Australian Health & Medical Law

Issue 112

16 September 2009

Editor: Anne Wardell BA, LLB


Failure to warn: Is it about me, you or everyone else?

By BILL MADDEN* and TINA COCKBURN** and as reproduced with permission of the Law Society Journal (Vol 47, No 3, pp 74–77).

Recent cases, including G & C v Down and Neal v Ambulance Service of NSW, have implications for medical negligence litigation.

Judicial comment over the past two years has helped our understanding of the practical effect of the failure-to-warn provision in the NSW Civil Liability Act 2002, but there remain some pathways not well trodden some six years after the legislation’s enactment.1

The applicable provision remains s 5D(3), which provides that if it is relevant to the determination of factual causation to determine what the person who suffered harm would have done if the negligent person had not been negligent. The matter is to be determined subjectively in the light of all relevant circumstances, and any statement made by the person after suffering the harm about what they would have done is inadmissible except to the extent (if any) that the statement is against their interest.

In the context of warnings, however, a medical practitioner might well come to ponder the question: “Is it about me, you or everyone else?”.

Is it about me? The surgeon’s own experience and success rates

Chappel v Hart [1998] HCA 55; (1998) 195 CLR 232 is a decision that goes some way towards supporting the recognition of a duty on the part of a surgeon to advise the patient of their limited experience in a particular surgical field.2 Gaudron J stated (at [10]): “if the foreseeable risk to Mrs Hart was the loss of opportunity to undergo surgery at the hands of a more experienced surgeon, the duty would have been to inform her that there were more experienced surgeons practising in the field”.

Overseas authority of similar type is also somewhat sparse, but includes an interlocutory application in Johnson v Kokemoor 1996 Wis 2d 615; 545 NW 2d 495 (1996), a failure-to-warn case where the surgeon quantified but understated the risk of a surgical procedure. On appeal, the Wisconsin Supreme Court upheld the admissibility of evidence concerning the surgeon’s limited experience and the relative risks of morbidity and mortality as between experienced and inexperienced surgeons.

In this context, Brus v Australian Capital Territory [2007] ACTSC 83 is of interest. In that case, the amended pleadings alleged negligence in permitting a surgical registrar, Dr Cree, to perform a procedure, failing to inform the plaintiff that the operation was to be performed by Dr Cree, and failing to inform the plaintiff of Dr Cree’s qualifications and experience.3

The court found that the first defendant hospital was negligent in permitting Dr Cree to perform a procedure that was beyond the capacity of a second year trainee with adverse training assessments for surgical skills.4 Justice Connolly held: “I do not accept that there is a general duty of care in a public hospital to in effect provide public patients with a choice of doctor, or to appraise a patient as to the academic standing of a registrar. However, there is a duty on a hospital to ensure that it provides patients with suitably qualified staff. In this case, a registrar known to the hospital to have major difficulties in surgical techniques for a level 2 registrar was held out to Dr Heaton as a level 3 registrar, and he allowed her to perform a procedure that he would not have permitted a level 2 registrar to perform.”5

Although it was submitted that the complication was causally related to the defendant’s negligence on the basis that, had the procedure been performed by a consultant surgeon or properly qualified practitioner on another day, the extubation incident may not have occurred,6 given Justice Connolly’s comment that the case was not pleaded as a failure to warn a patient of a foreseeable complication and as all the experts agreed that this was not a complication about which a patient should be warned,7 the issue of whether the risk of complication was greater given the surgeon’s qualifications and experience did not directly arise for consideration.

More recently, a first instance decision at the District Court level in South Australia, G & C v Down [2008] SADC 135, raised some interesting issues regarding the content of the duty to warn in the context of disclosure of a surgeon’s personal “failure” rate as opposed to generic data8 on failed or other adverse outcomes.9 The first plaintiff G consulted with the defendant Dr Downs for the purpose of undergoing tubal ligation surgery, the objective of which was to obtain permanent contraception.10 A laparoscopic tubal ligation was performed in April 2002 but she became pregnant a few months after the surgery and gave birth to her fifth child in March 2003.11

The claim made by G and her de facto partner D was one of type generally known as “wrongful birth”, including a claim for damages for the cost of raising the child.12 The issue for consideration became the plaintiffs’ allegation of negligence in failing to adequately warn the first plaintiff regarding the risk that the tubal ligation procedure may not succeed in permanently sterilising her, and thus expose her to the risk of becoming pregnant again.13

Dr Downs told G of what he understood to be his personal failure rate for the sterilisation procedure, namely one in 2,000, rather than referring to a 1998 pamphlet published by the Royal Australian College of Obstetricians and Gynaecologists (RACOG) which, in relation to tubal ligation, said the risk of failure was extremely small but approximately 1 in 500 women who have had the operation will become pregnant.14 There was clearly a gap between the RACOG document failure rate of 1/500 and that cited by Dr Down as his personal failure rate of 1/2,000.

Although it was acknowledged by Justice Robertson that “the personal experience of the gynaecologist would be of great significance to the patient”,15 he held that Dr Down was in breach of his duty to warn as he did not make it clear that the numerical ratio of 1/2,000 related to his personal experience.16 In order to provide a proper balance for the patient, the literature failure rate should have been conveyed.17 Even if Dr Down had made clear that the figures related to his personal experience, that would not have spared him from the obligation to refer to the literature figures so as to provide a proper balance to complete the picture.18 The court commented that “the fact that the gynaecologist does not know whether there has been a pregnancy, or more than one, following a procedure which has not been reported to him provides support for the proposition that the personal figures should be balanced by furnishing the literature figures”.19

A secondary issue which arose in G & C v Down was whether the correct approach to the communication of risk warnings is “adjectival” or “arithmetic”.20 The trial judge held that the fact that the defendant also described the risk in an adjectival way (that is, “rare, remote”21) did not remedy his omission to make it clear that the numerical ratio related to his own experience.22 Furthermore, even if the defendant had made it clear that the figures related to his personal experience and described the risk adjectivally, that would not have spared him from the obligation to refer to the literature figures in an arithmetic (or statistical) way.23

Is it about you? The patient’s response to a risk warning

The plaintiff in G & C v Down gave evidence about what she would have done, which the court examined as hypothetical evidence.24

The Civil Liability Act 1936 (SA) does not exclude evidence from a plaintiff about what they would have done, as was proposed by the Review of the Law of Negligence 2002 in recommendation 29(g), which was enacted in NSW as s 5D(3) Civil Liability Act 2002, and also in Queensland,25 Tasmania26 and Western Australia.27

Of course, in a failure-to-warn argument, the outcome depends on what the individual plaintiff would have done had they been adequately warned. In G & C v Down, despite the court’s having found that Dr Down had breached his duty as set out above, on the question of causation, the claim failed, as G was unable to persuade the court that she would have declined the sterilisation even if told that the risk was 1/500 rather than 1/2,000.28

The response of the individual patient to a warning or the provision of information was also addressed recently in Neal v Ambulance Service of NSW [2008] NSWCA 346, a decision of Basten JA with Tobias JA and Handley AJA agreeing. Mr Neal, the plaintiff, suffered a serious blow to the head while walking alone in Newcastle. Police discovered him and called an ambulance. He rejected assistance from the ambulance officers. Since he was clearly inebriated, the police took him into custody under the Intoxicated Persons Act 1979 (NSW). The following morning, his condition was observed to deteriorate and, being unable to rouse him easily, the police had him taken to a hospital. A CT scan showed an extradural haematoma with a fracture to the skull, and Neal later underwent surgery to drain the extradural haematoma. The plaintiff in his claim alleged that certain persisting disabilities were allegedly caused by the delay in taking him to hospital.

Without traversing all of the issues on appeal, it is sufficient for present purposes to highlight the difficulty faced by the plaintiff in that, even accepting that the ambulance officers should have informed the police of the plaintiff’s need for medical assessment, and accepting that the police would have taken him to hospital, the plaintiff would still have failed to establish liability on the part of the NSW Ambulance Service unless he satisfied the court that he would have accepted medical assessment and treatment from a hospital.29

In that context, no direct evidence from the plaintiff as to what he might or might not have done would have been admissible by reason of s 5D(3)(b): any statement made by the person after suffering the harm about what they would have done is inadmissible except to the extent (if any) that the statement is against their interest.30 Such evidence as did exist was unhelpful to the plaintiff.31

The crucial portion of the Court of Appeal’s consideration of s.5D(3)(b) appears at [40]-[42] as follows:

“Whatever the real purpose of the provision, the issue for determination is how a court is now to identify what course the plaintiff would have taken, absent negligence. That assessment might include evidence of the following: (a) conduct of the plaintiff at or about the relevant time; (b) evidence of the plaintiff as to how he or she might have felt about particular matters; (c) evidence of others in a position to assess the conduct of the plaintiff and his or her apparent feelings or motivations; and (d) other matters which might have influenced the plaintiff.

Properly understood, the prohibition on evidence from the plaintiff about what he or she would have done is of quite limited scope. Thus, the plaintiff cannot say, ‘If I had been taken to hospital I would have agreed to medical assessment and treatment’. Indeed, as the Negligence Review32 recognised, such evidence would be largely worthless. However, the plaintiff might have explained such evidence along the following lines: ‘I recall on the trip to the police station that I began to feel less well; my state of inebriation was also diminishing; I began to worry about the pain in my head ...’

That evidence (entirely hypothetical in the present case) would not be inadmissible. If accepted, it might provide a powerful reason for discounting any inference as to future conduct drawn from the past refusal of treatment. It would constitute evidence as to the plaintiff’s position, beliefs and fears”.

The succinct remarks above echo the approaches taken in earlier decisions,33 though perhaps with a little more emphasis on what the plaintiff may have “felt” about particular matters.

Is it about everyone else? Objective and subjective approaches

In the context of evaluating what would have been the patient’s response to the risk had it been disclosed, the focus has traditionally been on a subjective approach — the particular patient.

The question arises as to the extent to which there may be an overlap between such a subjective consideration of the individual person, as required by Rogers v Whitaker [1992] HCA 58; (1992) 175 CLR 479 and by s 5D(3)(a),34 and objective considerations — the likely response of a reasonable person in such circumstances.

The Court of Appeal recognised a role for both objective and subjective considerations. Basten JA commented at [42]:

“Because an inference would need to be drawn from that evidence, no doubt the court would take into account the likely response of a reasonable person in such circumstances. That is consistent with the Act requiring that the matter be determined ‘subjectively in the light of all relevant circumstances’”.

In this context, Trindade comments that there may be an objectifying of the causation test.35 This has been previously recognised from time to time by the courts, such as in Micallef & Anor v Minister for Health of The State of Western Australia [2006] WASCA 98.36

The approach of the NSW Court of Appeal in developing an inference about an individual plaintiff by reference to the likely response of a reasonable person in such circumstances would appear to leave the pathway open for evidence of a statistical nature.37 The Canadian approach in such circumstances38 has been described as a modified objective test,39 so perhaps the Australian approach should now be described as a modified subjective test.40

Such an approach may be instructive and applicable in a claim by a woman as to whether she would have terminated a pregnancy if told that antenatal investigations indicated that her child would suffer Down syndrome — statistical evidence41 may be admissible on the percentage of women who elect termination in such circumstances.42 Loane Skene discusses the converse situation, in the context of survey data on the small number of potential cosmetic surgery patients who decline surgery when presented with adverse outcomes data.43

Support for the admissibility of statistical evidence may be found in Seltsam Pty Limited v McGuiness [2000] NSWCA 29 at [79] per Spigelman CJ:

“Evidence of possibility, including expert evidence of possibility expressed in opinion form and evidence of possibility from epidemiological research or other statistical indicators, is admissible and must be weighed in the balance with other factors, when determining whether or not, on the balance of probabilities, an inference of causation in a specific case could or should be drawn. Where, however, the whole of the evidence does not rise above the level of possibility, either alone or cumulatively, such an inference is not open to be drawn.44

Concluding remarks

In this consideration of hypothetical outcomes, we must now return to the opening hypothetical medical practitioner’s question: “Is it about me, you or everyone else?”

G & C v Down provides support for the proposition that, at least on occasion, the answer may well depend on “me”, the information given about the particular circumstances of the surgeon.45
• In the context of determining what would have been the patient’s response to the risk warning if given, there can be no doubt that the answer will depend on the patient (you) and their hypothetical subjective response had a proper warning been given.
• However, identification of what “everyone else” (the reasonable person) may have done in the particular circumstances now appears to be of potential relevance, albeit within the framework of the development of an inference as to what a certain individual would have done in the circumstances of the plaintiff.

Over time, it seems likely that these issues will be further ventilated in the trial and appellate courts throughout Australia.

Editor’s note

The decision at first instance in G & C v Down has been upheld on appeal by the full court of the Supreme Court of South Australia in G, PA & C, P v Down [2009] SASC 217. The court held that G would have proceeded with surgery even if Dr Down had provided her with complete information as to the risks of pregnancy. The breach of duty was therefore not causative of G’s loss. The court confirmed that the “causation test in these circumstances is a subjective one”.

* Bill Madden is the national practice group leader (medical negligence) at Slater & Gordon Lawyers and a part-time lecturer at the University of Western Sydney.

** Tina Cockburn is a senior lecturer at the Queensland University of Technology.


An earlier discussion in the context of medical treatment warnings can be found in our article “What The Plaintiff Would Have Done” (NSW Law Society Journal, October 2006).


Clarke & Oakley, Informed Consent and Clinician Accountability; The Ethics of Report Cards on Surgeon Performance, Cambridge University Press, Cambridge, 2007 at pp.116-117.




At [59]-[60].


At [62].


See [66].


Discussed at [10]; see also [67].


See at [88] the discussion of generic data as statistics are produced from the results of surgery conducted by surgeons of all levels of experience and operative skills. The failure rates of sterilisation depend on many factors including the experience of the operator, technique, operating theatre equipment, as well as patient factors such as obesity, adhesions and scarring.


For a comprehensive discussion see Clarke & Oakley, op cit.






Such a claim would now be limited by Civil Liability Act 1936 (SA) s 67; see also Civil Liability Act 2002 (NSW) s 71 and Civil Liability Act 2003 (Qld) s 49A, 49B. The background to and framework of this legislation is discussed in Madden & McIlwraith, Australian Medical Liability, LexisNexis Butterworths Sydney 2008 at Ch 14.






At [140].


At [144]-[145].


At [141].






For a more detailed discussion of this issue see Earlam R., et al, “Obtaining consent for an operation: a choice of words or numerical probabilities?” Clinical Risk 2007; 13:45–52. The authors conclude that however inadequate words are in communicating risk, they should be used as a discourse opener and backed up by numerical figures.


Examples of such words are contained in [137].




At [143]. Cf. Carey-Hazell v Getz Bros & Co (Aust) Pty Ltd [2004] FCA 853 at [224]: “I do not however consider that it was necessary that the applicant be advised of the rate of risk of thromboembolisms developing on a St Jude Medical or other mechanical valve. A patient would need to know that the risk is a real one, but I do not think reference to statistics such as two to five per cent would inform them further in any meaningful way.”; see also Jeffrey v Witherow [2007] WADC 20 at [153]: “In my view, whether or not it is necessary to express a risk in percentage terms will depend on the particular facts of each case. Where the risk is that an operation may not succeed, as distinct from a risk that a new pain will develop, it is less likely to be necessary to express that risk in percentage terms.”


See [166]ff.


Civil Liability Act 2003 (Qld) s 11.


Civil Liability Act 2002 (Tas) s 13.


Civil Liability Act 2002 (WA) s 5C.








At [16]: “The ambulance officers gave evidence that, in the circumstances, they could not examine or treat the plaintiff or take him to hospital, without his consent. They clearly did not have his consent. Although there was a case run at trial that greater efforts should have been taken to obtain his consent and carry out a more complete investigation, her Honour rejected that complaint and it was not reagitated on appeal. ”


Review of the Law of Negligence 2002, Final Report. Available online at


See for example Hoyts Pty Limited v Burns [2003] HCA 61; (2003) 201 ALR 470, Richards v Rahilly [2005] NSWSC 352, Elbourne v Gibbs [2006] NSWCA 127 as discussed Madden & McIlwraith, Australian Medical Liability, LexisNexis Butterworths, Sydney, 2008 at p 245ff.


Review of the Law of Negligence at [7.40].


Trindade, Cane and Lunney, The Law of Torts in Australia 4th edition, Oxford University Press 2007 p.542. The authors helpfully refer to an article by Addison, “Negligent Failure To Inform” [2003] 11 Torts Law Journal 165.


At [64]. “The test of causation is subjective although objective factors are used to test the patient’s reliability: Chappel v Hart (1998) 195 CLR 232; Rosenberg v Percival (2001) 205 CLR 434. It is the case that the appellants’ evidence as to their hypothetical decision must be assessed in light of the circumstances at the time rather than after the risk had materialised in the form of another pregnancy ... .”


Conversely the absence of such evidence may be relevant: “There was no evidence of the efficacy of a warning sign of the type which the judge found to be required, nor any real consideration of such efficacy ... The topic of efficacy overlaps with the issue of causation, but it may be said in general that the duty of care does not require occupiers to take precautions that are unlikely to be efficacious” at [51] in Secretary to the Department of Natural Resources & Energy v Harper [2000] VSCA 36.


See Reibl v Hughes [1980] 2 SCR. 880; Arndt v Smith [1997] 2 SCR 539; see MH. Tse, “Tests For Factual Causation: Unravelling the mystery of material contribution, contribution to risk, the robust and pragmatic approach and the inference of causation”, (2008) 16 TLJ 249 at 267 fn 63.


Discussed in Rosenberg v Percival per Kirby J at [20] as cited in G & C v Down [2008] SADC 135 at [153]. Referred to in a different context in Bayley v Police [2007] SASC 411 at [45].


A similar approach may in practice be applied in England. See, for example, Birch v University College London Hospital NHS Foundation Trust [2008] EWHC 2237 (QB) at [81] where the court, in accepting the evidence of the plaintiff as to what she would have done if provided with information on the comparative risks of two procedures, also looked at her medical history of previous ischaemic lesions of the same nature in the immediate past, which had resolved spontaneously without the need for risk-carrying tests. (The assistance of Dr Jock Mackenzie, Leigh Day & Co Solicitors, London in drawing the authors’ attention to this decision is gratefully acknowledged.)


The helpful remarks on this issue by Mr Richard O’Keefe, Barrister, Maurice Byers Chambers, Sydney and Dr John O’Loughlin, Adelaide, are acknowledged with thanks. The authors understand that 15% of Australian women elect to continue with a pregnancy after being diagnosed with aneuploidy; hence 85% elect termination of the pregnancy. See also the article by MJ. Korenromp et al, “Maternal decision to terminate pregnancy in case of Down syndrome”, Volume 196 Issue 2, (2007) American Journal of Obstetrics & Gynecology at p 149.


Section 5D(3)(b) was applied in KT v PLG [2006] NSWSC 919, where Simpson J (at [44]) excluded that part of the evidence which amounted to a statement by the plaintiff that, had she known the true duration of her pregnancy, she would not have undergone termination of the pregnancy. However, the outcome of the matter ultimately did not turn on that issue.


Skene, L., Law and Medical Practice, 3rd edition LexisNexis Butterworths 2008 at pp 207–208.


For a commentary on this decision see Freckelton, “Epidemiology Evidence” (2000) 8 JLM 133. See also Danne, “McTear v Imperial Tobacco: Understanding the role and implications of expert epidemiological evidence in scientific litigation”,(2006) 13 JLM 471.


Subject of course to the well accepted principles in Rogers v Whitaker [1992] HCA 58; (1992) 175 CLR 479 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.

IMPORTANT DISCLAIMER. No person should rely on the contents of this publication without first obtaining advice from a qualified professional person. This publication is sold on the terms and understanding that (1) the authors, consultants and editors are not responsible for the results of any actions taken on the basis of information in this publication, nor for any error in or omission from this publication; and (2) the publisher is not engaged in rendering legal, accounting, professional or other advice or services. The publisher, and the authors, consultants and editors, expressly disclaim all and any liability and responsibility to any person, whether a purchaser or reader of this publication or not, in respect of anything, and of the consequences of anything, done or omitted to be done by any such person in reliance, whether wholly or partially, upon the whole or any part of the contents of this publication. Without limiting the generality of the above, no author, consultant or editor shall have any responsibility for any act or omission of any other author, consultant or editor.


101 Waterloo Road, North Ryde, NSW 2113.

Postal address: GPO Box 4072, Sydney, NSW 2001.

Document Exchange: Sydney DX 812.

Head Office North Ryde Ph: (02) 9857 1300

Fax: (02) 9857 1600

CCH Customer Support Ph: 1 300 300 224

Fax: 1 300 306 224. For the cost of a local call anywhere within Australia.

All rights reserved. No part of this work covered by copyright may be reproduced or copied in any form or by any means (graphic, electronic or mechanical, including photocopying, recording, recording taping, or information retrieval systems) without the written permission of the publisher.