COURT OF APPEAL FOR ONTARIO

CITATION: Shantry v. Warbeck, 2015 ONCA 395

DATE: 20150604

DOCKET: C58247

Cronk, Pepall and Benotto JJ.A.

BETWEEN

Melissa Anne Shantry and Taylor Anne Miller, a minor by her Litigation Guardian, Melissa Anne Shantry

Plaintiffs (Appellants)

and

Paul Warbeck

Defendant (Respondent)

AND BETWEEN

David Blake, Donna Miller, Harold Miller, Denise Post-Blake, Burkley Blake and David Blake Jr.

Plaintiffs (Appellants)

and

Paul Warbeck

Defendant (Respondent)

Paul Pape and Joanna Nairn, for the appellants

Michael Royce and Dena Varah, for the respondent

Heard: February 18, 2015

On appeal from the judgment of Justice Terrence A. Platana of the Superior Court of Justice, dated December 20, 2013, with reasons reported at 2013 ONSC 7805.

Benotto J.A.:

[1] A 22-year-old man arrived at a hospital emergency room in the early evening, presenting with severe back pain. Intravenous morphine throughout the night had little effect. In the morning, he came under the care of the respondent, who continued the morphine. There was still no relief. At around 5 p.m. the respondent switched the medication to Dilaudid, a stronger opioid. By 3 a.m. the following morning, the young man finally said he had no pain. At 4:30 a.m., hospital staff found him not breathing. He was pronounced dead at 5:09 a.m., less than 35 hours after he arrived at the hospital. The coroner’s investigation concluded that the cause of death was mixed-drug overdose.

[2] The man’s family brought an action in negligence alleging that the respondent ordered an excessive dose of Dilaudid and this caused the man’s death.

[3] The trial judge determined that the dose was appropriate in the circumstances and that the respondent’s conduct did not fall below the standard of care. The trial judge also rejected the coroner’s finding as to cause of death.

[4] The family members appeal. They submit that the trial judge’s approach to liability was flawed because he addressed the standard of care before causation, and then wrongly rejected the coroner’s conclusion as to the cause of death. This error, they allege, led him into further error with respect to the standard of care.

[5] I have concluded that the trial judge did not err in his approach to liability. Nor did he err in his findings with respect to the standard of care and causation. Accordingly, I would dismiss the appeal.

FACTS

[6] On Wednesday, July 11, 2007, Shane Miller – a 22-year-old man living with his common-law partner Melissa Anne Shantry and their young daughter – slipped on a pinecone in a parking lot and fell onto his back. He lay on the ground for a few minutes, but then got up and went shopping with his family. The following Saturday, July 14, Mr. Miller did not wake up until around 5 p.m. Ms. Shantry testified that, “he seemed out of it. He, he was mumbling about some dreams that he had had.” Mr. Miller was lying on the couch but would “jump up from the couch” with “little convulsions” and “jerking movements” that were not normal. He was complaining about his back hurting and “going on about ... his dreams”. Ms. Shantry summoned Mr. Miller’s father to take him to the hospital. According to his father, Mr. Miller “seemed to be in a lot of pain”.

[7] Mr. Miller arrived at the Thunder Bay Regional Health Sciences Centre at 6:19 p.m. on Saturday, July 14, complaining of intense pain in his upper back. He described it as both aching and sharp between his shoulder blades and sometimes radiating to his lower back. He had a fever and a rapid pulse.

[8] Mr. Miller was examined in the emergency department and administered a single dose of 5 milligrams (“mg”) of morphine intravenously at 6:50 p.m. It did not alleviate his pain. Mr. Miller received three more doses of morphine between 7:40 p.m. and 9 p.m. The total for the evening was 19-20 mg. His pain continued unabated.

[9] The emergency department personnel could not determine the cause of Mr. Miller’s pain. Shortly after midnight, he was admitted to the medical ward for investigation. The pain and fever did not subside. He received additional doses of morphine throughout the night and was also started on an antibiotic. By morning, he was still in pain. At 8:30 a.m. on the morning of Sunday, July 15, a nurse assessed him and found him to be experiencing throbbing and radiating back pain, which he rated as 10 on a scale of 10, as well as a fever and headache. He was itching, vomiting, moaning, crying and restless. He was screaming in pain.

[10] Through the morning and afternoon of Sunday, July 15, Mr. Miller continued to receive doses of morphine. At trial, the attending day nurse was asked to describe Mr. Miller’s pain levels:

A. He was, he was in horrible, terrible pain.

Q. Do you have a specific recollection of this patient?

A. Yeah. Like I work on [an] oncology patient clinic and I do see people with lots of pain and this young gentleman, he sticks with me because of his pain level and the young age.

Q. Okay.

A. He was in horrible pain.

[11] Mr. Miller’s family members also testified that he was in pain that Sunday morning and afternoon. At 5:09 p.m., Mr. Miller said he was still in pain. He described throbbing in his back and rated it as an 8 on a scale of 10. The day nurse did not consider this to be an improvement in pain level; he was still moaning, crying and suffering.

Involvement of the Respondent

[12] The respondent is a physician who was working as a hospitalist at the Health Sciences Centre. In his trial testimony, he described a hospitalist’s role as providing general medical care, guiding investigations and requesting consultations for patients without a family physician able to care for them in hospital.

[13] The respondent first saw Mr. Miller between 9:00 a.m. and 9:30 a.m. on the morning of Sunday, July 15. He reviewed notes from Mr. Miller’s attending physicians and nurses, issued orders for laboratory tests and noted that Mr. Miller had a pustular rash, a significant fever and severe back pain. He believed that Mr. Miller had a serious viral infection, possibly shingles, in the discs of his backbone.

[14] At around 5 p.m. that day, the day nurse told the respondent that Mr. Miller had received 20 mg of morphine over a four-hour period. Mr. Miller was clear-headed, but still in pain. The respondent changed his medication to 5-10 mg of Dilaudid (hydromorphone) every four hours. Dilaudid is an opioid that, according to the appellants’ expert witness Dr. Alan Drummond, is 3-8 times more potent than morphine.

[15] At 5:30 p.m. Mr. Miller received his first 5 mg dose of Dilaudid. The day nurse recorded in her “Shift Summary” that, after the new drug was administered, Mr. Miller was in less pain. The respondent left the hospital at 9 p.m.

Mr. Miller’s Death

[16] At 9:30 p.m., Mr. Miller received another 5 mg dose of Dilaudid. The following morning, at 2:00 a.m., the attending night nurse found him in his room awake, standing and urinating on the side of his bed and the floor. He was “very angry” that he had not been administered more Dilaudid while he was sleeping so that he would not have to wake up in pain. He was yelling and screaming that he was in a lot of pain. The night nurse, along with two other nurses, changed the sheets and assisted Mr. Miller back into bed. The night nurse administered another 5 mg dose of Dilaudid at 2:30 a.m. Within 15 minutes, Mr. Miller was calm.

[17] At around 3:00 a.m., Mr. Miller rang for the night nurse. He apologized for urinating on the floor and asked her to turn off the television and the lights. For the first time since his arrival at the hospital, he said he had no pain.

[18] Mr. Miller was not seen again by the night nurse until 4:30 a.m., when he was found not breathing. A Code Blue was called, but Mr. Miller could not be resuscitated. He was pronounced dead at 5:09 am.

Summary of the Opioids

[19] Mr. Miller received the following doses of morphine and Dilaudid:

• July 14: • 6:50 p.m. : 5 mg morphine • 7:40 p.m. : 2 or 3 mg morphine • 7:50 p.m. : 2 mg morphine • 9:00 p.m. : 10 mg morphine

• July 15: • 1:00 a.m. : 10 mg morphine • 3:00 a.m. : 5 mg morphine • 8:20 a.m. : 5 mg morphine • 11:30 a.m. : 5 mg morphine • 12:25 p.m. : 5 mg morphine • 3:40 p.m. : 10 mg morphine • 5:30 p.m. : 5 mg Dilaudid • 9:30 p.m. : 5 mg Dilaudid

• July 16: • 2:30 a.m. : 5 mg Dilaudid

The Coroner’s Report

[20] The tragic circumstances led to an extensive coroner’s investigation. In the Medical Certificate of Death, the investigating coroner cited “mixed drug overdose” as the immediate cause of death, more particularly “unexpected toxicity from potent opiates/other medications.” (The words “opiate” and “opioid” are synonymous.)

[21] A forensic pathologist provided a second opinion as to the cause of death. She concluded that the source of Mr. Miller’s pain and fever was likely a viral infection, but that the cause of death was “acute mixed hydromorphone and morphine intoxication”. These conclusions were again confirmed by another pathologist. The continuing investigation was overseen by the Regional Supervising Coroner. In all, at least eight diagnostic or analytical reports were written regarding the cause of Mr. Miller’s death.

THE TRIAL

[22] At trial, the appellants argued that the respondent failed to meet the standard of care in two respects: by prescribing an excessive dosage of Dilaudid and by failing to attend personally to Mr. Miller more than once on Sunday, July 15, 2007, relying instead on the observations of nurses. Only the first issue, the dosage of Dilaudid, is raised in this appeal.

[23] The correct dosage of an opioid is informed by the patient’s status as either “opioid naïve” or “opioid tolerant.” Patients who are opioid tolerant can tolerate higher dosages because their bodies have become used to the drugs. Patients who are opioid naïve generally cannot tolerate higher dosages. The experts agreed that opioid tolerance generally develops after at least a week of regular doses.

[24] Whether Mr. Miller was opioid tolerant or opioid naïve was the subject of extensive expert evidence. The appellants led expert evidence that he was opioid naïve and the dosage of Dilaudid was consequently too high. The respondent led expert evidence that, given the ineffectiveness of morphine on Mr. Miller’s pain, the dosage of Dilaudid that the respondent ordered was appropriate. The respondent’s experts testified that although Mr. Miller did not meet the strict definition of opioid tolerant – in that it was unlikely that he could have become tolerant in the span of a day or so – he nevertheless exhibited resistance to morphine such that the Dilaudid dosage was justified.

[25] The experts also disagreed as to the cause of death. While the appellants’ experts agreed with the coroner’s conclusion, the respondent’s did not. The latter testified that, had Mr. Miller’s death been caused by an overdose of opioids, he would have exhibited symptoms of such an overdose before he was last observed alive by the night nurse at 3:00 a.m. on the night of his death.

[26] The trial judge accepted the evidence of the respondent’s experts on both the dosage issue and the cause of death. With respect to the dosage of Dilaudid, the trial judge wrote:

All medical witnesses agreed that the proper dose is one which relieves pain. That being so, recognizing that the initial, and subsequent, 5 mg dose of Dilaudid did not relieve Mr. Miller’s pain, and there were no adverse side effects observed, it is difficult for me to find that it was other than a proper dose.

[27] With respect to causation, the trial judge rejected both the evidence of the appellants’ experts and the conclusions stated in the coroner’s report that a mixed drug overdose caused Mr. Miller’s death. Rather, he stated:

I consider that there is no reason for me not to accept that if toxicity caused the event, then it should certainly have occurred before 03:00, based on the clinical history of Mr. Miller from 17:30 to 03:00. I therefore accept [the respondent’s expert’s] evidence that it is “highly improbable” that the spectrum [of symptoms of opioid toxicity] could have occurred between 03:00 and 4:30.

THE ISSUES ON APPEAL

[28] The appellants submit that the trial judge’s errors began when he addressed the standard of care before causation. They submit that he erred in law and fact when he rejected the appellants’ evidence as to the cause of death. This error as to causation, they allege, led him to err in determining the standard of care with respect to the appropriate dosage of Dilaudid.

ANALYSIS

[29] I begin by addressing the appellants’ submission that the trial judge erred by deciding the standard of care before causation. It is submitted that a finding that the cause of death was mixed drug toxicity – as determined by the coroner - would have impacted the trial judge’s finding on standard of care. The trial judge, they argue, treated causation as a “throw away” and then made mistakes that infected his determination of negligence.

[30] The appellants rely on Meringolo v. Oshawa General Hospital (1991), 46 O.A.C. 260 (C.A.), leave to appeal refused, 50 O.A.C. 159 (S.C.C.), to support the position that causation must be determined before duty of care. In that decision, Osborne J.A. wrote, at para. 56:

The [cause of the brain damage] is, in my view, preliminary to the traditional negligence causation question, that is, whether it has been established that a tortfeasor’s conduct (breach of duty) caused a plaintiff’s injury. In this case, before the breach of duty issue can be appropriately addressed, what happened in the operating room ... must be determined in order to provide an answer to the question, what caused the appellant’s brain damage, or, to put it more broadly, how did the appellant sustain brain damage. It is only after that issue is examined that the issue of the respondents’ alleged negligence should be considered. [Emphasis added.]

[31] In a later case, Grass v. Women's College Hospital (2001), 2001 CanLII 8526 (ON CA), 144 O.A.C. 298 (C.A.), leave to appeal refused, [2001] S.C.C.A. No. 372, this court applied Meringolo, with Catzman J.A. writing, at para. 12:

[T]he trial judge in the present case disposed of the action by finding that the appellants had failed to establish negligence on the part of Dr. Weisberg and thereby finding it unnecessary to assess the conflicting theories relating to causation. In doing so, he fell into the error articulated in Meringolo. The resolution of the question of causation might have led to different findings of fact with respect to what transpired in the labour room and to a different conclusion with respect to negligence. As in Meringolo, the appellants were entitled to the benefit of the trial judge's findings with respect to causation or to his determination, on the basis of a consideration of all of the evidence that such findings could not be made, before he came to address the issue of negligence. [Emphasis added.]

[32] In both Meringolo and Grass,the trial judge had not resolved the factual disputes surrounding causation, and this court held that findings of fact about causation, had they been made, could have influenced the determination of whether the standard of care had been breached.

[33] The concerns expressed in Meringolo and Grassare fact-specific and, in any event, are not at issue in this case. In each of those cases, the plaintiff’s theory of causation was a crucial component to a determination of the standard of care. In each of those cases, the trial judge, having first considered the standard of care, failed to address or make findings on causation. This is not such a case. The trial judge here considered causation and rejected the appellants’ theory in relation to it.

[34] That the trial judge addressed causation after the standard of care is not an error. The Supreme Court has established that there are requirements that must be met to establish negligence. The requirements are sequential and separate. The plaintiff must demonstrate that:

1. the defendant owed a duty of care;
2. the defendant’s behaviour breached the standard of care;
3. the plaintiff sustained damage; and
4. the damage was caused, in fact and in law, by the defendant’s breach.

(See: Cleveland v. Whelan, 2011 ONCA 244, 277 O.A.C. 178, at paras. 32-33, citing Mustapha v. Culligan of Canada Ltd., 2008 SCC 27, [2008] 2 S.C.R. 114, at para. 3).

[35] The trial judge’s analysis targeted each of these requirements in sequence. I therefore do not agree that he erred in his approach to the standard of care and causation issues. He considered the standard of care first, then causation. I will do the same.

Standard of Care

[36] Four physicians testified as expert witnesses. The experts agreed that the switch from morphine to Dilaudid at 5 p.m. on July 15 was appropriate. They agreed that Dilaudid is a more powerful opioid than morphine and that the conversion factor between the two is generally accepted to be 1:5. They disagreed about the appropriateness of the dosage of Dilaudid prescribed to Mr. Miller. The appellants’ witnesses said that Mr. Miller was opioid naïve and therefore should have been started at a lower dose. The respondent’s witnesses said that the dose was appropriate because Mr. Miller had not responded to increasing doses of morphine and this had to be taken into account in determining the starting dose of Dilaudid.

[37] Dr. Brian Goldman and Dr. Alan Drummond testified for the appellants. Dr. David Juurlink and Dr. Gerald Rowland testified for the respondent.

Appellants’ Expert Evidence

[38] Dr. Goldman was qualified as “an expert to give an opinion with respect to the dosages of morphine and Dilaudid to be used”. He has practised in the area of chronic pain management for about 10 years and has prescribed morphine and Dilaudid intravenously “many thousands of times”.

[39] Dr. Goldman explained the practice of titration, by which physicians slowly increase opioid dosage while monitoring the patient in order to find the optimum dose for that particular patient. He testified that, in switching a patient from morphine to Dilaudid, the appropriate clinical practice is to start at a low dose and titrate upwards. Dr. Goldman opined that it is not safe to assume that, because a patient has not had an adverse reaction to morphine, they will not have one to Dilaudid – they are different opioids and work in different ways on the human body. In establishing the starting dose of Dilaudid, the physician should consider only the last dose of morphine, not the amount administered over the last four hours. In Dr. Goldman’s opinion, it was not possible for Mr. Miller to have become tolerant to the analgesic effects of opioids during the less than 35 hours that he was hospitalized.

[40] Dr. Goldman opined that the 5 mg dose of Dilaudid given to Mr. Miller was “an egregiously excessive dose of … Dilaudid compared to usual practice.” As Mr. Miller’s last dose of morphine was 10 mg, the respondent should have started him on 1-2 mg of Dilaudid every three hours and titrated upward from there. Dr. Goldman’s opinion was that the respondent failed to meet the standard of care by prescribing the 5-10 mg dose of Dilaudid every 4 hours. According to Dr. Goldman, the respondent did this because he mischaracterized Mr. Miller as being opioid tolerant when he was in fact opioid naïve.

[41] Dr. Drummond was qualified as an expert in general family medicine and the dosing of Dilaudid. Dr. Drummond, a family physician, has a certification in emergency medicine, experience as an anaesthetist and 20 years’ service as a coroner. He has prescribed intravenous morphine or Dilaudid on thousands of occasions.

[42] Dr. Drummond’s evidence was that, in switching a patient from morphine to Dilaudid, the physician should start with a lower dose of Dilaudid and titrate up (or down) to achieve pain relief. He shared Dr. Goldman’s view that it was not appropriate, in determining the starting dose of Dilaudid, to consider the amount of morphine given in the previous four hours and that a patient’s lack of response to morphine cannot justify a higher starting dose of Diluadid – they are “two distinctly different medications”.

[43] Dr. Drummond’s opinion was that Mr. Miller was opioid naïve and the dosage of 5-10 mg of Dilaudid was “shocking” and not in accordance with the appropriate standard of care, which would have been to give Mr. Miller a 1-2 mg starting dose of Dilaudid.

Respondent’s Expert Evidence

[44] Dr. Juurlink is the Director of Clinical Pharmacology and Toxicology at the University of Toronto’s Faculty of Medicine. He is board certified in both medical toxicology and internal medicine. He was qualified as an expert in the administration and the effect of opioids, in particular morphine and Dilaudid.

[45] Dr. Juurlink disagreed with the appellants’ experts on the issue of opioid tolerance. Although he agreed that it is unlikely a person would become opioid tolerant in the span of a day or two, he stated that a person who has received multiple doses of morphine cannot be considered opioid naïve in the true sense. He said:

[W]here someone has received a drug for a day or so, you couldn’t rightly call them naïve because you have some sense of how they have responded to the drug in positive or negative ways.

[46] Dr. Juurlink explained that the term “opioid naïve” can be used in different ways. On the one hand, a patient who has not previously received the drug and therefore might experience a more dramatic effect would be considered naïve. On the other hand, the description “naïve” is not entirely accurate when a patient has been receiving opioids, albeit for a shorter time than that required to become opioid tolerant. With respect to Mr. Miller, Dr. Juurlink stated that it was not correct to say he was opioid naïve because he had received morphine and there had been no overt signs of toxicity, and minimal benefit.

[47] Dr. Juurlink’s testified that the morphine given to a patient before the switch to Dilaudid must inform the proper starting dosage of Dilaudid. He opined that, in this case, 5 mg of Dilaudid was a “perfectly appropriate starting dose” because Mr. Miller had received about 20 mg of morphine in the five to six hours prior to the switch, which had not afforded him much pain relief. He therefore disagreed with the appellants’ experts that the starting dose of Dilaudid should be calculated based on the last individual dose of morphine. He did agree that administering 10 mg would have been excessive. However, it would have been “inappropriate” and would have caused “undue suffering” to have switched Mr. Miller to an “inadequate” dose of 1-2 mg Dilaudid. In his view, the change from morphine to Dilaudid was itself an act of titration.

[48] Dr. Rowland was qualified as an expert “with respect to the standard of care of a general practitioner”. He has been a general practitioner in Tillsonburg, Ontario since 1979, and has been on the active staff of the Tillsonburg District Memorial Hospital since 1980. He has treated patients with morphine and Dilaudid on several occasions, most recently three weeks before he testified. He has held a number of roles with the College of Physicians and Surgeons of Ontario pertaining to physician oversight and discipline, including having served as President of the College.

[49] Dr. Rowland testified that the respondent met the standard of care. He opined that it would not be appropriate to determine the correct dosage of Dilaudid without regard to the quantities of morphine that had been administered without effect. That would be akin to treating the patient as though he had just come in “off the street.” The physician should give enough of the medication to relieve the patient’s pain.

[50] Dr. Rowland’s opinion was that the respondent’s order of 5 mg of Dilaudid as the starting dose met the standard of care and was in keeping with the principles of pain management with opioids. He called it “a wise thing to do”.

[51] Dr. Rowland agreed with Dr. Juurlink that the 20 mg of morphine administered in the four hours prior to the first dose of Dilaudid was the appropriate amount upon which to base the conversion to Dilaudid. He stated:

[O]ver the previous four hours he had 20 milligrams of morphine. By all of the clinical experience that I have had, that is the appropriate figure to use when calculating a conversion, not the 10 milligrams given an hour and a half before or whatever. So … the equivalent dose would be 4 milligrams, it wasn’t effective, it wasn’t showing adverse effect[s], it was reasonable to increase the dosage [to 5 mg] at that point in time.

[52] Although Dr. Rowland described Mr. Miller as “opioid tolerant” in his report, he testified that he was using this phrase in a clinical, rather than pharmacological sense:

What I was attempting to do … was to indicate what was happening clinically rather than strictly physiologically or pharmacologically to Mr. Miller.... [W]hat was happening was that he was resistant to the pain relieving effects of the morphine that he was originally administered and that was what I was intending to imply. I was not intending to imply that he had developed opioid-tolerance for whatever reason previous to his arrival at the hospital.

[53] Dr. Rowland said that Mr. Miller demonstrated resistance to morphine. He was not receiving effective pain relief and he was not showing any of the adverse side effects of opioid over-dosage, such as sedation or the loss of sensorium.

Trial Judge’s Conclusions

[54] The trial judge accepted the respondent’s expert evidence. He concluded that the respondent had met the standard of care and that the dosage of Dilaudid was appropriate.

[55] The appellants submit that the trial judge found that Mr. Miller was opioid tolerant and that this was an error. They point to this statement in the reasons: “I accept Dr. Juurlink’s evidence that an inference can be drawn that Mr. Miller was opioid tolerant”.

[56] When the trial judge’s reasons are read in their entirety, it is clear that he was using the term not in the “strict definition” but in accordance with the qualifications explained by Drs. Juurlink and Rowland. He recognized that a person cannot become opioid tolerant in the amount of time that Mr. Miller had received morphine. However, he also accepted the respondent’s expert evidence that the prior doses of morphine should inform the starting dose of Dilaudid. In doing so, the trial judge was not, as the appellant submits, making a finding that Mr. Miller was opioid tolerant. Rather, he was speaking in a different sense, as is evident when the sentence the appellants seize upon is set out in its full context. The trial judge wrote:

I find it reasonable to find that Mr. Miller was not necessarily, in the strict definition, opioid naïve at the time Dilaudid was administered, and therefore, as the evidence of Drs. Goldman and Drummond is based on him being naïve, their evidence should be given less weight than that of Drs. Rowland and Juurlink. I accept Dr. Juurlink’s evidence that an inference can be drawn that Mr. Miller was opioid tolerant. [Emphasis added.]

[57] To further demonstrate this point, the trial judge acknowledged later in his reasons that Mr. Miller was not “technically” opioid tolerant:

Regardless of the strict definition, I can only find that with no reaction of pain relief from the morphine, it is clear that it was having no effect on Mr. Miller. From a non-medical perspective, that suggests to me that his body was, in fact, tolerant to the morphine and he could not be considered opioid naïve when the Dilaudid was administered. [Emphasis added.]

[58] The appellants argue that the above statements by the trial judge show that he erroneously conflated opioid tolerance with analgesic effect. Yet, it clearly was open to the trial judge to accept the evidence of Dr. Juurlink and Dr. Rowland that the dosages of morphine and their lack of effect on Mr. Miller properly informed the starting dosage of Dilaudid. The fact that Mr. Miller had been receiving morphine and that it had not been effective in relieving his pain was information that the respondent properly could use in establishing the starting dose of Dilaudid. The appellants’ and respondent’s witnesses specifically spoke to the need to treat the individual. As the trial judge wrote:

What is apparent from the evidence, and as acknowledged by all of the medical witnesses, is that treatment involving the use of opiates is highly individualized. It is a matter of clinical judgment and physicians can differ in clinical judgment.

[59] All the medical witnesses agreed that the proper opioid dose is one which relieves pain.

[60] The appellants submit that “[t]he threshold standard of care issue in this case was whether [Mr. Miller] was opiate naïve or whether he was opiate tolerant.” This, they submit, determines the appropriate dosage of Dilaudid.

[61] I disagree. By framing the issue in this way, the appellants ignore the evidence, accepted by the trial judge that Mr. Miller was treated as opioid naïve on his admission and that, by 5 p.m. on the following day, there was additional information upon which to base the dosage. That is, whether one characterizes Mr. Miller’s clinical response to morphine as “tolerant” or “resistant”, it is undisputed that morphine was having little to no analgesic effect. The trial judge accepted the evidence of two highly qualified expert witnesses that, based on Mr. Miller’s lack of response to morphine, including no symptoms of respiratory depression, the respondent’s prescription for a starting dose of 5 mg of Dilaudid met the standard of care. This finding was open to the trial judge and, in my view, does not evince any palpable and overriding error.

[62] In summary, the trial judge clearly understood the divergence in the medical evidence and accepted the opinions of Dr. Juurlink and Dr. Rowland in concluding that the respondent’s Dilaudid prescription did not fall outside the reasonable range of judgment calls that would be made by a prudent and diligent hospitalist in similar circumstances. I see no basis on which to interfere with this determination.

Causation

[63] The appellants submit that the trial judge erred in law and made factual errors when he rejected the coroner’s determination as to Mr. Miller’s cause of death. The appellants’ theory of causation was that a mixed overdose of morphine and Dilaudid led to respiratory depression and death.

[64] The appellants submit that, because they put forth extensive evidence to support their theory and the respondent did not lead sufficient evidence to support an alternative, the trial judge was required to accept the appellants’ theory of causation. By rejecting their theory, the appellants argue, the trial judge erred in law. They rely on Ediger v. Johnston, 2013 SCC 18, [2013] 2 S.C.R. 98, to support the following submission in their factum:

If the experts for [the] plaintiff and the defendant have conflicting theories [of causation], the court must consider if the defendant has introduced sufficient evidence to displace the plaintiff’s theory of causation – not just another theory but evidence in support of that theory.

[65] I do not read Ediger to support the appellants’ submission. The Supreme Court of Canada’s decision in that case makes it clear that the drawing of inferences against a defendant is permissive, not mandatory. As Rothstein and Moldaver JJ. wrote for the court, at para. 36:

The trier of fact may, upon weighing the evidence, draw an inference against a defendant who does not introduce sufficient evidence contrary to that which supports the plaintiff’s theory of causation. [Emphasis added.]

[66] The burden nonetheless remains with the plaintiff, as stated at para. 39:

Faced with … conflicting expert testimony on the feasibility of [the plaintiff’s theory of causation] and evidence of other potential causes, it was incumbent upon [the trial judge] to weigh the evidence before her and determine whether [the plaintiff] had proven causation on a balance of the probabilities. [Emphasis added.]

[67] Dr. Juurlink testified about the cause of death and disagreed with the coroner’s conclusion that a mixed opioid overdose was the cause. He described the “spectrum” of “predictable features” that a person exhibits before dying of an opioid overdose:

[T]he person would initially become sleepy. They would then – and might be rousable, but they would then effectively be asleep and then unrousable…. In conjunction with that their breathing would slow. It might still be a deep breath, but it would tend to slow, 10, 8, 6 breaths per minute. And these things happen invariably before someone dies of opioid toxicity. And so the primary reason why I disagree … with the coroner and the others … is that it seems very clear to me from reviewing the record that [Mr. Miller] did not exhibit those features. I cannot envision … any way in which someone dies of an opioid overdose without exhibiting those features.

[68] Dr. Juurlink later stated:

Even 30 minutes after he is given his final dose of opioid, [Mr. Miller] is communicative. He is apologetic. He seems rational. People who are about to die of opioids cannot be roused…. They cannot be awakened. And so the picture [of Mr. Miller] is vastly different from the picture of a patient who is about to succumb to the accumulative effects of opioids administered in the hours beforehand.

[69] Dr. Juurlink rejected the suggestion that Mr. Miller could have progressed through the full spectrum of symptoms of an opioid overdose in the time in which he was unobserved, between 3:00 a.m. and 4:30 a.m., on the day of his death. In his opinion, had Mr. Miller died of an opioid overdose from the Dilaudid administration, some of the toxicity symptoms would have manifested themselves earlier that night. He said:

[N]ot only would he have been expected to die fairly soon, at least some of the signs of toxicity would manifest before that and as far as I can tell they didn’t.

[70] Dr. Juurlink also opined that the post-mortem levels of Dilaudid in Mr. Miller’s blood were too low to suggest an overdose.

[71] Dr. Rowland testified that, in his experience, adverse effects from intravenous opioids are usually seen within 20 minutes after administration. His view was that monitoring a patient 15-30 minutes after administration of Dilaudid for signs of over-sedation, such as loss of alertness or breathing difficulties, was a “completely reasonable approach” as adverse side effects would be detectable in this time period.

[72] This evidence provided an adequate basis for the trial judge to conclude that the appellants had failed to prove causation on a balance of the probabilities. Although this was contrary to the conclusion of the coroner, I note that the coroner’s conclusions were also based on a balance of probabilities and not a higher standard of proof. The trial judge clearly preferred the evidence of the respondent’s experts on causation and rejected the appellants’ theory on that basis. In my view, it was open to him to do so.

[73] In this case, most of the facts underlying the respondent’s attack on the appellants’ theory of causation were not in dispute: (i) Mr. Miller presented with excruciating pain and fever before a single opioid had been administered; (ii) the pain continued unabated until shortly before his death; and (iii) Mr. Miller was never observed with symptoms of respiratory depression after two 5 mg doses of Dilaudid, nor in the 30 minutes after the third, tragically final, dose. In my view, it was open to the trial judge on these undisputed facts, given the respondent’s expert evidence, to reject the appellants’ theory of causation.

[74] The appellants argue that the trial judge made two factual errors that undermine his findings on causation. First, the trial judge wrote that “Mr. Miller came into the hospital after 2 days of acute pain”. The appellants correctly point out that Ms. Shantry’s evidence was that Mr. Miller was in acute pain for less than 2 hours on July 14, 2007 before arriving at the emergency room. The appellants argue that the trial judge’s factual error underlay his inference that Mr. Miller died from a significant pre-existing medical issue.

[75] I would reject this submission. When Mr. Miller arrived at hospital, and throughout his time there, it was clear that he was very ill, possibly with a viral infection. He was in excruciating pain and had a fever. The trial judge’s rejection of the appellants’ theory as to cause of death does not depend upon Mr. Miller being in pain for two days. The respondent’s theory of causation – an unidentified condition, possibly a viral infection – was in no way dependent on a finding that he was in pain for days prior to his arrival at the hospital.

[76] The second factual error alleged was that the trial judge wrote that Mr. Miller’s clinical presentation did not accord with death by overdose, as there were no symptoms of overdose up until 3 a.m. on the night of his death. This statement, it is argued, ignores the evidence that an adverse reaction to Dilaudid could occur 20 minutes to 3.5 hours after a dose. Mr. Miller was last observed alive by the night nurse about 30 minutes after his third dose of Dilaudid, meaning that his death occurred squarely within the time range indicated in the medical evidence. The appellants submit that, by making his impugned comment, the trial judge ignored or forgot about the last dose of Dilaudid.

[77] I would reject this submission. The trial judge’s statement is supported by the evidence of Dr. Juurlink, who rejected the suggestion that an adverse reaction to Dilaudid could have taken place after 3 a.m. Dr. Juurlink said this:

It’s extraordinarily improbable because it presupposes that all of his toxicity manifested, all of the harmful effects that I’ve described to you earlier manifested after a delay not just following the final dose of [Dilaudid], but after an even more considerable delay after the first and second doses. So it would … be tantamount to suggesting that all of the toxicity of all of the opioids he had received in the 12 to 24 hours before he died was somehow brought to bear on him in this period after the delay that you’ve just outlined and that is very difficult for me to accept.

[78] This evidence supports the trial judge’s implicit conclusion that the effects of an overdose would have manifested by 3 a.m. on the day of Mr. Miller’s death.

[79] I conclude that the trial judge did not err in his finding on causation.

DISPOSITION

[80] I do not agree that the trial judge’s approach to liability was flawed. Nor do I accept that he erred in his determinations of the standard of care and causation.

[81] I would dismiss the appeal. The parties advised the court that they would resolve costs on their own.

Released: “S.E.P.” June 4, 2015

“M.L. Benotto J.A.”

“I agree E.A. Cronk J.A.”

“I agree S.E. Pepall J.A.”