Laferrière v. Lawson, [1991] 1 S.C.R. 541
Dr. Ray Lawson Appellant
v.
Me Nicole Laferrière in her capacity as
testamentary executor of the
late Mireille Fortier‑Dupuis Respondent
indexed as: laferrière v. lawson
File No.: 21334.
1990: March 22; 1991: March 21.
Present: Lamer C.J.* and La Forest, L'Heureux‑Dubé, Sopinka, Gonthier, Cory and McLachlin JJ.
on appeal from the court of appeal for quebec
Civil responsibility ‑‑ Medical malpractice ‑‑ Causation ‑‑ Theory of loss of chance ‑‑ Doctor performing biopsy but failing to inform patient of cancerous condition found or follow up on patient's health in appropriate manner ‑‑ Patient later dying of generalized cancer ‑‑ Whether action against doctor can succeed even where it is not proven that patient's fate would have been different absent the doctor's fault ‑‑ Whether theory of loss of chance should be introduced into Quebec civil law in matters of medical responsibility.
Physicians and surgeons ‑‑ Medical malpractice ‑‑ Causation ‑‑ Theory of loss of chance ‑‑ Doctor performing biopsy but failing to inform patient of cancerous condition found or follow up on patient's health in appropriate manner ‑‑ Patient later dying of generalized cancer ‑‑ Whether action against doctor can succeed even where it is not proven that patient's fate would have been different absent the doctor's fault ‑‑ Whether theory of loss of chance should be introduced into Quebec civil law in matters of medical responsibility.
Respondent is testamentary executor of the late D, who consulted appellant about the presence of an abnormal lump in her breast. Appellant performed a biopsy and removed the lump, which was found to be cancerous, but did not inform D of the cancerous condition or arrange any follow‑up treatment. D's health later deteriorated; she was diagnosed as having generalized cancer and underwent surgery and various treatments. She brought an action against appellant for damages, but died before the proceedings had been completed. Respondent continued the suit, which the trial judge dismissed on the merits. The Court of Appeal reversed the decision. The majority of the court found that appellant's fault resulted in the loss of a real and serious chance to benefit from proper medical care, which should give rise to damages.
Held (La Forest J. dissenting): The appeal should be allowed in part.
Per Lamer C.J. and L'Heureux‑Dubé, Sopinka, Gonthier, Cory and McLachlin JJ.: Respondent's claim for damages for anguish and frustration must be accepted, since appellant's fault was directly related to the great, unnecessary psychological stress D faced upon learning that she had gone four years without knowing of her cancerous condition and without undergoing the follow‑up, monitoring and treatment that was appropriate in her case. The loss of chance analysis used to support the claim for damages for physical pain and suffering and premature death is inappropriate, however, at least in cases where death or sickness has already occurred. In such cases classical principles of causation suffice and are essential in order for individual responsibility to attach.
The independent recognition of a lost chance is questionable in all but the exceptional classic cases, and there is certainly no reason to extend such an artificial form of analysis to the medical context, where faults of omission or commission must be considered alongside other identifiable causal factors in determining what has produced the particular result in the form of sickness or death. In France and Belgium, while there is little doctrinal resistance to the application of loss of chance in the classic cases, the theory has been severely criticized when applied in the medical context as being an attempt to avoid the requirement of certain causation. In Quebec, in cases which have dealt with loss of chance in the medical context, the courts have generally focussed on the actual damage which has occurred rather than the lost chance itself, perhaps because of less rigid rules regarding causation. Such an approach, which ensures that the causal link between the fault and the actual situation now experienced by the plaintiff is established at least on the balance of probabilities, is preferable. The plaintiff is aided in establishing his case by presumptions and factual and statistical evidence. In some cases, where a fault presents clear danger for the health and security of the patient and where such a danger materializes, it may be reasonable for a judge to presume the causal link between the fault and such damage, unless there is a strong indication to the contrary.
The rules of civil responsibility require proof of fault, causation and damage. Both acts and omissions may amount to fault and both may be analyzed similarly with respect to causation. Causation in law is not identical to scientific causation, and must be established on the balance of probabilities, taking into account all the evidence: factual, statistical and that which the judge is entitled to presume. Statistical evidence may be helpful as indicative but is not determinative. Even where statistical and factual evidence do not support a finding of causation on the balance of probabilities with respect to death or sickness, such evidence may justify a finding of causation with respect to lesser damage, such as shorter life or greater pain. If, after consideration of all the factors, a judge is not satisfied that the fault has, on his or her assessment of the balance of probabilities, caused any real damage, then recovery should be denied.
While the evidence here amply supports the trial judge's finding that appellant's fault could not be said to have caused D's death seven years after the first diagnosis of breast cancer, she experienced a type of psychological suffering directly related to appellant's failure to inform his patient of her condition. Further, appellant's failure to follow up on his patient probably denied D the benefit of earlier treatment which would have translated into some real improvement in her condition, and this deprivation should be recognized and compensated.
Per La Forest J. (dissenting): The appeal should be dismissed for the reasons given by Jacques J.A. Civil law recognizes that the damage resulting from loss of chance is compensable in this case. Despite an inclination to award more modest damages, the opinion of the majority of the Court of Appeal should stand on this question.
Cases Cited
By Gonthier J.
Referred to: Brabander v. Goulet, [1985] C.A. 36; Wilson v. Rowswell, [1970] S.C.R. 865; Cass. civ. 1st, May 31, 1988, Bull. civ. 1988, I, no 165, p. 114; Cass. civ. 1st, October 11, 1988, Bull. civ. 1988, I, no 281, p. 192; Cass. civ. 1st, March 29, 1989, Bull. civ. 1989, I, no 147, p. 97; Cass. civ. 1st, June 7, 1989, Bull. civ. 1989, I., no 230, p. 154; Cass. civ. 1st, January 10, 1990, Bull. civ. 1990, I, no 10, p. 8; Cass. civ. 1st, February 7, 1990, Bull. civ. 1990, I, no 39, p. 30; Brussels, February 12, 1957, Pas. 1958, II, 1; Liège, 7th Ch., November 24, 1971, unreported, R.G. no 1201/70; Brussels, January 19, 1965, Pas. 1966, II, 13; Brussels, 2nd Ch. civ., October 22, 1982, unreported, appeal dismissed Cass. 1st Ch., January 19, 1984, Pas. 1984, I, 548; Tardif v. Laverrière, Sup. Ct. Mégantic, No. 6210, November 10, 1976 (summarized at [1976] C.S. 1803); Grenier v. Gervais, [1950] Que. K.B. 60; Lacourcière v. Laplante, [1976] C.A. 433; Beaupré v. Joly, [1971] C.S. 199; Sol‑Air B.G. Inc. v. Marsh & McLennan, [1988] R.R.A. 206; Lachambre v. Perreault (1983), [1990] R.R.A. 397; Zuk v. Mihaly, [1989] R.R.A. 737; Gburek v. Cohen, [1988] R.J.Q. 2424; Lapointe v. Hôpital Le Gardeur, [1989] R.J.Q. 2619; Snell v. Farrell, [1990] 2 S.C.R. 311; Shawinigan Engineering Company v. Naud, [1929] S.C.R. 341; Morin v. Blais, [1977] 1 S.C.R. 570; Laurentide Motels Ltd. v. Beauport (City), [1989] 1 S.C.R. 705; J.E. Construction Inc. v. General Motors du Canada Ltée, [1985] C.A. 275; Dodds v. Schierz, [1986] R.J.Q. 2623; Hotson v. East Berkshire Area Health Authority, [1987] A.C. 750.
Statutes and Regulations Cited
Canadian Charter of Rights and Freedoms, s. 7.
Civil Code of Lower Canada, art. 1205.
Authors Cited
Aubry et Rau. Droit civil français, 8e éd., t. VI‑2, Responsabilité délictuelle. Par Noël Dejean de la Bâtie. Paris: Librairies Techniques, 1989.
Anrys, Henri. La responsabilité civile médicale. Bruxelles: Ferdinand Larcier, 1974.
Baudouin, Jean‑Louis. La responsabilité civile délictuelle. Cowansville: Yvon Blais Inc., 1985.
Boré, Jacques. "L'indemnisation pour les chances perdues: une forme d'appréciation quantitative de la causalité d'un fait dommageable", J.C.P. 1974.I.2620.
Chabas, François. Annotation of Cass. crim., January 9, 1979, J.C.P. 1980.II.19272.
Chabas, François. "Vers un changement de nature de l'obligation médicale", J.C.P. 1973.I.2541.
Chartier, Yves. La réparation du préjudice dans la responsabilité civile. Paris: Dalloz, 1983.
Dorsner‑Dolivet, Annick. Contribution à la restauration de la faute, condition des responsabilités civile et pénale dans l'homicide et les blessures par imprudence: à propos de la chirurgie. Paris: L.G.D.J., 1986.
Durry, Georges. "Responsabilité civile" (1967), 65 Rev. trim. dr. civ. 153.
Hennau‑Hublet, Christiane. L'activité médicale et le droit pénal: Les délits d'atteinte à la vie, l'intégrité physique et la santé des personnes. Bruxelles: Émile Bruylant, 1987.
Honoré, A. M. Causation and Remoteness of Damage, ch. 7, International Encyclopedia of Comparative Law, vol. XI, Torts. Edited by André Tunc. Tubingen, West Germany: J.C.B. Mohr, 1971.
Mémeteau, Gérard. "Perte de chance en droit médical français" (1986), 32 McGill L.J. 126.
Molinari, Patrick. "La responsabilité civile de l'avocat" (1977), 37 R. du B. 275.
Penneau, Jean. La responsabilité médicale. Paris: Sirey, 1977.
Pineau, Jean et Monique Ouellette. Théorie de la responsabilité civile, 2e éd. Montréal: Thémis Inc., 1980.
Savatier, René. "Le droit des chances et des risques, dans les assurances, la responsabilité civile dans la médecine, et sa synthèse dans l'assurance de responsabilité médicale" (1973), 44 Rev. gén. ass. terr. 457.
Savatier, René. "Une faute peut‑elle engendrer la responsabilité d'un dommage sans l'avoir causé?", D.1970.Chron.123.
Tancelin, Maurice. Des obligations: contrat et responsabilité, 4e éd. Montréal: Wilson & Lafleur, 1988.
Tourneau, Philippe le. La responsabilité civile, 3e éd. Paris: Dalloz, 1982.
Vacarie, Isabelle. "La perte d'une chance", [1987] 3 R.R.J. 903.
Viney, Geneviève. Traité de droit civil, t. IV, Les obligations: la responsabilité‑‑conditions. Paris: L.G.D.J., 1982.
APPEAL from a judgment of the Quebec Court of Appeal, [1989] R.J.Q. 27, 20 Q.A.C. 52, 49 C.C.L.T. 309, reversing a decision of Deslongchamps J. dismissing respondent's action for damages against appellant. Appeal allowed in part, La Forest J. dissenting.
Allan R. Hilton, Julie Chenette and Serge Gaudet, for the appellant.
Lynne Kassie and Jean‑Pierre Sheppard, for the respondent.
//Gonthier J.//
The judgment of Lamer C.J. and L'Heureux-Dubé, Sopinka, Gonthier, Cory and McLachlin JJ. was delivered by
Gonthier J. -- This case deals with the legal consequences of a doctor's failure to inform his patient of a cancerous condition and, subsequently, to follow up on the patient's health in the appropriate manner. The Court is asked to consider whether an action can succeed even where it is not proven that the patient's fate would have been different absent the doctor's fault. In particular, the Court must examine the theory of "perte de chance", or loss of chance, and determine whether it should be introduced into the civil law of Quebec in matters of medical responsibility.
I -- Facts and Proceedings
The respondent, Me Nicole Laferrière, acts in her capacity as testamentary executor of the late Mrs. Mireille Fortier-Dupuis. Mrs. Dupuis began proceedings against the appellant, Dr. Ray Lawson, in November 1975, claiming the sum of $150,000 as damages arising out of the alleged non-fulfillment by Dr. Lawson of his obligations as a medical professional. Mrs. Dupuis died in 1978, before these proceedings had been completed. She was, however, able to testify before her death and did so, at the request of the appellant, on June 21, 1976. Following a judgment of Guérin J. on September 23, 1981, Me Laferrière was authorized to continue the suit on behalf of Mrs. Dupuis. A re-amended declaration, dated November 23, 1983, claimed a total of $250,000.
In 1970, Mrs. Dupuis, at that time 48 years of age, became increasingly concerned about the presence of an abnormal nodule or lump in her right breast. She consulted her gynecologist who recommended a mammogram, the results of which proved negative. Apparently, these results did not allay Mrs. Dupuis' anxiety. In March 1971, she read an article about Dr. Ray Lawson in the weekend magazine of a Montréal newspaper. She noted that he was an international authority on the treatment of breast cancer, that he used the most up-to-date equipment for the detection of breast cancer and that his progressive approach to treatment of such cancer avoided any unnecessary removal of breast tissue. Mrs. Dupuis contacted Dr. Lawson at his Westmount Breast Centre and made an appointment for March 10, 1971.
At the Centre, Mrs. Dupuis underwent a number of diagnostic procedures, including mammogram and thermogram. As a result of these tests, Dr. Lawson recommended an excisional biopsy, or surgical removal of the abnormal mass for more accurate diagnosis. Mrs. Dupuis accepted this proposal.
On April 12, 1971, Mrs. Dupuis was admitted to the Royal Victoria Hospital. Two days later, Dr. Lawson performed the surgery, described in the hospital records as "breast biopsy and excision of lump of right breast". The pathology report, dated April 20, 1971, indicated "intraductal carcinoma with infiltrative growth", that is, cancer of the breast.
Mrs. Dupuis was discharged from the hospital on April 15, 1971. She later saw Dr. Lawson at an out-patient clinic and discussed routine matters such as the healing of the scar left after surgery. According to the trial judge's findings, she was not informed that the lump in her breast was cancerous, and she was not advised as to the appropriate post-operative treatments. No long-term follow-up was arranged for her by Dr. Lawson.
From 1971 to 1974, Mrs. Dupuis underwent regular gynecological check-ups and had no cause during that time to worry about her health. In the fall of 1974, her health began to deteriorate and by March 1975, an eyelid disorder (known as the Claude Bernard-Horner syndrome) developed which caused her doctors to suspect that a cancerous condition had taken hold. One of her doctors looked more closely into his patient's history and by obtaining records from the Royal Victoria Hospital discovered the 1971 diagnosis of cancer of the breast. This information was made known to Mrs. Dupuis in April, 1975.
Subsequently, Mrs. Dupuis underwent surgery to remove nodules which had newly appeared on the right breast. This intervention revealed systemic metastases or generalized cancer requiring removal of the ovaries. Following surgery, Mrs. Dupuis received various treatments, including chemotherapy. She died on January 27, 1978.
II -- Judgments of the Courts Below
The Superior Court
Deslongchamps J. first reviews the facts and concludes that the appellant breached two of the duties traditionally expected of a medical professional: the duty to inform the patient and the duty to follow up on the patient's condition.
As to the first duty, the learned trial judge states that the appellant neither informed the patient that she had cancer of the breast nor advised her as to the various ways in which it could be treated:
[translation] I accordingly conclude that the defendant did not inform Mrs. Dupuis that she had cancer. In the case at bar this duty to inform was all the more crucial as there were various ways of treating breast cancer in 1971. By depriving Mrs. Dupuis of this information on her state of health, the defendant also deprived her of her right to choose the treatments which might have been available to her at that time.
Moreover, even if I had concluded that the defendant informed Mrs. Dupuis, there is nothing in the evidence on which I can conclude that the defendant discussed with Mrs. Dupuis, or informed her of, the various treatments available.
The defendant accordingly breached his duty to inform.
He observes with regard to the second duty that the appellant never initiated nor suggested any follow-up, despite the fact that such would have been the normal procedure regardless of the method of treatment selected by doctor and patient.
Turning to proof as to the causal link, Deslongchamps J. approaches the problem in the following way:
[translation] As the defendant's fault has been established, it remains to be determined whether that fault was the cause of the damage alleged, namely the premature death of Mrs. Dupuis and the pain, anxiety, hardship and loss of enjoyment of life she may have undergone.
He then reviews the expert evidence and notes that in 1971 there were at least two schools of thought on how to treat breast cancer. The standard procedure at that time was to perform a radical mastectomy, removing the tumor along with surrounding breast tissue and nodes, and to follow up with irradiation. A non-conformist but medically recognized procedure favoured by the appellant involved removing the tumor alone and employing no irradiation. Expert opinion at the time was, in the opinion of the trial judge, divided as to the relative merits of these two procedures and showed no statistically relevant difference in survival rate.
Deslongchamps J. concludes as follows:
[translation] All this expert evidence, including the statistical results obtained regarding the survival of patients suffering from breast cancer, cannot persuade me that it was probable in 1971 that Mrs. Dupuis' chances of survival would have been greater if she had undergone a treatment other than that given by the defendant.
Unfortunately, all the scientific evidence only confirms the insidious and unforeseeable nature of the development of a cancer, including the degree and duration of the suffering and hardship associated with such an illness.
The Court accordingly concludes that there is no causal link between the defendant's fault and the damage claimed by the plaintiff.
The learned trial judge accordingly rejects the respondent's action on the merits. He acknowledges, however, the injustice of requiring the respondent to pay the substantial costs of the action given the clear fault of the appellant, and, accordingly, he accedes to the respondent's claim for costs.
The Court of Appeal
The appeal was heard by a bench composed of Vallerand and Jacques JJ.A. and Moisan J. (ad hoc) (Laferrière v. Lawson, [1989] R.J.Q. 27). All three judges gave reasons, and together they provide an excellent discussion of the complicated issues which we encounter in this case. I will first summarize the concurring opinions of Jacques J.A. and Moisan J. and will then treat the dissenting opinion of Vallerand J.A.
Jacques J.A. reviews the facts of the case without significant variation from the account provided by the trial judge. He notes that the trial judge based his decision on the absence of a causal link between the fault of the appellant and the damage suffered by Mrs. Dupuis. This prompts Jacques J.A. to reconsider the case by focussing on the head of damages referred to as "la perte d'une chance" or the loss of a chance, a characterization which was not directly considered by the trial judge.
Quoting French and Quebec sources, Jacques J.A. points out that loss of chance was initially viewed with circumspection but is now a recognized head of damages. He states that the classic case of loss of chance arises where a lawyer or notary allows a client's rights to lapse (he cites, by way of example, Brabander v. Goulet, [1985] C.A. 36), a situation which is also recognized in the common law (and here he cites Wilson v. Rowswell, [1970] S.C.R. 865).
Turning to the facts of this case, Jacques J.A. states that it is clear that Mrs. Dupuis lost a chance to obtain the treatment that her condition required and that this loss was definitive. More specifically, the failure of her doctor to inform her properly and to follow up on her condition meant that she could not opt for the potential benefits of irradiation and chemotherapy treatments and could not benefit from the earliest discovery and treatment of any recurrence of the cancer. The fault of Dr. Lawson therefore resulted in the loss of a real and serious chance to benefit from proper medical care.
In the opinion of Jacques J.A., the extent of the damage is to be determined according to the likelihood of success of such proper medical care. Had cure been a certainty, the damages would have been pitched at the maximum level. In the present case, medical science held out some hope, however small. After acknowledging the difficulties inherent in quantifying damage, he estimates the value of the lost chance at $50,000.
In coming to this conclusion, he stresses that his observations are not inconsistent with the findings of the trial judge. First, he notes, at p. 36, that the fault which deserves attention is not the choice of a surgical technique in 1971 but rather the failure to inform the patient and follow up on her condition:
[translation] The fault alleged against Dr. Lawson is not that he chose one type of procedure rather than another. The fault is that he failed to inform the patient of her condition, with the result that she was unable to take advantage of the available care.
Furthermore, Jacques J.A. stresses that the trial judge dealt only with one aspect of the damage, that which was said to be linked to the choice of the proper surgical procedure in 1971. The conclusion regarding the absence of a causal link is not disputed at this point in the analysis. Jacques J.A. differentiates his approach from that of the trial judge, at pp. 36-37:
[translation] [The trial judge's] conclusion is not challenged on appeal, as the appeal deals only with a particular aspect of the damage. The probability that available care would have been more or less successful is only one of the factors and relates only to assessing one aspect of the damage. Apart from that aspect, there is what was described by Dr. Crile as follows:
And I would like to say in conclusion that it is worthwhile to have patients seen and treated early -- it is the only thing we know that if [sic] of any good -- and I think that with the simplification of the treatment of breast cancer, women are beginning to loose [sic] their fear of the treatment of breast cancer and, perhaps, they will come earlier. And that will be a significant way of improving the survival.
This damage is all that is at issue on appeal. As this specific aspect was not dealt with by the trial judge, I am of the view that there is a basis for intervening.
Mrs. Dupuis lost this opportunity to have follow‑up care and take advantage of available treatment.
As noted earlier, Jacques J.A. then proceeds to establish the extent and value of the chance which Mrs. Dupuis lost, focussing not on the consequences of the surgical procedure chosen by Dr. Lawson in 1971 but on the benefits which would have accrued to her had the proper information and follow-up been provided.
I turn now to the concurring reasons of Moisan J.
He introduces his reasons by noting that the evidence indicated that Mrs. Dupuis was the type of person who, properly informed of the nature of her illness, would have done everything possible to protect her health. In his view, she would have ensured that she had adequate follow-up for herself and would have accepted such appropriate treatments as were available. The difficulty in this case, in his view, lies not in the facts but in the law, particularly as concerns the notion of loss of chance.
Moisan J. takes care first to define what he understands by the word "chance". He puts aside a definition of chance as "blind luck" and chooses instead a definition which can be described as an opportunity whose realization is linked to factors such as hard work, determination, information and decision-making. Such a chance is larger or smaller depending on whether the appropriate action is taken at the opportune moment. And it can only be taken where one is aware of the existing situation and the risks which surround it.
He observes that Mrs. Dupuis was unaware of the true state of her health, of the serious risks associated which applied to her case, and of the means at her disposal in order to ward them off. Accordingly, she could not take advantage of the chance which presented itself for early detection and treatment of any new cancer. This amounted to damage linked to the fault of the doctor.
He then addresses the objections to this approach. It was argued that there was neither certainty nor probability that, even with all the information and means at her disposal, Mrs. Dupuis would have succeeded in prolonging her life. He agrees that there was no certainty, but he questions the view that positive results were simply possible and not probable. He notes that even in 1971 medical science had many tests and treatments at its disposal and that these represented a chance, or an opportunity in his terms, to be cured, or at least to prolong one's life. To deny this would be to assume that the methods had no value or, at best, that they represented no more than pure chance or blind luck, and Moisan J. is not prepared to assume either. At page 38, he states:
[translation] A prognosis is difficult to make in situations where a person's health and life are at stake. Maintaining health, a return to health, the prolonging of life are influenced by a host of factors such as heredity, early treatment, positive or negative reaction to treatment and undoubtedly a great many other factors. Even assuming that all the factors involved may have complicated, even conflicting or negative, influences on the outcome, no one can absolutely exclude the probability of greater longevity or even the possibility of a cure . . .
The difficulty of determining the more or less great possibility or, if you like, the more or less slim probability -- these concepts are in my view so close to each other that they sometimes overlap -- results from the fact that one has to speculate on the consequences of the events which might have occurred but did not occur and necessarily will not occur.
I hope that I do not misrepresent the learned judge in interpreting these statements as meaning that whatever the likelihood of possible cure or probable improvement one cannot deny that an opportunity presented itself, and that it was lost as a result of the doctor's fault. Further, Moisan J. acknowledges the difficulty inherent in speculating on what might have been, a particularly demanding task when dealing with events which are future and uncertain, but he stresses that courts regularly perform this task. In such cases, judges refer to well-known notions such as [translation] "the behaviour of a reasonable and prudent man, the natural order of things, the usual sequence of cause and effect and, in short, the normal and ordinary course of events" (p. 39), and they use such notions to establish, at p. 39:
[translation] . . . what the situation would have been if the event involving fault had not given it a different direction. Rejecting extreme positive or negative scenarios, seeking an objective, broad and general assessment and not a subjective estimate closely connected to the victim, they tend to arrive at the situation that would reasonably and probably have prevailed.
He also acknowledges that the task is particularly difficult in medical cases where so many unknown factors are at work. However, this does not, in his view, license the courts to abandon the task, adopt the most pessimistic scenario and deny that the patient had any chance or opportunity to improve her condition.
Moisan J. criticizes the trial judge for giving insufficient consideration to the opportunity which Mrs. Dupuis missed as a result of Dr. Lawson's inaction. There was too much attention paid to the choice of a particular surgical method in 1971 and to the insidious and often fatal character of the disease. Moisan J. states emphatically that in his view Mrs. Dupuis had a reasonable and probable chance to improve her lot had she been informed of her condition and correctly advised as to the proper preventive and curative measures to take.
By way of final comments, he deals with the question of whether the loss of a chance or opportunity can by itself give rise to damages, that is, independently of any prediction as to what the situation might have been in the absence of the alleged fault. In his view, this question should be answered in the affirmative. The loss of a chance -- understood in his terms as an opportunity -- is a result or a consequence. It is not a cause, as some have argued.
In his view, the cause in this case was Dr. Lawson's failure to inform and follow up, and the consequence of this fault was the loss of Mrs. Dupuis' right to make an informed decision regarding the action she should take to promote her health. Moisan J. concludes that this was [translation] "an unlawful deprivation of a right and of a freedom of choice, caused by fault" (p. 40), and he relates this deprivation to s. 7 of the Canadian Charter of Rights and Freedoms.
He summarizes his reasons at p. 40:
[translation] In my view loss of chance is first and foremost loss of the exercise of a right, in this case the right to dispose of one's person, the right to take decisions affecting one's life and health oneself. The Court must provide a penalty for such a loss: it must order at least nominal compensation.
It must then go further and consider the possible positive results that might follow from the situation which would have existed if the fault had not been committed.
In coming to a decision on the damages appropriate in such a case, Moisan J. points to the great frustration and anguish which Mrs. Dupuis must have felt upon learning that her cancer was in an advanced stage and that it might have been effectively treated and perhaps cured had it been dealt with properly at an earlier time. Taking this psychological damage together with the loss of chance, he expresses his agreement with the global award of $50,000 suggested by Jacques J.A.
The dissenting reasons of Vallerand J.A. address the arguments put forward in both majority opinions. The learned judge begins by setting out the factual findings at trial, by which he feels bound, and he relates these to the standard civil law approach to civil liability, at p. 29:
[translation] However appropriate the follow‑up and treatment that might have been undertaken at the outset may have been, it is not probable and, needless to say, not certain that a remission, much less a cure would have been the outcome.
Those are the facts which the trial judge accepted at the conclusion of a careful study of the evidence, and disregarding the strenuous objections of the respondent physician, I can find no reason to intervene.
Having thus found that the surgeon committed two clear faults, the judgment a quo nevertheless dismisses the action essentially on the ground that in civil law that which is neither certain nor probable but merely possible -- here, a remission or cure -- is not proven, and as such cannot be the basis for a causal connection between the fault and the damage. [Emphasis in original.]
It is not surprising, then, that he is unable to subscribe to the view of Moisan J. who was inclined to believe that, appropriately informed and correctly followed, Mrs. Dupuis would have had a reasonable and probable opportunity to improve her condition. He notes that Jacques J.A., unlike his colleague Moisan J., saw no probable improvement or cure absent the doctor's fault. On this point, Vallerand J.A. agrees. His argument diverges on the question of whether to accept the theory of loss of chance.
He notes that the loss of chance theory has been applied by the Cour de cassation in France in sometimes ambiguous fashion and that it has attracted mixed response from French doctrinal writers. He summarizes the theory in the following way, at p. 29:
[translation] In short, if I have understood this theory correctly, it concedes that a party cannot obtain compensation for damage which is not probably or certainly the consequence of the fault, but may on the other hand obtain compensation for the quite different damage represented by the simple loss of a chance for a better outcome. Once a mere possibility is accepted, that possibility is transformed into a loss of chance which is itself connected with the original fault according to the traditional rules of causation. [Emphasis in original.]
While admitting that the theory is not without seductive qualities, he asserts that the courts of Quebec are not bound by developments in the civil law of France, however interesting they might be. The civil law of Quebec, in his view, accepts nothing less than proof of the causal link on the balance of probabilities. He points out that in the Brabander and Wilson cases, supra, referred to by Jacques J.A., the successful realization of the chance was probable, and classical rules regarding causation therefore applied. If loss of chance is to be accepted at all, it can only be as a head of damage, not as a means of shoring up inadequacies in the proof regarding the causal link.
Vallerand J.A. cites at some length and adopts the arguments of René Savatier ("Le droit des chances et des risques, dans les assurances, la responsabilité civile dans la médecine, et sa synthèse dans l'assurance de responsabilité médicale" (1973), 44 Rev. gén. ass. terr. 457, and "Une faute peut-elle engendrer la responsabilité d'un dommage sans l'avoir causé?", D.1970.Chron.123, at pp. 124-25) and Geneviève Viney (Traité de droit civil, vol. IV, Les obligations: la responsabilité--conditions (1982), at p. 437). Both of these respected French authors are critical of the unnecessarily artificial nature of the loss of chance theory.
Taking into consideration the unchanged status of the classical rules of causation in the civil law of Quebec and the uncertain reception of the new theory in France, he declines the opportunity to innovate by recognizing loss of chance, preferring to leave that radical step to the legislator.
He then looks briefly at loss of chance as a head of damages and notes that it is properly recognized here. He points out that Jean-Louis Baudouin (La responsabilité civile délictuelle (1985)) discusses loss of chance in the chapter on "le dommage" (damage) and makes no mention of it under his discussion of "lien de causalité" (causal link).
Turning finally to the question of damages in the case before him, Vallerand J.A. states that even if he had been inclined to award damages based on the loss of chance he would not have awarded as much as his colleagues. He would, however, be prepared to recognize the frustration and anguish which Mrs. Dupuis no doubt felt on learning that but for the omissions of her doctor things might have been different. This damage is not the loss of a chance but the actual psychological suffering. He values this damage at $5,000.
III -- Issues
In this action the respondent seeks the following damages, as described in the re-amended declaration, dated November 23, 1983:
[translation] Mrs. Dupuis suffered incalculable damage, including permanent and constant anguish, suffering and discomfort;
For pain, anxiety, hardship and loss of enjoyment of life, [the respondent] is claiming [from the appellant] the sum of $150,000;
For the premature death of Mrs. Dupuis, [the respondent] is claiming [from the appellant] the sum of $100,000.
I am inclined to agree with Vallerand J.A. and, effectively, a unanimous Court of Appeal that the claim for anguish and frustration must be accepted. This aspect of the damages was not dealt with by the trial judge. There is no doubt in my mind that the appellant's fault was directly related to the great, unnecessary psychological stress which Mrs. Dupuis faced upon learning that she had gone four years without knowing of her cancerous condition and without undergoing the follow-up, monitoring and treatment that was appropriate in her case. I will discuss the quantum of this head of damages later in my reasons.
On appeal, the damages relating to physical pain and suffering and premature death were, given the trial judge's findings, argued primarily on the basis of loss of chance. As I have already noted, a majority of the Court of Appeal was willing to recognize damages on this basis, and it is necessary for this Court to reconsider such apparently innovative reasoning.
I propose to analyze the matter in the following way. I will first attempt to define loss of chance, and this will necessarily involve a consideration of the voluminous authority on the subject which has emerged, principally in France and Belgium. Recently the courts and commentators in Quebec have had occasion to comment on loss of chance, and I will be particularly interested in those opinions before coming to a decision on whether, or perhaps to what extent, loss of chance now forms a part of the law of Quebec. This will, of course, involve a critical analysis of loss of chance, all the while keeping in mind what has, until now, been considered the classical approach to civil responsibility.
In the end, I conclude that the loss of chance analysis recommended by the respondent is inappropriate, at least in cases where death or sickness has already occurred. In such cases, classical principles of causation suffice, and, further, are essential in order for individual responsibility to attach. I deal then with the facts of this case and consider the question of appropriate damages.
IV -- Analysis
1. Loss of Chance
France and Belgium
Loss of chance has been the subject of a great deal of analysis in France and Belgium where it has also been admitted in numerous cases of medical responsibility. Accordingly, I will explore loss of chance largely through the excellent literature on this subject which has been published in those countries. It discloses two principal schools of thought on the matter. Their differences turn essentially on whether loss of chance is to be envisaged as merely a type of damage or whether its true effect in medical liability cases is to undermine causality as an essential element of civil liability and it is therefore to be rejected as contrary to established principle.
Indeed, loss of chance is a type of damage. It is the damage which results from the loss of an opportunity either to realize a benefit or to avoid an injury. It certainly poses special difficulties regarding the analysis of the causal link, but it is first and foremost a type of damage, and I will deal with it initially in that context.
Even when examined as a type of damage, loss of chance is acknowledged to be theoretically troublesome. As a rule, the damage must be direct, present and certain (Philippe le Tourneau, La responsabilité civile (3rd ed. 1982), at p. 167). Courts are hesitant to award damages which are hypothetical, possible, future and uncertain. In cases where the damage is ongoing, that is where it has manifested itself at the time of judgment but is likely to develop further, the courts are generally willing to recognize the future aspect of the damage. Though a part of the damage develops in the future, it is saved by the fact that it is capable of ascertainment in the present. This is the familiar case of a physical injury which has the potential to deteriorate or worsen.
In cases of loss of chance, the damage is future or hypothetical and clearly not certain. It is distinguished by the fact that it is contingent, or dependent on an element of chance which must be evaluated in terms of probabilities. This contingent or probabilistic aspect provides the potential for ascertainment of damages in the present. It is the special characteristic of loss of chance.
Our analysis of this concept will be made easier by means of a number of examples. The classic examples of loss of chance are described by Viney, op. cit., at p. 341:
[translation] 278. -- It very frequently happens that an act which can make its perpetrator liable will deprive another of a chance to make a profit or avoid a loss.
Let us take the case of an accident causing bodily injury and resulting in disability. It may prevent the victim from entering a competition or examination on which his career depends or from marrying the person he wishes to wed, or it may deprive him of the expectation of professional promotion or the possibility of returning to work after retirement. For the victim's immediate dependants, there is the loss of expectation of support, and so on. However, many other possibilities can be imagined: a solicitor or barrister neglects a time limit or formal requirement and this prevents the trial from taking its usual course: the client thereby loses a chance of winning. A horse which should have run cannot because of a delay caused by a carrier: the owner loses a chance that it will win. . . .
Viney explains, at p. 342, the theoretical tension which loss of chance creates:
[translation] First, the damage in question is future damage. In the French system, the purpose of liability is primarily to compensate for existing damage. . . . Further, this future damage is also contingent, since by definition a chance does not always materialize. Finally, it is easy to see the extreme difficulty of determining the amount of money that will compensate for such damage.
Viney observes that these obstacles are not insurmountable. She affirms that the courts regularly provide compensation for damage which is ongoing -- and therefore in part future and hypothetical -- notably in cases involving personal injuries. She points out, regarding the contingent or chance-based element of the damage, that the law recognizes in many ways the reality of this type of conjecture, especially in the insurance field. As for the difficulty in evaluating the damage of a lost chance, she observes that other heads of damage, such as psychological injury, are notoriously difficult to quantify, and yet are regularly allowed.
Viney stresses also that the courts have set up limits or guidelines relating to when loss of chance analysis can be applied. The first condition is that the lost chance have been real and serious. Other writers have concluded that this condition amounts to a requirement that the realization of the chance have been at least likely (Yves Chartier, La réparation du préjudice dans la responsabilité civile (1983), at p. 50). The second condition is that the courts must take into account the level of likelihood that the chance would have been realized in evaluating the damages (Viney, op. cit., at pp. 347 and 350).
My reading of French and Belgian authorities on loss of chance leads me to believe that there is little doctrinal resistance to the application of the theory in the classical cases, set out earlier. A strong division of opinion appears, however, when the theory is applied to medical situations, or, more precisely, situations where the actual damage has already manifested itself and where the chance is therefore exhausted.
Viney provides early examples of this newer, controversial trend in medical responsibility, at p. 436:
[translation] The starting‑point of this line of authority is a judgment of the Grenoble Court of Appeal on October 24, 1961. A person whose wrist was injured was X‑rayed, but as no break was found immediately resumed ordinary activities. Some years later, however, when handling a heavy object she felt some pain and the physician consulted discovered that a fracture without dislocation could be seen on the X‑ray taken at the time of the initial accident. The victim then sued the physician who had misinterpreted the X‑ray and the Grenoble Court found that if the diagnosis had been correct, treatment could have been given that would have avoided "the formation of pseudo‑arthritis with a continuing weakness in the wrist". The Court accordingly concluded that "not giving the treatment in question" had "definitely deprived the victim of a chance of recovery which she would ordinarily have had", and it held the physician liable. . . .
Two years later, it was the first Civil Chamber of the Court of Cassation which itself formally recognized the concept of "loss of chance of recovery" and on this basis upheld a judgment against a physician whose diagnostic error had also led to the giving of inadequate treatment. These judgments were soon supported by two further decisions in even clearer cases. The first approved the Court of Appeal when, while recognizing that it had not been established that the surgeon's fault was the cause of the patient's death, it admitted that that fault "had deprived the latter of a chance of survival", which according to the supreme court was sufficient to justify compensation. The second case held that by not making sure of the immediate intervention of an anesthesiologist a surgeon "had compromised the patient's chance of survival" and owed the patient's mother compensation for this head of damage.
The courts in France and Belgium have followed this initiative in cases of medical responsibility, though the Belgian courts have shown more caution. (See Henri Anrys, La responsabilité civile médicale (1974), at pp. 39-40 and pp. 192-216; Christiane Hennau-Hublet, L'activité médicale et le droit pénal: Les délits d'atteinte à la vie, l'intégrité physique et la santé des personnes (1987), at pp. 231-75, and cases discussed therein.)
When applied in the medical context, the loss of chance theory has been severely criticized. It is said to be an artifice, or an attempt to avoid the requirement of certain causation. Doctrinal writers in France and Belgium have gradually come to acknowledge loss of chance as a valid type of damage in cases of medical responsibility, but this is largely due to the courts' acceptance of the theory. Academic writing in many instances still reveals a great deal of unease with the extension of the theory to these cases. In order to understand this unease, some elaboration on the loss of chance theory will be necessary.
I have already set out, quoting Viney, the difficulties which doctrinal writers see in the recognition of loss of chance as a type of damage. Loss of chance has come to be accepted despite these difficulties. The ongoing debate seems to stem, at least in part, from disagreement as to the basis on which the theory has been accepted.
One group of writers lays emphasis on the fact that the lost chance is a type of damage which is fully independent from the actual, "future damage" which can be found in the benefit which has been missed or in the injury which has not been avoided. It is the lost chance itself which is compensated, and this can occur in all instances where a chance has been lost, whether medical or other, and whether a chance of cure or a chance of career success.
Another group of writers sees loss of chance as a theory which is only available in situations where the actual, "future damage" has not occurred and remains for that reason contingent. The result of the appeal is never known due to the lawyer's neglect, and the ability of the child to support his or her parents is unconfirmed as a result of the negligence of a third party. In these cases, loss of chance analysis allows the courts to recognize and evaluate the damage despite the fact that it has not yet been made real and must still be seen as contingent. The theory is not available, according to this view, where the chance has already been exhausted, that is, where the final damage has already taken hold and where it is no longer, by definition, contingent. This is most obviously the case in the medical context when a plaintiff claims that he or she lost a chance of cure or improvement, and where, ex hypothesi, death or illness has already resulted. In such cases, those who reject a generalized theory of loss of chance claim that the plaintiff must establish the causal link between the defendant's fault and the death or illness. Accordingly, it is not permissible to cover up doubts as to the full causal link by claiming that the defendant's fault at least caused the loss of chance of cure or improvement.
I propose to look at some of the writers who favour a generalized or all-embracing theory of loss of chance before examining those who are reluctant to apply it in the medical context. Two considerations should be borne in mind. First, the partisans of each approach divide on whether loss of chance is merely a way of evaluating contingent damages where responsibility is otherwise clearly established; or whether it is, through the recognition of the loss of chance itself as a type of damage, a legitimate way of attenuating the causal link in all cases where the fault in question can at least be said to have caused the loss of the chance. Secondly, it should be noted that the debate in France and Belgium does not merely concern weak chances which are, for example, as likely as not to be realized; much of the discussion focusses on chances which, according to the evidence (statistical and other), are probable though not fully certain. The probability may be strong, but there is nonetheless concern over the possibility that the chance would not be realized. I will point out later in this judgment that this is a type of theoretical difficulty which is considered by our own courts to be less troublesome.
One of the most elaborate and persuasive versions of a broadly applicable theory of loss of chance is provided by Jacques Boré ("L'indemnisation pour les chances perdues: une forme d'appréciation quantitative de la causalité d'un fait dommageable", J.C.P. 1974.I.2620). Boré stresses that loss of chance is appropriate to deal with matters which are inherently uncertain and which are best understood in terms of statistical probability. He points out, for example, that in a case where a judge is faced with a twenty or forty percent level of uncertainty and consequent failure of the action in responsibility, it should be open to him or her to focus instead on a retrospective assessment of the certain statistical chance of a favourable result:
[translation] 2. -- . . .
And this probability calculation performs its function of reducing uncertainty: instead of being paralyzed by the 20 percent or 40 percent uncertainty that persists, the judge makes use of the 80 percent or 60 percent probability he finds to conclude that there is partial but certain damage.
Loss of chance provides a type of damage which is certain, but Boré is under no illusions as to the true nature of this form of analysis. He readily admits that the chance has no entirely independent status. It is intimately related to the final damage:
[translation] 12. -- Why is the order against the party liable thus limited to the value of the chance lost and by what standard is that limitation judged?
One could simply answer that there is no departure from the principle of complete compensation for damage, since by definition the only damage proven, and so subject to compensation, is the loss of a "chance" to obtain a result which is certainly desired, but the obtaining of which was not in any way certain; and this would be true.
However, that would not be the entire truth, as this answer raises most delicate questions. It is said that the only certain damage is the loss of a chance; but what is a chance if not simply a probability excluding any certainty? Should this probability not then be raised to the level of a praesumptio hominis that will condemn the party liable to compensate for the entire damage, or on the contrary deny it this evidentiary force and rule out all compensation if there is no certain damage? [Emphasis added.]
Boré, of course, prefers to focus on the reality of statistical probability where certainty of the entire damage is unavailable. His comments do acknowledge, however, the strain which loss of chance theory places on traditional approaches to causation and damage.
This admitted strain is most evident, in Boré's view, in the recognition of partial damages in loss of chance cases and in the attenuation of the causal link. He states, at para. 14:
[translation] If therefore the compensation awarded is only partial, it is not because the death damage was itself partial: a dead person is not partly alive, and the victim or his heirs suffered the damage in its entirety; it is because the compensation is determined in relation to the probable causal link found to exist between the fault and the final damage. It is partial causation, determined in accordance with a probability calculation taking into account the influence of possible external factors, which gives rise to partial liability in this case; and it should be noted that once again there is a departure from the principle of indivisibility of causes. [Emphasis in original.]
And again, at para. 15:
[translation] The damage resulting from loss of chance is thus damage curtailed in accordance with the probable causal link between the fact giving rise to liability and the final damage. [Emphasis in original.]
Having isolated the matter of causation as crucial to the defence of loss of chance, Boré goes on to point out that the French Civil Code does not specify a theory of causation. He rejects the two most accepted theories of causation -- l'équivalence des conditions (equivalence of conditions) and la causalité adéquate (adequate causation) -- and adopts instead a theory which obeys both causal and probabilistic laws. He acknowledges that the theory of adequate causation imports some considerations of probability, but in his view this is insufficient. His comments are revealing as to the difference between a traditional approach and the newer theory which he proposes:
[translation] 19. -- . . .
Undoubtedly [the system of adequate causation] has in common with [loss of chance] that it takes into consideration probability, likelihood, in determining causation; but it immediately differs in an essential respect: this probability is not used for its own sake but as a means of achieving a certainty, regarded as indispensable, as to the existence of the causal link considered. On the contrary, in the theory of loss of chance the presence of significant contingencies means it is impossible to bridge the gap separating probability from certainty: once it is impossible to say whether without the physician's fault a seriously ill patient would have survived, it cannot be said either that this fault was a sine qua non of the death or a fortiori that it was the proximate cause thereof. [Emphasis in original.]
At another point in the article, Boré makes the distinction between the two approaches clear, this time in the context of a medical example:
[translation] 24. -- . . .
. . . when the judge estimates the value of the chance lost, he is statistically assessing the correlation between the fact giving rise to liability and the damage. For example, he asks the medical experts: "Does this therapy frequently lead to a cure?" If the experts answer: "This treatment usually leads to a cure", the judge then has a reasonable certainty which, though not the same as metaphysical certainty, is sufficient for him to say by the praesumptio hominis that the deprivation of treatment was the proximate cause of the damage and that the fault committed by the physician justifies his being completely liable. If, on the contrary, the experts answer: "Of one thousand patients treated with this medication, five hundred recovered", the judge then has available sufficient scientific certainty to say that a patient treated in this way had one chance out of two of recovering. Accordingly, the probability is only 50 percent that omitting the treatment contributed to causing the damage. We have already seen that this reasoning is objective and scientifically valid. The progress of science has shown beyond any doubt that statistics are not, as they used to be jestingly described, one of the three forms of lying. [Emphasis in original.]
Clearly, Boré's endorsement of the loss of chance theory is motivated by a desire to see statistical chances taken seriously. He goes so far as to say that the potential benefits of medicine form part of a "scientific patrimony" to which, presumably, the patient has a full claim:
[translation] 38. -- . . .
. . . in medical as in any other matters, chance definitely has a particular and pre‑existing value: that value is the capacity of the patient to be cured taking into account, first, the nature and progress of the disease from which he is suffering, and secondly medical, pharmaceutical or surgical techniques available to the physician. These various therapies are a "scientific patrimony", apart from the physician, who merely implements them. [Emphasis in original.]
This view of medicine would have the potential of placing a very heavy duty on the doctor, approaching an obligation of result. Boré insists, however, that the doctor's fault must be clearly established, and that other factors unrelated to the doctor clearly have the potential to eliminate chances of cure or healing: [translation] "It must be emphasized that there are numerous chances of recovery which may be lost through no fault of the doctor" (para. 42). And furthermore, Boré notes that the courts will recognize only chances which are real and appreciable, i.e. those which represent more than a mere hope though less than a true certainty: [translation] "a sufficient probability" (para. 37).
Boré wrote his article in 1974, and since that time the French Cour de cassation has continued to recognize loss of chance in medical cases. The courts must be considered among the most ardent supporters of the theory in France. (See Aubry and Rau, Droit civil français, 8th ed., vol. VI-2, Responsabilité délictuelle (1989), p. 125, footnote 17 and the cases listed therein. See also Cass. civ. 1st, May 31, 1988, Bull. civ. 1988, I, no 165, p. 114; Cass. civ. 1st, October 11, 1988, Bull. civ. 1988, I, no 281, p. 192; Cass. civ. 1st, March 29, 1989, Bull. civ. 1989, I, no 147, p. 97; Cass. civ. 1st, June 7, 1989, Bull. civ. 1989, I, no 230, p. 154; Cass. civ. 1st, January 10, 1990, Bull. civ. 1990, I, no 10, p. 8; Cass. civ. 1st, February 7, 1990, Bull. civ. 1990, I, no 39, p. 30.)
More recently, Isabelle Vacarie has provided an analysis of loss of chance which, for the most part, approves the application of the theory in all cases, including medical ("La perte d'une chance", [1987] 3 R.R.J. 903). Her article shows both the commitment of the French courts to the theory and the difficulties of application which they have encountered.
She begins, like Boré, by stressing the distinctiveness of the loss of chance approach to responsibility. She points out, at pp. 906-7, that the courts use this approach to acknowledge the role of chance as opposed to will in human affairs, and that this represents a new development in law:
[translation] [The various applications of loss of chance] are all part of a trend in the law in the direction of recognizing the part chance plays in human affairs. This recognition has shaken the foundations of the 1804 Civil Code, which viewed man "as a noble being, master of himself and his destiny". The mystical nature of this concept has been attacked, and as chance has made its appearance in the legal field there has been a parallel limitation of the function of will in the law of obligations. This opposite trend is due to the antinomy of chance and will: chance excludes will and will marks the limits of chance. The legal solutions have reflected this opposition. [Emphasis in original.]
As we shall see, she regards the distinction between will and chance as the key to understanding the limits which the courts have placed on loss of chance.
At the most general level, Vacarie observes, at p. 907, that the French courts have invoked loss of chance analysis in two situations, the first being most relevant to the medical context:
[translation] . . . a close review of the judgments [of the Court of Cassation] indicates that loss of chance is relied on in two situations that need to be distinguished, because they do not involve the same aspects of the law of liability.
-- Either the judge uses loss of chance when he is ruling on the liability of the defendant, and admission of this damage is then a condition of liability; for example, on reading the expert reports the court will consider that it is not certain that such medical fault is the cause of the patient's disability, but rather that that fault deprived him of a chance of recovery.
-- Or the judge invokes loss of chance only at the stage of assessing the damage, when the defendant's liability has already been established; two examples are when an appeal has not been brought within the time limits as the result of negligence by counsel, and the judge awards his client an indemnity to compensate for loss of the chance of winning the case; and following an accident, compensation is awarded for loss of the chance of promotion or support.
Whatever the situation, the use of loss of chance indicates the judge's uncertainty. However, that uncertainty relates to the causal link or extent of the damage. [Emphasis added.]
As this comment indicates, Vacarie sees loss of chance in cases of medical responsibility as relating to uncertainty with respect to the causal link. In her opinion, loss of chance analysis is acceptable and appropriate here, though only to deal with faults of omission. This is a new distinction, and it is worth exploring briefly.
She states clearly, at p. 908, that loss of chance is misapplied in cases where the doctor's (or defendant's) act is alleged to have produced the damage:
[translation] In the absence of a causal link between the fault and the material damage, compensation for loss of chance is certainly questionable when the fault results from an action or positive act of the defendant. In such a case, he can only be held liable if it is established that his wrongful act materially contributed to the causing of the damage. Such proof may be made by any means: it may be inferred from serious, specific and concurring presumptions, but it is essential. Doubt in the judge's mind as to the causal link cannot justify partial compensation through awarding an amount for loss of chance: the rule is one of full compensation for direct damage. The Court of Cassation makes sure . . . that this rule is observed. [Emphasis added.]
According to Vacarie, the approach is not the same when the courts are presented with a fault of omission. Here, apparently, different rules must apply, and recourse to loss of chance analysis is fully justified, at pp. 908-9:
[translation] It is certain that causation is not determined in the same way depending on the nature of the fault. When the fault results from an omission, by definition it has not physically contributed to causing the damage; however, the courts and commentators now unanimously consider that it can be legally causal if it is shown that the defendant's action would have prevented the damage from being caused. This reasoning results in a shifting of the evidence. It is no longer a question of reconstituting the chain of events but one of imagining what the chain of events would have been if the defendant had acted, if he had met his obligations. This imaginary prognosis is based on the effects that could be expected from the act omitted by the defendant. However, whenever no necessary cause‑and‑effect relationship exists, no one can say that the act would have had beneficial consequences: while there is a chance that it might have avoided the damage, it is not a certainty. This element of uncertainty explains the recourse to the concept of loss of chance to penalize a wrongful omission. This solution has been developed in the main over a twenty‑year period, in the field of medical liability, bringing about compensation for loss of chance of survival or recovery.
She then adds a distinction and a qualification which are relevant in medical cases involving faults of omission. The distinction is that whereas loss of chance analysis is appropriate in cases where the doctor omits to apply or administer a particular treatment, it is not necessarily suited to cases where the doctor simply neglects to inform the patient of such a treatment. In the first situation, the treatment is inherently risky and loss of chance can deal with the risks in terms of probabilities. In the second situation, the causal link is affected and potentially broken by uncertainty as to will of the patient, a factor which loss of chance cannot consider. She notes, however, that there may be cases where there is no doubt as to the patient's willingness to accept any treatment.
The qualification which Vacarie makes, at p. 914, regarding cases of medical omissions is that loss of chance analysis is not appropriate where the evidence indicates that the disease or illness was irreversible:
[translation] The physician will not be held liable if the pathological process was irreversible, if the patient's condition could not help getting worse. His liability is also excluded when the causal influence of the fault seems too uncertain to the courts because it appears from the expert reports that there is nothing to contradict the assumption that death or infirmity is unavoidable, nothing to say that prior examination or treatment would have helped prevent the damage.
Vacarie's qualification here brings to the forefront the recurring question of what level of chance (i.e. from highly probable to hardly possible) will be acceptable to courts wishing to follow a loss of chance analysis. She seems to favour chances which are at least likely to succeed, although she concedes a large discretion to the trial judge, at pp. 916-17:
[translation] Loss of chance may thus be regarded as direct and certain damage whenever the chance of survival or recovery is greater than the risk of death or disability. That of course is a question of fact which is exclusively a matter for the triers of the merits.
In the end, once distinctions and qualifications have been made, Vacarie approves the resort to a broadly applicable theory of loss of chance, and asserts that it provides a place for "la justice par division" ("justice by division") and "la justice de Salomon" ("the justice of Solomon"), at p. 932.
There are numerous other writers who approve the application of loss of chance to medical cases (e.g. Chartier, op. cit., at pp. 31 and 35; François Chabas, Annotation of Cass. crim., January 9, 1979, J.C.P. 1980.II.19272; Georges Durry, "Responsabilité civile" (1967), 65 Rev. trim. dr. civ. 153, at p. 181). Some are more forthcoming than others in acknowledging the innovative and theoretically difficult nature of this recourse. I note in particular Annick Dorsner-Dolivet (Contribution à la restauration de la faute, condition des responsabilités civile et pénale dans l'homicide et les blessures par imprudence: à propos de la chirurgie (1986)).
Dorsner-Dolivet appears generally to approve of a theory of loss of chance in its classical application, i.e. the oft-cited examples referred to at the outset of this analysis. With regard to the new tendency in France to apply loss of chance in the medical context, she agrees with those who see it as a theoretical tangle, especially regarding the traditional rules of causation. For reasons of policy, however, she approves of the extension of loss of chance in this limited sphere, at p. 427:
[translation] The recent acceptance of loss of chance has remained limited to the field of medical, and in particular surgical, liability. The reason for this is that factors particular to this field dictated its development. However, although we approve of these criticisms made of the theory of loss of chance in terms of legal logic, we feel because of the uncertainty of scientific causation that maintaining the theory is essential. Even admitting these imperfections, we think that it is the lesser of evils which, if it were eliminated, would give rise to much more serious disruption.
What, then, are the considerations which prompt her to approve a theory which admittedly offends the previous logic of the legal system? With Chabas ("Vers un changement de nature de l'obligation médicale", J.C.P. 1973.I.2541), she notes three general factors, at p. 416:
[translation] The first factor is the unusual risk run by the patient in undergoing surgery. Even at the present time the most optimistic physicians recognize that the sequelae of anesthesia or operating shock are still to a large extent unpredictable.
The second factor is the economic inequality between the victim who is a recipient of social insurance and the practitioner who is always insured.
Finally, the most important factor is the fleeting nature of the evidence . . .
It is to offset all these factors working against the victims of surgical fault that judges have given a new face to the theory of loss of chance. In this interpretation, certainty as to the existence of a fault influences the concept of damage since it leads to the creation of a new type of damage, an intellectual abstraction designed to alleviate the difficulties of proving with certainty a causal relationship.
Chabas, who inspired some of Dorsner-Dolivet's comments, is another writer who has shown an ambivalent attitude towards the use of loss of chance analysis in medical cases. In his 1973 article noted above, he questioned the theoretical underpinnings of the new developments and added that they would likely have the result of hurting medical reputations and encouraging defensive medicine. By 1980, his attitude had changed, and he was willing to approve of loss of chance in certain medical situations (J.C.P. 1980.II.19272). He states that where the victim is already affected by a terminal condition, and where the doctor's fault completely deprives the patient of such chances as remain, then the doctor should be responsible for the loss of a chance of survival. He adds, however, that where the victim is in normal health prior to the doctor's fault, it is the full value of the life which must be considered; and where the doctor's fault is not at least a causa sine qua non of the loss of that life, loss of chance analysis cannot be used to provide a proportional recovery corresponding with the judge's doubts.
I turn now to that group of writers who, in France as well as Belgium, oppose the application of loss of chance in the medical context.
The most severe and oft-cited criticism of the recent evolution of loss of chance is that of Dean Savatier ("Le droit des chances et des risques, dans les assurances, la responsabilité civile dans la médecine, et sa synthèse dans l'assurance de responsabilité médicale", op. cit.; "Une faute peut-elle engendrer la responsabilité d'un dommage sans l'avoir causé?", op. cit.). Vallerand J.A. quotes large portions of these writings, and I will not reproduce them in their entirety though generally I approve of their purport.
The gist of Savatier's argument is that in medical cases, one is not faced with a true loss of chance: the factors bound up in the earlier chance have already played themselves out. In the presence of a death or infirmity, there is no longer any need to calculate a future chance; the judge must consider events which have now passed and for which the cause is uncertain: [translation] "What is incorrectly called chance is insufficient information on the part of the judge about how destiny took its course" (D.1970.Chron.123, at p. 126). Destiny having taken its course, loss of chance analysis is no longer appropriate or necessary, and traditional rules regarding causation and damage must apply (at p. 124):
[translation] The reasoning we find unacceptable places . . . chance in the past where the event to which it applies has already occurred. The analysis is outdated! . . . Destiny has played its part. By definition, infirmity or death has occurred. One is no longer dealing with chance but with a fait accompli. When the national lottery is drawn, tickets lost through the fault of another are no longer chances but vested rights or no right at all, as the case may be. We know the cause of the loss of the vanished chances: it is the draw. On the other hand, the problem for solution once the damage complained of by the victim has occurred . . . is whether that damage would or would not have occurred absent the fault alleged against the physician.
In the event that by not committing the fault the physician would have avoided the damage, the fault is causal. If on the other hand the death or infirmity would have occurred in any case without the fault, it is unrelated to the damage. What is at issue is no longer a future chance but a past event the cause of which is uncertain. It is this uncertainty that has to be overcome. If the judge is persuaded that the fault caused the damage, he must order its perpetrator to compensate for it; but he may not make such an order, even one reduced in proportion to any doubts he may have, if he is not personally persuaded that the fault was a necessary condition for the damage. He cannot tailor the order to his doubts, even in an inversely proportional manner. [Emphasis in original.]
In the end, then, Savatier sees theoretical difficulties in applying loss of chance in the medical context due principally to two factors: the damage in medical cases has already occurred and is no longer based on a chance or a contingency; and the question of whether the doctor caused that damage cannot be answered simply with information on what the patient's chances of cure or improvement were prior to the doctor's act or omission.
The comments of Jean Penneau (La responsabilité médicale (1977)) are of a similar nature. His analysis examines more closely the difference between the medical situation and the classical loss of chance situations. He contrasts the valid application of loss of chance in cases of "crystallized" factual development with those medical cases where the chance has evolved and in fact acquired a real and final form. In such cases, the causal analysis must be established between the defendant's fault and that real, final damage (at pp. 115-16 and p. 119):
[translation] The classic concept of loss of chance is concerned with damage; the concept of loss of chance of survival relates to causation.
As we know, in certain situations which have crystallized once and for all as a result of the defendant's fault, at an intermediary stage which will never again develop the gains or losses that might have been expected, the certainty of the damage is nevertheless apparent. The defendant's fault has interrupted a process the elements of which can be assessed, and if the positive aspects (chances of gain) predominate, damage exists, the certainty of which -- though it may be only virtual -- requires compensation. Loss of a chance is thus regarded as certain damage, and compensated for as such provided a chance really existed. This solution, consistently adopted by the courts, has been unanimously approved by the commentators, and in all logic must be commended.
The situation is quite different with the recent decisions on loss of chance of survival in medical cases.
. . .
. . . it is based on an obvious confusion, as such chances which have been lost and have now lapsed are . . . inseparable from the final damage itself at which the legal compensatory machinery is alone directed: loss of chance can no longer be the subject of future calculation since the event has actually occurred and it is inseparable from that event; as Dean Savatier very aptly puts it, "one is no longer dealing with chance, but with a fait accompli".
Loss of chance thus appears here, by a confusion of language, with a new face: that of causation. [Emphasis added.]
In order to make his argument clearer, Penneau transfers the chance analysis temporarily to the classical example of a student who, due to an accident, is prevented from taking an important career-related exam. According to the classical loss of chance analysis, the individual who caused the accident will be required to indemnify the student for the loss of a real and serious chance of passing the examination and embarking on a career. The situation is entirely different, in Penneau's view, where, despite the accident, the student attends the examination and fails: [translation] "Here, the chance has been tried, and we know that it has been lost; loss made concrete by the final damage -- the failure, death or infirmity -- with which it is precisely and necessarily confused" (p. 118). To speak, nonetheless, of a loss of chance in this situation would be to speak of the same damage in different and now irrelevant terms, according to Penneau.
A more recent and no less forgiving critic of the application of loss of chance to medical cases is Gérard Mémeteau ("Perte de chance en droit médical français" (1986), 32 McGill L.J. 126). He adds to the criticisms noted above the fear that recourse to loss of chance here brings the intensity of the doctor's obligation that much closer to one of result as opposed to the traditionally recognized obligation of means or diligence (at p. 148):
[translation] The state of illness was already creating a risk and compromising the patient's future. The physician promised to do all he could to avoid the danger, but as his obligation was one of means he did not undertake to bring about such a result. He is to be held liable on the ground that by acting otherwise he might perhaps have attained that result. The paradox is that doubt as to what perhaps might have happened leads to the equivalent of an obligation of result. Once a pathological condition is by definition capable of developing various sequelae which the best efforts of a physician cannot entirely control (the standard of uncertainty), the final fate of the patient can only be determined by this "retrospective prognosis" condemned by academics. Only one event is certain: the patient did not recover, and without this certain result, how can it be said that chances were wasted? A chance is only a hope, an idea, not a thing of with the existence is certain.
A number of points can be extracted from this criticism of loss of chance. First, as noted above, Mémeteau does not claim that the doctor's obligation has actually been transformed into an obligation of result. He simply points out that the element of risk which is an inescapable part of the practice of medicine becomes, through loss of chance analysis, automatically attributed to the doctor once his or her fault has been established. Secondly, in the classical loss of chance cases, the chance is clearly lost as a result of the defendant's fault, whereas in the medical cases the chance is lost only when, subsequent to the doctor's fault, fate plays out the morbid rather than the recuperative aspect of the original risk. Finally, Mémeteau clearly rejects Boré's view of chance as something to which the patient has a near-possessory right.
Mémeteau ends his commentary with a parting criticism: if the distortion created by loss of chance in medical cases is intended to favour patient-plaintiffs, then such a purpose should be openly admitted by its proponents. He states, at p. 150:
[translation] Loss of chance . . . is and can only be a contortion of reasoning made for the sole purpose, which should at least be clearly admitted, of compensating the most avid victims of medical accidents or those with the best advice. This theory, a reconstruction of the future from the past, a placing of greater importance on one assumption among many, a wager on probability, is merely an artifice.
A careful and balanced appraisal of new developments in loss of chance can be found in Viney's work. Viney places her discussion of loss of chance in medical cases in the chapter on causation and entitles the section "L'utilisation de la notion de "perte d'une chance" pour alléger la charge de la preuve de la causalité" ("Use of the notion of `loss of chance' to alleviate the onus of proof of causation") (p. 436). She immediately distinguishes the medical cases from the classical applications of loss of chance:
[translation] 370. -- As we know, the courts now generally recognize that "loss of chance" is damage for which compensation may be obtained. A series of very celebrated judgments rendered in proceedings involving medical liability has shown that the usual meaning of this concept has sometimes been stretched so as to shift the object of proof of causation in the victim's favour and so make that proof easier.
In her opinion, this new trend causes theoretical difficulties, though not nearly as many or as serious difficulties as its critics suggest. She rejects, for example, the claim that the intensity of the doctor's obligation has been raised to one of result, noting that all decisions to that point had taken pains to establish the doctor's fault in the nature of error, want of care, imprudence or negligence.
Regarding causation, however, Viney sees that there is clearly a problem in many of the medical cases, at p. 437:
[translation] However, what is beyond question . . . is that this concept has been used by the judgments considered to justify partial compensation in situations where the causal link between the medical fault and the damage suffered by the patient was not established with certainty. To avoid denying all compensation the courts have ignored the concrete damage (the death or worsening of the patient's condition) the causal connection of which with the fault was not certain and have substituted an abstract diminutive of that damage (loss of chance to recover or not to die), which is by definition associated with the medical fault on account of the danger to which it necessarily exposes the patient's health. Accordingly, it cannot be denied that this is a reasoning device invented in order to get around the obstacle which the requirement of a definite causal link is likely to present to certain claims for compensation brought against the perpetrator of (proven) medical fault by the victim of damage (also established). [Emphasis added.]
In Viney's view, the clear objective of this trend is to reduce the burden of proof on plaintiffs attempting to establish a causal link in medical cases. She notes that often it is expert medical reports that raise causal doubts due to the tentative way in which they are often expressed. Viney is not necessarily opposed to the facilitating role which loss of chance serves in medical cases, but she favours other, more direct ways of accomplishing the task, at pp. 438-39:
[translation] . . . while we feel it is acceptable to take a broad approach to the causal link between medical fault and the damage which the latter made likely, and this is in accordance with the general trend in the law of civil liability as a whole, we are less convinced that use of the technical procedure of "loss of chance" is a completely adequate means of achieving this objective. This means does not make it possible . . . to compensate in full for the actual and concrete damage (the death or worsening of the patient's condition). It requires the judge to cut down on the compensation to take into account "chances of recovery or survival" which the victim still had despite the physician's fault. Clearly, in making such a reduction the court has no objective criterion: it is thus led to adopt arbitrary solutions and, in particular, as Dean Savatier has very aptly remarked, to project its own doubts regarding the likelihood of causation onto its assessment of the compensation.
Viney favours the concept of "création fautive d'un état dangereux" ("wrongful creation of a dangerous state") which facilitates courts in bridging the gap in causal certainty created by medical evidence falling below the one hundred percent level. This concept has been applied in a number of cases in France, and Viney recommends this approach over that of loss of chance, at p. 441:
[translation] This new approach by the courts accordingly reduces the importance given after 1960 to "loss of chance" in establishing causation in medical fault. Proof of causation is now directly and openly expanded and liberalized on the basis of the danger inherent in fault committed in diagnosis, medical prescription or care.
Since Viney wrote in 1982, the courts in France continue to employ loss of chance in medical cases.
I will turn now, briefly, to the status of loss of chance in Belgium. An early assessment of the introduction of loss of chance to medical cases appeared in a book by Anrys (La responsabilité civile médicale, op. cit.). Anrys recognizes that the courts have accepted loss of chance as a type of damage in the classical, non-medical cases and have evaluated the damage according to the likelihood of its realization. He notes, however, that problems arise [translation] "when it has become impossible by definition to know whether or not the chance would have been realized" (p. 39). He points out, at p. 40, that Belgian courts have recognized lost chances where the causal link was clear, but that French courts have used the theory as a causal substitute:
[translation] In the medical field [in Belgium], the courts have accepted loss of chance of recovery due to not making examinations required by the current state of medical science, the making of which would without any possible question have allowed for an accurate diagnosis and consequently effective treatment. However, there has also been a distortion of the concept since 1965 in France, where the courts have transposed it from the level of damage to that of causation in order to dispense with proof of this connection. [Emphasis added.]
He rejects the French trend as tending to establish the principle that every fault of a doctor, if it has not caused the actual damage, will at least have caused the loss of a chance of cure or improvement. This trend violates the rule that: [translation] "[t]he causal link must be established and it is established when the damage in its concrete form would not have occurred without the fault" (p. 40). He points out that the Belgian courts have applied loss of chance only with the greatest of caution ("avec énormément de prudence") (p. 192). He cites numerous cases where the courts have rejected loss of chance arguments due to weaknesses in the proof as to the causal link. (See, e.g., Brussels, February 12, 1957, Pas. 1958, II, 1; Liège, 7th Ch., November 24, 1971, unreported, R.G. no 1201/70.)
Hennau-Hublet has also provided a discussion of loss of chance in both France and Belgium in a recent book. Her book deals with the theory as it has developed in both the civil and penal jurisdictions.
Hennau-Hublet situates the problem as follows: "La problématique du lien causal en matière médicale. Premier contexte: une faute médicale (le plus souvent d'omission) se greffe sur un processus morbide préexistant ou concomitant" ("The problem of the causal link in medical cases. First context: a medical fault (usually one of omission) occurs in addition to a pre-existing or concomitant morbid process") (p. 231). She notes first that, in a few cases, both the Belgian criminal and civil courts have cautiously followed their French counterparts in applying loss of chance in medical cases (e.g., Brussels, January 19, 1965, Pas. 1966, II, 13, and Brussels, 2nd Ch. civ., October 22, 1982, unreported, appeal dismissed Cass. 1st Ch., January 19, 1984, Pas. 1984, I, 548).
Hennau-Hublet is critical of the new developments. First, she points out that statistics showing less than certain efficacy of a medical procedure should not dissuade courts from linking the doctor's fault to the actual damage (death or infirmity) which results (at pp. 246 and 248):
[translation] . . . this statistical assessment of the chances that the damage will or will not occur is made on the basis of a theoretical or ideal situation: even in cases of sound medical practice there is the possibility that the disease will still develop unfavourably towards lesion or death. This possibility is calculated. It is only an abstract or theoretical average.
In actual fact the judge has to cope with a quite different specific situation: in treating a disease, the physician did not adopt the specific measure that was required, which could have halted the progress of the disease.
Accordingly, the results of the probability calculation made on the assumption of adequate medical treatment are modified by contact with the actual situation of medical fault.
Thus, how are we to assess the causal power and function of the physician's fault of omission in terms of a statistical causal average calculated on the basis of premises contrary to the factual situation which the judge must consider?
. . .
The fact that an area of risk remains does not ipso facto exclude the existence of a causal link between established fault by the physician and the damage suffered by the patient. [Emphasis in original.]
According to Hennau-Hublet, an unwarranted reliance on statistical doubt explains the reluctance of the criminal courts (in their civil jurisdiction) to find a causal link in medical cases involving omissions; but this attitude also explains the civil courts' recourse to loss of chance analysis where causation is apparently (i.e. statistically) uncertain.
In order to avoid unnecessary difficulty with statistical evidence, she recommends that courts ensure that the causal analysis is accomplished with full awareness of the particular facts of the case. This requires a concrete appreciation of the causal power of the fault and of the whole nature of the damage. She then goes on to show that in numerous cases where the court rejected the causal link and instead applied the theory of loss of chance, a more classical analysis was available (see p. 267). She also claims that the requirement that a fault have entirely eliminated any chance of survival -- suggested by Chabas, for example, and sometimes applied by the Belgian courts -- is equally unnecessary (p. 273).
Quebec
I propose first to deal with the limited doctrinal discussion of loss of chance in Quebec before moving on to the more detailed treatment of the subject which can be found in the decisions of the courts of the province.
Both the Court of Appeal and the parties in this case refer to an extract from Baudouin (now of the Court of Appeal) in his La responsabilité civile délictuelle, op. cit. In the chapter on damages he discusses the question of certainty of damages as well as problems of the fixedness of future damages, that is, the problems associated with ascertaining and assessing those damages. It is in this context that loss of chance is discussed, and it will be useful to cite the passage in its entirety, at pp. 106-7:
[translation] 183. -- Loss of chance -- It also compensates for what can be called loss of chance. In this case, it is not so much the actual loss which the court is seeking to compensate but rather the disappearance, because of the occurrence of the fault, of the chance either to avoid a loss or to make a profit. The theoretical difficulty is that the damage which is being claimed for is not only future but, by definition even, uncertain. No one can say with certainty whether the loss could have been avoided or the profit made if the fault had not occurred. However, the courts, recognizing that merely being deprived of a possibility is a direct damage, have applied this concept. Thus, when following the death of a child the parents are awarded a sum of money for loss of future support, this must clearly be regarded as loss of a chance. It is not certain that the parents would have needed such support or that the child would have been able to provide it. However, the fault has extinguished this possibility, this hope, this legitimate expectation of support. This is also the case in contractual matters when a lawyer whose services have been retained allows prescription to intervene and so loses the action which his client could have brought. In some cases no one can say with certainty whether the client would have been successful and how much he might have been awarded. The court will compensate him, however, because the fault has extinguished all possibility of recovering his debt. The courts have also applied this concept in a case where paramedical treatment caused an individual to lose the chance of improving his physical condition. [Emphasis in original.]
Most of the examples which Baudouin cites are classical, in French and Belgian terms. The last medical example refers to the case of Tardif v. Laverrière, Sup. Ct. Mégantic, No. 6210, November 10, 1976 (summarized at [1976] C.S. 1803), and I will have occasion to discuss this case later.
Maurice Tancelin (Des obligations: contrat et responsabilité (4th ed. 1988)) sets out the general rules regarding damages and notes in particular, at p. 336, the difficulty in distinguishing between damage which is real and certain and damage which is future and only possible:
[translation] 553. Actual damage -- Possible, virtual or potential damage is not sufficient to give rise to civil liability . . .
However, in practice it is not so easy to distinguish between actual and possible damage. The case law includes borderline cases where loss of chance is regarded as damage for which there should be compensation. This is true of "expectation of future support" given by a child to its parents. [Emphasis added.]
In Théorie de la responsabilité civile (2nd ed. 1980), Jean Pineau and Monique Ouellette also deal with loss of chance in their discussion of whether the damage must indeed be real and certain. They state first that future damage will always be compensated where it is certain that it will occur. They note, at p. 19, that courts also tend to award damages in cases of damage which is less than certain:
[translation] However, we know that the law does not always require certainty and sometimes accepts likelihood. Accordingly, compensation may be obtained for future damage. The lack of certainty regarding future damage has the consequence of giving the judge a rather wide discretion in assessing that damage.
They go on to observe at pp. 19-20 that damage which is only possible and therefore hypothetical is not compensable, and that loss of chance appears to be an exception to this rule:
[translation] On the other hand, damage which is possible and so only hypothetical will not be compensated for: if there is any doubt as to whether or not the damage will occur, no compensation will be allowed; it is impossible to prove that such damage will definitely occur some day . . .
. . .
Though compensation may not be had for possible damage, all the authors admit that loss of chance can constitute actual damage for which compensation may be obtained. Someone who, for instance, is prevented by the fault of another from entering a competition which he would have had a chance of winning (loss by the seller of a lottery ticket stub) thereby suffers damage for which he may obtain compensation. Someone who through his lawyer's fault loses a case he had some chance of winning (the lawyer allowed prescription to run) suffers a damage that is certain or "a certain damage" . . . the chance existed, but it was lost. [Emphasis added.]
It is not clear to which authors Pineau and Ouellette refer, though clearly, as we have seen, French authorities provide plentiful support for this view.
Before turning to the medical context, it might be useful to look briefly at some of the cases referred to in the footnotes to the texts quoted above in order to understand loss of chance as it has developed in Quebec in less controversial cases. In these cases, I observe that the courts are certainly presented with difficult questions -- lost chances which require some speculation on the courts' part -- but where they have sufficient proof to establish a causal link on the balance of probabilities, they conclude that the fault and damage are fully connected, and they do so without recourse to the elaborate loss of chance analysis which was explained in the last section. The causal link is established between the fault and the final damage, and while the "intermediate" damage represented by the loss of chance may be mentioned in the course of the analysis, the courts' preference is to follow the link through to its fullest extent. The damage is not the chance itself but rather the benefit unrealized or the injury unavoided. Where such damage is a condition of civil responsibility (in the fault, causation and damage equation), the courts probably require proof on the balance of probabilities. Effectively, the damage is considered present and certain and, accordingly, the damages can be evaluated at their full value. In cases where the responsibility has already been established regarding some concrete initial damage, and where the loss of a desired future result is an alleged consequence of the initial damage, then the court's attention will properly be focussed on the real damage represented by the loss of the future result, and questions of probability will only be considered, along with other relevant facts, in the evaluation of damages. In both instances, it is the real damage and not the chance which is the preoccupation of the court. Naturally, in the present case, we are most concerned with cases where the analysis of the loss of chance situation is vital to the initial finding of causation and, hence, responsibility.
In his brief discussion of loss of chance, Baudouin refers to the familiar case of parents, following the death of their child, claiming loss of future support. In the case of Grenier v. Gervais, [1950] Que. K.B. 60, for example, the trial judge awarded a sum of $2,000 for "deprivation of future services" following the death of a fifteen year old who was about to start working. St. Jacques J.A. approved this award of damages noting that [translation] "in all probability", and according to the available proof, the boy would have earned a salary similar to that of his older brothers and that the damages were therefore reasonable. Four judges of the Court of Appeal accepted these damages, although two dissented as to the quantum. I note that in this case the defendant's responsibility for the child's death had already been established.
Baudouin also mentions the example of a lawyer who neglects or mishandles his or her client's case and allows it to be prescribed or blocked. In Lacourcière v. Laplante, [1976] C.A. 433, a lawyer, Lacourcière, had been instructed by Laplante to sue two defendants, but through his own fault sued only one. At trial it was decided that the person who had not been named as a defendant had been responsible. Laplante sued Lacourcière and obtained judgment at first instance. On appeal, Bernier J.A. reviewed the evidence and concluded that had the proper defendant been named, Laplante would have succeeded in the first trial: [translation] "there is no doubt that the Superior Court would have been justified in holding him responsible for the accident and would have ordered him to compensate the respondent" (p. 435) (emphasis added). The causal link was clearly established between the lawyer's neglect and the fully realized chance, as represented by the damages which would have been recovered in the negligently handled first action.
The analysis used by the Court of Appeal in this case is typical of the way Quebec courts approach such situations. Patrick Molinari ("La responsabilité civile de l'avocat" (1977), 37 R. du B. 275) describes the usual procedure in cases of lawyer's responsibility, at pp. 291-92:
[translation] 3.4. Causal link and damage suffered
When a fault has been committed and the victim complains of damage, he must show a direct relationship between the two, must to some extent prove that the damage is the consequence of the act of fault. In cases involving a lawyer's liability the question of causal link is quite clearly settled. The plaintiff in an action must establish that he would have succeeded in the first proceeding but for the fault of his counsel . . . Thus, if the plaintiff alleges that he lost his case because of his counsel's negligence, the court will have to decide on the eventual outcome of the first proceeding and determine whether the plaintiff would have succeeded.
This type of analysis is necessarily speculative, but the courts do not generally or overtly adjust damages according to the probability of the chance in question. The fault is linked to the final damage -- that which the party would have obtained but for the lawyer's fault -- and the causal analysis proceeds accordingly. This analysis is made easier by the availability of evidence from the first trial and by the judge's expertise in matters of law. Molinari points out that courts will not always award the full amount sought in the first action, but this appears to be a function of the strength of the case rather than the chance of success. This distinction may seem overly nice, but it is important. We are interested in determining whether lost chance situations are dealt with as matters of chance or as matters requiring an assessment of what actually would have occurred. The courts of Quebec appear to prefer the latter approach.
An example of this approach can be found in the case of Beaupré v. Joly, [1971] C.S. 199, which is cited by both Baudouin and Molinari. In that case, Beaupré mandated his lawyer, Joly, to sue a municipality following a motorcycle accident. Joly neglected the file and allowed the claim to become prescribed. Beaupré sued Joly to recover the damages which he allegedly lost as a result of the lawyer's fault and the trial judge granted his action. The judge noted, however, that the proof revealed that Beaupré had been responsible in the motorcycle accident to the extent of forty percent, and, accordingly, damages valued at $23,120 were decreased to $13,872. This type of analysis proceeds from the basis of the actual damage which, in the court's view, the client would have received but for the lawyer's neglect. The damages are discounted not on the basis of an assessment of probabilities, but according to the principles of law which would have actually governed the case had it been heard in the first place.
Jacques J.A. in the case before us cites Brabander v. Goulet, supra, which is a case of notarial responsibility. The notary, Brabander, failed to register a hypothec in favour of Goulet and thereby deprived her of a means of ensuring payment of a loan. In an action for recovery of the loan, Goulet sued Brabander and three others for inexecution of contractual obligations and Brabander alone for delictual responsibility (based on conflict of interest, abuse of functions and fraud). The trial judge found in favour of Goulet and held the four defendants jointly and severally liable for the debt of $27,000. On appeal, Brabander claimed, inter alia, that no causal link had been established between [translation] "the actions and schemes of the appellant and the damage alleged by the respondent" (p. 38). The unanimous Court of Appeal answered that claim as follows, at p. 39:
[translation] . . . there is no doubt that a cause‑and‑effect relationship exists between the appellant's acts and the damage incurred. As the debt is outstanding and unpaid, the mere fact that the notary Brabander neglected to see that the creditor was protected by a hypothec . . . is sufficient in itself to conclude that, in all probability, she has by her representative's negligence been deprived of a means of reimbursement. That, apart from all the other circumstances disclosed by the evidence, is something which justifies the conclusion that there is a relationship between the fact and the damage. [Emphasis added.]
As stated above, the notary was held jointly and severally liable for the whole amount of the damages. I conclude that the Court viewed the causal link as having been fully established, i.e. the link between the fault of the notary and the actual damage represented by the whole amount of the defaulted loan. This is not then a case which applies a loss of chance analysis along the lines discussed in the last section.
In a recent case involving the responsibility of an insurance broker, the full loss of chance analysis was applied in dissent by Jacques J.A. (Sol-Air B.G. Inc. v. Marsh & McLennan, [1988] R.R.A. 206 (C.A.)). I think that it is worthwhile to consider this case briefly.
Sol-Air provided airline services. Its airplanes were insured through Marsh & McLennan, insurance broker. The policy listed the planes which were insured and the pilots who were authorized to fly them. Wishing to add a pilot to the list, Sol-Air contacted its broker and requested the change to the policy. The broker dealt with the request negligently, notably failing to return the client's calls. Before obtaining confirmation from the broker, Sol-Air assigned the pilot to flying duties, and one of those flights crashed, resulting in the destruction of the airplane. Sol-Air claimed the full value of the airplane and alleged that the broker's negligence had caused the loss.
The trial judge and a majority of the Court of Appeal (Rothman J.A. and Chevalier J. (ad hoc)) found that the broker was indeed negligent, but that it was Sol-Air's own negligence in allowing Allard to fly prior to confirmation of policy coverage that had truly caused the loss in question. The trial judge added that Sol-Air could have claimed for the down time during which the airplane or pilot were unavailable for service had the proof been made along those lines.
Jacques J.A. dissented, stating that the trial judge had not appreciated the direct and immediate consequence of the broker's negligence: [translation] "The direct and immediate consequence of the broker's negligence was the loss of a real and serious chance to obtain the necessary insurance coverage" (at p. 208). He then noted that it was not certain but highly probable that the new pilot would have been added to the list coverage. He repeated the discussion of loss of chance which appears in his reasons in the present case and concluded that Sol-Air lost a near certain chance, a chance so close to certainty that it should in fact be evaluated at full value, i.e. the value of the airplane. He then considered the fault of Sol-Air in failing to confirm the new insurance coverage and judged that this was another, equal cause of the damage. Accordingly, he would have required the broker to pay half of the damage represented by the value of the airplane.
In my opinion, loss of chance analysis adds unnecessary confusion to such a case, notably by the manner in which it diverts attention from the full and proper connection between fault and damage. As I understand the case, the trial judge had more to consider than the chance of obtaining insurance coverage. Had that been the only issue, it appears that the result would have been clear given the high probability of new coverage being granted. However, the causal analysis involved more complicated questions such as whether, and if so at what time, the insured was entitled to presume that insurance was in force and to permit its pilot to fly. It seems to me that the loss of chance analysis is artificial here and in fact hides a break in the causal link. A full analysis requires, as far as possible, a consideration of the entire factual situation, from the fault to the actual damage.
I will now turn to the cases which have dealt with loss of chance in the medical context. In these cases, the court is often placed in a situation where evidence is scarce and where the trial judge must rely on medical expertise, often presented in tentative or statistical form. Both of these factors lend themselves, as we have seen in France and Belgium, to clear difficulties in the establishment of the causal link and the assessment of damages. There is a temptation to overcome these difficulties by removing the analysis of the case to a more abstract, statistical level. In general, however, perhaps because of less rigid rules regarding causation, Quebec courts focus on the actual damage which has occurred and the actual benefit which would have been conferred.
Loss of chance situations arise, notably in the case of omissions, but the analysis does not appear to change here. I note, however, that some difficulties emerge in situations where the court must consider the prejudice which a patient has suffered as a result of unprofessional care which a doctor has administered, along with the benefit which the patient has forgone due to a failure to use a more appropriate treatment. I will look now at the cases cited by the parties, in order to expand on these preliminary comments.
The respondent cites the case of Tardif v. Laverrière, supra, which is also referred to by Baudouin, op. cit., in the section on loss of chance referred to earlier. In Tardif, a young hydrocephalic boy had been treated by a chiropractor over the course of seven years. It was found that the treatments employed by the chiropractor were medically unsound. They had neither worsened nor improved his condition, but they had involved a certain amount of pain and suffering for the boy. There was also some question as to whether the boy could have benefited during those seven years from an operation to cure him or from various treatments which might have led to some improvement in his condition.
The court held that the chiropractor had committed contractual and delictual faults and should be held liable for damage which he could be said to have caused. Regarding the operation which might have cured the boy of his ailment, the trial judge noted that expert opinion predicted certain benefits in terms of comfort and cosmetics but concluded, at p. 13, that:
[translation] . . . this is an uncertain operation which, with all the difficulties it entails, would probably not result in any significant mental improvement; the injury to the brain is congenital and incurable.
It is thus not surprising if one physician recommends the operation and another does not.
It followed that if the plaintiff had failed to show (at p. 14):
[translation] . . . that it was the defendant's conduct which was the cause of the fact that the chance for improvement which could have resulted from an operation was not taken.
The trial judge was also of the opinion that the boy's father was not likely to have consented to the operation in any event (at p. 14):
[translation] In short, it is very doubtful that the operation did not take place owing to the fault of the defendant; that remains only a possibility, just as success was only a possibility.
The trial judge went on, however, to say that the unsound treatment administered by the chiropractor caused the boy pain and suffering and also caused him to miss out on other, beneficial treatments (at pp. 14-15):
[translation] The fact remains, however, that while it is doubtful that the defendant caused the child to miss the chance of improving his physical or mental condition by an operation, there is no doubt that the fact of subjecting him for seven years to repeated, stressful and sometimes painful treatments that were useless was a source of hardship and stress from which he undoubtedly suffered . . . It is also probable that this might have made him miss other treatments that could have improved his condition.
The trial judge noted that damages in such a case were extremely difficult to quantify. In the end, he saw fit to compensate the plaintiff (at p. 15):
[translation] . . . taking into account on the evidence presented all the hardship, stress, delays, discomfort, suffering (not forgetting the burns caused by thermotherapy) and loss of chance of a minimal improvement which the defendant's presence may have caused, I feel it is fair to assess the damage resulting to the child at $5,000. [Emphasis added.]
In addition to the damage which the doctor caused by his unprofessional acts, this case clearly recognizes a lost chance. We are not given enough information to know whether the judge recognized the chance itself and compensated according to its probability or whether he recognized a small lost benefit and deemed it probable. Both solutions could have produced a similar result.
That which is clear is that the trial judge refused to recognize the potential benefits, or the lost chance, of the operation, though clearly such benefits were possible. If he had done so, this case would be a strong precedent in favour of the use of loss of chance in medical cases along the French model. In fact, the trial judge viewed this damage as too "aléatoire" ("uncertain") which suggests that he was not tempted to use that model.
The appellant cites a number of more recent cases in which the loss of chance theory is more directly addressed, and ultimately rejected.
In Lachambre v. Perreault (1983), [1990] R.R.A. 397 (C.A.), a worker, Lachambre, suffered an accident in which his leg was crushed. The leg was treated by Dr. Perreault who immobilized the leg and placed it in a cast. After a few days, the patient was taken to another hospital which was better equipped to deal with the vascular problems associated with the injury. At that time, doctors concluded that revascularization of the lower leg was impossible and that amputation was necessary. Lachambre claimed that the amputation was made necessary due to the fault of Perreault, who, so it was claimed, should have noticed and dealt with the circulatory problem at an earlier time.
The trial judge concluded that Perreault was not at fault and that there was probably no causal link between the doctor's alleged fault and the damage incurred by Lachambre. The Court of Appeal dealt only with the issue of the causal link. McCarthy J.A. stated first, at p. 398, that the law of Quebec requires proof of causation on the balance of probabilities:
[translation] Of course, the causal relationship between the fault and the damage does not need to be shown with certainty, only on a balance of probabilities: Lévesque v. Comeau.
He concluded on the basis of expert testimony that even the most prompt and appropriate medical attention would not have saved the leg, though he could not categorically exclude that possibility. His concluding remarks, at p. 399, address the argument based on loss of chance:
[translation] The evidence therefore seems to me to be far from showing on a balance of probabilities that there is a causal connection between the fault of Perreault, if there was any, and the damage complained of by Lachambre. Can it still be said, always assuming that Perreault should have dealt with the blood circulation problem much sooner, that such a fault deprived Lachambre of the chance, however small, of avoiding the amputation and that there should be compensation for the loss of this chance?
I do not think so, despite some French case law and scholarly commentary to this effect. This case law has been vigorously criticized by, Professor Savatier, among others. Since Lachambre did in fact have the amputation, there is in my opinion no longer any question of compensating him for loss of the chance of avoiding it. The only question is as to compensation for the amputation itself, if the evidence shows that it was probably due to Perreault's fault. As I have already indicated, this condition has not been met. [Emphasis added.]
Bélanger J.A. and L'Heureux-Dubé J.A., as she then was, concurred with these reasons.
It seems to me that the approach used in Lachambre is typical of Quebec courts. For example, while causation must be clearly established, it can be proved on the balance of probabilities. Furthermore, in cases involving lost chance, it is not the chance itself -- possible (as here) or probable -- which is considered, but the outcome of that chance in so far as it bears on the patient's present condition. Where the loss of chance alone is considered to be the damage, it is theoretically consistent to admit possibilities as well as probabilities. In France, we have seen that this logical extension is masked to a certain degree by the notion of a "real and serious" chance. By focussing on the outcome of the chance, the courts in Quebec ensure that the causal link between the fault and the actual situation now experienced by the plaintiff is established at least on the balance of probabilities.
The respondent, however, cites a more recent case which appears to go against the grain of the Quebec approach described above. In Zuk v. Mihaly, [1989] R.R.A. 737 (Sup. Ct.), the wife and daughter of the deceased, Mr. Zuk, claimed $126,982 in damages allegedly caused by the fault of an elderly doctor, Mihaly. Mr. Zuk had experienced some discomfort and had gone, accompanied by his wife, to Dr. Mihaly's office which was situated in the neighbourhood. Dr. Mihaly's wife met the Zuks and informed them that the doctor's office was not open. The Zuks left and, only moments later, Mr. Zuk had a heart attack and collapsed in the Mihalys' front yard. Dr. Mihaly was informed and came to the side of Mr. Zuk, without, however, providing any medical care other than placing his hands on the victim's chest. The ambulance arrived twenty-five minutes later, and the attendants immediately provided the victim with oxygen. Once at the hospital, Mr. Zuk was pronounced dead.
The trial judge found that Dr. Mihaly was clearly at fault in failing to provide the victim with oxygen. Regarding causation and damages, Martineau J. had this to say, at p. 738:
[translation] However, and the evidence is in no way contradicted on this point, despite his presence (and the increasing presence of curious members of the public), despite the knowledge that professionally he could not have been unaware of the immediate needs and care which alone could have given the victim a chance in five to six minutes to survive his attack, the defendant physician's failure to act deprived Mr. Zuk of his only chance of survival.
For these reasons, the Court:
Allows the plaintiffs' action in part;
Orders the defendant to pay . . . $50,000 . . .
In my view, this case involved a loss of chance analysis which is consistent with the French model. The judge appeared to link the doctor's fault only to the loss of the chance itself. He did not ask whether this chance would probably have produced a more favourable result. Instead, he awarded partial damages, presumably in order to take into account that the chance was less than certain, and perhaps not even probable. If the lost chance itself is the damage, then it is sufficient that it be real and serious, and this, at least in the French understanding, can include probable as well as some borderline possible chances. Where, as appears to be the case in Quebec, a court must satisfy itself not only that a chance was lost but that, on the balance of probabilities, that chance would have been somehow realized (e.g., in the form of recovery, slightly longer life, less pain etc.), it is insufficient to halt the analysis at the lost chance.
The Zuk case was decided in the midst of the recent debate over whether the French version of loss of chance has been accepted in Quebec, and it may well have been decided on the assumption that the debate had been resolved in favour of the French approach.
A series of Court of Appeal decisions, including the decision appealed from, have considered loss of chance in very thorough fashion, and I will turn to them now (Gburek v. Cohen, [1988] R.J.Q. 2424; Laferrière v. Lawson, supra; Lapointe v. Hôpital Le Gardeur, [1989] R.J.Q. 2619). I will not go into the facts of these cases in any detail but will simply extract the comments which the learned justices of the Court of Appeal made with respect to loss of chance.
In Gburek, only Beauregard J.A. commented on the appropriateness of the French approach to loss of chance. At page 2445, he stated:
[translation] I wondered whether the appellant could take advantage of the theory of "loss of chance" developed in France over the last twenty‑five years. . . .
As we know, between 1962 and 1982 the Court of Cassation appears to have approved a theory put forward by some appellate courts that, if there is no evidence of a causal connection between fault and final damage, a victim could obtain partial compensation corresponding to the value of the chance lost to avoid the final damage. . . .
It is difficult for a Quebec lawyer to know exactly what the Court of Cassation feels on the point. . . . Personally, I humbly concur in the opinion of those who think that, if the causal link between fault and damage is not proven, the victim cannot be given partial compensation on the ground that there is a possibility that the fault could have caused the damage. [Emphasis added.]
I have already summarized the views of the judges in Laferrière v. Lawson, and I will not repeat those views here.
In the Lapointe case, four of the five judges expressed an opinion on loss of chance, the only refraining judge being Beauregard J.A., who had already expressed a disapproving view in Gburek.
Monet J.A. expressed what can be described as cautious acceptance of the theory, at p. 2625:
[translation] As to whether the theory of loss of chance has been accepted in Quebec medical law, I would have preferred not to express a formal opinion for the time being, in view of the fact that Laferrière v. Lawson is now pending in the Supreme Court of Canada. Nevertheless, as my colleagues have dealt with it, I will make a few comments. First, if one is going to rely on French scholarly analysis, there are those nowadays who would not say that René Savatier is the greatest authority on the particular area which concerns us, namely civil liability. Second, decisions of the French courts supported by legal writers have clearly laid the groundwork of the theory of loss of chance, which be it said in passing is not in any way a fashion, still less the most recent one. Thus, Prof. Geneviève Viney writes:
But the courts have not simply applied these general conditions of civil liability to limit the right to compensation for "loss of chance". They have formulated specific requirements by saying, first, that the chance lost must be real, and serious, and second, that the assessment of the compensation must take into account the uncertainty affecting realization of the lost chance.
. . .
Accordingly, by the taking of precautions to avoid abuses, the courts have managed to keep compensation for loss of chance within reasonable limits. These decisions have also been generally approved by legal writers.
(I note that the last-quoted passages from Viney refer to loss of chance in its classical French application. As noted earlier, her comments regarding loss of chance in medical cases were less approving.)
LeBel J.A., at pp. 2642-43, adopted the position of Vallerand J.A. in Lawson and clearly rejected loss of chance theory in the medical context:
[translation] 4. Nature of damage suffered
Jacques J.A. concludes that the damage suffered is loss of the chance of recovery. In my opinion, the damage suffered is otherwise. Under the rules of proof, fault, the damage suffered and the relationship between them are determined on the basis of probability. If it is not possible to establish, using this traditional formula of a balance of probabilities, that a fault was the cause of such damage, there is no liability.
It does not seem to me that the so‑called loss of chance theory, which has been applied by the French Court of Cassation from time to time, and which certain writers have at times vigorously recommended, should be accepted, especially with respect to causation. In our system, one encounters here an obstacle having to do with positive law requirements and the system of civil proof. The causal link must be established in accordance with the ordinary rules of civil proof or, in some cases, in keeping with the presumptions established by law.
. . .
It will perhaps be added that the theory of loss of chance could be applied with respect to the development of damage. It would provide a means to assess possible damage such as the loss of a lottery ticket, to take the classic example. Where medical or hospital liability is concerned, the situation is very different. The damage has not remained simply a potential event. The damage has occurred. The court must determine whether compensation should be awarded for it. [Emphasis added.]
Mailhot J.A., at p. 2647, expressed herself similarly, also adopting the reasons of Vallerand J.A. in Laferrière v. Lawson:
[translation] For the reasons given by Vallerand J.A. in Laferrière v. Lawson, I consider that the loss of chance theory should not be used to supply a causal link between fault and damage.
As Vallerand J.A. put it, I am of the view that loss of chance should only be considered at the time damages are being considered. [Emphasis added.]
In the Lapointe case Monet, LeBel and Mailhot JJ.A. agreed that causation had been established on the balance of probabilities and awarded full damages.
Jacques J.A. dissented in part. On the subject of loss of chance, he repeated the remarks which he made in Lawson regarding the nature of the theory, and concluded that a chance of recovery had been lost. He judged the probability of that chance at fifty percent and, accordingly, awarded half of the full damages that would have been payable had causation been certain, and exactly half the amount awarded by the majority.
Critical Analysis of Loss of Chance
Following on the division employed in the discussion of loss of chance in France and Belgium, I will deal with the theory first in its classical applications as a type of damage or a method of evaluating damages, and secondly in its more controversial application to medical cases where it is said to be an attenuation of the causal link.
This analysis began with the assertion that a loss of chance situation is theoretically troublesome even as a type of damage. The actual damage is contingent, hypothetical or future and, accordingly, it is uncertain and difficult to evaluate. The consensus in France and Belgium is that damages are nonetheless recoverable when the lost chance itself is considered the damage and where certain limits are respected: the chance must be "real and serious" (likely or probable) and the probability of the chance being realized must be taken into account in quantifying the damages.
I cannot help but note that among the classical examples of loss of chance, certain cases are more contingent, or probability-based, than others. When one speaks of a lottery ticket there is little or nothing other than probability in play. When one speaks of the chance of winning a legal appeal or benefiting from the support of a child, many other factors come into consideration and are weighed by the judge. In France and Belgium, the judge may prefer to focus on the chance itself and to compensate in accordance with its degree of probability. In Quebec, the judge will more commonly focus on the actual result of the lost chance, having satisfied himself or herself that such damage can be considered present and certain, at least according to his or her assessment of the factual context and the balance of probabilities, or that such damage is a consequence of some concrete initial damage and must therefore be taken into account in the evaluation of damages (where the degree of probability will be considered along with other relevant factors). In many of the classical loss of chance situations, the outcome of these differing judicial approaches will be quite similar, and, furthermore, the recognition and evaluation of such damages will often be less controversial in these situations: either because the realization of the chance was probable or, in cases where its realization was only possible, no other identifiable factor competes with the defendant's fault as a probable cause of the hypothetical damage (as in the case of the loss of a lottery ticket); or because the defendant's responsibility has already attached to some concrete initial damage of which the lost chance is an extension.
It is difficult to explain why different legal systems approach a problem in different ways. Mindful of the dangers of comparative law unequipped with full information and understanding of other legal systems, I will nonetheless hazard a few observations. It does seem to me that the French and Belgian regimes of civil responsibility have in the past relied to a greater degree than our own on the principles of the equivalence or conditio sine qua non theory of causation (see A. M. Honoré, Causation and Remoteness of Damage, ch. 7, International Encyclopedia of Comparative Law, vol. XI, Torts, ed. André Tunc (1971), p. 36). According to this theory, every condition is a cause of the harm which would not have occurred without it (Honoré, op. cit., at p. 33). If, as in France and Belgium, causal certainty is generally required, this theory becomes at once permissive (it admits even minor factors as causes provided their causal role is certain) and restrictive (it denies even major factors whose causal role is less than certain). It is perhaps as a result of reliance on this theory and emphasis on causal certainty, or perhaps for other systemic reasons, that French and Belgian courts have been especially alert to the contingency of lost chances. It may be that, according to their view of causation, the causal link was more safely anchored where the certain loss of the chance was itself considered the damage rather than the contingent result which that chance in fact represented.
In Quebec, as we have seen, the courts are satisfied with proof of causation on the balance of probabilities, perhaps influenced to a greater degree by the so-called theory of adequate causation (Baudouin, op. cit., at p. 183). According to the various formulations of this theory, an act or omission is an adequate cause if it is of a type apt, in accordance with experience, to be followed by a harm of the sort which occurs; or, put another way, if the consequence follows in the normal course of events (Honoré, op. cit., at p. 49). Because probability is already taken into account in this theory, it is no obstacle to recovery that the damage in question was not certainly caused by the alleged fault; it is sufficient that it is made real and credible by a combination of factual and statistical evidence. Accordingly, a judge may, without violating the general rules of causation, link the fault to contingent damage, provided that he or she is satisfied that such a link has been established with sufficient evidence as to that which would normally occur or that which follows in the normal course of events. It is in part for this reason, in my view, that loss of chance situations, at least in their classical contexts, are less theoretically troublesome in Quebec than in France and Belgium. While it is acknowledged that loss of chance situations must be analyzed with the aid of a judge's assessment of probabilities, it is not considered necessary to rely on the artificial notion of an intermediate damage called the "lost chance". The loss of chance situation always presents evidentiary problems, but provided that adequate evidence has been furnished through facts and statistics relating to the particular case, the judge will attempt to assess the actual or final damage, in effect, the chance as realized. Even in cases where statistical evidence is heavily relied upon the evaluation of damages is still aimed at approximating the chance as realized.
As I have suggested, it is only in exceptional loss of chance cases that a judge is presented with a situation where the damage can only be understood in probabilistic or statistical terms and where it is impossible to evaluate sensibly whether or how the chance would have been realized in that particular case. The purest example of such a lost chance is that of the lottery ticket which is not placed in the draw due to the negligence of the seller of the ticket. The judge has no factual context in which to evaluate the likely result other than the realm of pure statistical chance. Effectively, the pool of factual evidence regarding the various eventualities in the particular case is dry in such cases, and the plaintiff has nothing other than statistics to elaborate the claim in damages. Where the fault of the defendant has already been established and where no other identifiable competing causal factors have been identified, it may be open to the judge to evaluate the damages according to the chance alone. To transform this exceptional case into the theoretical basis for recovery in all loss of chance situations would be unnecessarily abstract, and, more importantly for the case before us, would give the mistaken impression that the court is more interested in the certainty of statistical chances than in the probable results which those chances represent.
With these considerations in mind, I turn to the role of loss of chance in the medical context.
Loss of chance becomes critically difficult, in France, Belgium and Quebec, as elsewhere, when it is employed as a method of analysis in the complex cases of medical responsibility. In the most difficult cases, such as the one which concerns us, the defendant doctor's fault cannot be easily attached even to any initial actual damage suffered by the plaintiff patient. Accordingly, it is analysis of the lost chance itself which will determine whether the doctor is at all responsible. The lost chance can be analyzed in two ways.
In France and Belgium, as we have seen, it is the chance itself which is considered, usually described as a chance of recovery or survival. The chance must be "real and serious", and this is said to include chances which are likely or probable (Chartier, op. cit., at p. 50), or chances where recovery or improvement is more likely than death or illness (Vacarie, op. cit., at pp. 916-17). The damages are, of course, awarded in relation to the chance itself, and therefore such damages are only partial. According to a recent commentator, loss of chance analysis is said to be appropriate in cases involving faults of omission; faults of commission must be analyzed according to a method which connects the fault to the actual death or sickness (Vacarie, op. cit., at p. 908). It is acknowledged even by supporters of the full loss of chance analysis in the medical context that by focussing on the lost chance rather than the actual damage which that chance represents the judge is effectively permitted to translate his or her doubts as to the causal link between the fault and the final damage into a reduced award for the patient (Boré, op. cit., para. 14; Vacarie, op. cit., at p. 907).
In Quebec, courts are more inclined to examine the damage which has already occurred, and to consider whether that damage was caused by the doctor's fault or by other identifiable factors. If the fault was causal, then full damages are awarded. Faults of commission are treated in the same manner as faults of omission, and, clearly, in more complicated cases both types of fault may be present, as was seen in the Tardif case, supra. The judge attempts, in effect, to determine whether and to what extent the doctor's acts and omissions are responsible for the situation in which the patient now finds himself or herself. A positive result which should have been produced and a negative result which should have been avoided are considered on the same terms, whether they correspond with acts or omissions.
As I have stated earlier, I am inclined to favour an approach which focusses on the actual damage which the doctor can be said to have caused to the patient by his or her fault, and to compensate accordingly. First, as I have said, I can see no basis for treating acts and omissions differently. Accordingly, there is no theoretical imperative directing courts to abandon traditional causal analysis and to adopt instead an essentially artificial loss of chance analysis. Secondly, while I concede that loss of chance analysis is less objectionable when used to evaluate damages in cases where the defendant's responsibility is otherwise clearly established or, perhaps, where no other causal factors can be identified, this type of analysis must be viewed with extreme caution in cases where there are serious doubts as to the defendant's causal role in the face of other identifiable causal factors. Even though our understanding of medical matters is often limited, I am not prepared to conclude that particular medical conditions should be treated for purposes of causation as the equivalent of diffuse elements of pure chance, analogous to the non-specific factors of fate or fortune which influence the outcome of a lottery. Thirdly, as has been pointed out frequently, in the medical context the damage has usually occurred, manifesting itself in sickness or death. In Savatier and Penneau's terms, the chance is not suspended or crystallized as is the case in the classical loss of chance examples; it has been realized, and the morbid scenario has necessarily played itself out. It can and should be analyzed by means of the generally applicable rules regarding causation.
Overall, then, not only do I question the independent recognition of a lost chance in all but the exceptional classical cases (such as the case of the lottery ticket), but I can certainly see no reason to extend such an artificial form of analysis to the medical context where faults of omission or commission must be considered alongside other identifiable causal factors in determining that which has produced the particular result in the form of sickness or death. As far as possible, the court must consider the question of responsibility with the particular facts of the case in mind, as they relate concretely to the fault, causation and actual damage alleged in the case. While probabilities are unquestionably a part of the assessment of these elements in the finding of responsibility, I am very reluctant to remove the analysis from the concrete to the probabilistic plane.
It is important to recognize that, in cases where the proof indicates that a particular procedure or treatment would probably (though not certainly) have produced a positive result, the patient will usually be able to recover damages under both of the methods described above. If the chance itself is compensated, however, damages will only be measured according to the level of probability. If the actual damage which has been caused is compensated, then the full value of the actual damage will be accorded. These differing results were seen in the majority and minority decisions in the Lapointe case.
Cases in which the evidence is scarce or seemingly inconclusive present the greatest difficulty. It is perhaps worthwhile to repeat that a judge will be influenced by expert scientific opinions which are expressed in terms of statistical probabilities or test samplings, but he or she is not bound by such evidence. Scientific findings are not identical to legal findings. Recently, in Snell v. Farrell, [1990] 2 S.C.R. 311, this Court made clear (at p. 328) that "[c]ausation need not be determined by scientific precision" and that "[i]t is not . . . essential that the medical experts provide a firm opinion supporting the plaintiff's theory of causation" (p. 330). Both this Court and the Quebec Court of Appeal have frequently stated that proof as to the causal link must be established on the balance of probabilities taking into account all the evidence which is before it, factual, statistical and that which the judge is entitled to presume. (See, e.g., Shawinigan Engineering Company v. Naud, [1929] S.C.R. 341, at pp. 343-45; Morin v. Blais, [1977] 1 S.C.R. 570, at p. 580; Laurentide Motels Ltd. v. Beauport (City), [1989] 1 S.C.R. 705, at p. 808; J.E. Construction Inc. v. General Motors du Canada Ltée, [1985] C.A. 275, at p. 278; Dodds v. Schierz, [1986] R.J.Q. 2623 (C.A.), at pp. 2635-36.)
If one takes, for example, a case in which a doctor neglects to employ a recommended procedure which is said to have a fifty percent chance of complete cure, a judge would not necessarily be bound by expert opinion which declined to conclude that application of the procedure to the patient would have avoided the patient's present worsened condition. The judge might well be justified in finding that the procedure in question would probably have benefited the patient, if other factors particular to that plaintiff support that conclusion. The judge's duty is to assess the damage suffered by a particular patient, not to remain paralysed by statistical abstraction.
If one moves then to a procedure which is recommended despite a mere twenty-five percent chance of success according to expert evidence, it is still not a foregone conclusion that the doctor's fault in not using this procedure must be said to have had no causal role in the patient's death or sickness. If the experts are examined properly, a judge might well find that he or she is justified in concluding that the omission of that procedure did not cause the death or sickness, but that it caused other lesser but clearly negative results (e.g. slightly shorter life, greater pain). The doctor's fault could then be judged causal to the extent of the aggravation of what was otherwise an inevitably terminal or morbid condition.
The plaintiff is aided in establishing his or her case by presumptions (as provided by art. 1205 C.C.L.C.) and by such factual and statistical evidence as will aid the judge in appreciating what Moisan J. described properly as [translation] "reasonable and prudent behaviour", "the natural order of things", "the sequence of cause and effect" and, generally, "the normal and ordinary course of events". The judge will want to pay especially close attention to the various causal properties of the doctor's fault as well as the particular character of the damage which has manifested itself. In some cases, where a fault presents clear danger for the health and security of the patient and where such a danger materializes, it may be reasonable for a judge to presume the causal link between the fault and such damage, "unless there is a demonstration or a strong indication to the contrary" (Morin v. Blais, supra, at p. 580, per Beetz J.). If, after all has been considered, the judge is not satisfied that the fault has, on his or her assessment of the balance of probabilities, caused any actual damage to the patient, recovery should be denied. To do otherwise would be to subject doctors to an exceptional regime of civil responsibility.
In conclusion, then, and with all due deference to those who have expressed other opinions, I do not feel that the theory of loss of chance, at least as it is understood in France and Belgium, should be introduced into the civil law of Quebec in matters of medical responsibility. In the Court of Appeal, Jacques J.A. states without elaboration that loss of chance is recognized in the common law. I have taken note of the vigorous debate which is taking place in the United States and can find no dominant jurisprudential position favouring loss of chance in that country. In the United Kingdom, the House of Lords has expressed reservations about loss of chance analysis, but has not, as yet, reached a settled conclusion about its possible application: Hotson v. East Berkshire Health Area Authority, [1987] A.C. 750. I have also made note of this Court's recent decision in Snell v. Farrell, supra, which I take to endorse traditional principles of causation, properly applied.
By way of summary, I would make the following brief, general observations:
.The rules of civil responsibility require proof of fault, causation and damage.
.Both acts and omissions may amount to fault and both may be analyzed similarly with regard to causation.
.Causation in law is not identical to scientific causation.
.Causation in law must be established on the balance of probabilities, taking into account all the evidence: factual, statistical and that which the judge is entitled to presume.
.In some cases, where a fault presents a clear danger and where such a danger materializes, it may be reasonable to presume a causal link, unless there is a demonstration or indication to the contrary.
.Statistical evidence may be helpful as indicative but is not determinative. In particular, where statistical evidence does not indicate causation on the balance of probabilities, causation in law may nonetheless exist where evidence in the case supports such a finding.
.Even where statistical and factual evidence do not support a finding of causation on the balance of probabilities with respect to particular damage (e.g. death or sickness), such evidence may still justify a finding of causation with respect to lesser damage (e.g. slightly shorter life, greater pain).
.The evidence must be carefully analyzed to determine the exact nature of the fault or breach of duty and its consequences as well as the particular character of the damage which has been suffered, as experienced by the victim.
.If after consideration of these factors a judge is not satisfied that the fault has, on his or her assessment of the balance of probabilities, caused any real damage, then recovery should be denied.
2. Application to the Facts in This Case
I have reviewed the detailed evidence submitted by the parties and have considered the findings of the trial judge. I agree with the majority of the Court of Appeal that the trial judge focussed primarily on the surgical technique employed by the appellant in 1971 rather than on those procedures and treatments which should have been available to the patient had she been appropriately informed and followed up on by her doctor. I cannot, however, endorse the conclusions of the Court of Appeal as to the extent of the damages.
Jacques J.A. bases his award of $50,000 in damages on the small hope of survival which medical science held out to Mrs. Dupuis had she been informed, followed more closely and consequently treated in a timely and proper fashion. For reasons which I have set out earlier, I do not feel that it is appropriate to focus on the degree of probability of success and to compensate accordingly; it is at least necessary that such a probability, or here, at most, a small possibility, translate into a concrete benefit for the patient which she can be said to have lost as a result of the doctor's fault.
In my view, the evidence amply supports the trial judge's finding that the appellant's fault could not be said to have caused Mrs. Dupuis' death seven years after the first diagnosis of cancer of the breast. Unfortunately, I must agree with the trial judge that all the evidence clearly confirms the stubborn and virulent nature of this disease.
I am less convinced, however, by the judge's implicit findings regarding the pain, anguish and suffering which the appellant's fault allegedly caused to Mrs. Dupuis. I have decided as a matter of law that proper causal analysis requires the judge to examine closely all elements of the damage, and with the greatest respect for the trial judge who was faced with a complex and difficult case, I find it necessary to reconsider his findings in this more narrow respect.
First, I am convinced that Mrs. Dupuis experienced a type of psychological suffering which was directly related to the appellant's failure to inform his patient of her condition. From 1975 until her death, she experienced the horrible rhythms of her disease and the regular and seemingly ineffectual treatments and medications in the knowledge that things might have been different had she known earlier and been treated earlier. Her chances may not have been sufficient for the law, but they were very real to her, no doubt. I think that it was also probable that the pain which she experienced as a result of the advancing disease was all the more distressing given that knowledge. While she was a person who was concerned with her health and prepared to seek the best medical evidence and abide by it, she was denied the opportunity and choice of doing so by reason of the appellant's failure to inform her. This led, for instance, to her continued use of contraception pills which she would have been advised against had her condition been known. I would therefore agree with Vallerand J.A.'s specific recognition of psychological damages, and, given the exceptional nature of this case, I would increase the amount in question to $10,000.
Furthermore, I am of the view that while the death caused by cancer was not caused by the appellant's failure to follow up on his patient, it is probable that Mrs. Dupuis was denied the benefit of earlier treatment which would have translated into some real improvement in her admittedly terminal condition. I note, for example, that expert witnesses for both parties stated that proper and timely treatment would probably have provided for the patient a better quality of life even in the face of such a malignant condition.
Dr. Jacques Cantin, expert witness for the respondent, stated the following in the course of his testimony:
[translation] There is a second aspect to patient follow‑up, namely the notion of quality of life.
Certainly, patients have been followed very regularly and die all the same, but if they are properly followed, and recurrences are detected early, the patient may have a better quality of life. I will explain.
If, for example, you have a patient whose first recurrence of the disease is one or two (2) pulmonary metastases, if the problem is taken at the start, when there are two (2) or three (3) small images on the lung, it is an easier problem for us to handle, the situation is not as serious for the patient as if, for example, the patient comes to us with water in her lungs: her quality of life will not be the same. We will prevent her having a lot of problems.
If, for example, this patient -‑ I suggest it is possible that if we catch it when she only has these small local recurrences, well, the problems that she has are fewer than theoretically if we wait until she has an invasion of the cervical plexus with a Claude Bernard Horner, and the other dimension that I would like to pass on to the judge, it is not true that this will necessarily cure more people, but in any case often the quality of patients' lives at the time they suffer recurrences can be better.
Dr. Roger Poisson, one of the appellant's expert witnesses, was even more confident on this score:
[translation] . . . [patients] live for several years in a more comfortable way than in the past, without the rate of survival being necessarily improved that much. I think it is mainly . . . the quality of life which is improved. . . .
. . .
. . . Now, to answer the judge . . . it has to be admitted that survival rates in breast cancer . . . have unfortunately not improved very much . . . and, you know, it is ironic because despite the advance in surgery, radiotherapy and so on . . . survival rates unfortunately remain too stable. On the other hand, quality of life has I think improved a great deal.
The improvement in the quality of life may not have been great, but I believe that deprivation of such a real and probable improvement should be recognized and compensated. I believe that $7,500 would be appropriate for the damage represented by the diminished quality of life which Mrs. Dupuis experienced as a result of the appellant's fault.
V -- Disposition
I would allow the appeal in part and order the appellant to pay respondent $17,500 with interest and the additional indemnity pursuant to art. 1078.1 C.C.L.C. beginning April 1, 1983 and with costs throughout.
//La Forest J.//
English version of the reasons delivered by
La Forest J. (dissenting) -- For the reasons given by Jacques J.A., I agree with the majority of the Court of Appeal that the civil law recognizes that the damage resulting from loss of chance is compensable in this case. Although I would have been inclined to award more modest damages, I do not think that I should substitute my opinion on this question for that of the majority of the Court of Appeal. I would accordingly dismiss the appeal with costs.
Appeal allowed in part with costs, La Forest J. dissenting.
Solicitors for the appellant: McCarthy Tétrault, Montréal.
Solicitors for the respondent: Robinson, Sheppard, Montréal.