, was "very slight". The conclusion that the respondent had not established that, if her attention had been drawn to the risk in question, she would not have gone ahead with the surgery, was justified by the evidence and supported by cogent reasons. It should not have been overturned by the Full Court.

  18. I would allow the appeal with costs, set aside the orders made by the Full Court, and order that the appeal to the Full Court be dismissed with costs.

    McHugh J

  19. The Full Court of the Supreme Court of Western Australia (Kennedy, Wallwork and Owen JJ) has ordered a new trial of an action for damages for breach of duty owed by a doctor to a patient. The breach alleged is the doctor's failure to warn the patient that, if she underwent a form of oral surgery, known as an osteotomy, she could suffer a temporomandibular joint disorder. The trial judge found that there was no duty to warn the patient of this risk. He also found that, even if the patient had been warned of the risk, she would have undergone the surgery. The Full Court took a different view. It held that the doctor owed a duty to the patient to warn her of the risk, that he had breached that duty and that the trial judge had erred in finding that the patient would have undergone the surgery even if warned. Accordingly, the Full Court allowed the appeal and made a declaration that the doctor was negligent. It ordered that the action be remitted to the District Court of Western Australia for a new trial to determine "whether the [patient] has suffered loss or damage which has been caused by the [doctor's] breach of duties" ("the causation issue") and, if so, "the amount of the loss and damage suffered".

  20. The issues in this appeal are whether the Full Court erred in setting aside the findings of the trial judge on the issues of breach of duty and causation.

  21. The action was tried in the District Court before Gunning DCJ. The facts of the case are set out in the judgment of Callinan J. Because of the view that I take of the causation issue, it is unnecessary to refer to them in any detail.

  22. In finding that the patient would have consented to the surgery even if she had been warned of the risk, Gunning DCJ said:

    I am quite satisfied in the circumstances that even if the plaintiff had been warned of the slight possibility, and certainly it was very slight, of complications she would have proceeded with the surgery in any event.

  23. Whether or not the Full Court was correct in reversing the trial judge on the issue of breach of duty, it erred in overturning his Honour's finding that the patient "would have proceeded with the surgery" even if she had been warned of the risk of temporomandibular joint disorder and its consequences.

    THE TEST FOR CAUSATION WHERE THERE HAS BEEN A FAILURE TO WARN A PATIENT OF RISKS

  24. Under the Australian common law, in determining whether a patient would have undertaken surgery, if warned of a risk of harm involved in that surgery, a court asks whether this patient would have undertaken the surgery. The test is a subjective test[2]. It is not decisive that a reasonable person would or would not have undertaken the surgery. What a reasonable person would or would not have done in the patient's circumstances will almost always be the most important factor in determining whether the court will accept or reject the patient's evidence as to the course that the patient would have taken. But what a reasonable person would have done is not conclusive. If the tribunal of fact, be it judge or jury, accepts the evidence of the patient as to what he or she would have done, then, subject to appellate review as to the correctness of that finding, that is the end of the matter. Unlike other common law jurisdictions[3], in this field Australia has rejected the objective test of causation in favour of a subjective test.

  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. In exceptional cases, the judge may even reject the patient's testimony as not credible and then infer from the objective facts that the patient would not have proceeded. The judge might find, for example, that the patient was a person whose general credibility was so poor that no reliance could be placed on that person's oral evidence. Yet, notwithstanding the rejection of the patient's oral testimony, the judge might infer that nevertheless this patient would not have undergone the procedure. That inference would ordinarily be based not only on the objective facts but also on the tribunal's assessment of the general character and personality of the patient.

    APPELLATE REVIEW OF THE CAUSATION ISSUE WHERE THERE HAS BEEN A FAILURE TO WARN

  26. When the tribunal of fact has accepted the patient's evidence that he or she would not have undertaken the surgery, an appellate court will often be in a stronger position to reverse that finding than when the tribunal of fact has rejected the patient's evidence. That is because, human nature being what it is, most persons who suffer harm as the result of a medical procedure and sue for damages genuinely believe that they would not have undertaken the procedure, if they had been warned of the risk of that harm. Thus, the demeanour of the patient in the witness box will often not restrain appellate review of this finding of fact to the extent that it does in respect of fact finding in many other appeals[4]. Nevertheless, where the tribunal of fact has accepted the evidence of the patient, an appellate court cannot entirely disregard the "subtle influence of demeanour"[5] on the tribunal of fact.

  27. When the tribunal of fact has rejected the patient's evidence that he or she would not have proceeded with the surgery, however, the ordinary restrictions on appellate review of fact finding apply. If the tribunal of fact is a judge, as in the present case, an appellate court must respect the advantage that the judge has had over the appellate court in seeing and hearing the patient give evidence. Ordinarily, the appellate court cannot reverse the finding of the judge unless it is satisfied "that any advantage enjoyed by the trial judge by reason of having seen and heard the witnesses, could not be sufficient to explain or justify the trial judge's conclusion"[6]. Unless that condition is satisfied or the judge has misdirected himself or herself or has misapprehended the evidence or has indicated that the demeanour of the patient played no part in the finding, the appellate court cannot reverse it. These restrictions on appellate review also apply when the trial judge makes a positive finding that the patient would have undertaken the procedure if warned of the relevant risks.

    FINDINGS ON CAUSATION IN THE PRESENT CASE

  28. In the present case, the trial judge took a very adverse view of the patient's credibility on a number of issues including the causation issue. In evidence, the patient said: "If there had been any risk I would not have had the surgery." Gunning DCJ described her evidence as to what she would have done "had she been advised of the slightest possibility of complications with her temporomandibular joint .... as unbelievable". But the judge went further than rejecting her evidence as to what she would have done if she had been warned of the risk. He found positively that she would have proceeded with the surgery. That finding was no doubt influenced by the judge's assessment of the patient. Absent some misdirection or misapprehension that undermined the reliability of the judge's assessment of the patient's credibility, character and personality, his findings placed an insurmountable hurdle in the way of the Full Court reversing him on the causation issue and substituting its own finding on that issue. Correctly, the Full Court did not substitute its own finding on that issue for that of the trial judge. Instead, perceiving error in the trial judge's approach to the factual assessment of the causation issue, the Full Court ordered a re-trial of that issue.

  29. Wallwork J thought that Gunning DCJ had erred in not determining the reasons why the patient lacked credibility. In his reasons for judgment, Wallwork J said that the patient's counsel was correct in submitting "that in coming to an adverse view of the [patient's] credibility the learned trial Judge did not take into account the important matter of the severe effects which the complications have had upon the [patient], both physically and mentally". Wallwork J also said that counsel for the patient had "made the valid point that had the [patient] been properly warned .... she may not have been carrying the obvious depression, the chronic anxiety and the collapse of her career which had occurred." Building on the acceptance of these submissions, Wallwork J commented that the patient "may not have been as prone to make [the] alleged extravagant statements" that Gunning DCJ had found she made.

  30. With great respect to Wallwork J, he should not have accepted these submissions of counsel. The trial judge had to determine whether the patient should be believed when she asserted that she would not have proceeded with the surgery, if she had been warned of the risk. The reasons for the patient's lack of credibility were irrelevant. Wallwork J said that "the reasons for judgment [of Gunning DCJ] should have identified and articulated the actual disabilities which the [patient] had and attempted to dissect how they were contributing to her performance and behaviour." But dissecting how the patient's disabilities contributed to her performance in the witness box could not improve her credibility. Nor could it assist the judge in determining whether to accept her evidence. The patient's credibility could not be improved by determining the cause or causes that led to her lack of credibility. Whether that failing arose from the consequences of the surgery, the desire to win her case or otherwise, the result was the same: her evidence on the causation issue was unacceptable.

  31. Moreover, in the circumstances of this case, the patient could not succeed on the causation issue unless her evidence concerning that issue was credible. The objective facts in her favour were not so strong that a court could infer from them that, if warned, she would not have proceeded with the surgery even though the court found that no reliance whatsoever could be placed on her oral testimony. All members of the Full Court implicitly recognised that this was so by sending the causation issue for a re-trial instead of making their own findings on that issue.

  32. No doubt it was for this reason that Wallwork J did not suggest that, even if the patient's evidence on the causation issue was rejected, it was still open to the trial judge to find that she would not have had the surgery. His Honour merely said that the judge had "rejected the proposition that the [patient] would have not had the surgery had she been warned of the possible risks." That is true. But the trial judge did more than reject this proposition. He also made a positive finding that

    even if the [patient] had been warned of the slight possibility .... she would have proceeded with the surgery in any event.

    Given the evidence and the advantage that he had of seeing and hearing the patient give evidence, this finding was also open to him.

  33. The finding was the product of a number of matters. They included:

    • the rejection of the patient's testimony on the causation issue;

    • the assessment of the patient's character and personality;

    • the 20 years experience that the patient had had as a qualified nurse with a doctorate of philosophy in nursing and a senior lectureship in nursing at a university;

    • the patient knowing that surgical operations carry inherent risks of harm;

    • the patient suffering from a worsening condition of malocclusion for a number of years;

    • the consulting of several specialists for the purpose of remedying the condition and getting the best result;

    • the osteotomy procedure being the operation most likely to produce the best result in her case;

    • the osteotomy being a common operation;

    • the risk of suffering the harm that the patient suffered being very small; and

    • the patient subsequently undergoing another operation to correct the consequences of the temporomandibular joint disorder.

    Some of these matters were dealt with expressly by Gunning DCJ; others were necessarily involved in the judge's determination or, at all events, are not shown to have had no influence in the making of his finding.

  34. Owen J agreed with the reasons of Wallwork J on the causation issue, saying that there was nothing that he could usefully add to the examination by Wallwork J. Owen J also agreed that the action "should be remitted to the District Court for further consideration in the manner and to the extent suggested by Wallwork J."

  35. The approach of Kennedy J to the causation issue differed sharply from that of Wallwork and Owen JJ. Kennedy J referred to a number of matters that supported the probability that the patient would not have proceeded with the surgery if she had been warned of the risk of temporomandibular joint disorder.

    His Honour went on to say:

    With respect, his Honour's consideration of these important aspects was simply dismissive to the extent that any consideration was given to them, although each required careful attention. Any adverse finding on this issue should not have been made, as it appears to have been, essentially on the basis of the credibility of the [patient].

  36. But Gunning DCJ saw and heard the patient in the witness box. He had the advantage, denied to the judges of the Full Court, of assessing her character, fortitude and intelligence and whether a warning would have deterred her from proceeding with the surgery.

    THE ADVANTAGE OF THE TRIAL JUDGE

  37. As Brennan and Gaudron JJ and I said in Devries v Australian National Railways Commission (1993) 177 CLR 472 at 477, "the trial judge had the advantage, which was denied to the judges of the Full Court, of being able to judge the true character and intelligence of the plaintiff". In addition, as I have pointed out above, there were many objective matters that supported the trial judge's finding that the patient would have had the operation even if the risk had been drawn to her attention. None of the matters referred to by Kennedy J pointed so overwhelmingly to the patient not proceeding with the surgery that they negatived the advantage that Gunning DCJ had in seeing and hearing the patient. Even if the matter is looked at on a purely objective basis, the matters against the patient's claim were as weighty as the matters in her favour, perhaps more so. Once it is accepted that the matters in her favour could not overwhelm either the matters against her or the judge's assessment of her character, fortitude and intelligence, the Full Court's opportunity to review the judge's finding on the causation issue was very limited.

  38. Wherever the boundary of review lay, in the circumstances of this case, the Full Court could not set aside the trial judge's finding on the bare ground that he did not give sufficient weight to matters that the judges of the Full Court thought assisted the patient's case. In Devries (1993) 177 CLR 472 at 479, Brennan and Gaudron JJ and I also said (citations omitted):

    More than once in recent years, this Court has pointed out that a finding of fact by a trial judge, based on the credibility of a witness, is not to be set aside because an appellate court thinks that the probabilities of the case are against - even strongly against - that finding of fact. If the trial judge's finding depends to any substantial degree on the credibility of the witness, the finding must stand unless it can be shown that the trial judge 'has failed to use or has palpably misused his advantage' or has acted on evidence which was 'inconsistent with facts incontrovertibly established by the evidence' or which was 'glaringly improbable'.

  39. To similar effect are remarks of Deane and Dawson JJ in Devries. Their Honours said ((1993) 177 CLR 472 at 482-483) that, "consistently with the obligation to make full allowance for the advantage which the trial judge had enjoyed, the Full Court could properly overturn the trial judge's finding only if it was vitiated by some error of principle or mistake or misapprehension of fact or if the effect of the overall evidence was such that it was not reasonably open" to make the finding that he did.

  40. In this case, the trial judge's finding was based on the credibility of the witness and on facts that were not "inconsistent with facts incontrovertibly established by the evidence" or "glaringly improbable". That being so, it is impossible to conclude that he failed to use or has palpably misused "his advantage" because he did not give to countervailing matters the weight that the Full Court thought they deserved.

  41. One of the consequences of the "advantage" of seeing and hearing the witnesses is that the trial judge is in a far better position than an appellate court to know what individual weight should be assigned to the various factors - credibility, matters for and matters against - that must be evaluated in making the ultimate findings of fact in the case. Where a finding is based on credibility and other facts support the finding, the case would need to be exceptional before an appellate court could set aside the finding on the ground that, judging by the transcript, the trial judge gave insufficient weight or consideration to other facts and circumstances in the case. The common law tradition is an oral tradition. Trial by transcript can seldom be an adequate representation of an oral trial before a judge or an oral trial before a judge and jury.

  42. No suggestion could reasonably be made that Gunning DCJ acted on an erroneous principle or mistook or misapprehended the facts of the case. Nor was the overall effect of the evidence such that it was not reasonably open to find that the patient would have proceeded with the surgery, if she had been warned of the possibility of a temporomandibular joint disorder.

  43. No doubt the trial judge erred in one respect. He regarded the claims of the patient and her husband concerning the deteriorating condition of their garden as dishonestly made. No suggestion of dishonesty in respect of this matter had been put to them in cross-examination. But it is impossible to conclude that this issue played a decisive part in the judge's assessment of the patient's credibility. It was merely one of many matters that led Gunning DCJ to conclude that the patient's testimony was generally unreliable.

    CONCLUSION

  44. Australian law is committed to a subjective test in determining whether a patient would have refused to undergo a medical procedure if that person had been warned of the risk of relevant injury. If the patient is believed, he or she succeeds even though the objective facts point the other way. If the evidence of the patient is rejected, he or she carries the heavy evidentiary burden of persuading the court to make a favourable finding on the causation issue solely by reference to the objective facts and probabilities. Courts frequently make findings as to states of mind based on nothing more than the objective facts and probabilities of the case. But usually such findings refer to the mental state that simultaneously accompanied some act or omission of a person. Thus, courts often have to determine whether an act causing harm was done with intent to kill or inflict grievous bodily harm, whether a representation was made with knowledge of its falsity, whether an omission to account for trust moneys was done with intent to defraud or whether a publication was made maliciously. The very doing of the act or the omitting to do the act, when considered in its context, generally throws much light on the mental state of the person concerned. But these cases are different from the "What if?" situation that arises when the court has to determine what a patient would have done if warned of a risk. In the "What if?" situation, the relevant mental state has not accompanied any act or omission of the patient.

  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. The answer concerning the hypothetical mental state provides the answer to the hypothetical course of action. The onus is on the patient to prove that he or she would have decided not to have the operation if given a warning of the risk of harm. That means that the patient must prove what he or she would have decided to do. When the direct testimony of that person on the causation issue has been rejected, it is unlikely, as a matter of fact, that the patient will succeed on that issue unless the objective evidence in favour of the patient is very strong.

  46. In the present case, as the judgment of Kennedy J shows, a number of matters supported the patient's case. But neither individually nor collectively were they strong enough for the patient to succeed without the trial judge accepting her evidence that she would have decided not to have the operation. Gunning DCJ rejected her evidence and, given the limits of appellate review in respect of findings of fact based, wholly or partly, on the assessment of a witness's credibility, character or personality, it was not open to the Full Court to set aside his finding on the causation issue.

  47. The appeal must be allowed.

    Gummow J

    THE HISTORY OF THE LITIGATION

  48. The appellant is a dental surgeon who on 6 December 1993 performed a surgical procedure upon the respondent at St Anne's Hospital, Mt Lawley in Western Australia. The procedure is known as a sagittal split osteotomy. The respondent subsequently suffered from severe temporomandibular joint ("TMJ") complications; her symptoms include chronic (ie permanent), severe and disabling pain. A further procedure was performed on her at the Glengarry Hospital on 25 February 1994.

  49. The respondent brought an action against the appellant in the District Court of Western Australia. The relationship of medical practitioner and patient is contractual in nature (Breen v Williams (1996) 186 CLR 71 at 102) although statutory schemes now play a part in that relationship. However, in this case, as in most of the recent authorities, the action was brought in tort. The respondent sued in negligence, on two grounds. The first was that the appellant was negligent in performing the operations. The second was that he was negligent in failing to advise her beforehand of risks inherent in the surgery. The action was tried without a jury. The trial judge (Gunning DCJ) found that the appellant conducted both operations with the required skill and care.

    His Honour continued:

    The important issue in this case is the question of whether the [appellant] should have, in the circumstances, warned of any complications that could develop and, if he had done so, whether the [respondent] would have declined to have had the 1993 surgery.

  50. In respect of this claim, Gunning DCJ reviewed the evidence and found:

    It follows that in the circumstances the [appellant], prior to the operation, was not negligent in his interpretation of any of the [respondent's] complaints following a thorough examination, or of his interpretation of the x-rays and therefore it follows that there was no known problem to him that could develop that he could communicate to the [respondent].

    It follows he was not negligent in not warning the [respondent] of any material problem that might develop.

    The next question to be dealt with is that of causation ....

    I am quite satisfied in the circumstances that even if the [respondent] had been warned of the slight possibility, and certainly it was very slight, of complications she would have proceeded with the surgery in any event.

    Accordingly, the respondent's action was dismissed. However, his Honour, for the purposes of taxation of costs, fixed damages in the sum of $350,000.

  51. It is to be emphasised that the trial judge's decision that the failure to warn of the risk was not causative was based, in no small measure, on his Honour's findings on the credibility of the respondent. In assessing the respondent's credibility his Honour stated:

    However, it very rapidly became apparent that the [respondent] was most anxious to tell her story in a way in which she thought would benefit her case and to play down anything that she thought might be to the contrary and at the end of the cross-examination I can only say that this view was reinforced and I was far from satisfied that the [respondent] was a reliable witness.

  52. Counsel for the respondent sought and obtained leave to recall the respondent at the close of the appellant's case for further examination in chief directed to the issue of causation. In those circumstances, the trial judge considered the assertion by the respondent that "[i]f there had been any risk I would not have had the surgery", as of "no evidentiary value whatsoever".

  53. An appeal to the Full Court of the Supreme Court of Western Australia (Kennedy, Wallwork and Owen JJ) was successful. The principal judgment was given by Wallwork J. Kennedy J and Owen J gave further reasons, the former dealing particularly with the findings as to the credit of the respondent. The conclusions of the trial judge were overturned and a new trial was ordered on the issue of causation. The approach taken by the Full Court is best understood from order (3) of its orders. This was as follows:

    (3)

    In lieu of the order of the District Court of 13 November 1997 it be ordered, declared and adjudged that:

    (i)

    the [appellant] was in breach of his duty of care to the [respondent] in failing to warn of the risks of [TMJ] problems and symptoms arising after the procedure which the [respondent] underwent.

    (ii)

    the action be remitted to the District Court of Western Australia for a new trial before a different Judge on the remaining issues, namely:

    (+)

    whether the [respondent] has suffered loss or damage which has been caused by the [appellant's] breach of duties aforesaid;

    (+)

    if so, what is the amount of the loss and damage suffered by the [respondent];

    and for judgment to be entered after the re-trial of those issues accordingly.

    THE APPEAL TO THIS COURT

  54. The appellant seeks from this Court orders which would set aside the Full Court orders and restore his success at trial.

  55. There are two major issues arising out of this appeal. The first is whether the Full Court was correct in holding that the appellant had breached his duty of care ("the materiality issue"); and, secondly, whether the Full Court erred in principle in overturning the trial judge's findings on credibility and in ordering a re-trial of the causation issue.

  56. At first blush, it may appear that, if the second issue were to be decided in the affirmative, the appeal to this Court could be allowed without the need to investigate the materiality issue. However, in this case, the issues of materiality and causation are so interrelated that this is not the proper approach. Observations by Gaudron J in Bennett v Minister of Community Welfare (1992) 176 CLR 408 at 422 indicate the point. Her Honour said "questions of the sufficiency of the precaution to avert the harm are inevitably subsumed in the finding that there was a duty". 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.

  57. Of course, notwithstanding the logical difficulties involved, cases may be conducted on the basis, or may reach this Court on the footing, that materiality or causation but not both are in issue. Rogers v Whitaker (1992) 175 CLR 479 at 492 was fought on the issue of materiality not causation, whereas in Chappel v Hart (1998) 195 CLR 232 the issue was the other way around. Here, as indicated, both issues are involved.

  58. The appellant submits to this Court that the Full Court failed properly to identify the risk in question and had erred in determining its materiality. He further submits that the Full Court erred in principle in overturning the findings of the trial judge concerning the credibility of the respondent's evidence. These submissions should be accepted and the appeal allowed.

  59. The path to be followed in these reasons to reach that conclusion involves consideration of the identification of the risk to the patient, the materiality of that risk, and the causation of the injury she suffered.

    THE IDENTIFICATION OF THE RISK

  60. It is established by Rogers (1992) 175 CLR 479 at 490 that a medical practitioner owes a duty "to warn a patient of a material risk inherent in the proposed treatment". However, that proposition in turn poses further questions.

  61. The first question is "what 'risk' is being spoken of here?" Put another way, it is "what are the facts and circumstances, the possibility of the occurrence of which constitutes that 'risk'?" Once that question is answered one may turn to consider whether the risk is "material". 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[7]. It is not, for example, the risk that the patient will make an uninformed decision or choose the wrong option, although that may well underpin the rationale behind the duty (see Rogers v Whitaker (1992) 175 CLR 479 at 489).

  62. It is important to understand the decision in Rogers in the context in which it was decided. Before Rogers dealt with the law in Australia, the law in England was that evidence of medical practice was the sole determinant of the appropriate standard. Thus, if evidence that it was the practice of a respected body of medical practitioners not to give a warning in the circumstances of the case were accepted by the tribunal of fact, any such failure to warn would not be negligent. This approach was described as the Bolam[8] test. The decision of this Court in Rogers rejected the Bolam test. The Court held that the standard to be observed by medical practitioners was not to be determined solely or even primarily by medical practice. Rather, it was for the courts to judge what standard should be expected from the medical profession. In the joint judgment that standard was identified and fixed. Hence the statement by Gaudron and McHugh JJ in Breen v Williams (1996) 186 CLR 71 at 114 that:

    Rogers took away from the medical profession in this country the right to determine, in proceedings for negligence, what amounts to acceptable medical standards.

  63. The structure and sequence of their Honours' reasoning can be understood from the following six passages (1992) 175 CLR 479 :

    (i)

    In Australia, it has been accepted that the standard of care to be observed by a person with some special skill or competence is that of the ordinary skilled person exercising and professing to have that special skill. But, that standard is not determined solely or even primarily by reference to the practice followed or supported by a responsible body of opinion in the relevant profession or trade.

    [at 487 (footnotes omitted)]

    (ii)

    [I]t is for the courts to adjudicate on what is the appropriate standard of care after giving weight to 'the paramount consideration that a person is entitled to make his own decisions about his life'.

    [at 487 (footnote omitted)]

    (iii)

    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. However, the factors according to which a court determines whether a medical practitioner is in breach of the requisite standard of care will vary according to whether it is a case involving diagnosis, treatment or the provision of information or advice; the different cases raise varying difficulties which require consideration of different factors.

    [at 489 (footnote omitted)]

    (iv)

    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. In diagnosis and treatment, the patient's contribution is limited to the narration of symptoms and relevant history; the medical practitioner provides diagnosis and treatment according to his or her level of skill. However, except in cases of emergency or necessity, all medical treatment is preceded by the patient's choice to undergo it. In legal terms, the patient's consent to the treatment may be valid once he or she is informed in broad terms of the nature of the procedure which is intended. But the choice is, in reality, meaningless unless it is made on the basis of relevant information and advice.

    [at 489 (footnote omitted)]

    (v)

    Whether a medical practitioner carries out a particular form of treatment in accordance with the appropriate standard of care is a question in the resolution of which responsible professional opinion will have an influential, often a decisive, role to play; whether the patient has been given all the relevant information to choose between undergoing and not undergoing the treatment is a question of a different order. Generally speaking, it is not a question the answer to which depends upon medical standards or practices.

    [at 489-490 (original emphasis)]

    (vi)

    We agree that the factors referred to in F v R (1983) 33 SASR 189 at 192-193 by King CJ must all be considered by a medical practitioner in deciding whether to disclose or advise of some risk in a proposed procedure. 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. This duty is subject to the therapeutic privilege.

    [at 490 (emphasis added)]

    Thus, the Court, in (vi), sets the standard that the law demands of medical practitioners in relation to the provision of information. This standard does not deal with the foreseeability of the risk in question, save to the extent that the risk must be "inherent" in the procedure. In this respect the general law of negligence still applies. Support for this view can be found in the concurring judgment of Gaudron J. Her Honour stated that a real and foreseeable risk was required to found a duty to warn and, further, saw "no basis for treating the doctor's duty to warn of risks .... as different in nature or degree from any other duty to warn of real and foreseeable risks.

    [at 494]

  64. A risk is real and foreseeable if it is not far-fetched or fanciful, even if it is extremely unlikely to occur (Wyong Shire Council v Shirt (1980) 146 CLR 40 at 48). The precise and particular character of the injury or the precise sequence of events leading to the injury need not be foreseeable. It is sufficient if the kind or type of injury was foreseeable[9], even if the extent of the injury was greater than expected. Thus, in Hughes v Lord Advocate [1963] AC 837, there was liability because injury by fire was foreseeable, even though the explosion that actually occurred was not.

  65. A misunderstanding as to what is involved in the notion of risk in cases such as this case may lead to an exaggerated view as to the nature of the burden imposed upon defendants. That appears to have infected some of the reasoning of the Full Court in this case. Wallwork J said:

    The question of the risk in this case was not dependent on it being known to the operating surgeon or the other experts concerned that there were pre-operative signs of a [TMJ] disorder. On the overwhelming evidence it was a risk which existed because of the likelihood of some patients having complications after the operative procedure. Once there is a risk which is generally known to the profession, there is a duty to warn.

  66. What is lacking here is an identification of the "risk" with which the court should be concerned. For example, there is no reference to the severity of the potential injury or the likelihood of its occurrence. Indeed, in the present case it appears that the risk which the Full Court had in mind was a general risk of the occurrence of "TMJ complications". However, such a broad definition is of little use in determining either materiality or causation. Nor did the Full Court identify or give content to the notion of materiality, a matter dealt with later in these reasons.

  67. One of the factors relevant to, but not decisive of, the question of what a reasonable medical practitioner ought to have foreseen is the state of medical knowledge at the time when the duty should have been performed (Rogers v Whitaker (1992) 175 CLR 479 at 486-487, 492-493). A reasonable medical practitioner cannot be expected to have foreseen an event wholly uncomprehended by medical knowledge at the time.

  68. This reflects the fundamental proposition that the law demands no more than what was reasonable in all the circumstances of the case. In the words of Barwick CJ in Maloney v Commissioner for Railways (NSW) (1978) 52 ALJR 292 at 292; 18 ALR 147 at 148:

    [T]he respondent's duty was to take reasonable care .... It is easy to overlook the all important emphasis upon the word 'reasonable' in the statement of the duty. Perfection or the use of increased knowledge or experience embraced in hindsight after the event should form no part of the components of what is reasonable in all the circumstances.

    (original emphasis)

    More recently, it was said in Perre v Apand Pty Ltd (1999) 198 CLR 180 at 249 [186] (footnote omitted) that:

    it needs to be kept in mind .... that the criterion is 'reasonable foreseeability'. Liability is to be imposed for consequences which [the respondent], judged by the standard of the reasonable man, ought to have foreseen.

    (original emphasis)

  69. With this in mind, the first step must be to define the relevant risk. It is appropriate in this context to define the risk by reference to the circumstances in which the injury can occur, the likelihood of the injury occurring, and the extent or severity of the potential injury if it does occur. These factors are to be considered from the point of view of what a reasonable medical practitioner in the position of the defendant ought to have foreseen at the time. This approach directs attention to the content of any warning that could have been given at the time.

  70. In this case, as a result of the osteotomy performed by the appellant, the respondent suffered extremely severe TMJ complications, including chronic disabling pain in the jaw. In 1993, the relevant time, it was known that an osteotomy could aggravate a pre-existing TMJ problem. However, the literature on the subject was equivocal as to the likelihood and potential severity of such complications. Such severe symptoms were certainly not widely known to result from osteotomy procedures. The only evidence concerning knowledge of such a severe occurrence was given by Professor Goss, who had seen one prior case where an osteotomy had aggravated a pre-existing TMJ condition to the same level as the present case.

  71. The trial judge found that neither the specialist dentists nor the appellant found that the respondent presented any TMJ problem prior to the osteotomy. However, the appellant admitted in cross-examination that he had detected signs of what he took to be a minor TMJ disorder. The appellant also admitted that in 1993 he appreciated that "such subtle non-painful subclinical [TMJ disorder] could possibly be aggravated by a bilateral sagittal osteotomy". The trial judge accepted the appellant's evidence that he was not aware of any patient suffering chronic debilitating pain as a result of an osteotomy.

  72. From the facts as found at trial, it does not follow that a reasonable practitioner ought to have foreseen that the osteotomy could lead to a TMJ problem manifesting the severe symptoms that the respondent suffered. However, that is not the end of the inquiry; a reasonable practitioner should have foreseen the risk of some kind of TMJ complications. This was the basis on which the Full Court proceeded. As has been pointed out, the error of the Full Court was in failing to identify the content of the risk. The broad umbrella of "TMJ complications" does not assist the Court in reaching a decision on materiality and causation.

  73. The problem is in identifying with some precision from the evidence the nature and severity of the complications that should have been foreseen. The appellant gave evidence that in his experience about 10 per cent of patients suffer from some sort of TMJ complications. For about half of those, the symptoms would involve temporary pain in the joints. Others would experience some jaw movement difficulties that respond to conservative treatment, while a few might experience more serious problems requiring referral to a specialist. Clinical features or symptoms of TMJ disorders were known to include pain / tenderness in the muscles of mastication, pain/tenderness in the TMJ, TMJ noises, limitation of jaw movement and incoordination/deviation of jaw movement.

  74. Therefore, a reasonable practitioner in 1993 could only be expected to give a warning containing reference to the following: TMJ problems are known to occur and be aggravated by this procedure (ie osteotomy); the likelihood of such problems developing is about 10 per cent; the likely symptoms are as listed above and any such symptoms are likely to be temporary and non-serious in nature. This identification of the risk is to be used in determining both its materiality and whether failure to warn of it was causative.

    MATERIAL RISK

  75. The next question is whether the risk, in the above sense, was "material". Under the Rogers test set out in par (vi) above, a risk is material if ((1992) 175 CLR 479 at 490):

    1. in the circumstances of the case, a reasonable person in the patient's position would be likely to attach significance to it ("the objective limb"); or

    2. the medical practitioner was, or should have been, aware that the particular patient would be likely to attach significance to it ("the subjective limb").

  76. Both Rogers and Chappel v Hart were concerned with the subjective limb; in this case, it is the objective limb. However, what is of immediate importance is that the key to the formulation in both limbs is the phrase "likely to attach significance to". Some of the difficulty which appears to be encountered in the application of the passage in par (vi) may stem from a failure to read what is said there with the judgment of King CJ in F v R, with which the joint judgment in Rogers expressly agreed in that very passage. In F v R, the Chief Justice of South Australia discussed five factors that are useful in determining whether a reasonable patient or the particular patient respectively, would be "likely to attach significance to" the risk. This discussion merits careful attention. However, by way of summary, it can be seen that these factors are in effect an adaptation of the criteria set out by Mason J in Wyong Shire Council v Shirt (1980) 146 CLR 40 at 47-48 for determining breach of duty. Those criteria were "the magnitude of the risk and the degree of the probability of its occurrence", balanced against "the expense, difficulty and inconvenience of taking alleviating action and any other conflicting responsibilities which the defendant may have".

  77. Clearly, in cases of a medical practitioner's failure to warn, the extent or severity of the potential injury is of great importance in applying the test in Rogers of "likely to attach significance to", as is the likelihood of the injury actually occurring. These two matters, the extent or severity of the potential injury and the likelihood of it coming to pass, are to be considered together. A slight risk of a serious harm might satisfy the test, while a greater risk of a small harm might not. It is also important to note that, in considering the severity of the potential injury, that severity is judged with reference to the plaintiff's position. Thus, the risk of blindness in one eye would ordinarily be considered serious; if however, as in Rogers, the patient is already blind in one eye and stands to lose sight entirely, that risk becomes one of an altogether greater magnitude.

  78. These considerations need to be weighed against the circumstances of the patient. The patient's need for the operation is important, as is the existence of reasonably available and satisfactory alternative treatments. A patient may be more likely to attach significance to a risk if the procedure is elective rather than life saving. As will be seen, these factors merge with the issue of causation.

  79. The second, or subjective, limb of the test in Rogers for material risk requires further discussion. The second limb recognises that the particular patient may not be a "reasonable" one; he or she may have a number of "unreasonable" fears or concerns. These will be given full weight under the second limb if the medical practitioner was or should have been aware of them. One way of satisfying that condition is if the patient asked questions revealing the fear or concern. However, that is not the only means of satisfying the second limb. There are a multitude of potential circumstances in which a court might find that the medical practitioner should have known of a particular fear or concern held by the patient. Courts should not be too quick to discard the second limb merely because it emerges that the patient did not ask certain kinds of questions.

  80. The phrase "likely to attach significance to" as it appears in both limbs does not present a threshold issue of the same nature as that presented by the issue of causation. In the authorities, reference has been made to "information that is relevant to a decision or course of action" (Rogers v Whitaker (1992) 175 CLR 479 at 494 per Gaudron J) and "matters which might influence the [decision]" (F v R (1983) 33 SASR 189 at 192 per King CJ). It is not necessary when determining materiality of risk to establish that the patient, reasonable or otherwise, would not have had the treatment had he or she been warned of the risk in question. The test is somewhat lower than that. However, it is necessary that the reasonable patient or particular patient respectively would have been likely seriously to consider and weigh up the risk before reaching a decision on whether to proceed with the treatment. The authorities referred to above should be read in that way.

  81. This case does not fall within the second, or subjective, limb of the Rogers test. The respondent did not ask questions identifying a particular area of concern and there is no indication of any relevant physical or mental characteristics peculiar to the respondent of which the appellant should have been aware. Therefore, the test of whether the risk was material falls to be determined by reference to the first, or objective, limb and thus to the reasonable person in the patient's position. From the findings of fact made at trial, the relevant factors to consider include: the temporary nature of any potential harm; the possibility of some pain, but not severe pain; the 10 per cent chance of such complications arising; the fact that the respondent was concerned about her malocclusion, she wanted the "best result" and the osteotomy was the most effective way to remedy the problem; the respondent was an experienced and knowledgeable nurse who was certainly aware that all surgery carried some risk and the fact that the respondent had received advice from a number of sources indicating that she should proceed with the treatment.

  82. In light of these factors, it was open to the trial judge to conclude that a reasonable person in the respondent's position would not, in the sense of Rogers, be "likely to attach significance to" the risk. Therefore the risk was not a material risk.

    CAUSATION

  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" (See Chappel v Hart (1998) 195 CLR 232 at 257 [66]). This can be seen as an example of a situation where the application of the "but for" test would lead to an unacceptable result (March v Stramare (E&MH) Pty Ltd (1991) 171 CLR 506 at 516).

  84. At the second level, there must be a causal connection, in the legal sense, between the failure to warn of the material risk and the occurrence of the injury. Cases involving a failure to warn of a risk encounter difficulties of causation that do not arise in cases of, for example, a negligent physical act "causing" injury. The failure to warn the patient of the risk can never amount in the same sense to the cause of the injury. Moreover, the issue of failure to warn usually arises when the performance of the physical cause of the injury was not negligent. Indeed, the present is such a case, given the finding that the appellant conducted both operations upon the respondent with the required skill and care.

  85. It is well understood that the legal concept of causation differs from notions of causation which appear in the speculations of philosophers and the perceptions by scientists of the operation of natural laws. This is because the legal concept of causation is primarily concerned with attributing responsibility. It has been said that the test of causation is one of common sense (March v Stramare (E&MH) Pty Ltd (1991) 171 CLR 506), but, as Lord Hoffmann has observed[10]:

    [C]ommon sense answers to questions of causation will differ according to the purpose for which the question is asked. Questions of causation often arise for the purpose of attributing responsibility to someone, for example, so as to blame him for something which has happened or to make him guilty of an offence or liable in damages. In such cases, the answer will depend upon the rule by which responsibility is being attributed.

    Again, in Chappel v Hart (1998) 195 CLR 232 at 238 [7], Gaudron J said:

    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.

  86. 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. In Bennett v Minister of Community Welfare (1992) 176 CLR 408 at 420 (footnote omitted), Gaudron J put the matter this way[11]:

    [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'.

    From this consideration, it can be seen that 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 been given, the injury would have been averted, in the sense that the relevant "patient" would not have had the treatment in question.

  87. In Australia the relevant "patient" for the purposes of the second criterion is the particular patient and, in that sense, the criterion is a subjective one[12]. The question is whether the particular patient would not have had the treatment had a warning been given. This subjective criterion, it has been recognised, involves practical questions of proof. The court must deal with hypothetical considerations as to what the patient in question would have done had a warning been given. In Gover v State of South Australia & Perriam (1985) 39 SASR 543 at 566, Cox J said:

    The court has to reach a decision about a topic to which the patient, in most cases, will not have addressed his mind at the time that matters most. His evidence as to what he would have done is therefore hypothetical and is very likely to be affected, no matter how honest he is, by his own particular experience.

    His Honour added (at 566):

    It will often be very difficult to prove affirmatively that a plaintiff would not have taken a risk, say, that the evidence shows that many other people freely take. I am not sure that the application of the subjective test will always be disadvantageous to a plaintiff, rather than the other way round.

  88. The way in which the law deals with some of these difficulties was explained by McHugh J in Chappel v Hart (1998) 195 CLR 232. His Honour said (at 247-248 [34]):

    [T]he onus of proving that the failure to warn was causally connected with the plaintiff's harm lies on the plaintiff. However, once the plaintiff proves that the defendant breached a duty to warn of a risk and that the risk eventuated and caused harm to the plaintiff, the plaintiff has made out a prima facie case of causal connection. An evidentiary onus then rests on the defendant to point to other evidence suggesting that no causal connection exists. Examples of such evidence are: evidence which indicates that the plaintiff would not have acted on the warning because of lack of choice or personal inclination .... Once the defendant points to such evidence, the onus lies on the plaintiff to prove that in all the circumstances a causal connection existed between the failure to warn and the injury suffered by the plaintiff.

  89. These matters also have been discussed in recent English authority. In Smith v Barking, Havering & Brentwood Health Authority (1994) 5 Med LR 285, Hutchison J said (at 289):

    [T]here is a peculiar difficulty involved in this sort of case - not least for the plaintiff herself - in giving, after the adverse outcome of the operation is known, reliable answers as to what she would have decided before the operation had she been given proper advice as to the risks inherent in it. Accordingly, it would, in my judgment, be right in the ordinary case to give particular weight to the objective assessment. If everything points to the fact that a reasonable plaintiff, properly informed, would have assented to the operation, the assertion from the witness box, made after the adverse outcome is known, in a wholly artificial situation and in the knowledge that the outcome of the case depends upon that assertion being maintained, does not carry great weight unless there are extraneous or additional factors to substantiate it. By extraneous or additional factors I mean, and I am not doing more than giving examples, religious or some other firmly-held convictions; particular social or domestic considerations justifying a decision not in accordance with what, objectively, seems the right one; assertions in the immediate aftermath of the operation made in a context other than that of a possible claim for damages; in other words, some particular factor which suggests that the plaintiff had grounds for not doing what a reasonable person in her situation might be expected to have done.

  90. His Lordship's reference to the giving of particular weight by the tribunal of fact to what he called "the objective assessment" should not be taken too far. At a jury trial, as the litigation in Hocking v Bell (1947) 75 CLR 125 at 132  strikingly demonstrated, and as the decision in Puntoriero v Water Administration Ministerial Corporation (1999) 199 CLR 575 at 586 [26]-[28] recently emphasised, the ultimate question on any appeal will be whether it was open to the jury to return a verdict for the plaintiff; the jury, for example, may prefer the evidence of a lay plaintiff to that of experts. Where the tribunal of fact is a judge sitting without jury, the reasoning process will be disclosed, or should be apparent, from the reasons for judgment.

  91. In this case, the trial judge rejected the respondent's evidence. His Honour, as has already been noted, based this decision primarily on his assessment of her credibility. However, his Honour found support for this assessment from a number of objective considerations. These included the low probability of the risk occurring, the fact that the likely consequences were not very severe, the desirability of the treatment and the respondent's awareness, as an experienced nurse, of the risks inherent in any surgery. There was no evidence of substantial fears or concerns that could counter these objective considerations. Consequently, the trial judge held that, even if the risk was material, the failure to warn was not causative. Subject to the observations made above as to the dangers involved in considering causation divorced from materiality, his Honour's reasoning did not disclose appealable error.

    WAS THE FULL COURT ENTITLED TO REJECT THE TRIAL JUDGE'S FINDINGS BASED ON CREDIBITILY?

  92. I agree with what is said by McHugh J upon this issue.

    THE DECISION IN CHAPPEL v HART

  93. There is one further matter to which reference should be made, although it is not determinative of the present appeal. It became apparent from the submissions to this Court that there was some uncertainty as to the basis on which rest the judgments of the majority in Chappel v Hart. It is appropriate to deal with the matter. The starting point is that the decision was concerned with the issue of causation not materiality; it was accepted that the risk in question was material under the second or subjective limb of the Rogers test.

  94. The risk that occurred was inherent in an operation of that sort and could occur irrespective of the experience or skill of the surgeon performing the operation. The plaintiff would have had to undergo the operation at some stage in any event, but it was accepted that, had she been warned of the risk, she would have sought out the most experienced surgeon to perform the treatment. One of the divisions between the majority and minority was that the minority (McHugh and Hayne JJ) thought that the degree of risk was the same no matter who performed the operation, while two members of the majority (Gaudron and Kirby JJ) thought that the risk was less if a more experienced surgeon performed the operation.

  95. However, this is not the only basis on which the decision of the majority rests. In argument, there had been discussion of the notion of the loss of a chance ((1998) 195 CLR 232 at 235-236  ). However, the judgments of the majority express the view that the damage sustained by the plaintiff was not the exposure to the risk of harm, but rather the actual harm that eventuated ((1998) 195 CLR 232 at 240 [12] per Gaudron J, 260 [76] per Gummow J, 277 [96] per Kirby J). Thus, there could be no action for loss of a chance because the plaintiff was not suing for the loss of an opportunity to utilise a more experienced surgeon; she was suing for the injury that she actually suffered. Gaudron J put the matter this way ((1998) 195 CLR 232 at 239 [9]):

    If [there is] evidence .... to the effect that the injured person would have acted to avoid or minimise the risk of injury, it is to apply sophistry rather than common sense to say that, although the risk of physical injury which came about called the duty of care into existence, breach of that duty did not cause or contribute to that injury, but simply resulted in the loss of an opportunity to pursue a different course of action.

  96. In their work on professional negligence, Sir Rupert Jackson of the English High Court and Mr John Powell reflect these sentiments when they observe of Chappel v Hart[13]:

    It was wrong to analyse the plaintiff's damage as the loss of a chance. On the evidence, had she been properly warned, she would not have undergone the operation when she did. She would not have suffered loss of her voice when she did. The fact that she would have undergone the operation, bearing a risk of the same nature, at a later date did not prevent a finding that the defendant's failure to warn had caused her to suffer the injury of which she complained.

  97. Had the warning been given, the plaintiff in that case would have had the operation at a different time by a different surgeon. Given the very low probability of the risk occurring, it would have been extremely unlikely that the harm would have eventuated. That was so, even if the view of the minority was correct and the likelihood of the injury occurring was the same irrespective of who performed the operation ((1998) 195 CLR 232 at 249-250 [41] per McHugh J, 286 [129] per Hayne J). Therefore, in a legally sufficient sense, the failure of the defendant to warn of the risk caused the harm.

    CONCLUSION

  98. The risk that the appellant reasonably could have warned against prior to the osteotomy was not "material" in the necessary sense. The Full Court erred in holding otherwise. The Full Court also erred in overturning the findings of the trial judge with respect to causation.

  99. I agree with the orders proposed by the Chief Justice.

    Kirby J

  100. This appeal[14] takes this Court once again into three areas of the law that have recently engaged its attention.

  101. The first area concerns the duty of healthcare providers (including medical practitioners and dental surgeons) to inform patients contemplating invasive procedures (such as surgery) of the material risks involved in the treatment proposed, and of any available alternatives. Any "choice" by the patient, in respect of such procedures, without the provision of such information, is "meaningless": Rogers v Whitaker (1992) 175 CLR 479 at 489 ("Rogers").

  102. The second area, where a patient is not provided with information about material risks, concerns whether such an omission is the cause of the damage which the patient claims. If, for example, the patient would have accepted the risk anyway and gone ahead with the procedure, any damage will not be found to have been caused by the proved omission: Chappel v Hart (1998) 195 CLR 232 at 271-272 [93] ("Chappel").

  103. The third area raises an issue that has had to be considered by every appellate court since the procedure of appeal was created by statute[15]. Depending upon their precise statutory powers, appellate courts are ordinarily bound by findings of fact made in the trial court where such findings are, expressly or by implication, dependent on the primary judge's assessment of the credibility of material witnesses[16]. This principle is itself subject to various qualifications and exceptions. However, such qualifications and exceptions modify the basic rule without abolishing or casting doubt upon it: State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (In Liq) (1999) 73 ALJR 306 at 331-332 [93]; 160 ALR 588 at 620-622 ("SRA"). It remains a serious step for an appellate court to disturb a primary judge's finding of fact based upon considerations of credibility.

  104. The Full Court of the Supreme Court of Western Australia ("the Full Court") concluded that the present was "one of the rare cases" ([1999] WASCA 31 at [108] per Owen J) requiring such intervention. Unanimously, it set aside a judgment entered at trial in favour of a dental surgeon. It did so despite conclusions, adverse to the patient, expressed by the primary judge. On several of the relevant issues in the case, those conclusions were stated to have been based on the primary judge's view that the patient's credibility as a witness was unreliable. The critical, and ultimately determinative, question before this Court is whether, in the circumstances, the Full Court was entitled to take that step.

  105. Reversals on appeal of the judgments of trial courts in leading cases of negligence by healthcare workers are not unknown. They have occurred in several "landmark" cases. These include Canterbury v Spence 464 F 2d 772 at 779 (1972)[17], in the United States, and F v R (1983) 33 SASR 189 at 196, in Australia. The latter decision was highly influential in the development of this Court's reasoning in Rogers. But in those cases the appellate court reversed the decision at trial on grounds of legal principle. That principle related to the scope of the duty which the law imposed on the healthcare provider concerned. This appeal presents an additional complication. Although there are important issues of legal principle raised by the facts, those facts must be approached consistently with the findings of the primary judge based on his assessments of credibility, unless a legal basis is established for the appellate court to substitute its own conclusions on such matters.

    THE FACTS AND ISSUES

  106. The reasons of Callinan J recite the background facts in terms that I adopt. Subsequent to a sagittal split osteotomy performed by Dr Ian Rosenberg ("the appellant"), Dr Patricia Percival ("the respondent") suffered pain and loss of function. The extent of her damage was strongly contested but the existence of damage was not. A second operation performed by the appellant produced no reported relief.

  107. The expert evidence called at the trial was to the effect that the actual performance of the operations accorded with the requisite standard of care and skill and that the second operation, and follow-up procedures, were necessary and proper. Earlier disputes about these matters, found against the respondent at trial, were not pressed in this Court. The respondent's case was confined to the one upon which she had succeeded in the Full Court. This was that the appellant had failed to discharge his duty to warn the respondent of the risk of developing temporomandibular joint ("TMJ") disorders as a consequence of the osteotomy which he had recommended and performed.

  108. So confined, the issues in this Court were reduced to:

    1. Whether the Full Court erred in concluding that, in the circumstances, the appellant was under a duty to warn the respondent more effectively than he did about the risk of TMJ disorders of the kind that she subsequently developed;

    2. Whether the Full Court was justified in setting aside the finding of the primary judge (Gunning DCJ) that, had she been warned by the appellant, prior to the osteotomy, of the risk of developing TMJ disorders as a result, she would nevertheless have proceeded with the surgery and, consequently, had suffered no relevant loss as a result of the failure to warn; and

    3. Whether, for the provision of relief, the Full Court was authorised in law, on appeal, to disturb the finding of the primary judge expressed to be based, in part, on his assessment of the credibility of the respondent.

    THE EVIDENCE AT TRIAL

  109. The trial was extensive. Expert witnesses of high professional repute were called on both sides. The respondent had herself qualified as a nurse, had practical nursing experience over a long period, had completed a doctorate in nursing and was a senior lecturer in the subject at a Western Australian university. The respondent's cause of action, as pleaded, included a claim based on the appellant's alleged failure to alert her to the inherent risks involved in the operative procedure[18]. It was initially left to inference from the respondent's evidence in chief that, had she been warned of the risk of the kinds of complications that eventuated, she would not have undergone the operation. Apparently sensing that this might produce a gap, potentially fatal, in the respondent's case, the respondent's counsel recalled her after cross-examination. She was resworn to give evidence of what, hypothetically, would have been the case had she been properly warned of the risks of osteotomy. Evidence of this kind has, conventionally, been permitted in Australia although its self-serving character and other defects have been commented upon[19].

  110. Conscious of the fact that the expert evidence would probably suggest that TMJ complications of the exact kind that the respondent had suffered were quite rare, the questioner approached the added questions cautiously:

    [Counsel]

    :

    Dr Percival, before you had your osteotomy in late 1993, if you had been advised by Dr Rosenberg that there was some - perhaps very small - chance of an adverse outcome, what would you have done?

    [Respondent]

    :

    If there had been any risk I would not have had the surgery.

    [Counsel]

    :

    I gather that you were anxious to obtain a satisfactory result for your dental problems. What would you have done if faced with advice that there was some small irreducible risk of adverse complications or a bad outcome?

    [Respondent]

    :

    I would have [had] the orthodontic treatment to get the best I could out of the situation and have [had] some prosthodontic treatment and been very happy with the result. I have lived with my malocclusion all my life.

  111. The cross-examiner pressed the respondent with the suggestion that her malocclusion was such that, with a small risk, she would have undertaken surgery anyway:

    [Counsel]

    :

    You would have remained keen to undergo the treatment and you would in fact have undergone the treatment had you known that there was a risk of you developing symptoms emanating from the temporomandibular joint after the procedure?

    [Respondent]

    :

    Never. Never.

    ....

    :

    [Counsel]

    :

    But I suggest to you .... that had you been informed that you might get temporomandibular joint problems, that there was every expectation that those problems would resolve themselves either without treatment or with conservative treatment and that there was a very, very unusual risk that you might have longer term problems, you still would have undergone the operation?

    [Respondent]

    :

    If I had been informed there was any risk to my temporomandibular joint I would never have had the surgery, .... any risk.

  112. Much of the testimony of the expert witnesses (oral and written) was addressed to the nature and extent of the risk of TMJ complications following an osteotomy. The effect of the evidence, stated very broadly, was that any disturbance, surgical or manipulative, of the kind involved in the procedures to which the respondent submitted, might cause TMJ problems. The extent to which they would be likely to do so would be influenced by whether or not symptoms of TMJ disfunction were already present in the patient. Ordinarily, the symptoms resulting from such an operation would be of short duration. Long-lasting and serious symptoms, of the kind suffered by the respondent, were extremely rare. But they were not unknown.

  113. That such complications were not unknown was established in a number of ways. Professor Goss, Professor of Oral and Maxillofacial Surgery at the University of Adelaide, was called in the respondent's case. He gave the following answer in cross-examination:

    [Counsel]

    :

    [Y]ou mentioned in your evidence that you are aware, with the exception of Dr Percival, [that] two other cases with a similar presentation to Dr Percival's, had developed problems of the kind that she has developed in the last decade?

    [Professor Goss]

    :

    That's correct. In an attempt to determine the incidence of this problem I looked at the situation in South Australia over the last decade and I identified that there were two patients who had temporomandibular joint disorder. Subsequently they had a facial bone osteotomy and then developed severe chronic pain .... I then calculated the number of osteotomies which would have been performed over that decade and the number was 5000 osteotomies in South Australia .... I have then attempted to do the same thing for Australia and at the moment I have ended up that the total number of osteotomies in Australia over the last decade would be of the order of 19 to 20 thousand .... On that basis, if you take three out of 20,000 then you end up, I guess, with an incidence of 1 in 6000. So I think that one can confidently say that the incidence of this problem is certainly measured 1 in thousands and my best guess is somewhere between 1 in 2 and a half and 1 in 6.

  114. A textbook description of "Temporomandibular Dysfunction: Considerations in the Surgical-Orthodontic Patient" by Tucker and Proffit, Surgical-Orthodontic Treatment (1991) 660 , tendered for the respondent, states (at 662-663):

    Malocclusion has frequently been cited as a cause or contributing factor in the occurrence of TM joint pain or dysfunction. This relationship, however, remains extremely controversial. ...

    In summary, it appears that some types of malocclusion may slightly predispose patients to TM joint problems, but the relationship is weak enough to be questionable.

  115. Later, in dealing with "TM Joint Problems Related to Surgical-Orthodontic Treatment", those authors add (at 684):

    The chance that TM joint pain / dysfunction can arise as a result of surgical-orthodontic treatment is always of concern.

  116. Specific to the kind of operative procedure undertaken in relation to the respondent, the authors suggest (at 684):

    Several factors related to surgical correction of dentofacial deformities can be associated with postoperative TM joint pain. Limitation of motion postsurgically is observed at least transiently in all patients who have mandibular ramus surgery and in a significant number long-term. Immobilization of the jaws during healing is a major contributor to postsurgical limitation of motion and may contribute to other effects as well.

    (Emphasis added)

  117. On this point, the respective cases of the appellant and the respondent at trial appeared to pass each other like ships in the night. The appellant accepted (and most of the expert witnesses agreed) that the respondent had minor pre-operative signs of TMJ disfunction, although asymptomatic. Differences of opinion existed as to whether it could be described as a "disorder" or a "condition". But it was detected clinically. It also appeared in the respondent's radiographs of November 1993, prior to the osteotomy. The appellant also accepted that generally TMJ is a potential area for concern and specifically that an osteotomy of the kind he performed, with or without manipulation under general anaesthesia with arthrocentesis, could cause a deterioration of a quiescent TMJ disorder.

  118. The appellant conceded that he did not warn the respondent of the risk of developing or aggravating TMJ problems. The entries in his clinical notes confirm this. They do not record any specific attention to, or warning of, that possible complication. The materials received by the respondent, relevant to the postoperative complications, included an information pamphlet prepared for patients undergoing orthognathic surgery. Far from referring to the possibility, even low possibility, in a small number of cases, of prolonged TMJ pain and discomfort following such surgery, the tone of the pamphlet was highly optimistic. Relevantly it stated:

    Following on surgery there will be a degree of swelling of the facial tissues, but with modern day management and medication this is not excessive.

    ...

    A complete return to normal function may take several weeks or months. It is advisable that contact sporting activities should not be undertaken for a period of six months, during which time the bone healing strengthens. The period of maximum inconvenience will be for approximately two to three weeks after surgery.

  119. The only serious warning in the pamphlet of complications was of "the risk involved in the patient undergoing general anaesthesia". There was no explicit reference to TMJ problems. Nor was that risk mentioned in a consent form proffered by the appellant and signed by the respondent. Like the information pamphlet, the consent form's only emphatic warning related to the well-known but small risk of undergoing general anaesthesia.

  120. In response to allegations of such failings, the appellant emphasised the fact that, although he had performed some 450 osteotomies, his patients never experienced the complication that developed in treating the respondent. His case laid emphasis on the extremely small incidence of long-term pain and disability. Ultimately the appellant argued that a warning to the respondent of the type of complications she eventually suffered was not called for because the risk was not "material". The appellant focussed on the precise complications. The respondent's case focussed on the general incidence of postoperative TMJ disorders. She complained that the appellant had omitted to warn her of the general risk of such disorders. Although that risk might be low, it was her assertion that she was entitled to be warned of it so that she could make her own decisions in respect of it.

  121. Other evidence adduced at the trial demonstrated that bilateral sagittal osteotomy is a relatively common procedure. It is one quite often recommended to improve dental malocclusion. However, it is not an essential intervention. Nor is it the only remedy available to address such a problem. Other options include undertaking crown and bridge work or prosthodontic measures to reduce the malocclusion. Alternatively, a patient might elect to continue to live with the problem and do nothing. In the respondent's case, the concern was not cosmetic but functional. The avoidance of future problems developing in the jaw was therefore of importance. It is in this context that the relevance of a warning about the risks in the course of treatment proposed fell to be determined.

    THE DECISION OF THE PRIMARY JUDGE

    Obligation to warn

  122. Most of the expert witnesses at the trial supported the contention that the respondent should have been given a warning that there was a small risk of complications developing with TMJ problems. Professor Goss was of this view. So were Dr Punnia-Moorthy, Professor Levant, Dr Delcanho and Dr McNamara. Only the appellant, to whom it "did not occur", and Professor Norman, who did not regard a warning as necessary in the circumstances, expressed a contrary opinion.

  123. The primary judge "unreservedly" accepted Professor Norman's evidence. He also accepted that the appellant was a "very experienced surgeon" for whom complications of the kind that had affected the respondent were unique. He noted that such complications were, on all of the evidence, "very rare indeed". He concluded:

    It is for these reasons that I do not accept the evidence of Professor Goss in one aspect where he considered that a warning should have been given in the circumstances as he saw them. However he was operating with hindsight and none of the other specialists who saw the plaintiff before the operation and examined the x-rays before the operation came to a conclusion that there was or could be any temporomandibular problem. ...

    It follows that in the circumstances the defendant, prior to the operation, was not negligent in his interpretation of any of the plaintiff's complaints .... or of his interpretation of the x-rays and therefore it follows that there was no known problem to him that could develop that he could communicate to the plaintiff.

    It follows [that] he was not negligent in not warning the plaintiff of any material problem that might develop.

  124. It is inherent in this passage in the primary judge's reasons that he adopted the narrow view of the obligation to warn propounded for the appellant. It was an obligation addressed to relatively well-known risks of long-term disability, not to a risk, concededly very rare, of ongoing pain and discomfort caused by the operative disturbance of a pre-existing, but previously asymptomatic, TMJ condition.

    Rejection of causation

  125. Similarly, the primary judge rejected the respondent's case that, had she received the warning propounded, she would not have gone ahead with the operation. He pointed out, correctly, that the respondent was not in the position of an "ordinary layman". This was because of her own professional training and the fact that she had made her own enquiries. She knew of the three options available to correct her malocclusion, only one of which was osteotomy.

  126. Early in his reasons, the primary judge indicated his disbelief of evidence given by the respondent. In this regard, he singled out her contested attribution to the appellant of an admission that he could not read the results from an MRI (magnetic resonance imaging) examination. The primary judge rejected this evidence and other evidence given by the respondent in connection with an MRI examination. He also rejected as "unbelievable and of no assistance in the circumstances" the respondent's assertion that, "had she been advised of the slightest possibility of complications with her [TMJ]", she would not have proceeded with the surgery.

  127. After describing the respondent as a "very intelligent lady", the primary judge proceeded to state a conclusion that she was "most anxious to tell her story in a way in which she thought would benefit her case and to play down anything that she thought might be to the contrary". The respondent had, perhaps understandably, undertaken research into the complications which she developed following her osteotomy. Unfortunately, this knowledge seeped into her oral testimony, to the apparent irritation of the primary judge. A question in cross-examination gave rise to a response on her part that included reference to the respondent's reading of research literature from the United States. Her answer prompted the judge to give the respondent a rebuke designed to discourage her "dissertation on American temporomandibular problems". The primary judge ultimately formed an unfavourable view of the respondent's evidence:

    [A]t the end of the cross-examination I can only say that this view was reinforced and I was far from satisfied that the plaintiff was a reliable witness. This impression was gained after examining the plaintiff for quite a long time when she was in the witness box in examination-in-chief and re-examination over a period of some days and there were particular instances where I considered her evidence to be unreliable.

  128. One specific instance of unreliability to which the primary judge referred concerned the respondent's application for appointment as a professor at her university which, it had been suggested for the appellant, was incompatible with her complaints of constant pain and disability. His Honour remarked:

    [It] only came out in cross-examination and she had not even discovered the application which contained all her CV over the years and was very pertinent to the case and her attempts to explain it away as I have mentioned, or highlighted, in this judgment, were unsatisfactory.

  129. The primary judge likewise commented adversely on the fact that the respondent was recalled, after the close of the appellant's case, to remedy the gap in her testimony which I have already mentioned. In the circumstances, because her then answers were provided with knowledge of the evidence that had emerged during the trial as to the extremely small incidence of complications such as she had suffered, the primary judge stated that what she had to say would have to be subject to "very strict scrutiny" and its value judged accordingly. He then referred to the repeated assertions of the respondent that, had she been warned of the small risk of such complications, she would not have undertaken surgery. He was unimpressed:

    [T]he answer is one that would be expected in the circumstances and of course, again in the circumstances, [is] of no evidentiary value whatsoever.

  130. In expressing his conclusions on the issue of causation, the primary judge stated that all that the appellant could have told the respondent was "that there was always the possibility of problems with any surgery with anaesthetic, a fact she already knew". He pointed out that she had not asked questions herself and concluded:

    I am quite satisfied in the circumstances that even if the plaintiff had been warned of the slight possibility, and certainly it was very slight, of complications she would have proceeded with the surgery in any event.

  131. The closing paragraph of the primary judge's reasons, which contained an assessment for the purposes of the taxation of costs, added a parting shot. He assessed the costs at little more than the out-of-pockets, taking into account that her damages depended "almost, if not entirely, on .... credibility".

    THE DECISION OF THE FULL COURT

  132. The Full Court was unanimous in reversing the judgment entered for the appellant. The leading opinion was written by Wallwork J. Kennedy J and Owen J added some observations of their own. Because Callinan J has identified in his reasons the chief considerations that influenced the Full Court, I will not repeat them (Reasons of Callinan J at [200]-[205]).

    The duty to warn issue

  133. It is clear enough that, on the scope of the duty to warn, all members of the Full Court considered that the standard established by this Court in Rogers had not been complied with by the appellant nor properly applied by the primary judge. Upon this first issue it was open to the Full Court to reach a conclusion on the facts different from that of the primary judge. This was so because the Full Court was conducting an appeal by way of rehearing (Rules of the Supreme Court 1971 (WA), O 63 r 2(1)). The evaluation of the respective medical opinions, and the applicable professional knowledge and standards, did not depend, as such, on credibility factors. Nevertheless, given the length of the trial, the detail and complexity of the evidence, and the conflicting views expressed by the witnesses, an appellate court would have to pause, and be convinced of error, before substituting its own opinion for that reached by the primary judge (SRA (1999) 73 ALJR 306; 160 ALR 588). Yet having come to a different view on the facts, if no question of credibility of witnesses or other impenetrable barrier stood in the way of reversal, the Full Court was entitled, indeed obliged, to give effect to its own clear conclusion (Warren v Coombes (1979) 142 CLR 531 at 551).

  134. Clearly enough, the judges constituting the Full Court ([1999] WASCA 31 at [3] per Kennedy J, [99] per Wallwork J, [105]-[107] per Owen J) were of the opinion that the standard required by Rogers obliged the appellant to give warnings beyond those actually provided. Wallwork J put it this way (at [97]-[99]):

    On the overwhelming evidence [the risk of postoperative TMJ disorder] was a risk which existed because of the likelihood of some patients having complications after the operative procedure. Once there is a risk which is generally known to the profession, there is a duty to warn. It was not necessary to establish that the [appellant] should have been alerted to any disorder which existed in the [respondent's] jaw joint.

    It was put for the [appellant] that although the [respondent] had been given the pamphlet and had received a letter from the [appellant], she had not raised any matters which had resulted in the [appellant] warning her of possible complications. That has been held in Rogers v Whitaker not to be determinative of the relevant question.

    [T]he learned trial Judge erred in finding that the [appellant] was not required to warn the [respondent] of the risks of TM joint problems and symptoms arising after the procedures which the [respondent] underwent.

  135. In his reasons, Owen J also addressed the materiality of the risk. He noted that some commentators had been critical of the implications of the stringent requirements about warnings established by this Court's decision in Rogers [1999] WASCA 31 (at [106] referring to Olbourne). He acknowledged that, in evolving "away from 'medical paternalism' the law [should] not place an unreasonable burden on members of the medical profession" (at [106]). However, Owen J concluded, with Wallwork J, that "there was a sufficient evidentiary base to support a finding that the risks were material in the relevant sense and that the [appellant] was under a duty to warn [the respondent] of them" (at [107]). In failing to so find on the evidence, he held, the primary judge had fallen into error.

    The causation issue

  136. Resolution of the second issue was, as the Full Court recognised, more difficult because the finding of the primary judge was based, in this respect, on his assessment of the credibility of the respondent. The reasons given respectively by Kennedy J and by Wallwork J, for detecting error in the primary judge's reasoning in this regard, are identified by Callinan J (at [200]-[205]). Explicit mention was made by Owen J of the restraint required by the decisions of this Court in disturbing conclusions of such a kind[20]. But having concluded that the primary judge had been wrong on this issue, the Full Court made the orders described by Callinan J (at [205]).

  137. It is possible to detect a feeling on the part of the judges constituting the Full Court, that the primary judge had overreacted to the respondent's case and had underestimated the extent to which prolonged pain and consequent depression, caused by the operative complications, may themselves have influenced the testimony to which he had reacted adversely.

  138. The orders ultimately made by the Full Court presented certain difficulties which counsel for the respondent properly conceded before this Court. Because negligence includes establishment of the fact that a breach of duty has caused the damage alleged, the reservation to the retrial of causation and damage amounted, in effect, to half a finding on liability. The failure of the Full Court to give reasoned consideration to all of the issues argued would make it extremely difficult for a trial judge to determine the remaining issues, including that of causation, in accordance with the Full Court's order.

  139. Doubtless the Full Court was anxious to contain the scope, and thus the costs, of any retrial. Where liability is found or confirmed on appeal, it is not unusual to confine a retrial of an action, such as this, to questions of damages. However, that is not what the Full Court did. In the event, the form of the orders does not present a crucial problem. Yet those orders indicate something of the ambivalence which is contained in the Full Court's reasoning.

    THE DUTY TO WARN THE PATIENT

    The rule in Rogers

  140. The rule established by this Court in Rogers is undoubtedly a strict one. Adopting it involved following the lead of the Supreme Court of Canada[21] and earlier decisions in Australia[22]. Those decisions took Australian law away from the test, in respect of the duty to warn, established by English law and expressed in Bolam v Friern Hospital Management Committee [1957] 1 WLR 582; [1957] 2 All ER 118 ("Bolam")[23]. In Rogers this Court did not explain the duty to warn in terms of "informed consent", because it was there concerned with a cause of action framed in negligence, not trespass. "Consent" (or the lack of it) was not, as such, a defence. Nevertheless, there is no doubt that the rule that the Court expressed in Rogers was addressed to the concerns that are commonly dealt with, in legal and medical literature, as relevant to securing the "informed consent" of a patient to invasive treatment. Accordingly, much of the discussion in that literature is relevant to any suggested reconsideration of the ambit of the rule. The rule, as stated in Rogers (1992) 175 CLR 479 at 490., is clear:

    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.

  141. The test is not, therefore, expressed in terms of what a reasonable healthcare provider would give by way of a warning[24]. Nor is it dependent upon questions asked by the patient who might be completely unaware of the issues to which his or her mind should be addressed. It is an aspect of the duty of care owed to the patient by the service provider. Its content is decided by the application of an objective criterion: the needs of a reasonable person in the patient's position. Such needs may be enlarged in the case of particular patients because of perceived features special to them.

  142. Fundamental to the formulation adopted by this Court in Rogers is a recognition, expressed much earlier in the United States cases, that a patient "has a right to determine what shall be done with his own body"25]. Thus "it is the prerogative of the patient, not the physician, to determine .... the direction in which his interests seem to lie"[26]. No one in this appeal challenged the holding in Rogers. No one suggested that its authority should be reconsidered. Even in England, the former Bolam principle has recently been modified[27] and continues to be criticised[28]. Indeed, few today argue against the line of authority recognised by Rogers[29].

    Confining the rule in Rogers

  143. Nevertheless, various arguments of principle have been mounted from time to time, in response to Rogers and decisions like it, suggesting that the legal rule established is, or should be, confined, as far as authority permits, by reference to practical considerations. These typically include arguments addressed to the following considerations:

    1. That some patients do not wish to be unsettled by unnecessary disclosures by professional experts whom they trust, or about risks and concerns that, in any case, they will only understand imperfectly[30];

    2. That it is impossible, within sensible time constraints, for a professional person to communicate the detail of every tiny complication that may accompany medical procedures[31]. A full appreciation of these only emerges out of specialist training and a lifetime's experience. Complications that are very rare could not be explained effectively in the time typically available for a healthcare consultation. To attempt such explanations, as is arguably the "ideal" expressed by this Court in Rogers, would involve the expenditure of time and effort that would not be cost effective[32];

    3. That the efficacy of warnings against slight risks has not been objectively established[33];

    4. That belief in the efficacy of warnings is a lawyer's fancy which other lawyers then seek to circumvent by drafting substantial consent and waiver forms[34];

    5. That the principle tends to view patient concurrence in important medical procedures as depending upon a single instance of warning and "consent" which consigns patient participation in decision-making to an unacceptably passive role[35]. It fails to view the healthcare relationship, as best practice contemplates, as involving a continuous relationship[36]. In a continuous dialogue, fears and concerns are explored and experience effectively communicated so that a meaningful choice may be made which takes into account the considerations personal to the patient[37] as well as the experience of the practitioner and the advancing knowledge of his or her discipline;

    6. That in practice the strict standard is contradicted by everyday professional experience which is that, alerted to slight risks, patients ordinarily elect to accept them if otherwise the intervention is considered justifiable or necessary[38];

    7. That the test can easily become a prop to disappointed patients, resulting in the imposition by legal decisions of unrealistic and unreasonable professional obligations[39]; and

    8. That the standard of care demanded, and any resulting increase in malpractice litigation, will lead to defensive practices and even the needless retirement from their professions of healthcare providers, disappointed by adverse legal decisions[40].

  144. In a particular case, consideration of the foregoing arguments may help to contribute to an expression of the content of the duty to warn that is realistic and achievable. But neither as a matter of legal authority, nor in terms of legal principle or policy, do any of the stated considerations cast doubt on the duty to warn expressed by this Court in the terms used in Rogers. As a matter of authority, that decision has stood for a decade. No application was made here to reopen it. Doubtless, to some degree, healthcare practice throughout Australia has already adjusted to its requirements. No reason has been shown to reformulate more narrowly the rule stated or to apply it in a way that would be inconsistent with the rule stated in Rogers.

  145. Moreover, reasons of principle and policy support the stringency of that rule:

    1. Fundamentally, the rule is a recognition of individual autonomy that is to be viewed in the wider context of an emerging appreciation of basic human rights and human dignity[41]. There is no reason to diminish the law's insistence, to the greatest extent possible, upon prior, informed agreement to invasive treatment, save for that which is required in an emergency or otherwise out of necessity[42];

    2. Whilst it may be desirable to instil a relationship between the healthcare professional and the patient, reality demands a recognition that sometimes (as in the present case) defects of communication demand the imposition of minimum legal obligations so that even those providers who are in a hurry, or who may have comparatively less skill or inclination for communication, are obliged to pause and provide warnings of the kind that Rogers mandates;

    3. Such obligations have the added benefit of redressing, to some small degree, the risks of conflicts of interest and duty which a provider may sometimes face in favouring one healthcare procedure over another[43];

    4. Also, to some extent, the legal obligation to provide warnings may sometimes help to redress the inherent inequality in power between the professional provider and a vulnerable patient[44]; and

    5. Even those who are dubious about obligations, such as those stated in decisions such as Rogers, commonly recognise the value of the symbolism which such legal holdings afford. Thus, Professor Katz, a noted writer in this field, whilst accepting that the use of the principle of "informed consent" may sometimes mislead "patients into thinking that they are making decisions when indeed they are not"[45], also accepts that such principles can "nag and prod and disturb and ultimately bring about some change"[46]. I agree with this opinion. It is reinforced by reports about healthcare practice in Australia[47] and by common experience. In so far as the law can influence such practice, it should tend, as Rogers does, towards the provision of detailed warnings so that the ultimate choice, to undertake or refuse an invasive procedure, rests, and is seen to rest, on the patient rather than the healthcare provider. To the extent that this result is upheld, it seems likely that recriminations and litigation following disappointment after treatment will be diminished.

    THE FULL COURT DID NOT ERR ON LACK OF WARNING

  146. The foregoing reasoning requires the application to the evidence, taken at the trial, of the rule in Rogers so as to uphold the principles inherent in its reasoning. In my opinion, there was ample evidence in the present trial to show that, at the time of the operation, it was known, or ought reasonably to have been known, by an oral and maxillofacial surgeon such as the appellant, that bilateral sagittal osteotomy entailed a small risk of causing TMJ problems or of exacerbating or triggering symptoms in a previously asymptomatic TMJ condition.

  147. That possibility, as a complication of osteotomy, was mentioned in textbooks. It was recognised by experts in the profession. It had been observed in clinical practice in Australia, as Professor Goss deposed. It had even given rise to some litigation in the Supreme Court of South Australia in which Professor Goss is recorded as having given evidence: Hribar v Wells (1995) 64 SASR 129 at 131. Whilst that matter did not come to appeal until 1995 (and the law report would scarcely be amongst the reading of a dental surgeon such as the appellant), it was an illustration of the incidence in Australia of the complication which Professor Goss described in his evidence. I would agree with the judges of the Full Court that, having regard to the evidence at trial, the respondent clearly established that the appellant had failed in his duty to provide her, as his patient, with a significant warning of a material risk inherent in the proposed treatment.

  148. By the standards established in Rogers, and, I believe, by the standards of the dental profession in 1993 as disclosed by the evidence, the appellant's warning fell short of what was required. In fact, his warnings were rather perfunctory and substantially confined to offering written documents that were optimistic in tone and largely confined to notifying the risks of anaesthesia. While the provision of such written documents is to be commended, as it allows a patient time to reflect on the procedures described and to ask questions on issues left unanswered, such forms are no substitute for dialogue between patient and surgeon. Such dialogue, inherent in informed decision-making, must, to some extent, be "shared"[48] so that it secures consent by a patient to a medical procedure that is truly understood[49]. Optimism may be an admirable, even necessary, quality in the performance of a surgeon's art. However, by the standard of Rogers, the patient has an entitlement to be warned of all material risks and, in this case, she was not.

  149. It is not the case that the risks of which a patient must be warned are confined to those that are commonplace (such as anaesthesia). The risks of quadriplegia in Ellis v Wallsend District Hospital (1989) 17 NSWLR 553 ("Ellis"), of mediastinitis in Chappel, of impotence and bladder malfunction considered in Smith v Tunbridge Wells Health Authority (1994) 5 Med LR 334 ("Smith") or of sympathetic ophthalmia examined in Rogers itself ((1992) 175 CLR 479 at 482), were all rare outcomes. As found, the relevant risks existed and were undisclosed to the respective patients. In Rogers (at 482), according to the evidence, the risk involved was "once in approximately 14,000 such procedures, although there was also evidence that the chance of occurrence was slightly greater when, [as in that case], there had been an earlier penetrating injury to the eye operated upon". The importance of all of these cases is that they emphasise that it is the patient who ultimately carries the burden of the risks. Therefore, unless such risks may be classified as "immaterial", in the sense of being unimportant or so rare that they can be safely ignored, they should be drawn to the notice of the patient. Only then can an informed choice be made by the person who alone, in law, may make that choice, namely the patient.

  150. It is true that there are differences in this case from the facts found in Ellis, Rogers and Chappel. In each of the latter cases, it was concluded that the patient had asked questions or communicated explicit concerns that should have enlivened an appreciation on the part of the surgeon of the requirement to communicate information about the risk, albeit small, that existed. Moreover, in Ellis and in Rogers, the disabilities that could follow from the risks were so profound (quadriplegia and total blindness) as to impose particular duties of detailed communications and warnings. But in the present case, as was obvious to all who dealt with her, the respondent was also person who gave attention to detail. She stressed that she wanted her dental malocclusion to be properly attended to. In the circumstances, the failures of communication, about the possible range of risks of the operative procedure as revealed by the evidence, are fairly obvious.

  151. I do not believe that this conclusion represents the unreasonable judgment of hindsight. Handing a patient, such as the respondent, the written documents that were provided fell short of the proper communication and adequate warnings that were required, as much by appropriate dental practice as by authority of law. To the extent that this is relevant, the risk of the type of complication occurring, although less devastating than quadriplegia or blindness, was significantly greater than the risk disclosed in Rogers. To the extent that the practice of the relevant healthcare profession assists in defining what the law reasonably requires, the evidence called at the trial was overwhelmingly supportive of the respondent's case. No consideration of credibility would, in my opinion, warrant deciding otherwise (cf Ahmedi v Ahmedi (1991) 23 NSWLR 288 at 291).

  152. Put another way, I can see no error in the conclusion which the Full Court reached on this point. It is a conclusion that I too would have reached on the evidence. The primary judge erred in dismissing the complaint about the appellant's failure to provide an adequate warning. He was led into that error by accepting the appellant's argument that he, the appellant, could not have envisaged the precise complications that occurred to the respondent and thus need not have warned against them. But the appellant ought to have envisaged as material in the respondent's case, the real, although small, risks of TMJ complications with long-term symptoms, as a result of the procedure that he advised. A general warning to that extent was therefore necessary. This is the standard that Rogers lays down. The Full Court was correct to insist on that standard. On that point, their Honours were entitled to substitute their conclusion for that of the primary judge.

    EVALUATING A PATIENT'S ASSERTION OF THE EFFECTS OF A WARNING

  153. Although decisions, of which Rogers is the local example, are said to have caused "alarm" in some circles[50], occasioned warnings about floods of "'informed consent' litigation"[51] and given rise to assertions that the "pendulum may have swung too far"[52], more temperate opinions seem rather more persuasive[53]. In Canada, on the basis of a close analysis of decisions following the rejection of the Bolam test in that country, the conclusion has been expressed that what the courts gave in the duty of disclosure, they often took away by their approach to the issue of causation[54]. However, in respect of causation of medical misfortunes, Canadian courts rejected the approach of enquiring into what the particular patient would have done if a proper warning had been given (the subjective approach). Influenced by United States cases, which asked what a "prudent person" provided with such a warning would have done (the objective approach - Canterbury v Spence 464 F 2d 772 at 791 (1972)), the Canadian Supreme Court adopted the test of what a reasonable person in the patient's position would have done if properly warned[55]. By majority, the Supreme Court of Canada has recently affirmed this test (the modified objective approach) (Arndt v Smith [1997] 2 SCR 539).

  154. In the way in which Rogers was argued, it was strictly unnecessary for this Court to resolve this aspect of the controversy. However, the opinions in Chappel appear consistent with the subjective criterion. The law in England has consistently accepted the subjective test for the issue of causation in this respect[56]. Certainly, it is the one which I have accepted (Chappel, at 272-273 [93]). It had earlier been adopted by the New South Wales Court of Appeal in Ellis (1989) 17 NSWLR 553 at 560-561, 579-582. It is an approach more consistent with the traditional principles of tort law[57]. It is more respectful of the entitlements of patients, whose privilege of choice this area of the law is intended to reinforce[58]. Furthermore, it avoids undermining the social objectives to which the obligation to provide effective warnings is directed[59].

  155. The practical problems presented by adopting a subjective criterion must, however, be recognised. They have been noted in many earlier decisions. When Chappel was before the New South Wales Court of Appeal, Mahoney P commented on the element of "unreality" that was involved in considering an injured person's ex post assertion of what he or she would have done if given a warning judged later to have been necessary[60]. Allowing that the patient concerned is sufficiently disappointed with the outcome of some healthcare procedure that he or she has ventured upon expensive, time-consuming and stressful litigation to obtain redress, it is scarcely conceivable that such a patient would destroy the case by equivocating in evidence over such a matter.

  156. It is the inherent unreliability of such self-serving testimony that has persuaded courts in most parts of the United States and the majority of the Supreme Court of Canada to adhere to some form of modified objective standard. Otherwise, as those courts have indicated, the only available evidence which can bear upon the hypothetical state of mind of patients, had they been warned of the material risks prior to the surgery, is the retrospective evaluation by that aggrieved patient who only stands to gain by affirming that he or she would not have proceeded with the treatment. Clearly, this approach risks converting a standard of reasonable care, inherent in the tort of negligence, effectively into strict liability for treatment that is, or is regarded as, unsuccessful[61]. In this appeal, the appellant conceded that the test for causation was subjective. However, he argued that objective facts had to be "called in aid" to establish whether the patient in question would have had, or would have declined, the operation[62].

  157. Notwithstanding the foregoing criticisms of the subjective test, which the appellant reflected in his submissions to this Court, I am not persuaded that this Court should now adopt a variant of the objective standard. To do so would be incompatible with general doctrine. It would be difficult to reconcile with the test which the Court has laid down in Rogers (1992) 175 CLR 479 at 490 that expressly mentioned the need for attention to the considerations to which a particular patient may be likely to attach significance. And it would be unnecessary to stem any flood of meritless claims, so long as the courts observe the need for "great care", as suggested by Samuels JA in Ellis (1989) 17 NSWLR 553 at 582., in evaluating the patient's own assertions.

  158. In Chappel, I expressed confidence that the dangers of attaching too much weight to the ex post claims of plaintiffs should not be overstated. It was my opinion then, as it is now, that "[t]ribunals of fact can be trusted to reject absurd, self-interested assertions" ((1998) 195 CLR 232 at 273 [93]). A similar confidence was expressed in an article comparing the approach to this topic in a number of jurisdictions. The author concluded in words that I would endorse[63]:

    When determining whether a reasonable patient would attach significance to a risk, the court should bear in mind that reasonable people accept not insignificant risks on a regular basis, eg driving cars and playing contact sport.

    When deciding whether the patient would have undergone treatment if the risk had been disclosed, the court should assess the plaintiff's testimony carefully. Naturally, the reasonableness of refusing treatment is a consideration here. If a reasonable person would have undergone treatment, regardless of disclosure, then in the absence of personal characteristics or circumstances which would explain a refusal, it must be difficult for a court to conclude that the plaintiff would have rejected the treatment no matter what the plaintiff now genuinely believes that he or she would have done. It should be remembered that causation in other areas of negligence presents similar difficulties, eg would an injured employee have used the safety equipment if his or her employer had provided it?

  159. The price of putting faith in tribunals of fact to conclude such issues of causation in a sensible and practical way is restraint in appellate reversal of such decisions. This is especially so where those decisions rest on the assessment by the primary judge of the truthfulness of the patient in cases where that issue has been contested. That this was so was accepted where the judge believed the patient's denial in Ellis, Rogers, Chappel and Smith. The same rule must be observed where, as here, the judge disbelieved the patient, unless the case falls into an exceptional category where the appellate court is justified in substituting its own, different, conclusion.

    THE REVERSAL OF THE DECISION ON CAUSATION WAS UNJUSTIFIED

  160. The respondent argued that the Full Court properly reversed the primary judge's finding, based on an evaluation of her credibility, that she would not have had the operation if properly warned. This was justified by reference to the unsatisfactory features of the primary judge's findings about her credibility. These, she submitted, were sufficient to cast doubt on the correctness of the primary judge's conclusion.

  161. Like Callinan J (at [222]) I have a sense of disquiet about some parts of the primary judge's reasons. From the transcript, it appears that sometimes his Honour found the respondent irritating. The importance he attached to the suggested inconsistency of her undisclosed application for promotion to professor and her asserted disabilities is, with respect, rather unconvincing. People frequently act in inconsistent ways. The respondent, for example, may simply have hoped that promotion would provide a distraction to her depression and pain. Perhaps she thought that promotion was her due. Similarly unsettling was the primary judge's use of the deterioration in the garden of the respondent's home, as suggestive of the fact that she and her husband were seeking improperly and dishonestly to enlarge the verdict. This was unfair given that no such suggestion was ever put to either of them in cross-examination. These considerations, which Callinan J has identified, leave me with a sense of unease about the credibility findings. The Full Court's decision may have been influenced by these factors.

  162. However, with respect to the Full Court, if the deficiencies in the primary judge's evaluation of the respondent's credibility were the foundation of their Honours' intervention on the issue of causation, they should have said so. The reasons which they did express are not convincing of error on the part of the primary judge[64]. Certainly, they do not rise to the level that would have allowed the Full Court to place this case amongst the comparatively rare exceptions that justify an appellate court overturning a credibility-based assessment of a trial judge, especially one expressly said to be founded on scrutiny of a witness giving evidence[65].

  163. I have stated in SRA that, in my opinion, judges of trial should be reticent to rely upon impressions derived from such observations, given the very strong doubt that is cast on the safety and reliability of such impressions by a now substantial body of scientific evidence[66]. Contemporary judges in Australia know about such dangers where the English judges of the nineteenth century, who voiced their faith in the reliability of judicial observation, did not[67]. This is why most judges, and equivalent decision-makers, today endeavour to rest their conclusions, as far as possible, on objective considerations, contemporaneous facts and logical inferences, rather than a self-claimed capacity to tell truth from falsehood by visual or aural impression.

  164. The primary judge's reliance on impression creates a significant hurdle for an appellant to overcome. In part, this is because of the advantages which the trial court is taken to have in observing witnesses and hearing the evidence unfold in its entirety. These advantages are denied to the appellate court (SRA (1999) 73 ALJR 306 at 330-331 [89]-[92]). However, occasionally, this Court (See eg Voulis v Kozary (1975) 180 CLR 177) and other Australian appellate courts (See eg Beale v Government Insurance Office of NSW (1997) 48 NSWLR 430) have considered that a case falls within an exception to the primary rule. Such cases warrant the substitution, on appeal, of a conclusion of fact different from that expressed by the primary judge. Such cases are comparatively rare (Jackamarra v Krakouer (1998) 195 CLR 516 at 543 [66]). According to the primary judge, the respondent was, in a relevant respect, simply not to be believed. Yet, even if the judge's explicit findings on credibility were doubted, there were a number of objective factors that supported the rejection of the respondent's evidentiary assertions. These included the facts that the respondent had sought out specialist dental advice; that she was concerned to secure the best result and ordinarily, in a case such as hers, that would have meant having an osteotomy; that she had better access to medical and dental knowledge than the ordinary patient, including in relation to risks of operative intervention; that because of her professional training and knowledge she would have known that any operative procedure at all carried some, exceptional, risks inherent in the disturbance of the human physiology; and, most importantly, that, as the evidence showed, the risk of the kind of complications which happened in her case was extremely small. The complications were so rare, indeed, that a number of the expert witnesses, as well as the appellant, had never experienced them. On the other hand, osteotomy was a comparatively common operative procedure. Ordinarily, it quickly produced good results.

  165. As a matter of inherent probabilities and logic, therefore, these objective facts cast real doubt on the respondent's evidence that, had she been warned of such a very small risk, she would not have proceeded with the operation. Causation often presents a problem to plaintiffs in negligence actions (Crimmins v Stevedoring Industry Finance Committee (1999) 200 CLR 1 at 87 [238]). This is certainly so in cases alleging medical malpractice. A completely dispassionate approach to the respondent's claim might therefore have warranted rejection of it upon the basis of the inherent probabilities disclosed by the evidence. But had that been done, the Full Court would have had a freer hand. Because of the way the primary judge reasoned, the Full Court was constrained by binding authority. It suffices to say that I do not believe that an exception was established to the application of the principle of restraint required by the credibility rule. It follows that the appellant is entitled to succeed on the second issue argued in this Court.

  166. The respondent filed a notice of contention which raised the argument that the Full Court was entitled to order a new trial because of the unsatisfactory features of the primary judge's reasons and his failure to make findings concerning facts (principally in relation to the issue of damages) which had been established by evidence at the trial[68]. Although, with respect, I would accept some of the criticisms which the respondent has voiced of the reasons of the primary judge in this regard, I do not believe that such criticisms touch the central issue upon which the appellant is entitled to succeed. For example, although I consider that the respondent has made good her criticism of the primary judge's suggestion that the respondent and her husband deliberately and dishonestly allowed their garden to deteriorate, this was only one of a number of findings made adverse to the respondent and her credibility. It was not specifically invoked by the primary judge in his consideration of the causation issue. Nothing in the notice of contention displaces the conclusion which I have reached as to the disposition of the appeal.

    ORDERS

  167. I therefore agree in the orders proposed by Gleeson CJ.

    Callinan J

  168. The respondent, Dr Percival, is an experienced nurse who had obtained a doctorate of philosophy in nursing some years before she consulted the appellant, Dr Rosenberg, who was, and is, an oral and maxillofacial surgeon, on 6 February 1993. By then the respondent was 42 years old and working as a senior lecturer in nursing at Edith Cowan University.

  169. Dr Percival had suffered a worsening condition of malocclusion for some years. Her general dentist Dr Davies referred her to a prosthodontist who told her of the different forms of treatment that she might undergo: restorative dentistry; replacement of crowns; realignment of teeth by orthodontic treatment without any interference with the jaw bone; and, a combination of these and orthognathic surgery involving surgical movement of the jaw. The respondent stressed to the prosthodontist that she wanted the best result.

  170. The respondent, after a consultation with another specialist dentist, Dr Mezger, was referred to the appellant. While she was waiting to see the appellant, she found an information pamphlet produced by, or on behalf of the appellant. The pamphlet spoke of risk in these terms:

    In accepting this surgery, there is always the risk involved in the patient undergoing general anaesthesia. Although these risks are small with modern day anaesthesia, nevertheless this responsibility should be clearly understood. The risks are in fact no greater than for any other surgical procedure.

    The patient will be under the surgeons [sic] care for at least two months, at which time he/she will be referred back for orthodontic management. During this period, a plastic splint, which is a guide and template for closure of the jaws into the new position, will be retained over the maxillary teeth.

    A complete return to normal function may take several weeks or months. It is advisable that contact sporting activities should not be undertaken for a period of six months, during which time the bone healing strengthens. The period of maximum inconvenience will be for approximately two to three weeks after surgery.

    It is stressed that post surgical management by the Orthodontist is required to achieve a good final result. Other Dental Specialists may also be called upon to finalise the case.

  171. The appellant at a second consultation took casts of the respondent's teeth and made other investigations. The respondent had a third consultation with the appellant on 23 April 1993 at which time he recommended that the respondent undergo orthognathic surgery, that is, an osteotomy. It is common ground that the appellant did not advise the respondent of risks, either specific or general, of temporomandibular joint problems following surgery of that kind.

  172. The appellant then wrote to the respondent setting out the approximate cost of the surgery for correction of the malocclusion. The respondent did not make her decision to undergo orthognathic surgery at the consultation on 23 April 1993 but she did subsequently obtain a form of a letter of consent seven or ten days before the surgery and thereafter read and signed it, and gave it to the appellant. Before the surgery a pre-operative radiological examination was carried out and a radiology report was obtained. The radiologist Dr Young reported that the respondent's temporomandibular joints were normal. The respondent read that report. The appellant performed the sagittal split osteotomy that he had earlier recommended, on 6 December 1993.

  173. The respondent began to suffer persistent pain after the operation. Further radiological investigations were undertaken. The appellant wrote to the respondent on 21 February 1994 suggesting that the respondent undergo another operation to mobilise both joints and an arthrocentesis, that is, a "washout" of the joints. The respondent adopted that suggestion and a second operation was carried out by the appellant on 25 February 1994 at Glengarry Hospital.

  174. In March 1994 the appellant referred the respondent for physiotherapy. There was no improvement in the respondent's condition. Accordingly, the appellant recommended that the respondent consult Dr McNamara for a further opinion and treatment.

  175. The respondent had MRI (magnetic resonance imaging) scans in May and September 1994. There was a further meeting between the respondent and the appellant and Doctors McNamara and Mezger in October 1994. These doctors made joint recommendations for a plan of treatment which included further surgery. The respondent rejected all of these recommendations.

  176. From January 1995 Dr Henry took over the management and co-ordination of the treatment of the respondent. She was then treated by Doctors Williamson, Shannon, and Delcanho. When the possibility of further surgery was canvassed, the respondent, not surprisingly, sought a guarantee that the surgery would be successful.

  177. There is no doubt that the respondent has been left with a number of disabling conditions. She has chronic pain and cannot speak loudly, because to do so hurts her jaws. She is unable to eat hard food and her vegetables have to be cooked until they are soft. When the pain is really bad she takes only fluids. She is embarrassed when eating in restaurants because she loses food from her mouth. She has muscle spasms and, once every two or so weeks she is unable to open her mouth. In order to relieve the pain she applies hot-packs and ice-packs. She has had to wear a splint. The respondent takes anti-inflammatory drugs and strong painkillers when necessary, and lesser ones when she can, because she wishes to avoid strong medications. When she has not been able to cope with the pain Dr Shannon has prescribed pethidine. All of this has reduced her enjoyment of life. Some degree of malocclusion remains. She has had to undergo psychiatric treatment. Her capacity to earn income has clearly been adversely affected.

    THE TRIAL

  178. The respondent sued the appellant in negligence in the District Court of Western Australia. The respondent alleged that the appellant was negligent in diagnosing and treating her in various respects. The respondent's pleading also contained a paragraph in these terms:

    24.

    Had the Plaintiff been alerted to the inherent risks involved in the oral surgery and the manipulation of the temporomandibular joints post surgery and had in particular the possible risks and complications of these procedures been brought home to her, she would not have consented to either of these procedures and her submission thereto was therefore obtained not on the basis of informed consent.

  179. The action was heard by Gunning DCJ in August 1997. It is unnecessary to refer to much of the dental and medical evidence given at the trial because the only issues with which this Court is concerned are those raised by par 24 of the Statement of Claim. It is relevant however to refer to evidence which was given at the trial of the incidence of the post-operative conditions which now beset the respondent. Professor Goss, an expert in the field, who was called by the respondent gave evidence in chief as follows:

    HEENAN, MR: Have you seen similar outcomes in comparable surgery? --- In comparable circumstances, namely, a patient who has a mild degree of temporomandibular joint disorder who has a facial osteotomy and ends up with severe temporomandibular joint problems, I have seen two in the last decade and with Dr Percival that makes three.

    Yes. Do you know of other instances from your reading of literature or your position as [a] leading surgeon in Australia? --- Firstly in regard to - those three cases are the only cases of which I am aware in Australia. It doesn't mean to say that they haven't occurred but I am in a position where I very commonly get asked either clinically or legally to give advice on such cases and at the moment I'm not aware of other ones.

  180. In cross-examination he said this:

    To my knowledge there are certainly the two in South Australia and certainly Dr Percival. I am not aware of any others in Australia but would be the first to admit that there probably are although given my interest in chronic pain and given that I am quite commonly asked to give medico-legal advice on these matters, I don't think that there's very many. On that basis, if you take three out of 20,000 then you end up, I guess, with an incidence of 1 in 6000. So I think one can confidently say that the incidence of this problem is certainly measured 1 in thousands and my best guess is somewhere between 1 in 2 and a half and 1 in 6.

  181. The possibility of temporomandibular disorders ("TMD") had been adverted to in dental literature in evidence before the Court. One example of this was to be found in a defensive paper written by Dr Delcanho, addressed to the profession, and designed to protect dentists and oral surgeons against suits by dissatisfied patients. In that paper published in April 1994 he wrote[69]:

    Seemingly in the 1990s, simple recognition of overt TMD signs and symptoms in patients complaining of jaw pain is not the only problem, rather dentists must also be able to assess subtle signs of a TMD in patients presenting for routine dental work who may be unaware of those signs. Such a subtle non-painful 'subclinical' TMD could possibly be aggravated due to trauma to jaw structures that may occur during routine dental treatment procedures. The uninformed patient sees a natural cause and effect relationship, blames the dentist for producing the problem and may be motivated to seek legal compensation. As part of the treatment planning process, the dentist must therefore be able to evaluate the relative risk for proposed dental procedures to either worsen TMD symptoms already present or aggravate an asymptomatic dysfunctional condition into a clinically symptomatic TMD. By performing a screening evaluation for TMD, the dentist is gathering important baseline information and can thereby record the pretreatment functional status of the patient's masticatory system. Following discussion with the patient, treatment can then proceed on an informed consent basis and be structured to minimize the possibility of aggravating either a symptomatic or a subclinical TMD.

    The TMD screening procedure is therefore not only performing a valuable diagnostic service for the patient, but enables the dentist to make informed decisions and then discuss with the patient the possible risks of exacerbating TMD symptoms. By obtaining informed consent from the patient prior to embarking on dental treatment procedures, any subsequent TMD problems that may develop will not be viewed by the patient as being the fault of the dentist with possible legal consequences.

  182. Dr Delcanho concluded as follows[70]:

    Because of their training in anatomy, function, and pathophysiology of the oral and facial structures, dentists have traditionally assumed responsibility for the diagnosis and treatment of TMD. Physicians (and many dentists) have no training in the diagnosis and treatment of these disorders. The recognition of TMD is now considered a professional responsibility on the part of the practising dentist and requires looking beyond morphological relationships of the teeth and jaws.

    Every patient should undergo a simple screening process to differentiate the completely healthy patient from the patient with one or more signs of a TMD. Of medico-legal importance in such cases is the documentation of baseline parameters of the patient's jaw functional status and the obtaining of informed consent before embarking on any dental treatment. This is of special significance where complex treatment plans involving occlusal relationships are envisaged. The predictability of extensive restorative, orthodontic or surgical treatment often rests upon the stability of jaw position which itself is highly affected by the presence of TMD. The screening described in this paper therefore has important implications from both dental treatment planning and medico-legal perspectives. If the described screening process reveals the presence of a TMD, then comprehensive evaluation is needed. This may be done either by the individual dentist, if familiar with TMD, or by referral to another practitioner competent in TMD diagnosis and management.

  183. Dr Delcanho gave evidence at the trial. He accepted that his paper was not directly concerned with warnings about the possible consequences of orthognathic surgery but about the need to take all possible steps before treatment to identify the possibility of their occurrence after it.

  184. Another paper in evidence, by Tucker and Proffit[71] was also concerned with the diagnosis and relative rarity of adverse consequences of the kind suffered by the respondent:

    Malocclusion has frequently been cited as a cause or contributing factor in the occurrence of TM joint pain or dysfunction. This relationship, however, remains extremely controversial ...

    Summarizing the available data, Greene and Marbach suggest that any relationship between malocclusion, morphology, and dysfunction is greatly exaggerated. They failed to document an increased percentage of Class II or III malocclusions among patients presenting for evaluation in two TM joint pain programs ...

    In summary, it appears that some types of malocclusion may slightly predispose patients to TM joint problems, but the relationship is weak enough to be questionable. The malocclusion types that have been indicated as potentially significant - Class II, division 2; Class III with anterior interference - seem to be those most likely to cause mandibular shifts on closure. There is no evidence that this triggers bruxism, but that outcome is possible. An obligatory shift on closure could increase muscle strain, lowering the threshold for hyperactivity. Whether this could lead to internal derangement of the joint remains unknown. In short, if there is a relationship between malocclusion and TM joint problems, the malocclusion probably makes it easier for a patient to hurt himself or herself while clenching or grinding. It is impossible to totally rule out the other mechanisms as well.

  185. It is right to say, as the respondent submits, that there was a considerable volume of expert evidence that the appellant should, out of prudence at least, have given the respondent a warning. It was put this way by Professor Goss in his written report:

    However there is a risk of TM joint problems from orthognathic surgery and, given that the problem did eventuate then under the current informed consent conditions, Dr Rosenberg should have mentioned the risk.

  186. Dr Punnia-Moorthy put the matter no higher than this:

    In my view, it is prudent to warn the patient of likely TMJ problems in the presence of clicking and condylar abnormalities, prior to jaw surgery such as a sagittal split osteotomy of the mandible.

  187. Dr Levant's evidence was relevantly as follows:

    Consequently, I should have expected the patient to have been warned of the possibility of a less than ideal prognosis and to be aware of the additional risks to the jaw joints.

  188. Dr Delcanho wrote this in his report which was tendered in evidence:

    Furthermore, in view of the likely pre-existing TMD, it is my opinion that the possible consequences of orthognathic surgery ie aggravating a pre-existing TMD should have been discussed with your client and signed informed consent obtained.

  189. It may be noted that he spoke of the "likely pre-existing TMD". He was speaking after the event.

  190. Another expert Dr McNamara gave evidence to a not dissimilar effect to that of Dr Delcanho.

  191. Only the appellant, and Professor Norman who did acknowledge a general duty to warn but not a specific one in the case of this respondent, were of a different opinion[72].

  192. The trial took an unusual course, in that, notwithstanding par 24 of the Statement of Claim and the body of expert evidence to be adduced with respect to warnings, the respondent in her evidence in chief and cross-examination, did not say what she would have done had she been informed of the possibility of the complications which in fact ensued. She had to be recalled, with leave, to give evidence about this. She then said:

    If there had been any risk I would not have had the surgery.

  193. In cross-examination this exchange occurred:

    MARTINO, MR: And you were very keen to obtain the best outcome, weren't you? --- I think anyone who said they weren't would be foolish, Mr Martino. Yes, of course I wanted the best outcome and I still want the best outcome.

    You would have remained keen to undergo the treatment and you would in fact have undergone the treatment had you known that there was a risk of you developing symptoms emanating from the temporomandibular joint after the procedure? --- Never. Never.

  194. The respondent's husband also gave evidence, directed mainly to the issue of damages. For reasons which will later appear, I mention that it was not put to him that he and the respondent deliberately allowed their garden to fall into a state of dereliction.

  195. Gunning DCJ formed an adverse view of the respondent's credibility. In a long judgment he summarised the evidence of each witness. He reached the conclusion that there was no evidence, before the first operation conducted by the appellant, that the respondent had any temporomandibular problems, which I take to mean, indications of them. After holding that the appellant had not been negligent in diagnosing or treating the respondent, his Honour turned to the case presented on the issue of the need for a warning.

  196. On this issue also the primary judge found against the respondent. His Honour said:

    It follows that in the circumstances the defendant, prior to the operation, was not negligent in his interpretation of any of the plaintiff's complaints following a thorough examination, or of his interpretation of the x-rays and therefore it follows that there was no known problem to him that could develop that he could communicate to the plaintiff.

    It follows he was not negligent in not warning the plaintiff of any material problem that might develop.

  197. Notwithstanding that those findings would of themselves have led to a verdict for the appellant, his Honour went on to make findings on the issue of causation. In respect of this he was heavily influenced by the respondent's intelligence, training and knowledge as an experienced nurse. Because of these and other factors he concluded as follows:

    I am quite satisfied in the circumstances that even if the plaintiff had been warned of the slight possibility, and certainly it was very slight, of complications she would have proceeded with the surgery in any event.

  198. In taking the adverse view that he did of the respondent's credibility his Honour referred with scepticism or disbelief to a number of matters: the respondent's unwillingness to submit to an MRI; her application for appointment as a professor; her failure to disclose documents relating to that application; her exposure as a nurse to injured people in the past and consequential acquaintance with the risks of surgery; and her husband's neglect of their garden and her claim for the cost of restoration of it.

  199. The trial judge thought it unnecessary to make an assessment of the respondent's damages and dismissed her action although he fixed a sum far short of her claim for the purposes of an assessment of the appellant's costs.

  200. From that judgment the respondent appealed to the Full Court of the Supreme Court of Western Australia (Kennedy, Wallwork and Owen JJ). Kennedy J agreed with Wallwork J that the appeal should be allowed and added some observations of his own. After referring to Rogers v Whitaker (1992) 175 CLR 479 his Honour said that the appellant was in breach of his duty to warn the respondent of a risk of temporomandibular problems and that the trial judge was in error in finding to the contrary.

  201. Kennedy J said that his Honour's conclusion with respect to the absence of any occasion to warn did not depend upon the respondent's credibility. He dismissed the primary judge's finding that even if the respondent had been warned she would have proceeded with the surgery by pointing to a number of factors which he said were objective factors militating against it. It is convenient to comment on these as I list them. The respondent undertook slow, deliberate and extensive investigations and preparations before agreeing to the first operation. As I read the evidence, in fact, all that the respondent did was submit to such examinations and preliminary procedures as were recommended by her general dentist and those other dentists and surgeons to whom he referred her. The next relevant factor, Kennedy J said, was that the procedure was elective and that other procedures only slightly less satisfactory were available. The first may be accepted, but the second accords little or no weight to the evidence that the respondent sought the best result and that would have been produced by the procedures actually adopted had the complications not developed. His Honour also thought it relevant that the first operation was not to correct a cosmetic defect but to prevent further deterioration in the state of her teeth and mouth. It seems to me that this factor tells as much against the respondent's case as it does in favour of it. A person might well be more willing to undergo a procedure to correct a functional problem than simply to improve his or her appearance. Finally, Kennedy J considered that the respondent's background and training inclined her to meticulousness, and that this, in effect, should have provoked an adequate warning rather than the express reassurance that she was given about a full return of function within a week or month. That training also served, as I have already suggested, to give her an acquaintance with surgical procedures, and the possibility that will often, perhaps even always exist, of undiagnosed or undiagnosable underlying conditions, and that tissue and bone may not always react in a predictable way.

  202. After reviewing some of the evidence at the trial Wallwork J said that the primary judge had applied the wrong test. He said that "the relevant risk" was known to the appellant and to the expert witnesses: once there is a risk which is generally known to the profession there is a duty to warn, and it is not necessary to establish that the appellant should have been alerted to any disorder in the respondent's joint.

  203. Wallwork J regarded the primary judge's rejection of the respondent's claim that she would not have had the surgery if she had been warned of the possible risks as erroneous, in view of both the severity of the complications which the respondent suffered, and his Honour's finding that the respondent's husband, together with the respondent, had allowed the garden to deteriorate notwithstanding that neither the respondent nor her husband had been cross-examined about that matter. Exaggeration by the respondent was probably explicable, Wallwork J thought, on the basis of the respondent's depression and chronic anxiety, matters to which the trial judge should have had regard in assessing the respondent's credibility.

  204. Owen J too expressed his agreement with the reasons of Wallwork J but also chose to add some comments of his own. His Honour said that there was a sufficient evidentiary basis for a finding that the risks were material in a relevant sense, and that the appellant was under a duty to warn the respondent of them.

  205. The appeal was accordingly allowed and declarations and orders inter alia as follows were made:

    (i)

    the Respondent was in breach of his duty of care to the Appellant in failing to warn of the risks of temporo-mandibular joint problems and symptoms arising after the procedure which the Appellant underwent.

    (ii)

    the action be remitted to the District Court of Western Australia for a new trial before a different Judge on the remaining issues, namely:

    (+)

    whether the Appellant has suffered loss or damage which has been caused by the Respondent's breach of duties aforesaid;

    (+)

    if so, what is the amount of the loss and damage suffered by the Appellant;

    and for judgment to be entered after the re-trial of those issues accordingly.

    THE APPEAL TO THIS COURT

  206. The appellant appeals to this Court upon the following grounds:

    The Full Court of the Supreme Court of Western Australia erred:

    2.1

    in holding that once there is a risk involved in dental treatment which is generally known to the dental profession, there is a duty to warn a patient of that risk;

    2.2

    in failing to hold that a dental surgeon is under a duty to warn a patient of a material risk of which he knew or ought to have known;

    2.3

    in failing to consider or give reasoned consideration to the specific complication suffered by the patient, the effects thereof, whether and why the risk of that complication was material and if material, the nature of the warning required in the particular circumstances of the case;

    2.4

    in failing to consider or give reasoned consideration to the Appellant's case that the post-operative condition the subject of the Respondent's claim was not a risk of which he knew or ought to have known at the material time;

    2.5

    in having no proper basis or no properly reasoned basis for overturning the learned trial judge's finding that had the Respondent been warned of the relevant risk, she would have proceeded with the surgery in any event;

    2.6

    in failing to hold that the trial judge was entitled to reject the evidence of the Respondent that had she been warned of the risk she would not have proceeded with the surgery in circumstances where her evidence was inherently improbable and not supported by the trial judge's factual findings and the evidence as a whole.

  207. The facts, arguments and division of judicial opinion in this case and Chappel v Hart (1998) 195 CLR 232 are examples of some practical difficulties that are emerging in the application of the decision of this Court in Rogers v Whitaker (1992) 175 CLR 479.

  208. In Rogers v Whitaker, Mason CJ, Brennan, Dawson, Toohey and McHugh JJ stated that a medical practitioner owed a single, comprehensive duty extending to the provision of information in an appropriate case ((1992) 175 CLR 479 at 483). Their Honours said that the standard of care is not to be determined solely, or even primarily, by reference to the practice followed or supported by a responsible body of opinion in the profession. This was, it was said, a matter for the courts to decide after giving weight to the paramount consideration that a person is entitled to make his or her own decisions about his or her life although evidence of acceptable medical practice in that regard is a useful guide for the courts ((1992) 175 CLR 479 at 487).

  209. Their Honours said that a number of factors had to be considered by a medical practitioner in deciding whether to disclose or advise of some risk in a proposed procedure. These included the nature and degree of risks of the proposed treatment. Their Honours then formulated the test in these terms ((1992) 175 CLR 479 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, 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.

  210. What their Honours propounded then were both an objective and a subjective test, that is to say, a universal test for an hypothetical reasonable person in the patient's position, and a test to be applied to the particular patient, even if, perhaps, she or he is an unreasonable one. What this in practice may mean is that the more inquisitive, or demanding, or less or more sophisticated perhaps, or obsessive, or suspicious, or hypochondriacal the patient may be, the greater the need for identification of and elaboration upon the slightest risks because such a patient may be likely to attach significance to them. The plaintiff in Rogers v Whitaker was an incessant questioner of her doctor, and that, taken with the possibility which eventuated, of the devastating disability of blindness, required that she be told of the risk which in fact eventuated ((1992) 175 CLR 479 at 491). Although in view of that conclusion alone of their Honours, the plaintiff was entitled to succeed, they went on to say that perhaps the objective test might also have been satisfied in that case ((1992) 175 CLR 479 at 491).

  211. The principal issue in the appeal to this Court in Chappel v Hart was one of causation. But the way in which that issue was resolved affected the issue of what was relevantly to be taken to be a material risk. There was no challenge to Rogers v Whitaker. That the principal issue was as I have stated it, appears not only from the judgments of the majority but also from the dissenting judgment of McHugh J who said ((1998) 195 CLR 232 at 251 [45]) that to hold the defendant liable on the basis that if the plaintiff had been given a warning of the risk she would have avoided a condition that developed, is simply to apply the test that had been rejected in March v Stramare (E & M H) Pty Ltd (1991) 171 CLR 506 that is, the "but for" test. Hayne J (in dissent) too was of the opinion that the test of common sense and experience for which March v Stramare (E & M H) Pty Ltd stands should be applied ((1998) 195 CLR 232 at 290 [148]).

  212. In Chappel v Hart (1998) 195 CLR 232 the Court held by a narrow majority (Gaudron, Gummow and Kirby JJ; McHugh and Hayne JJ dissenting) that a patient whose condition was relentlessly progressive and who had developed an infection after an operation which was performed with reasonable care and skill, and, which could have developed ((1998) 195 CLR 232 at 237 [2] per Gaudron J) no matter which qualified person might have performed the operation, was entitled to recover damages, because the surgeon had not advised the patient of the risk of infection. This followed, it was held by Gaudron and Kirby JJ, as the patient had said that she would have postponed the operation and sought the most experienced surgeon in the field to do it if she had been so informed, because the degree of risk would have been reduced if the operation had been performed by the most experienced surgeon available. Gummow J was of the opinion that the patient should succeed because she had specifically asked the surgeon about the risk in question, and would not have undergone the operation in his hands if he had warned her about it. Their Honours in the majority also declined to reduce the damages to allow for the possibility that the patient might later have suffered harm of the kind that eventuated independently of the doctor's breach. It was their opinion that this was a speculative possibility only.

  213. Whilst it cannot be doubted that patients should be entitled to be told, in detail, and in terms that they can understand, of the material risks of the procedures available to them, and of other relevant matters such as the costs, duration, and pain of those procedures and their aftermath, "material risk" presents difficulties of definition in practice. This case is an example of such a difficulty, a matter to which I will return.

  214. The decision in Rogers v Whitaker has been received with some consternation by the medical profession[73]. No doubt the manufacturers of bottled drinks viewed the reasoning of the House of Lords in Donoghue v Stevenson [1932] AC 562 in the same way. The common law does however evolve, albeit usually incrementally, with the result that practices and conduct may have to be changed to accord with it. But there is, in my opinion, a real doubt, whether a negative answer to a question, "Would you have had the operation?", artificially posed, years after the event, and answered, almost certainly, after the patient has suffered unexpected complications, and after repeated innocent rehearsal in making a statement and in conference with legal advisors, can ordinarily carry much conviction, or should provide the basis for an undiscounted award of damages, or indeed, damages at all. A disinterested bystander might well say of such an answer, "Of course, naturally she (or he) would say that". This is a case in which that doubt certainly existed. I will revert to it shortly.

  215. There is a related question which presented itself acutely in Chappel v Hart. It was whether the damages should be assessed on the basis that although the plaintiff would not have undergone the surgery then she would still have been at some risk on another occasion in any event. The majority were not prepared to make any discount of the plaintiff's damages on account of that risk. Gaudron J referred to the infection as an extremely rare one, indeed as both random and rare ((1998) 195 CLR 232 at 242 [20]). If, it might be asked, its random and rare incidence did not prevent it from being regarded as a "material risk" then equally why should the risk of its rare and random occurrence following surgery by another surgeon not be a factor in reduction of the plaintiff's damages? The answer may be that the application of a "but for" test allows the plaintiff not only to define the risk but also to define it as a material one. Objectively or statistically, in many cases, the difference in risk if the surgery were performed by a more experienced surgeon might be very small, that is to say non-material[74]. In the Court of Appeal of New South Wales, Handley JA with whom Mahoney P and Cohen AJA agreed, said no more than that, "superior skill and experience could reduce [the] risk"[75]. In fact the lesser degree of risk (if any) was not quantified. It was the plaintiff's assertion of her belief as to it that was effectively determinative of the case in her favour.

  216. Gummow J in Chappel v Hart contemplated the possibility that in some circumstances a defendant might be able to mitigate his or her damages, by proving, in an appropriate way, that independently of a defendant's negligence the plaintiff would or might (beyond speculation) ((1998) 195 CLR 232 at 263 [83] have suffered the relevant harm ((1998) 195 CLR 232 at 262-263 [82]). As his Honour explained, the decision of this Court in Malec v J C Hutton Pty Ltd (1990) 169 CLR 638[76] did not foreclose the possibility of that but directed attention to the real question, which was, whether the plaintiff there would have suffered that harm by the time that he in fact suffered it as a result of the defendant's negligence. It may also perhaps be inferred from his Honour's detailed reference to some of the expert evidence that he did not regard the incidence of perforation causing infection as rare ((1998) 195 CLR 232 at 260-262 [79]-[81]). A further consideration which appears to have been influential in his Honour's reasoning was that, as with Mrs Whitaker in Rogers v Whitaker, the plaintiff Mrs Hart made plain to the appellant that she regarded the obtaining of adequate advice as a matter of central concern (at 262 [81]). Accordingly Gummow J may well have thought the case satisfied the second, that is, the subjective test stated in Rogers v Whitaker.

  217. The other member of the majority, Kirby J, was of the opinion that the measure of damages, and accordingly whether they should be reduced or mitigated, turned upon the pleadings and the way in which the case had been conducted (at 278-279 [100]). These issues his Honour did not doubt would return for consideration to this Court on some other occasion (at 265 [88]).

  218. How far does a prudent professional have to go? Should a prudent doctor not only advise a patient of extremely rare and random risks, but also that there are more experienced surgeons than he or she is, who could be engaged to perform the operation, leaving it to the patient to decide whether to engage such a surgeon? Gaudron J in Chappel v Hart (1998) 195 CLR 232 at 239 [10] thought there was such an obligation in that case having regard to the foreseeability of the risk of infection. In saying, as Mrs Hart did, that she would have taken steps to have the surgery performed by the "most experienced [surgeon] with a record and a reputation in the field" (at 237 [2] per Gaudron J) she was making an assumption that surgery by such a surgeon would or might have reduced the risks of infection. That was an understandable assumption. Gaudron J accepted it as an accurate one ((1998) 195 CLR 232 at 239 [10]). But it may not necessarily be, by any means, a correct assumption in any or all cases. A less experienced surgeon might very well be able to compensate for any lesser experience by the taking of greater pains. A less experienced one might also be better versed in modern techniques. And reputation is not always a reliable guide to competence assuming that there are ways and means available to a patient to ascertain who has the best reputation. On one view, "material risk", certainly as the patient was allowed to identify and define it, was not simply the risk of perforation producing a rare and random infection: it was the risk that a less experienced, it is not clear how much less experienced, surgeon was more likely, albeit only marginally more likely, to make a perforation producing that rare and random infection than an unnamed more experienced surgeon. Views may well differ whether that risk was a material risk.

  219. The inclusion in the test for negligence of a need for foreseeability in a professional context, and perhaps others, does not make it any easier for a court to decide whether there has been negligence. With the growth in scientific awareness by informed people of ordinary imaginative power, it will almost always be possible to say, after an event that it was foreseeable. It should be kept in mind that the word "reasonably" has real work to do in testing whether an event was foreseeable. So too, what is required to establish negligence is a want of reasonable care and skill. In testing whether they are lacking, the word "reasonable" again has real work to do. Neither professionals nor other providers of services should have too onerous obligations of foresight or care and skill imposed upon them[77].

  220. Questions of the kind that I have just discussed were either latent or fell to be directly answered in this case, either at first instance or in the Full Court. I do not therefore have to express a final opinion on two matters that I seriously doubt:

    1. that in the absence of pre-operative signs of temporomandibular problems in Dr Percival's jaw not detected on examination, consultation or radiological examination, and in view of the rarity of the complications suffered, as deposed to by Professor Goss, what subsequently happened to her was a material risk of which she was required to be warned; and,

    2. that there was any absence of reasonable care and skill on the appellant's part in not warning Dr Percival of the possibility of post-operative temporomandibular problems.

    I am relieved from that task because the primary judge found, and he was entitled to find, that the respondent had not proved that she would not have undergone the osteotomy had she been told that there was a risk that she might develop temporomandibular problems following it. In trying to prove that matter the respondent was not assisted by the fact that she gave the evidence, almost as an afterthought, by leave, after she had been cross-examined and re-examined, and at a time when perhaps it might have been apparent that her primary case of negligence in diagnosis and treatment was looking unpromising. I have already said in my discussion of the reasons for judgment of the Full Court why I do not think that the factors that were referred to by the respondent as objective factors supporting her claim that she would not have had the operation did in fact tend to establish that matter.

  221. It is perfectly understandable that a person who has suffered what the respondent suffered here would say, and might also even have come to believe implicitly that she would not have had the operation had she known of the risk which has in fact materialised. That would usually be, and it probably was here an honest belief on the part of the respondent at the time that she gave her evidence. However, the true position is much more likely to be, no matter what a plaintiff may have honestly come to believe, that she cannot really say, in an absolute way, that she would have not had the operation. The much more likely position is that perhaps she might not have. And such an answer honestly given, might, in an appropriate case raise a question which was not decided in Chappel v Hart, whether a plaintiff should recover damages for the loss of a chance of not undergoing the operation and running the risks of adverse consequences of it.

  222. It does seem to me here that the primary judge was unduly critical of the respondent with respect to the caution which she adopted towards treatment now available to her to ameliorate her condition. He was also too critical of the way in which she dealt with her application for appointment as a professor. And the primary judge further erred, in treating the respondent's and her husband's claims regarding the deterioration of their garden, as dishonest claims when no suggestions to that effect were made in cross-examination of them. In Boland v Yates Property Corporation Pty Ltd (1999) 74 ALJR 209 at 270-271 [299] I expressed my concern about the artificiality in some situations, of excising clearly erroneous parts of a judgment from other parts in which no error is apparent. That process may assume an even more unconvincing air when the error relates to credibility and credibility is a significant issue in the trial.

  223. Concerns of this kind certainly gave me serious pause in this case. But in the end I formed the view that the primary judge's conclusion as to the incredibility of Dr Percival's answer that she would "never" have had the operation must be accepted notwithstanding them. In this case these concerns are outweighed by the fact of the respondent's awareness, as a highly experienced practising and teaching nurse, of risks going beyond those of frequent or regular occurrence, her stressed desire for the "best result" before the surgery, the inherent undispelled improbability of a categorical denial that she would have had the operation, and the time and circumstances in which she made the denial that the primary judge rejected.

    ORDERS

  224. For these reasons, I would allow the appeal with costs and order that the respondent pay the appellant's costs of the trial and the appeal to the Full Court.


[1] See, for example, Jones v Bartlett (2000) 75 ALJR 1 at 5-6 [19]; 176 ALR 137 at 141-142; Modbury Triangle Shopping Centre Pty Ltd v Anzil (2000) 75 ALJR 164 at 167 [17], 183-184 [109]; 176 ALR 411 at 415, 438.

[2] Rogers v Whitaker (1992) 175 CLR 479 at 490; Chappel v Hart (1998) 195 CLR 232 at 246 [32], 272 [93].

[3] United States: Canterbury v Spence 464 F 2d 772 at 791 (1972); Canada: Reibl v Hughes [1980] 2 SCR 880 at 898-899.

[4] Chappel v Hart (1998) 195 CLR 232 at 246, fn (64).

[5] Abalos v Australian Postal Commission (1990) 171 CLR 167 at 179.

[6] Watt or Thomas v Thomas [1947] AC 484 at 488; see also Paterson v Paterson (1953) 89 CLR 212 at 224.

[7] Chappel v Hart (1998) 195 CLR 232 at 239 [10] per Gaudron J, 260 [76] per Gummow J.

[8] After the decision of McNair J in Bolam v Friern Hospital Management Committee [1957] 1 WLR 582; [1957] 2 All ER 118, applied by the House of Lords in Sidaway v Governors of Bethlem Royal Hospital [1985] AC 871.

[9] Chapman v Hearse (1961) 106 CLR 112 at 120-121. See also Perre v Apand Pty Ltd (1999) 198 CLR 180 at 248-249 [185]-[186].

[10] Environment Agency v Empress Car Co (Abertillery) Ltd [1999] 2 AC 22 at 29. See also Marks v GIO Australia Holdings Ltd (1998) 196 CLR 494 at 532 [109]; Campomar Sociedad, Limitada v Nike International (2000) 74 ALJR 573 at 591-592 [98], 592-593 [103]; 169 ALR 677 at 702-703, 704.

[11] See also her Honour's statement to like effect in Chappel v Hart (1998) 195 CLR 232 at 238-239 [7]-[8] and the remarks of Dixon J in Betts v Whittingslowe (1945) 71 CLR 637 at 649.

[12] See Gover v State of South Australia and Perriam (1985) 39 SASR 543 at 566; Ellis v Wallsend District Hospital (1989) 17 NSWLR 553 at 559-560, 581.

[13] Jackson & Powell on Professional Negligence, Third Cumulative Supplement (2000) to the Fourth Edition, (1997), §6-175.

[14] From a judgment of the Supreme Court of Western Australia, Full Court: Percival v Rosenberg [1999] WASCA 31.

[15] The history is set out in State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (In Liq) (1999) 73 ALJR 306 at 322-325 [72]-[80]; 160 ALR 588 at 609-613.

[16] Jones v Hyde (1989) 63 ALJR 349 at 351-352; 85 ALR 23 at 27-28; Abalos v Australian Postal Commission (1990) 171 CLR 167 at 179; Devries v Australian National Railways Commission (1993) 177 CLR 472 at 479, 482-483.

[17] see Olbourne, "The Influence of Rogers v Whitaker on the Practice of Cosmetic Plastic Surgery", (1998) 5 Journal of Law and Medicine 334 at 337 ("Olbourne").

[18] The relevant paragraph of the respondent's pleadings is set out in the reasons of Callinan J at [178].

[19] See eg Chappel (1998) 195 CLR 232 at 272-273 [93].

[20] He cited Devries v Australian National Railways Commission (1993) 177 CLR 472; see [1999] WASCA 31 at [108].

[21] Reibl v Hughes [1980] 2 SCR 880; see also Hopp v Lepp [1980] 2 SCR 192 at 210.

[22] F v R (1983) 33 SASR 189 at 192-193; see Rogers (1992) 175 CLR 479 at 487-488; Fridman, "Judicial Independence of a Different Kind", in Mullany (ed), Torts in the Nineties (1997) 305 at 320.

[23] See also Sidaway v Governors of Bethlem Royal Hospital [1985] AC 871 at 895.

[24] As required by the Bolam test: Bolam [1957] 1 WLR 582; [1957] 2 All ER 118.

[25] Schloendorff v Society of New York Hospital 105 NE 92 at 93 (1914). See also Salgo v Leland Stanford Jr University Board of Trustees 317 P 2d 170 at 181 (1957); cf Monks, "The Concept of Informed Consent in the United States, Canada, England and Australia: A Comparative Analysis", (1993) 17 University of Queensland Law Journal 222 at 223 ("Monks").

[26] Canterbury v Spence 464 F 2d 772 at 781 (1972); see also Olbourne, (1998) 5 Journal of Law and Medicine 334 at 337; Kerridge and Mitchell, "Missing the Point: Rogers v Whitaker and the Ethical Ideal of Informed and Shared Decision-making", (1994) 1 Journal of Law and Medicine 239 at 240 ("Kerridge and Mitchell").

[27] Bolitho v City and Hackney Health Authority [1998] AC 232; cf Jones, "The Bolam Test and the Responsible Expert", (1999) 7 Tort Law Review 226 at 235-241.

[28] Jones, "Doctor Knows Best?", (1984) 100 Law Quarterly Review 355; Kennedy, "The Patient on the Clapham Omnibus", (1984) 47 Modern Law Review 454; Hodgkinson, "Medical Treatment: Informing Patients of Material Risks", (1984) Public Law 414; Teff, "Consent to Medical Procedures: Paternalism, Self-determination or Therapeutic Alliance?", (1985) 101 Law Quarterly Review 432; Kennedy and Grubb, Medical Law, 2nd ed (1994) at 200; Irvine, "The Patient, the Doctor, their Lawyers and the Judge: Rights and Duties", (1999) 7 Medical Law Review 255 at 258-259; Brazier and Miola, "Bye-Bye Bolam: A Medical Litigation Revolution?", (2000) 8 Medical Law Review 85. The approach in Bolam has now been abandoned in South Africa: Castell v De Greef 1994 (4) SA 408. As to Malaysia see Shuaib, "Rogers v Whitaker: The end of the Bolam's saga in medical negligence cases in Malaysia?", (2000) 16 Professional Negligence 25.

[29] Schuck, "Rethinking Informed Consent", (1994) 103 Yale Law Journal 899 at 959 ("Schuck").

[30] Kerridge and Mitchell, (1994) 1 Journal of Law and Medicine 239 at 243-244; see also Robertson, "Informed Consent Ten Years Later: The Impact of Reibl v Hughes", (1991) 70 Canadian Bar Review 423 at 431, 438 ("Robertson"); Fleming, The Law of Torts, 9th ed (1998) at 122-123.

[31] Schuck, (1994) 103 Yale Law Journal 899 at 904, 933-934.

[32] Schuck, (1994) 103 Yale Law Journal 899 at 905, 938-941, 942-948; Olbourne, (1998) 5 Journal of Law and Medicine 334 at 345.

[33] Schuck, (1994) 103 Yale Law Journal 899 at 906, 959; Herz, Looman and Lewis, "Informed Consent: Is It a Myth?", (1992) 30 Neurosurgery 453 at 455.

[34] Schuck, (1994) 103 Yale Law Journal 899 at 911 referring to Tunkl v Regents of University of California 383 P 2d 441 (1963).

[35] Olbourne, (1998) 5 Journal of Law and Medicine 334 at 342.

[36] Schuck, (1994) 103 Yale Law Journal 899 at 926.

[37] Kerridge and Mitchell, (1994) 1 Journal of Law and Medicine 239.

[38] Olbourne, (1998) 5 Journal of Law and Medicine 334 at 344.

[39] Schuck, (1994) 103 Yale Law Journal 899 at 919-920.

[40] Olbourne, (1998) 5 Journal of Law and Medicine 334 at 344; Mendelson, "The Breach of the Medical Duty to Warn and Causation: Chappel v Hart and the Necessity to Reconsider Some Aspects of Rogers v Whitaker", (1998) 5 Journal of Law and Medicine 312 at 317; see also McInnes, "Failure to warn in medical negligence - a cautionary note from Canada: Arndt v Smith", (1998) 6 Torts Law Journal 135 at 143 ("McInnes") citing Cory J in Arndt v Smith [1997] 2 SCR 539 at 553; Girgis, Thomson and Ward, "'The Courts Expect the Impossible': Medico-legal Issues as Perceived by New South Wales General Practitioners", (2000) 7 Journal of Law and Medicine 273.

[41] F v R (1983) 33 SASR 189 at 192-193, 196 per King CJ; Kirby, "Patients' rights: Have we gone too far?", (1993) 2 Australian Health Law Bulletin 38 at 40; Olbourne, (1998) 5 Journal of Law and Medicine 334 at 342.

[42] Englard, "Informed Consent - The Double-faced Doctrine", in Mullany and Linden (eds), Torts Tomorrow: A Tribute to John Fleming (1998) 152 at 153, 161 ("Englard").

[43] Schuck, (1994) 103 Yale Law Journal 899 at 927.

[44] Schuck, (1994) 103 Yale Law Journal 899 at 931; Chalmers and Schwartz, "Rogers v Whitaker and informed consent in Australia: a fair dinkum duty of disclosure", (1993) 1 Medical Law Review 139 at 148; cf Cassidy, "Malpractice - Medical Negligence in Australia", (1992) 66 Australian Law Journal 67.

[45] Katz, "Informed Consent - Must It Remain a Fairy Tale?", (1994) 10 Journal of Contemporary Health Law and Policy 69 at 84.

[46] Katz, The Silent World of Doctor and Patient (1984) at 60 cited in Robertson, (1991) 70 Canadian Bar Review 423 at 440.

[47] Law Reform Commission of Victoria, Informed Decisions About Medical Procedures, Report No 24 (1989) at 3 cited in Robertson, (1991) 70 Canadian Bar Review 423 at 439.

[48] Kerridge and Mitchell, (1994) 1 Journal of Law and Medicine 239.

[49] United States, President's Commission For the Study of Ethical Problems in Medicine and Biomedical and Behavioural Research, "Informed Consent as Active, Shared Decision-making", in Making Health Care Decisions (1982), vol 1 at 390 cited in Kerridge and Mitchell, (1994) 1 Journal of Law and Medicine 239 at 242-243. See also Katz, "Disclosure and Consent: In Search of Their Roots", in Milunsky and Annas (eds), Genetics and the Law II (1980) 121 at 122; Englard, in Mullany and Linden (eds), Torts Tomorrow: A Tribute to John Fleming (1998) 152 at 159-160.

[50] Trindade, "Disclosure of Risks in Proposed Medical Treatment", (1993) 109 Law Quarterly Review 352 at 356.

[51] Olbourne, (1998) 5 Journal of Law and Medicine 334 at 346.

[52] Olbourne, (1998) 5 Journal of Law and Medicine 334 at 347.

[53] Australia, Review of Professional Indemnity Arrangements for Health Care Professionals, Final Report (1995) cited in Reinhardt, "Compensation and professional indemnity in health care - the Final Report of the Tito Committee", (1996) 4 Torts Law Journal 173 at 173; Fleming, The Law of Torts, 9th ed (1998) at 123.

[54] Osborne, "Causation and the Emerging Canadian Doctrine of Informed Consent to Medical Treatment", (1985) 33 Cases on the Canadian Law of Torts 131 at 143 noted in Robertson, (1991) 70 Canadian Bar Review 423 at 433.

[55] Reibl v Hughes [1980] 2 SCR 880 at 928; see Robertson, (1991) 70 Canadian Bar Review 423 at 425.

[56] Bolam [1957] 1 WLR 582 at 590-591; [1957] 2 All ER 118 at 124; Chatterton v Gerson [1981] QB 432 at 445; McInnes, (1998) 6 Torts Law Journal 135 at 136.

[57] Honoré, "Causation and Disclosure of Medical Risks", (1998) 114 Law Quarterly Review 52 at 54-55; Monks, (1993) 17 University of Queensland Law Journal 222 at 231, 233; McInnes, (1998) 6 Torts Law Journal 135 at 137-138.

[58] Giesen and Hayes, "The Patient's Right to Know - A Comparative View", (1992) 21 Anglo-American Law Review 101 at 116-117.

[59] Osborne, "Causation and the Emerging Canadian Doctrine of Informed Consent to Medical Treatment", (1985) 33 Cases on the Canadian Law of Torts 131 at 142-144; Monks, (1993) 17 University of Queensland Law Journal 222 at 232-233.

[60] Chappel v Hart unreported, New South Wales Court of Appeal, 24 December 1996 at 7; cf Mendelson, "The Breach of the Medical Duty to Warn and Causation: Chappel v Hart and the Necessity to Reconsider Some Aspects of Rogers v Whitaker", (1998) 5 Journal of Law and Medicine 312 at 315. See also Olbourne, (1998) 5 Journal of Law and Medicine 334 at 342; Herz, Looman and Lewis, "Informed Consent: Is It a Myth?", (1992) 30 Neurosurgery 453 at 455-456; McInnes, (1998) 6 Torts Law Journal 135 at 142; Schuck, (1994) 103 Yale Law Journal 899 at 919.

[61] Fleming, The Law of Torts, 9th ed (1998) at 123; see also Englard, in Mullany and Linden (eds), Torts Tomorrow: A Tribute to John Fleming (1998) 152 at 157-159.

[62] Ellis (1989) 17 NSWLR 553 at 560, 581; Hribar v Wells (1995) 64 SASR 129 at 140; Chappel (1998) 195 CLR 232 at 245 [28], 246 [32], 247 [34].

[63] Monks, (1993) 17 University of Queensland Law Journal 222 at 233.

[64] See the reasons of Callinan J at [201] with which I agree.

[65] Jones v Hyde (1989) 63 ALJR 349; 85 ALR 23; Abalos v Australian Postal Commission (1990) 171 CLR 167; Devries v Australian National Railways Commission (1993) 177 CLR 472. The principal exceptions are collected in SRA (1999) 73 ALJR 306 at 331-332 [93]; 160 ALR 588 at 620-622.

[66] SRA (1999) 73 ALJR 306 at 328-329 [88], especially n 109; 160 ALR 588 at 617-618.

[67] In the United Kingdom there are now eight blind or partially sighted judicial officers, suggesting that faith in the essential ability of judges to tell truth from falsehood by visual appearance is on the decline: "Crime court debut for blind judge", (2000) 26 Commonwealth Law Bulletin 468.

[68] cf Jones v The Queen (1989) 166 CLR 409 at 411, 414; Palmer v The Queen (1992) 66 ALJR 270; 106 ALR 1; Bank of South Australia Ltd v Ferguson (1998) 192 CLR 248 at 263 [33]-[35].

[69] Delcanho, "Screening for temporomandibular disorders in dental practice", (1994) 39 Australian Dental Journal 222 at 222.

[70] Delcanho, "Screening for temporomandibular disorders in dental practice", (1994) 39 Australian Dental Journal 222 at 226-227.

[71] Tucker and Proffit, "Temporomandibular Dysfunction: Considerations in the Surgical-Orthodontic Patient", in Proffit and White, Surgical-Orthodontic Treatment, (1991) 660 at 662-663.

[72] No point was taken with respect to the opinion evidence as to what should or should not have been done. See Naxakis v Western General Hospital (1999) 197 CLR 269 at 306, fn 137.

[73] See Mendelson, "The breach of the medical duty to warn and causation: Chappel v Hart and the necessity to reconsider some aspects of Rogers v Whitaker", (1998) 5 Journal of Law and Medicine 312 at 317.

[74] cf the discussion of increase in risk as a non-material factor in Vetter v Lake Macquarie City Council [2001] HCA 12 at [32] per Gleeson CJ, Gummow and Callinan JJ.

[75] See (1998) 195 CLR 232 at 241 [18] per Gaudron J. But see her Honour's opinion at 241 [19]: "that the risk of injury would have been less if [the plaintiff] had retained the services of the most experienced surgeon in the field".

[76] See also Wilson v Peisley (1975) 50 ALJR 207 at 210; 7 ALR 571 at 576-577; Allied Maples Group Ltd v Simmons & Simmons [1995] 1 WLR 1602 at 1609-1610; [1995] 4 All ER 907 at 914-915; Athey v Leonati [1996] 3 SCR 458 at 470-471.

[77] cf Boland v Yates Property Corporation Pty Ltd (1999) 74 ALJR 209 at 273 [311]-[312] per Callinan J; 167 ALR 575 at 659-660.


Cases

Rogers v Whitaker (1992) 175 CLR 479; Bolam v Friern Hospital Management Committee [1957] 1 WLR 582; [1957] 2 All ER 118; Sidaway v Governors of Bethlem Royal Hospital [1985] AC 871; Devries v Australian National Railways Commission (1993) 177 CLR 472; Bennett v Minister of Community Welfare (1992) 176 CLR 408; Chappel v Hart (1998) 195 CLR 232; Breen v Williams (1996) 186 CLR 71; Hughes v Lord Advocate [1963] AC 837; Maloney v Commissioner for Railways (NSW) (1978) 52 ALJR 292; Perre v Apand Pty Ltd (1999) 198 CLR 180; F v R (1983) 33 SASR 189

Wyong Shire Council v Shirt (1980) 146 CLR 40; March v Stramare (E&M H) Pty Ltd (1991) 171 CLR 506; Gover v State of South Australia and Perriam (1985) 39 SASR 543; Smith v Barking, Havering and Brentwood Health Authority (1994) 5 Med LR 285; Hocking v Bell (1947) 75 CLR 125; Puntoriero v Water Administration Ministerial Corporation (1999) 199 CLR 575; State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (In Liq) (1999) 73 ALJR 306; Canterbury v Spence 464 F 2d 772 (1972); Hribar v Wells (1995) 64 SASR 129; Ellis v Wallsend District Hospital (1989) 17 NSWLR 553; Smith v Tunbridge Wells Health Authority (1994) 5 Med LR 334; Donoghue v Stevenson [1932] AC 562; Malec v J C Hutton Pty Ltd (1990) 169 CLR 638

Authors and other references

Tucker and Proffit, "Temporomandibular Dysfunction: Considerations in the Surgical-Orthodontic Patient", in Proffit and White, Surgical-Orthodontic Treatment, (1991)

Katz, "Informed Consent - Must It Remain a Fairy Tale?", (1994) 10 Journal of Contemporary Health Law and Policy 69 at 84.

Katz, The Silent World of Doctor and Patient (1984) at 60 cited in Robertson, (1991) 70 Canadian Bar Review 423 at 440

Delcanho, "Screening for temporomandibular disorders in dental practice", (1994) 39 Australian Dental Journal 222

Representations

C J McLure QC with D J Martino for the appellant (instructed by Clayton Utz)

E M Heenan QC with P A Monaco for the respondent (instructed by Godfrey Virtue & Co)


all rights reserved