(delivered the judgment of the court)
This appeal concerns a claim under an accident insurance policy by Mr. Gibbens who had unprotected sex with three women during January and February 2003. He thereby acquired genital herpes (HSV-2), which in turn caused inflammation of his spinal cord (transverse myelitis), a rare but known complication of herpes, which resulted in total paralysis from his mid-abdomen down. He claimed compensation under a group insurance policy on the basis that the paralysis resulted “directly and independently of all other causes from bodily Injuries occasioned solely through external, violent and accidental means, without negligence” on his part.
Accident insurance is not comprehensive health insurance. Mr. Gibbens contracted a sexually transmitted disease in the ordinary way through sexual intercourse. In most cases genital herpes is a minor irritant (if indeed there are any symptoms at all). I agree with the courts in British Columbia that Mr. Gibben’s paralysis was tragic and unexpected but I do not agree with them that it was caused by “external, violent and accidental means” within the meaning of the insurance policy. I would therefore allow the appeal.
Mr. Gibbens is a beneficiary under a Group Insurance Policy issued by the Co-operators Life Insurance Company, that covered certain specified risks as follows [emphasis added]:
“Covered Loss” means a Critical Disease Benefit, Accidental Death Benefit or an Accidental Disease/Dismemberment Benefit covered under this Benefit Provision. The Covered Loss must occur prior to the Member’s 71st birthday and while the Member is insured under this Provision. In the case of an accident, the Covered Loss must occur within 365 days after the date of the accident.
CRITICAL DISEASE BENEFIT
The Insurance Company will pay to the Member an amount equal to 10% of the Principal Sum provided the Member has been diagnosed with a Critical Disease while insured under this Provision and has been Totally Disabled from that disease for at least nine (9) months. Benefits are limited to the first covered Critical Disease in the Member’s lifetime.
ACCIDENTAL DISEASE/DISMEMBERMENT BENEFIT
If the Insurance Company is furnished with proof that a Member sustains one of the following losses, as a direct result of a Critical Disease or resulting directly and independently of all other causes from bodily Injuries occasioned solely through external, violent and accidental means, without negligence on the Member’s part, the Insurance Company will pay:
The policy does not contain any definition of “accident” or “accidental means”. It does, however, define “Critical Disease”:
“Critical Disease” shall mean any of the following diseases diagnosed after the Effective Date of the Member’s Insurance: Poliomyelitis, Parkinson’s Disease, Huntington’s Chorea, Multiple Sclerosis, Alzheimers Disease, Type 1 Diabetes (Insulin Dependent), Amyotrophic Lateral Sclerosis (ALS), Peripheral Vascular Disease and Necrotizing Fascitis.
Mr. Gibbens was employed in high pressure water blasting. The group policy was a benefit provided by his union. At the time of the trial he was 45 years old.
According to an agreed statement of facts Mr. Gibbens did not know that any of the three women with whom he had sexual intercourse during January and February 2003 had Herpes Simplex Virus Type-2 (HSV-2).
On February 17, 2003, Mr. Gibbens consulted his family physician complaining of headaches and myalgia (muscular pain) with back pain. His doctor diagnosed a viral illness and prescribed symptomatic treatment.
On February 21, 2003, Mr. Gibbens went to the emergency department at Eagle Ridge Hospital complaining of urinary retention for two days, some recent kidney/flank pain, supra pelvic discomfort, vomiting for two days, and poor physical coordination. On examination he was noted to be shaky/weak and it was hard for him to sit up. He was diagnosed with urinary retention, possible alcohol withdrawal and epigastric pain not yet diagnosed. Mr. Gibbens continued to deteriorate and by February 23, 2003, he was paraplegic.
In terms of causation, the parties agreed that by reason of the unprotected sex Mr. Gibbens had been infected with HSV-2 which caused the inflammation of his spinal cord (transverse myelitis) resulting in a permanent T3 ASIA C spinal cord injury (total paralysis from mid-abdomen down).
Under the policy, Co-operators Life was obligated to pay $200,000 if Mr. Gibbens suffered the paraplegia through “external, violent and accidental means”.
Mr. Gibbens acknowledged that he was aware of the risk of contracting a sexually transmitted disease when having unprotected sexual intercourse, but stated that he did not intend or expect to contract HSV-2 or to develop transverse myelitis.
II. JUDICIAL HISTORY
A. Supreme Court of British Columbia (Cole J.), 2007 BCSC 1076, 74 B.C.L.R. (4th) 393
Cole J. found for Mr. Gibbens. He concluded, based on Martin v American International Assurance Life Co., 2003 SCC 16,  1 S.C.R. 158, that to ascertain whether a given means is “accidental”, what must be considered is whether the consequences were unexpected. Accordingly, the question in this case was whether Mr. Gibbens expected to become a paraplegic as a result of having unprotected sexual intercourse. The answer was that he did not. The fact that his sexual conduct was foolish and risky did not preclude the consequences from being considered accidental within the meaning of the policy.
The trial judge rejected the insurer’s argument that Mr. Gibben’s paraplegia was caused by a disease and could not, in the circumstances, be considered to be the product of external, violent and accidental means. The trial judge held that “[d]iseases which do not result from a natural cause may be accidental” (para. 16). Mr. Gibbens was awarded $200,000 plus interest and costs.
B. British Columbia Court of Appeal (Newbury J.A. (Frankel J.A. Concurring)), 2008 BCCA 153, 77 B.C.L.R. (4th) 26
Newbury J.A. rejected the idea that the decision in Martin meant that the question of whether an event was accidental could be determined solely by considering whether it was unexpected. As the Martin decision made clear, it was equally important that the words “accident” and “accidental” be construed in accordance with the “ordinary person’s understanding”, or in “ordinary parlance” (citing paras. 19-20). Newbury J.A. agreed with the trial judge that in light of the decided cases, Mr. Gibbens’ transverse myelitis did not arise “naturally”. Rather, it arose from an external factor or “unlooked-for mishap” – the introduction of the HSV-2 virus into his body by a sexual partner. This was close to the line, but was sufficient to qualify as “accidental” in the ordinary meaning of that term.
Newbury J.A. added that if she were wrong in her view of Martin and the only question was whether Mr. Gibbens’ bodily injury was unintended or unexpected, it was obvious that it was, and applying that test the loss was caused by “accidental means” as required by the policy (paras. 14-23).
Saunders J.A., concurring, pointed out that Martin “did not concern the introduction of a pathogen into the insured’s body or the development of what by any standard must be considered a disease” (para. 35). “The world is populated with pathogens”, she observed, which are acquired through the “regular activities of living” (para. 37), and it may be difficult to perceive a principled difference between the unintended and unexpected contraction of a common ailment (which would not be considered an accident) from genital herpes acquired through sexual intercourse, yet the “state of jurisprudence weighs in favour of Mr. Gibbens” (para. 34). The prudent course is “to follow the view that has already been judicially taken” (para. 38), and leave it to the insurance industry to adapt the wording of its policies in light of jurisprudence.
A century and a half of insurance litigation has failed to produce a bright line definition of the word “accident”. Insurers have consistently declined to attempt to define the term in their policies. It has been left to the courts to interpret it, and the courts have found the analysis to pose, as an American court put it, “one of the more philosophically complex simple questions”: Fegan v State Mutual Life Assurance Co. of America, 945 F. Supp. 396 (D. N.H. 1996), at p. 399.
Generally speaking the cases exclude from the concept of accident a “bodily infirmity caused by disease in the ordinary course of events”, as summarized in Mr. Welford’s book:
The word “accident” involves the idea of something fortuitous and unexpected, as opposed to something proceeding from natural causes; and injury caused by accident is to be regarded as the antithesis to bodily infirmity caused by disease in the ordinary course of events.
(A. W. Baker Welford, The Law Relating to Accident Insurance (2nd ed. 1932), at p. 268)
The Welford definition was accepted as authoritative by this Court in Smith v British Pacific Life Insurance Co.,  S.C.R. 434, and in two other cases adopting reasons from the courts below, namely, Milashenko v Co-operative Fire & Casualty Co. (1970), 11 D.L.R. (3d) 128 (S.C.C.), and Columbia Cellulose Co. v Continental Casualty Co.,  I.L.R. ¶ 1-119 (S.C.C.). More recently, it was adopted by the English Court of Appeal in De Souza v Home and Overseas Insurance Co.,  L.R.L.R. 453, per Mustill L.J., at p. 458.
Welford argued that the element of accident may be manifested in the cause or in the result (p. 269). The distinction between accidental cause (or means) and accidental result has been controversial. Cardozo J. (in dissent) famously invoked Milton’s Paradise Lost in warning that “[t]he attempted distinction between accidental results and accidental means will plunge this branch of the law into a Serbonian Bog”: Landress v Phoenix Mutual Life Insurance Co., 291 U.S. 491 (1934), at p. 499. While the recent jurisprudence amply shows the existence (perhaps even the flourishing) of a Serbonian bog, we cannot entirely overlook the fact that Mr. Gibbens’ accident policy contains the time-honoured formula confining the risk to bodily injuries that are “occasioned solely through external, violent and accidental means” (emphasis added). The question is what meaning is to be given to those words in context.
It is evident that the parties did not expect the policy to cover all loss or bodily injury. Quite apart from the usual concept of “accident” as itself excluding a bodily infirmity caused by disease in the ordinary course of events, the policy provided separate coverage against the risk of enumerated “critical diseases”. Transverse myelitis is not listed among them (although poliomyelitis – which may also result in paraplegia – is so listed). Yet the decision of the B.C. Court of Appeal in this case effectively makes the insurer liable for all sexually transmitted diseases where, as here, the insurer does not contend that the claimant courted the disease either deliberately or negligently. Moreover, as will be seen, the decision would have considerable impact on liability for infectious diseases generally, which are spread in the usual course of events by viruses and bacteria passed from person to person, whether by sneezing in a bus, an unprotected cough in a crowded elevator, or a simple handshake. In my view, for the reasons that follow, the policy here excludes bodily injury from processes that occur naturally within the body in the ordinary course of events and, as well, from diseases that are transmitted in the ordinary way without any associated mishap or trauma except the spread (or inception) of the disease itself.
A. General Principles of Interpretation
The courts have developed a number of general interpretative principles that reflect a concern that customers not suffer from the imbalance of power that often exists between insurers and the insured but, on the other hand, that customers obtain no greater coverage than they are prepared to pay for. The exercise of interpretation should avoid “an unrealistic result or a result which would not be contemplated in the commercial atmosphere in which the insurance was contracted”: Consolidated-Bathurst Export Ltd. v Mutual Boiler and Machinery Insurance Co.,  1 S.C.R. 888, per Estey J., at p. 901.
(1) Words Like “Accident” Should Be Given Their Ordinary Meaning
In Mutual of Omaha Insurance Co. v Stats,  2 S.C.R. 1153, Spence J. stated that the word “accident” is “an ordinary word to be interpreted in the ordinary language of the people” (p. 1164). Such terms should be construed “as they would be understood by the average person applying for insurance, and not as they might be perceived by persons versed in the niceties of insurance law”: National Bank of Greece (Canada) v Katsikonouris,  2 S.C.R. 1029, at p. 1043. This approach was affirmed by McLachlin C.J. in Martin, at para. 19.
What then is the “ordinary meaning” of “accident”? In Canadian Indemnity Co. v Walkem Machinery & Equipment Ltd.,  1 S.C.R. 309, at pp. 315-16, Pigeon J. suggested that an accident is “any unlooked for mishap or occurrence” (p. 316). This definition was endorsed in Martin, at para. 20. Martin also quoted Stats which similarly held that “‘accident’ .... denot[es] an unlooked-for mishap or an untoward event which is not expected or designed” (ibid.). The verbal formula stems from Fenton v J. Thorley & Co.,  A.C. 443 (H.L.), at p. 448. Of course, as Welford points out, not every unexpected mishap is an accident. Someone who picks up a disease “in the ordinary course of events” would not ordinarily be described as having been in “an accident”.
(2) A Generous Interpretation Should Be Given to the Term “Accident”, Unless a Policy Clearly Restricts it
In Walkem, Pigeon J. observed that the jurisprudence assigns a generous meaning to “accident” in the absence of limiting language in the policy (p. 316). Yet, generosity has its limitations as a principle of contractual interpretation. Insurance is written to protect against certain defined risks. Care should be taken not to convert, for example, an accident policy into a general health, disability, or life insurance policy. Accident insurance is relatively cheap compared to the more comprehensive forms of insurance.
Different policies are written to address different needs, although inevitably there will be some overlapping of coverage. In Wang v Metropolitan Life Insurance Co. (2004), 242 D.L.R. (4th) 598 (leave to appeal dismissed,  1 S.C.R. xvii), discussed below, there was no doubt the husband was entitled to the death benefit under the life insurance policy where his wife died as a result of complications that arose in giving birth to their child. At issue was whether he was also entitled to a double indemnity in the form of an accidental death benefit. The Ontario Court of Appeal, Laskin J.A. dissenting, held that he was not so entitled. The added element of “accident”, the majority held, was absent.
(3) The Words of an Insurance Contract, When Ambiguous, Should Be Construed Against the Drafter (i.e. the Insurer) (Contra Proferentem)
Whoever holds the pen creates the ambiguity and must live with the consequences. In Brissette Estate v Westbury Life Insurance Co.,  3 S.C.R. 87, at p. 114, Cory J. wrote:
It is right and just to interpret the ambiguities in favour of the insured. It is the insurance company which draws up a contract of insurance. It is the company which determines the clauses which will go into a standard form of contract. It is that standard form of contract which is offered to the people in all walks of life on a take-it-or-leave-it basis.
(See also Consolidated-Bathurst, at p. 899.)
This doctrine is complemented by other rules of contractual interpretation which can assist courts where ambiguity is present. None is relevant here.
(4) Where a Policy is Ambiguous, Effect Should Be Given to the Reasonable Expectations of the Parties
In Consolidated-Bathurst, Estey J. wrote at pp. 901-2 that
[t]he courts should be loath to support a construction which would either enable the insurer to pocket the premium without risk or the insured to achieve a recovery which could neither be sensibly sought nor anticipated at the time of the contract.
Similarly, in Reid Crowther & Partners Ltd v Simcoe & Erie General Insurance Co.,  1 S.C.R. 252, McLachlin J. urged “[t]he desirability, at least where the policy is ambiguous, of giving effect to the reasonable expectations of the parties (p. 269).
(5) Continuity of Interpretation
As Newbury J.A. pointed out in the court below, “courts will normally be reluctant to depart from [authoritative] judicial precedent interpreting the policy in a particular way” (para. 30) where the issue arises subsequently in a similar context, and where the policies are similarly framed. Certainty and predictability are in the interest of both the insurance industry and their customers. See Algoma Steel Corp. v Allendale Mutual Insurance Co. (1990), 72 O.R. (2d) 782 (C.A.) (leave to appeal refused,  1 S.C.R. v); MacGillivray on Insurance Law (10th ed. 2003), at para. 11-3; G. R. Hall, Canadian Contractual Interpretation Law (2007), at pp. 93-94; and E. R. Hardy Ivamy, General Principles of Insurance Law (6th ed. 1993), at p. 355.
It was not suggested in the present case that these principles of interpretation should be modified because Mr. Gibbens’ claim is under a group policy. The contractual nature, and the process of negotiation, of group policies is somewhat different than that of individual policies: see generally D. Norwood and J. P. Weir, Norwood on Life Insurance Law in Canada (3rd ed. 2002), at p. 212. Nevertheless, in the absence of argument to the contrary, I proceed on the basis that the usual principles of interpretation apply here.
B. The Relationship Between Accident and Disease
Traditionally, the courts have carved out of the potential universe of “unlooked-for mishaps or untoward events which are not expected or designed” the sub-universe of bodily “infirmit[ies] caused by disease in the ordinary course of events”. In adopting this approach to the definition of “accident”, Welford was treading a well-worn path reaching back at least to 1861 when Cockburn C.J. decided Sinclair v Maritime Passengers’ Assurance Co. (1861), 3 E.L. & E.L. 478, 121 E.R. 521, at p. 523. In that case, a sea captain had sailed his ship The Sultan to the southwest coast of India, and while “superintending the turning of the said ship . . . was struck down by a sunstroke, and died the same day from the effects of the said sunstroke” (p. 523). The court held that sunstroke “is a disease to which persons exposing themselves to the sun in a tropical climate are more or less liable” (pp. 524-25), and the sea captain’s death must therefore “be considered as having arisen from a ‘natural cause’ and not from ‘accident’ within the meaning of this policy” (p. 525). A similar conclusion was reached in Wyman v Dominion of Canada General Insurance Co.,  2 D.L.R. 268 (Ont. H.C.), where a street cleaner died from heat exposure while working the streets of Toronto on the second hottest summer day of 1935. In England the more recent De Souza (1995) case reached the same conclusion on similar facts (tourist goes to hot climate, spends too much time in the sun, succumbs to sunstroke).
In the course of giving his judgment in Sinclair, Cockburn C.J. offered an illustration of how, in his view, the concepts of accident, disease and natural causes related to one another [p. 524]:
Thus disease or death engendered by exposure to heat, cold, damp, the vicissitudes of climate, or atmospheric influences, cannot, we think, properly be said to be accidental; unless at all events, the exposure is itself brought about by circumstances which may give it the character of accident. Thus (by way of illustration), if, from the effects of ordinary exposure to the elements, such as is common in the course of navigation, a mariner should catch cold and die, such death would not be accidental; although if, being obliged by shipwreck or other disasters to quit the ship and take to the sea in an open boat, he remained exposed to wet and cold for some time, and death ensued therefrom, the death might properly be held to be the result of accident.
For present purposes, keeping in mind the importance of continuity of interpretation, I make three observations on the interpretation of “accident” in an insurance policy.
Firstly, in both of Cockburn C.J.’s alternative fact situations, as well as in Wyman and De Souza, death occurred from natural causes (i.e. exposure). Such a death may be classified as the outcome of an accident, or it may not, depending on how the court characterizes the facts that gave rise to it.
Secondly, the traditional key to liability under an accident policy has been to identify a “mishap or untoward event” to which the disease or death can be attributed. Unless the bodily injury arose from a mishap of some sort there can generally be no liability under an accident policy.
Thirdly, in choosing the example of “catching a cold” in the context of deciding a “sunstroke” case, Cockburn C.J. made it doubly clear that by “disease” he did not limit himself to degenerative diseases brought about by age or wear and tear. Young and healthy people suffer sunstroke and catch colds. Such ailments can, when shown to be the product of an antecedent event correctly characterized as accidental, be covered. Sinclair is also an early example of the principle that the entire chain of events must be looked at, and not just the means or the end result, to determine if the loss can be considered accidental, an approach recently affirmed by this Court in Martin, at para. 14.
Disease for present purposes may be defined as:
In general, a departure from the normal state of health. More specifically, a disease is the sum total of the reactions, physical and mental, made by a person to a noxious agent entering his body from without or arising within (such as a micro-organism or a poison), an injury, a congenital or hereditary defect, a metabolic disorder, a food deficiency or a degenerative process. These cause pathological changes in organs or tissues which are revealed by characteristic signs and symptoms.
(Butterworths Medical Dictionary (2nd ed. 1978), at p. 523)
The distinction between “accidents” and “diseases contracted in the ordinary course of events” is often not an easy one to make in practice. The difficulty is illustrated by Brintons, Ltd. v Turvey,  A.C. 230. In that case a man died after working with wool in the defendant’s factory. A bacillus from infected wool had found its way into the man’s eye, and it developed into anthrax. The insurer argued that this was death from disease, not by accident. Lord Lindley in the House of Lords, in rejecting this position, held that “[t]he fact that an accident causes injury in the shape of disease does not render the cause not an accident” (p. 238), but the dissenting judgment of Lord Robertson supplied a pertinent caution. He expressed the concern that as a result of allowing such a claim “[I]n every case in which a man dies of any infectious disease .... all he has got to do is to get the doctor to prove (what could not be disputed) that a bacillus did it, and the accident is there”. Referring to illustrations given by his colleague the Earl of Halsbury L.C. in the same case of “tetanus, pneumonia or erysipelas ensuing on an accident”, he continued, “such illustrations differ from the present case in one point essential to the controversy, for in the illustrations there is postulated an accident distinct from the disease, while in the case before your Lordships the so-called “accident” is simply the inception of the disease” (pp. 235-36). In this, perhaps, Lord Robertson anticipated the dictum of Cardozo J. that “if there was no accident in the means, there was none in the result” (Landress, at p. 501).
Brintons was decided under a legislative scheme to compensate workers injured on the job. Such schemes are, as they ought to be, generously interpreted in favour of injured workers. Decisions about the meaning of “accident” under such legislative schemes often turn on the particular statutory text, purpose and legislative history. Accordingly such decisions should be viewed with caution when it is sought to apply them to private insurance policies. Nevertheless, the debate in Brintons about the general relationship between accident and disease helpfully illustrates an analytical issue at the heart of the present appeal. I quote it only for that limited purpose.
A more recent illustration of the problematic relationship between accident and disease is Toronto Professional Firefighters’ Association v Toronto (City) (2007), 223 O.A.C. 146 (Div Ct.). In that case a claim was brought on behalf of a firefighter who died of renal failure caused by his contact with toxic substances over his 20 years fighting fires. A claim for accident coverage was made under the terms of a collective agreement. A labour arbitrator decided that the insured did not die as a result of an “accident” but his decision was quashed by the Ontario Divisional Court. It was established that the firefighter’s renal cancer was caused by “exposure to toxic substances .... when the dangers were unknown and the safety equipment was unsafe” (para. 53). This was a case where an “unlooked for mishap or occurrence” caused a disease. It was not a case of the spread of the disease itself in the ordinary course of events. In the language of Brintons, there was “postulated” in Toronto Professional Firefighters an accident (or chain of events) distinct from the disease.
C. The Definition of “External, Violent and Accidental Means”
As Sinclair shows, the fact that Mr. Gibbens’ paraplegia was caused by a disease does not in itself bar his claim. The question according to the insurer, relying on Welford, is whether Mr. Gibbens acquired the disease “in the ordinary course of events”. The insurer says that nothing could be a more ordinary part of life than sexual intercourse with a willing partner. Mr. Gibbens responds that paraplegia, being a totally unexpected outcome of sex “in the ordinary course of events”, should be deemed to have come about through “accidental means” within the meaning of the policy.
Welford’s definition of accident is a good starting point. For ease of reference I repeat it [p. 268]:
The word “accident” involves the idea of something fortuitous and unexpected, as opposed to something proceeding from natural causes; and injury caused by accident is to be regarded as the antithesis to bodily infirmity caused by disease in the ordinary course of events.
Welford defines accident as much by what it is not, i.e. “bodily infirmity caused by disease in the ordinary course of events”, as by what it is, i.e. “something fortuitous and unexpected”. In the present appeal, as might be anticipated, counsel for Mr. Gibbens spent much time emphasizing the latter phrase at the expense of the former, and counsel for the insurance company did the opposite. The acquisition of a disease can frequently be considered an “unlooked for mishap”, broadly speaking. The challenge in this appeal is to relate different types of insurance risk and coverage in a way that makes sense “in the commercial atmosphere in which the insurance was contracted”: Consolidated-Bathurst, at p. 901.
For present purposes it is important to keep in mind that diseases are transferred from person to person through natural processes such as coughing or sneezing in someone’s presence “in the ordinary course of events”. The viruses thus transmitted may, in some situations, prove to have calamitous and unexpected consequences. Yet, if such transmissions are, viewed with hindsight, to be classified as accidents, then the accident policy becomes a comprehensive health policy. Moreover, such a theory comes close, as Professor Adam Scales writes, “to finding an accident simply by tracing the path of the bacillus. It is the fortuity of “knowing” how the disease reached its ultimate victim that lends a superficially accidental character to the transaction” (A. F. Scales, “Man, God and the Serbonian Bog: The Evolution of Accidental Death Insurance” (2000-2001), 86 Iowa L. Rev 173, at p. 288).
On the other hand, the insurance industry ought not use “disease” to extricate itself from valid accidental claims. As the Earl of Halsbury L.C. put it in Brintons, “It does not appear to me that by calling the consequences of an accidental injury a disease one alters the nature of the consequential results of the injury that has been inflicted” (p. 233-34).
D. The Expectation Test
Mr. Gibbens argued that Martin did away with the need to consider the “accidental means” if the resulting death or disease was “unexpected” but this is not so. In Martin, a doctor, who was addicted to Morphine and Demerol, killed himself by a self-administered drug overdose. The Court rejected the insurer’s argument that the category of death caused by accidental means was narrower than the general category of accidental death. However, Martin, not being a “disease” case, found it unnecessary to address the traditional distinction between “disease” and “accident”. Instead, the question “Did the insured expect to die?” went to the issue of whether a death brought about by the deliberate act of self-injection could be considered accidental (through his miscalculation of the dosage) or was it necessarily to be considered suicide. The result (death) was consistent with either hypothesis. However, it was reasoned, if the doctor’s miscalculation created an “unlooked-for mishap or an untoward event which was not expected or designed”, death by overdose would qualify as an accident. In Martin the policy language was for all practical purposes the same as in the present case.
In Martin, the Chief Justice, at para. 14, quoted Justice Cardozo’s observation that “if there was no accident in the means, there was none in the result”: Landress, at p. 501. “‘Means’”, she wrote, “refers to one or more actions or events, seen under the aspect of their causal relation to the events they bring about” (para. 13). Accordingly, in Martin, the end result taken in isolation did not determine the outcome of the claim. The core of the argument was whether the means (the injection) was deliberate (suicide) or accidental (i.e. a miscalculation or misjudgment by the insured doctor). Consideration of the means by which the loss was brought about was not excluded from the analysis. The circumstances surrounding the fatal injection were crucial. It could not be otherwise given the language of the accidental risk endorsement.
The focus of the Chief Justice’s reasoning in Martin is evident from the line of “miscalculation” cases on which she relied. The insured in Glenlight Shipping Ltd. v Excess Insurance Co., 1983 S.L.T. 241 (Sess. 2nd Div), died when he drove off a ferry into the sea, erroneously believing that the ferry had reached the pier. In Cornish v Accident Insurance Co. (1889), 23 Q.B.D. 453 (C.A.), the insured died when he was hit by a train as he crossed a railway track, having failed to notice the oncoming train. The insured in Brown v Continental Casualty Co., 108 So. 464 (La. 1926), regularly inhaled chloroform to relieve headache and insomnia, and killed himself when he mistakenly took more than the usual dose. The insured in Bertalan Estate v American Home Assurance Co. (1999), 68 B.C.L.R. (3d) 118 (S.C.), made regular use of nitrous oxide, and died when he failed to remove the mask that delivered the gas to him in a timely fashion. In all these cases, death was held to be accidental. In all of them death was unexpected. All of them resulted from a miscalculation or misjudgment of an event on the part of the deceased insured. None of them involved disease.
At the same time, a claimant who can establish that death was unexpected does not thereby, without more, establish a valid accident. Otherwise, every bad happening, natural or unnatural, whether caused by disease in the ordinary course of events or otherwise, would be classified as an accident. Take the case of an insured who is sitting on a couch in front of her television set when suddenly she suffers a stroke and dies. The tragedy is totally unexpected. Yet, there is no accident involved in any ordinary manner of speech.
E. Disease in the Ordinary Course of Events
We come now to the unsuccessful attempt in Wang to apply a simple “expectation test” to a disease/natural causes situation. In that case the deceased woman suffered a cardiac arrest soon after giving birth by elective caesarean section. The caesarean was performed in the standard fashion and was uneventful. Shortly thereafter, however, she encountered difficulty with her breathing, turned blue, and died of cardiac arrest. A post-mortem examination established the cause of death as “amniotic fluid embolism”, which was described as a rare, unpredictable and non-preventable complication of pregnancy. The insurance policy provided added compensation on “proof that the insured died, directly and independently of all other causes, as a result of an accident” unless the death was “caused or contributed to, directly or indirectly, by physical or mental illness or treatment for the illness” (para. 3). The husband of the deceased argued (relying on Martin) that the central question was whether a reasonable person in the position of his wife would have expected to die (para. 7). The answer obviously was no, but, in the majority view, the unexpected nature of the outcome did not, without more, establish an accidental loss. Charron J.A., as she then was, pointed out that Martin, at para. 23, stated that “[t]he expectation test can be applied generally to all cases in which death appears to be accidental” (Wang, at para. 25 (emphasis added)). A death from natural causes is not one that would “appear to be accidental” in ordinary speech. Thus [para. 27]:
.... a person’s unexpected death during his sleep, caused by aneurysm or other condition with fatal consequences, would not be described as an accident in ordinary and popular language.
In other words, expectations, while relevant, must be placed in the context of the other circumstances of the case. Neither the means nor the result should be viewed in isolation.
Wang illustrates that just because an outcome is unexpected does not establish the existence of an accident within the scope of the policy. In Martin, the Court was able to infer accidental means from the circumstances of the death that pointed to a miscalculation. The Court rejected the insurer’s argument that deaths that are the natural and predictable consequences of deliberate actions could never be considered accidental, but the Court, nevertheless, found it necessary to identify a miscalculation as a necessary step in finding the claim to be valid (para. 17). There is no necessary equivalence between “unexpected” and “accident”. If a man, sitting at a bus station, is hit by a bus that has careened out of control, that is unquestionably an accident – but it is not an accident by virtue of the fact that the man did not expect it.
According to the “expectations” test advanced by Mr. Gibbens, any unexpected loss becomes an accident. But this ignores the fact that the insurer did not issue a policy against the risk of “unexpected bodily injuries”.
Wang concerned death caused by the natural processes of the body in the ordinary course of the delivery of a baby. In the case of Mr. Gibbens we are dealing with a disease transmitted in the ordinary course of having sex.
F. The Scope of Coverage Provided to Mr. Gibbens
At this point, the question is to determine, in light of the general principles of interpretation already cited, what “the average person applying for insurance” would understand by the terms of Mr. Gibbens’ group policy. He or she would take notice not only of the written text but the juxtaposition of limited coverage for critical diseases and the separate and distinct coverage for bodily injuries occasioned “solely through external, violent and accidental means”. The prospective customer would thereby understand that he or she was not buying comprehensive health or disability insurance. On the other hand, unlisted diseases or other bodily infirmities might still be covered if attributable to some antecedent event or events that could, together with the unexpected result, be characterized as accidental.
A bit of reflection would suggest that in the case of an unlisted disease that pre-dates the event said to be the accident, there would be no coverage if the “event” simply provided an occasion on which the pre-existing disease manifested itself, e.g. the heart patient who goes out for a walk and, on being startled by the sound of a car horn, experiences an incapacitating cardiac arrest. On a common sense approach the incapacitation would be attributed to the pre-existing heart condition, not the sound of the car horn. Customers would understand that such matters properly relate to the province of life, disability or comprehensive health insurance.
The prospective customer would not, I think, have any difficulty in distinguishing that case from the situation of an unlisted disease or physical infirmity that is produced by the event or events said to be the accidental means, e.g. a heart condition that arises as one of the elements of injuries suffered in a car crash. In such a case the insurer would be expected to pay.
An area of difficulty, however, occurs where the event (e.g. sexual intercourse) is itself both an ordinary act of everyday life and the natural method by which the disease replicates, i.e. contracted in the customary manner in the ordinary course of events without any mishap or trauma other than contracting the disease itself. The onus, of course, is on the plaintiff to show that the loss is covered by the policy. In Martin, the Court added an important nuance [para. 30]:
The onus is on the plaintiff to establish a prima facie case that the death [or bodily injury] was accidental, at the risk of non suit. The plaintiff must therefore adduce evidence that permits the trier of fact to infer, on a balance of probabilities, that the insured’s death [or bodily injury] was accidental, within the ordinary meaning of that word. The tactical burden then shifts to the insurer to displace these inferences. The burden of proof never shifts, but remains squarely with the plaintiff.
Adapting this proposition to the present case, I would say that once the claimant leads evidence sufficient to establish a prima facie case that the bodily injury was caused by an “unlooked for mishap or an untoward event which is not expected or designed” (Stats, at p. 1164), the “tactical burden” then shifts to the insurance company to displace the “prima facie case” by some evidence that the bodily injury is not an accident but its “antithesis”, namely, the result of a “disease in the ordinary course of events”. The burden of proof however, “remains squarely with the plaintiff”. If he or she fails to establish on a balance of probabilities that the bodily injury resulted from “an accident”, the claim will fail.
The insurance insider will know, based on the vast repertoire of cases decided under accident policies, that all of these situations have given rise at the margins to fierce arguments. The cases necessarily involve value judgments related to the reasonable expectations of the parties. It seems to me the general nature of the risk covered by the policy outlined above is consistent with the understanding that a customer would think makes sense if he or she stopped to reflect on its terms. The disease element is carved out of the universe of unexpected mishaps. The ordinary use of language has placed well-understood (if not well defined) limitations on the scope of “accident” in the jurisprudence. The jurisprudence has informed the practice in the insurance industry and no compelling reason has been advanced to disturb it.
G. The Words “External” and “Violent” Add Little to the Definition of “Accident”
The insurer argues that the words “external” and “violent” in the definition of the risk should be given independent meaning. On its view, only a sub-category of accidents are covered. However, I agree with Newbury J.A. that the words “external” and “violent” have long since been subsumed into the concept of accident. In N. W. Commercial Travellers’ Association v London Guarantee and Accident Co. (1895), 10 Man. R. 537 (Q.B.), the insured died of cold while awaiting assistance in a wagon that had broken down. Bain J. reasoned that [pp. 543-44]:
Perhaps the meaning that is most obvious when we speak of something having been caused or effected by violent means is that a greater or less degree of physical force has been used to bring about the result. But the word “violent” is also used to express that which is unusual or unnatural or extreme, and I think it is reasonable to suppose that it was used in the policy in a somewhat wide and general sense.
On this point the court cited with approval Bacon v U.S. Mutual Accident Assn., 44 Hun. 599 (N.Y. Sup. Ct. 1887), where the court said, at p. 604: “We say a man dies a violent death, without necessarily implying anything more than that he dies, not in the ordinary course of nature and disease.”
Similarly, the editors of Couch on Insurance (3rd ed. 1995), write that [§139:18]:
“Violent” refers to some act not occurring in the ordinary run of things and may be fulfilled by any force whatsoever, however slight. It has been said that unnatural death, the result of an accident of any kind, imports an external and violent agency as the cause, within the meaning of an insurance policy limiting recovery to death caused through “external, violent, and accidental means.”
(Cited by Newbury J.A., at paras. 26-28.)
Historically, the addition of the word “external” was likely inserted to help distinguish accidents (that will support a claim) from losses caused by naturally occurring “internal” conditions such as diseases (which will not).
When more than one factor has contributed to the bodily injury the courts have also relieved claimants from a strict showing that the loss was “solely” the result of an accident “independently of all other causes”. The wording of the typical accident insurance policy has been softened in light of the practical difficulty in most cases of isolating a “unique” cause. In the insurance context the Court has held that:
The question of whether insurance applies to a loss should not depend on metaphysical debates as to which of various causes contributing to the accident was proximate. Apart from the apparent injustice of making indemnity dependent on such fine and contestable reasoning, such a test is calculated to produce disputed claims and litigation.
(Derksen v 539938 Ontario Ltd., 2001 SCC 72,  3 S.C.R. 398, at para. 36, citing C.C.R. Fishing Ltd. v British Reserve Insurance Co.,  1 S.C.R. 814, at p. 823)
Efforts by insurers to split hairs on causation issues have traditionally been rebuffed as in Re Etherington and The Lancashire and Yorkshire Accident Insurance Co.,  1 K.B. 591. In that case the insured fell off his horse onto a wet surface, and subsequently developed pneumonia, from which he died. He was covered by a policy that provided recovery in the event of death caused by “bodily injury caused by violent, accidental, external, and visible means” (p. 592). The insurer contended that the proximate cause of death was pneumonia, not the fall from the horse. The English Court of Appeal had no difficulty in holding that “The fatal pneumonia was a ‘sequela’ of the accident against the consequences of which it was the object of the policy to insure” (p. 603). In other words, the pneumonia could not be said to be an independent cause of death. The courts do not favour the self-serving isolation of a particular element in a chain of events that should be considered in its entirety. Such law office metaphysics would make nonsense of the reasonable expectation of the parties at the time the policy was entered into.
H. Consideration of Disease-Based Claims in the Case Law
As noted in Walkem, the jurisprudence assigns a generous meaning to “accident” (p. 316). However, the word is also “an ordinary word to be interpreted in the ordinary language of the people” (Stats, at p. 1164). In ordinary speech “accident” does not include ailments “proceeding from natural causes” (Welford, at p. 268). In Carroll v CUNA Mutual Insurance Society, 894 P.2d 746 (Colo. 1995), for example, death resulted from a ruptured aneurysm suffered during the exertions of sexual intercourse. An accident claim was denied because the insured’s death was essentially the result of her pre-existing aneurysm and hypertension, not the sex. The issue was one of causation. Sex did not cause the aneurysm. It simply provided the occasion on which the aneurysm happened to occur.
In the present case the evidence is that genital herpes is a sexually transmitted virus that spreads by sexual intercourse. Sex is its normal method of transmission. As such, unlike for example an internally developing condition leading to an aneurysm, its transmission requires an outsider’s participation. But the same could be said of infectious diseases generally. Viruses and bacteria pass, directly or indirectly, from person to person, and occasionally across species. In the “ordinary language of the people” (Stats, at p. 1164), an individual would not say on coming down with influenza that “I had an accident”. We come down with the flu “in the ordinary course of events”.
The “bodily malfunction” cases illustrate the proposition that mishaps not in the ordinary course of events may be classified as accidents. For example, in the American case of Peoples Life Ins. Co. v Menard, 117 N.E. 2d 376 (Ind. App. 1954), the insured choked on partially digested food that he had regurgitated. The court concluded at p. 379 that the death resulted from accidental means, because
[i]f the regurgitation had been a perfectly normal process as it normally occurs the decedent would have expelled the contents of his stomach in a perfectly normal manner and no unforeseen, unexpected or unusual contingency would have occurred, and decedent would not have suffered the asphyxiation which was the immediate cause of his death.
(See also American Accident Co. of Louisville v Reigart, 23 S.W. 191 (Ky. 1893), at p. 192.)
Similar reasoning was used to justify recovery in Koch v Empire Life Insurance Co. (1981), 29 A.R. 49 (Q.B.), where the insured died as a result of choking on partially digested food while asleep. The agreed statement of facts provided that “[t]here was no pathological reason for the vomiting” (para. 4). Disease was not an issue. The insured was in good health. However his body did not on that occasion function as would be expected in a perfectly “normal manner”. See also Reigart, at p. 192; Jones v Aetna Life Insurance Co., 439 S.W. 2d 721 (Tex. Civ App. 1969). These cases do not involve Welford’s “antithesis” between accident and “disease in the ordinary course of events”. An isolated instance of regurgitation, however unpleasant, is not a disease.
In the course of argument reference was made to a series of “bodily contortion” cases, and in particular Claxton v Travellers Insurance Co. of Hartford (1917), 36 D.L.R. 481 (Que. Ct. Rev), where the insured suffered a hernia while swinging a golf club, Voison v Royal Insurance Co. of Canada (1988), 66 O.R. (2d) 45 (C.A.), where the insured suffered an occlusion of the anterior spinal artery when he assumed an awkward position extending his neck while engaged in remodelling his house and Guillet v American Home Assurance Co. (2004), 72 O.R. (3d) 641 (C.A.), where the claimant twisted his neck playing basketball and suffered a stroke. These cases turn on rather peculiar fact situations and like the bodily malfunction cases, the analytical problem raised has nothing to do with the spread of a disease in the ordinary and natural course of events. A debate about the merits of these cases would not assist in the resolution of this case and I say no more about them.
I. The Natural Transmission of Disease
The causal chain that led to Mr. Gibbens’ bodily injury was sex that transmitted the herpes that led to transverse myelitis. Transverse myelitis was an unexpected consequence of the disease, but it occurs (though rarely) as a normal incident or consequence of that disease. I do not place emphasis on the fact the sex was unprotected. The insurance company does not contend that risky behaviour disqualifies him from coverage. It recognizes that one of the reasons people buy insurance is to provide protection against the consequences of risky behaviour. Rather its point is that the transmission followed the normal method by which sexually transmitted diseases replicate and thus the bodily injury “proceed[ed] from natural causes” (Welford, at p. 268).
To conclude that Mr. Gibbens’ acquisition of herpes was “an accident” despite the absence of any mishap or trauma other than the acquisition of a sexually transmitted disease in the ordinary way would simply serve to add sexually transmitted diseases to the list of Critical Diseases in the group policy contrary to the intent of the policy.
Mr. Gibbens relies on Kolbuc v ACE INA Insurance, 2007 ONCA 364, 85 O.R. (3d) 652, where an insured, a plasterer, was bitten by a mosquito carrying the West Nile virus and was rendered a paraplegic. He recovered compensation under an accident policy. I make no comment about the merits of that decision, which is not before us, but as Saunders J.A. noted in the court below, “the world is populated with pathogens” (para. 37). Various forms of bacteria and viruses constantly make their way into our bodies, sometimes with little effect, and we in turn spread them to others. Bacterial infections include, for example, tuberculosis, anthrax and typhoid. The Bubonic Plague was transmitted by fleas. Malaria is transmitted by mosquitoes. In ordinary speech, we would not say that the Bubonic Plague was the result of a pandemic of accidents, or that the inhabitants of warm climates are particularly “accident prone” to contracting Malaria. It cannot be correct that passengers sitting in an airliner who catch the SARS virus through the externality of the plane’s air circulation system, or riders on a bus who catch “swine flu” from an infected fellow passenger, or people who contract any number of infectious diseases because of a failure to wash hands in disinfectant, or to smack a circling mosquito, have valid claims under an accident policy.
In my view, with respect, such a conclusion would stretch the boundaries of an accident policy beyond the snapping point and convert it into a comprehensive insurance policy for infectious diseases contrary to the expressed intent of the parties and their reasonable expectations.
In Mr. Gibbens’ case, to borrow the words of Lord Robertson in Brintons, the so-called “accident” was simply the “inception of the disease” in the ordinary course of events.
I would therefore allow the appeal and dismiss the action. The appellant is entitled to the costs of this appeal.
 Milton describes “a gulf profound, as that Serbonian bog betwixt Damiata and mount Casius old, where armies whole have sunk”: John Milton, Paradise Lost (1767), Book II, at p. 44.
Martin v American International Assurance Life Co., 2003 SCC 16,  1 S.C.R. 158; Fegan v State Mutual Life Assurance Co. of America, 945 F. Supp. 396 (1996); Smith v British Pacific Life Insurance Co.,  S.C.R. 434; Milashenko v Co-operative Fire & Casualty Co. (1970), 11 D.L.R. (3d) 128; Columbia Cellulose Co. v Continental Casualty Co.,  I.L.R. ¶ 1-119; De Souza v Home and Overseas Insurance Co.,  L.R.L.R. 453; Landress v Phoenix Mutual Life Insurance Co., 291 U.S. 491 (1934); Consolidated-Bathurst Export Ltd. v Mutual Boiler and Machinery Insurance Co.,  1 S.C.R. 888; Mutual of Omaha Insurance Co. v Stats,  2 S.C.R. 1153; National Bank of Greece (Canada) v Katsikonouris,  2 S.C.R. 1029; Canadian Indemnity Co. v Walkem Machinery & Equipment Ltd.,  1 S.C.R. 309; Fenton v J. Thorley & Co.,  A.C. 443; Wang v Metropolitan Life Insurance Co. (2004), 242 D.L.R. (4th) 598, leave to appeal refused,  1 S.C.R. xvii; Brissette Estate v Westbury Life Insurance Co.,  3 S.C.R. 87; Reid Crowther & Partners Ltd. v Simcoe & Erie General Insurance Co.,  1 S.C.R. 252; Algoma Steel Corp. v Allendale Mutual Insurance Co. (1990), 72 O.R. (2d) 782, leave to appeal refused,  1 S.C.R. v; Sinclair v Maritime Passengers’ Assurance Co. (1861), 3 E.L. & E.L. 478, 121 E.R. 521; Wyman v Dominion of Canada General Insurance Co.,  2 D.L.R. 268; Brintons, Ltd. v Turvey,  A.C. 230; Toronto Professional Firefighters’ Association v Toronto (City) (2007), 223 O.A.C. 146; Glenlight Shipping Ltd. v Excess Insurance Co., 1983 S.L.T. 241; Cornish v Accident Insurance Co. (1889), 23 Q.B.D. 453; Brown v Continental Casualty Co., 108 So. 464 (1926); Bertalan Estate v American Home Assurance Co. (1999), 68 B.C.L.R. (3d) 118; N. W. Commercial Travellers’ Association v London Guarantee and Accident Co. (1895), 10 Man. R. 537; Bacon v U.S. Mutual Accident Association, 44 Hun. 599 (1887); Derksen v 539938 Ontario Ltd., 2001 SCC 72,  3 S.C.R. 398; C.C.R. Fishing Ltd. v British Reserve Insurance Co.,  1 S.C.R. 814; Re Etherington and The Lancashire and Yorkshire Accident Insurance Co.,  1 K.B. 591; Carroll v CUNA Mutual Insurance Society, 894 P.2d 746 (1995); Peoples Life Ins. Co. v Menard, 117 N.E. 2d 376 (1954); American Accident Co. of Louisville v Reigart, 23 S.W. 191 (1893); Koch v Empire Life Insurance Co. (1981), 29 A.R. 49; Jones v Aetna Life Insurance Co., 439 S.W. 2d 721 (1969); Claxton v Travellers Insurance Co. of Hartford (1917), 36 D.L.R. 481; Voison v Royal Insurance Co. of Canada (1988), 66 O.R. (2d) 45; Guillet v American Home Assurance Co. (2004), 72 O.R. (3d) 641; Kolbuc v ACE INA Insurance, 2007 ONCA 364, 85 O.R. (3d) 652.
Authors and other references
Butterworths Medical Dictionary, 2nd ed. Toronto: Butterworths, 1978, “disease”.
Couch on Insurance, 3rd ed. by Lee R. Russ and Thomas F. Segalla. Deerfield, Ill.: Clark Boardman Callaghan, 1995.
Hall, Geoff R. Canadian Contractual Interpretation Law. Markham, Ont.: LexisNexis, 2007.
Ivamy, E. R. Hardy. General Principles of Insurance Law, 6th ed. London: Butterworths, 1993.
MacGillivray on Insurance Law, 10th ed. by Nicholas Legh-Jones. London: Sweet & Maxwell, 2003.
Milton, John. Paradise Lost, Book II. Edinburgh: A. Donaldson, 1767.
Norwood, David, and John P. Weir. Norwood on Life Insurance Law in Canada, 3rd ed. Toronto: Carswell, 2002.
Scales, Adam F. “Man, God and the Serbonian Bog: The Evolution of Accidental Death Insurance” (2000-2001), 86 Iowa L. Rev 173.
Welford, A.W. Baker. The Law Relating to Accident Insurance, 2nd ed. London: Butterworths, 1932.
Bruce Laughton, Q.C., and Leah Terai (instructed by M/s Laughton & Company, Vancouver), for the appellant.
Guy J. Collette, Michael Sobkin and A. C. Richard Parsons (instructed by M/s Guy J. Collette Law Corporation, Vancouver), for the respondent.
Patricia D. S. Jackson and David Outerbridge (instructed by M/s Torys, Toronto), for the intervener.
all rights reserved