A first impression that is reasonable when it is formed can still become negligent if a physician refuses to revisit it as the clinical picture changes. That is the thread running through the Court of Appeal for Ontario’s decision in Brown v Meaney, released on June 19, 2026, which upheld a trial finding that two pediatric neurologists were negligent in their treatment of an infant with a rare form of epilepsy. I wrote about Justice Bordin’s trial decision when it was released. The appeal does not disturb it.
Skyler Brown was born in December 2002 and began experiencing seizures at about four months of age. She was transferred to McMaster Children’s Hospital in Hamilton and came under the care of Dr. Meaney and Dr. Ronen, both pediatric neurologists. The condition behind her seizures, pyridoxine-dependent epilepsy (PDE), was rare but known in 2003: an inability to metabolize vitamin B6 that ordinary anti-seizure medications cannot control, but pyridoxine can. On the expert evidence accepted at trial, PDE was diagnosed through an empirical trial in three phases: challenge (give pyridoxine), withdrawal (stop it and watch), and rechallenge (restart it if seizures return). There was no consensus in 2003 on the dose or on how long each phase should last.
On the trial judge’s findings, the physicians started the trial but did not finish it. They concluded only three to four days into the withdrawal phase that pyridoxine had not worked, ruled it out, and did not reintroduce it even after the seizures returned and escalated. Pyridoxine was finally restarted in mid-July 2003, after Skyler had been placed in a drug-induced coma. Her seizures then stopped. Genetic testing confirmed PDE in 2008. By the time of trial, Skyler was 21, non-verbal, and required assistance with all of her daily activities. The parties had agreed on damages, so the appeal concerned only liability and causation.
The appellants raised three grounds: that the trial judge created his own standard of care, that he mishandled informed consent, and that he erred on causation. Wilson J.A., writing for a panel that included Copeland and Pomerance JJ.A., rejected all three.
The standard of care came from the experts, not the judge
The most substantial argument was that the trial judge had invented a standard of care: a rigid, formulaic rule about observation periods that no expert had endorsed. The point matters because, in professional negligence, the standard of care is generally a question for expert evidence rather than something a judge can set on his own. The benchmark is that of a reasonably prudent specialist in the same field (ter Neuzen v Korn, [1995] 3 S.C.R. 674), judged without the benefit of hindsight, and expert evidence is normally required to establish it (Flood v Boutette, 2021 ONCA 515, 460 D.L.R. (4th) 648). The standard is not one of perfection, or even excellence. Proving what it was, and that it was breached, is part of what a plaintiff must establish in any medical negligence case.
The Court of Appeal accepted the legal premise but found it did not apply. The standard the trial judge used, that the physicians had to keep monitoring beyond the short window and reintroduce pyridoxine if the seizures returned, was the standard the experts themselves described. All of them agreed that a proper empirical trial runs through challenge, withdrawal, and rechallenge, and that no firm conclusion about the drug’s efficacy could be drawn without completing it. Experts on both sides, including one called by the appellants, testified that seizures could take weeks to return after withdrawal, particularly in atypical PDE like Skyler’s. Against that evidence, concluding after three to four days that pyridoxine had failed, and closing the door on it, was open to challenge.
The court also disposed of the narrower complaint that the judge had found the physicians “set” a fixed observation period. Wilson J.A. accepted the phrasing was imprecise but read the reasons as merely describing how long the physicians had taken to reach their conclusion. The reference to a period of observation was, in her words, “descriptive, not prescriptive.”
The defence had a genuine argument here, and it is worth stating at its strongest. PDE was rare, there were no clear dosing or timing guidelines in 2003, and the initial clinical impression that PDE was unlikely was itself reasonable. The court did not say otherwise, and it was careful to note that a merely wrong judgment call is not negligence. Liability attached only because the specific conclusion, that pyridoxine could be ruled out after a few days and never revisited, was found unreasonable on the expert evidence. That distinction is the practical core of the case. It also means the result rested on trial-level findings of fact, which an appellate court will not lightly disturb. That deference cuts both ways: the same standard that protected this judgment would have protected an adverse one.
A physician cannot escape the duty to disclose by dismissing the option
The second ground produced the decision’s most transferable holding. The appellants argued that if they did not consider pyridoxine a viable option, they could not be faulted for failing to mention it to Skyler’s parents. The Court of Appeal rejected the logic. The duty to obtain informed consent requires disclosure of the nature of a proposed treatment, its material risks and benefits, and reasonable alternatives (Denman v Radovanovic, 2024 ONCA 276, 495 D.L.R. (4th) 444), and it extends to what a physician knew or ought to have known, not only what they happened to believe (Hopp v Lepp, [1980] 2 S.C.R. 192). A negligent failure to appreciate that pyridoxine was a reasonable alternative could not double as a defence to not disclosing it. The same point about the scope of the disclosure duty runs through the Court of Appeal’s reasoning in Denman.
Materiality is measured objectively, by asking whether a reasonable person in the patient’s position would want to know of the risk or alternative (Ciarlariello v Schacter, [1993] 2 S.C.R. 119). The trial judge found that Skyler’s parents were never told why pyridoxine had been given, that she had been seizure-free while on it, or that it could be restarted if the seizures returned. That was information any reasonable parent would want, and the appellate court saw no reason to disturb the finding. The court also confirmed that the same facts can ground both a treatment-negligence claim and an informed-consent claim, although a plaintiff cannot recover twice for the same loss.
Causation without a precise mechanism
On causation, the appellants argued that there was no proof Skyler’s brain injury came from the July 2003 seizures, because the evidence was that only prolonged seizures, those lasting more than 45 minutes, cause damage, and none of hers was shown to have lasted that long. The court treated the 45-minute figure as one mechanism of injury, not the only one. Causation in negligence does not demand scientific precision, and a trier of fact need not pin down the exact mechanism (Snell v Farrell, [1990] 2 S.C.R. 311; Cheung v Samra, 2022 ONCA 195, 467 D.L.R. (4th) 708). The question was the commonsense “but for” one.
The evidence supporting it was substantial. Skyler was developing normally, with a normal MRI, before July 2003; her brain growth slowed afterward; and later imaging showed injury. The physicians’ own contemporaneous records reflected concern about brain injury at the time. And there was a striking comparator: Skyler’s sister, who also has PDE but was treated promptly with pyridoxine and developed normally. On the facts as found, that comparison functioned as a natural experiment, and the trial judge was entitled to rely on it without resolving whether the injury came from a single prolonged seizure, successive seizures, the induced coma, or some combination.
A note of caution for anyone reading this as a template. The comparator here was unusually clean. Most causation disputes do not come with a same-condition sibling who was treated differently and did well. The decision confirms the law’s tolerance for commonsense inference on causation; it does not promise that every plaintiff will have evidence this stark.
A pointed note on the length of reasons
Wilson J.A. closed with a comment unrelated to the merits. The trial reasons ran to 125 pages for a 12-day trial in which only liability was in issue, which the court considered unnecessarily long and repetitive, echoing earlier guidance that reasons should not become a data dump or an extended note to self (Welton v United Lands Corporation Limited, 2020 ONCA 322, 64 C.C.E.L. (4th) 265). It is a reminder, mostly for counsel and judges, that thoroughness and length are not the same thing.
What the decision means
Brown v Meaney is a binding Court of Appeal statement on three points that recur in medical negligence claims. A reasonable initial diagnosis does not license a physician to stop reconsidering it as new information arrives. A doctor cannot avoid the duty to discuss a reasonable alternative by having wrongly written it off. And causation can be proven on a commonsense basis, without identifying the precise biological mechanism of the injury. None of this lowers the bar a plaintiff must clear. The standard of care in any case still turns on expert evidence, and the result here rested on findings the trial judge made after hearing that evidence. In that respect the decision sits comfortably alongside other delayed-diagnosis appeals where the Court of Appeal has deferred to a trial judge’s findings grounded in the expert record.
Brown v Meaney, 2026 ONCA 445, affirming Brown v Meaney, 2024 ONSC 7256.
Brown v Meaney: Clinical Judgment, Informed Consent, and Commonsense Causation on Appeal
The Court of Appeal dismisses the appeal in Brown v Meaney, upholding findings that two pediatric neurologists breached the standard of care and the duty to obtain informed consent when they abandoned a pyridoxine trial in an infant with a rare epilepsy. The decision affirms that a defensible first impression does not excuse the failure to revisit it, and that causation need not be proven with scientific precision.
A first impression that is reasonable when it is formed can still become negligent if a physician refuses to revisit it as the clinical picture changes. That is the thread running through the Court of Appeal for Ontario’s decision in Brown v Meaney, released on June 19, 2026, which upheld a trial finding that two pediatric neurologists were negligent in their treatment of an infant with a rare form of epilepsy. I wrote about Justice Bordin’s trial decision when it was released. The appeal does not disturb it.
Skyler Brown was born in December 2002 and began experiencing seizures at about four months of age. She was transferred to McMaster Children’s Hospital in Hamilton and came under the care of Dr. Meaney and Dr. Ronen, both pediatric neurologists. The condition behind her seizures, pyridoxine-dependent epilepsy (PDE), was rare but known in 2003: an inability to metabolize vitamin B6 that ordinary anti-seizure medications cannot control, but pyridoxine can. On the expert evidence accepted at trial, PDE was diagnosed through an empirical trial in three phases: challenge (give pyridoxine), withdrawal (stop it and watch), and rechallenge (restart it if seizures return). There was no consensus in 2003 on the dose or on how long each phase should last.
On the trial judge’s findings, the physicians started the trial but did not finish it. They concluded only three to four days into the withdrawal phase that pyridoxine had not worked, ruled it out, and did not reintroduce it even after the seizures returned and escalated. Pyridoxine was finally restarted in mid-July 2003, after Skyler had been placed in a drug-induced coma. Her seizures then stopped. Genetic testing confirmed PDE in 2008. By the time of trial, Skyler was 21, non-verbal, and required assistance with all of her daily activities. The parties had agreed on damages, so the appeal concerned only liability and causation.
The appellants raised three grounds: that the trial judge created his own standard of care, that he mishandled informed consent, and that he erred on causation. Wilson J.A., writing for a panel that included Copeland and Pomerance JJ.A., rejected all three.
The standard of care came from the experts, not the judge
The most substantial argument was that the trial judge had invented a standard of care: a rigid, formulaic rule about observation periods that no expert had endorsed. The point matters because, in professional negligence, the standard of care is generally a question for expert evidence rather than something a judge can set on his own. The benchmark is that of a reasonably prudent specialist in the same field (ter Neuzen v Korn, [1995] 3 S.C.R. 674), judged without the benefit of hindsight, and expert evidence is normally required to establish it (Flood v Boutette, 2021 ONCA 515, 460 D.L.R. (4th) 648). The standard is not one of perfection, or even excellence. Proving what it was, and that it was breached, is part of what a plaintiff must establish in any medical negligence case.
The Court of Appeal accepted the legal premise but found it did not apply. The standard the trial judge used, that the physicians had to keep monitoring beyond the short window and reintroduce pyridoxine if the seizures returned, was the standard the experts themselves described. All of them agreed that a proper empirical trial runs through challenge, withdrawal, and rechallenge, and that no firm conclusion about the drug’s efficacy could be drawn without completing it. Experts on both sides, including one called by the appellants, testified that seizures could take weeks to return after withdrawal, particularly in atypical PDE like Skyler’s. Against that evidence, concluding after three to four days that pyridoxine had failed, and closing the door on it, was open to challenge.
The court also disposed of the narrower complaint that the judge had found the physicians “set” a fixed observation period. Wilson J.A. accepted the phrasing was imprecise but read the reasons as merely describing how long the physicians had taken to reach their conclusion. The reference to a period of observation was, in her words, “descriptive, not prescriptive.”
The defence had a genuine argument here, and it is worth stating at its strongest. PDE was rare, there were no clear dosing or timing guidelines in 2003, and the initial clinical impression that PDE was unlikely was itself reasonable. The court did not say otherwise, and it was careful to note that a merely wrong judgment call is not negligence. Liability attached only because the specific conclusion, that pyridoxine could be ruled out after a few days and never revisited, was found unreasonable on the expert evidence. That distinction is the practical core of the case. It also means the result rested on trial-level findings of fact, which an appellate court will not lightly disturb. That deference cuts both ways: the same standard that protected this judgment would have protected an adverse one.
A physician cannot escape the duty to disclose by dismissing the option
The second ground produced the decision’s most transferable holding. The appellants argued that if they did not consider pyridoxine a viable option, they could not be faulted for failing to mention it to Skyler’s parents. The Court of Appeal rejected the logic. The duty to obtain informed consent requires disclosure of the nature of a proposed treatment, its material risks and benefits, and reasonable alternatives (Denman v Radovanovic, 2024 ONCA 276, 495 D.L.R. (4th) 444), and it extends to what a physician knew or ought to have known, not only what they happened to believe (Hopp v Lepp, [1980] 2 S.C.R. 192). A negligent failure to appreciate that pyridoxine was a reasonable alternative could not double as a defence to not disclosing it. The same point about the scope of the disclosure duty runs through the Court of Appeal’s reasoning in Denman.
Materiality is measured objectively, by asking whether a reasonable person in the patient’s position would want to know of the risk or alternative (Ciarlariello v Schacter, [1993] 2 S.C.R. 119). The trial judge found that Skyler’s parents were never told why pyridoxine had been given, that she had been seizure-free while on it, or that it could be restarted if the seizures returned. That was information any reasonable parent would want, and the appellate court saw no reason to disturb the finding. The court also confirmed that the same facts can ground both a treatment-negligence claim and an informed-consent claim, although a plaintiff cannot recover twice for the same loss.
Causation without a precise mechanism
On causation, the appellants argued that there was no proof Skyler’s brain injury came from the July 2003 seizures, because the evidence was that only prolonged seizures, those lasting more than 45 minutes, cause damage, and none of hers was shown to have lasted that long. The court treated the 45-minute figure as one mechanism of injury, not the only one. Causation in negligence does not demand scientific precision, and a trier of fact need not pin down the exact mechanism (Snell v Farrell, [1990] 2 S.C.R. 311; Cheung v Samra, 2022 ONCA 195, 467 D.L.R. (4th) 708). The question was the commonsense “but for” one.
The evidence supporting it was substantial. Skyler was developing normally, with a normal MRI, before July 2003; her brain growth slowed afterward; and later imaging showed injury. The physicians’ own contemporaneous records reflected concern about brain injury at the time. And there was a striking comparator: Skyler’s sister, who also has PDE but was treated promptly with pyridoxine and developed normally. On the facts as found, that comparison functioned as a natural experiment, and the trial judge was entitled to rely on it without resolving whether the injury came from a single prolonged seizure, successive seizures, the induced coma, or some combination.
A note of caution for anyone reading this as a template. The comparator here was unusually clean. Most causation disputes do not come with a same-condition sibling who was treated differently and did well. The decision confirms the law’s tolerance for commonsense inference on causation; it does not promise that every plaintiff will have evidence this stark.
A pointed note on the length of reasons
Wilson J.A. closed with a comment unrelated to the merits. The trial reasons ran to 125 pages for a 12-day trial in which only liability was in issue, which the court considered unnecessarily long and repetitive, echoing earlier guidance that reasons should not become a data dump or an extended note to self (Welton v United Lands Corporation Limited, 2020 ONCA 322, 64 C.C.E.L. (4th) 265). It is a reminder, mostly for counsel and judges, that thoroughness and length are not the same thing.
What the decision means
Brown v Meaney is a binding Court of Appeal statement on three points that recur in medical negligence claims. A reasonable initial diagnosis does not license a physician to stop reconsidering it as new information arrives. A doctor cannot avoid the duty to discuss a reasonable alternative by having wrongly written it off. And causation can be proven on a commonsense basis, without identifying the precise biological mechanism of the injury. None of this lowers the bar a plaintiff must clear. The standard of care in any case still turns on expert evidence, and the result here rested on findings the trial judge made after hearing that evidence. In that respect the decision sits comfortably alongside other delayed-diagnosis appeals where the Court of Appeal has deferred to a trial judge’s findings grounded in the expert record.
Brown v Meaney, 2026 ONCA 445, affirming Brown v Meaney, 2024 ONSC 7256.
Paul Cahill
Partner, Davidson Cahill Morrison LLP | LSO Certified Specialist in Civil Litigation
Paul represents victims of medical malpractice across Ontario, with trial experience including a $11.5M jury verdict in a birth injury case. He is recognized in Best Lawyers in Canada and serves as trial counsel to other lawyers on complex medical negligence matters.
About PaulMore on medical malpractice in Ontario.
Other articles by Paul exploring the conditions, decisions, and systems behind preventable medical harm.
Ozempic, Wegovy, and the Standard of Care for GLP-1 Prescribing
GLP-1 drugs are now mainstream medicine in Canada. When prescribing falls below the standard of care, the harms can be serious and the legal questions are real.
Hacopian-Armen Estate v Mahmoud: An Appeal on Causation and the “Same Class or Kind” Foreseeability Analysis
The Ontario Court of Appeal upheld Paul Cahill’s trial verdict in Hacopian-Armen Estate v Mahmoud, clarifying the foreseeability test in delayed cancer cases.
Kotorashvili v Lee: When a Surgeon Deviates from His Own Treatment Plan
An Ontario orthopaedic surgeon was found liable after removing clavicle hardware six weeks early without revisiting his own documented treatment plan.