Most medical negligence trials turn into a contest between experts. Occasionally one turns on whether the court believes the patient about something she saw. Bradley v Wang, 2026 ONCA 570, released on August 7, 2026, is the second kind, and the Court of Appeal for Ontario’s short reasons dismissing the appeal are a useful reminder of how little room there is to attack that kind of finding.
Bonnie Bradley attended her family doctor’s office on December 1, 2015 for a tetanus vaccination. The injection was given by Dr. Richard Wang, then a first-year family medicine resident being supervised by her family doctor. She felt severe pain on injection, looked at the needle while it was still in her arm, and saw that it was too high and too far forward. She was a registered nurse who had given many vaccinations herself, so she knew where the needle was supposed to be.
The pain did not resolve. She tried to continue in her job as an emergency department nurse, stopped going in a little over a week after the vaccination, and has never returned to any employment. Specialists who assessed her concluded that she had suffered an adverse inflammatory response to the vaccination and developed adhesive capsulitis, or frozen shoulder. Justice Vallee found Dr. Wang negligent, found that the negligence caused the injury, and awarded $931,585 for loss of income and $361,600 for pension loss. I wrote about the trial decision and the injection landmarking evidence behind it when it was released in 2025. This appeal leaves all of it standing.
A word on framing, because this case will attract readers looking for something it is not. The finding was not that a vaccine caused an injury. It was that this injection was placed in the wrong part of the arm, and that a correctly placed one would very likely have caused nothing at all. Ms. Bradley’s own expert accepted that adverse reactions to properly administered vaccines occur, and testified that they are rare. The injury pattern is described in the immunization literature as shoulder injury related to vaccine administration, or SIRVA, although the Court of Appeal does not use that term in these reasons.
A concession that made the case a credibility contest
The single most consequential thing that happened at trial was an agreement. The parties agreed that if Dr. Wang had administered the vaccine where Ms. Bradley said he did, he had breached the standard of care. That agreement removed the clinical question from the case and replaced it with a purely factual one: where did the needle go in.
Dr. Wang could not recall the specifics of the appointment, which after nearly a decade is unremarkable. More significantly, his notes did not record the precise location of the injection. What he was left with was a denial grounded in his usual practice: he would not inject in the area Ms. Bradley described. The trial judge preferred her direct eye-witness account over his evidence of habit, and rejected the challenges to her credibility and reliability, which included arguments that she had no real opportunity to observe the needle, that her evidence was internally inconsistent, and that she was an unreliable historian.
On appeal, the argument was that the trial judge had not properly analyzed the competing evidence, had left material inconsistencies in Ms. Bradley’s testimony unaddressed, and had given reasons too thin to permit meaningful review. The panel of Huscroft, Dawe and Wilson JJ.A. rejected all of it, applying settled principles: reasons are read functionally and as a whole, in light of the evidence and submissions (Farej v Fellows, 2022 ONCA 254; Monk v Farmers’ Mutual Insurance Company (Lindsay), 2019 ONCA 616, 92 B.L.R. (5th) 1), and a trial judge need not canvass every piece of evidence, answer every argument, or resolve every alleged inconsistency in a witness’s testimony (Liu v Huang, 2020 ONCA 450). Where the judge has made a reasoned assessment and dealt with the key credibility challenges actually raised, the decision to believe or disbelieve attracts significant deference: Housen v Nikolaisen, [2002] 2 S.C.R. 235, at para. 72.
None of that is new. What makes it worth noting is the structural point the concession created. Once both sides accepted that liability turned entirely on a fact about needle placement, and the defence had no contemporaneous record of that fact, the case reduced to whether an experienced nurse’s account of watching a needle enter her own arm would be believed. That is a difficult position to defend, and an almost impossible one to appeal from. The same dynamic appears whenever a defendant’s answer to a specific allegation is general habit rather than a note, which is one reason the contents of the chart matter so much in these cases.
It is worth adding that the appeal never reached the more interesting question the trial raised. The trial reasons addressed the standard applicable to a first-year resident performing a routine procedure. The concession meant none of that was in issue on appeal, so the resident standard-of-care analysis survives untested. In a different case it may not be conceded, and this decision says nothing about it. On the standard of care being established without a duelling-experts contest, compare Drain v Ziesmann.
What counts as evidence of the but-for world
The causation ground is the part of the decision with the widest application. Dr. Wang argued that Ms. Bradley had led no evidence at all of what would have happened in the but-for world, and that the trial judge had therefore filled the void with common-sense inference and reasoned backwards from the fact of injury to its cause.
The Court of Appeal disagreed, and the reason it gave is the useful part. Dr. David Ruggles, the plaintiff’s expert, testified that it would be extremely unlikely for a patient to experience Ms. Bradley’s symptoms after a properly administered vaccination. He accepted that abnormal immune responses happen, but described them as rare, and testified that immediate, severe and unremitting pain of the kind Ms. Bradley had was not consistent with that mechanism. There was also evidence that she had received tetanus vaccinations before without incident. Contrary to the appellant’s submission, the court held, “this was evidence of what might have occurred in the ‘but for’ world.”
That sentence repays attention. The appellant’s premise was that counterfactual evidence has to be a positive account of the alternative history. The court’s answer is that evidence which excludes the competing explanation, combined with the plaintiff’s own uneventful history with the same intervention, does the same work. The counterfactual was assembled from a differential, not narrated.
The point becomes sharper when this decision is read alongside Ibrahimova v Cavanagh, 2026 ONCA 573, which the Court of Appeal released the same day. Different panel, different facts, and a far longer set of reasons, but the same appellate argument: the plaintiff proved no counterfactual, so the trial judge must have speculated. In both cases the court answered that evidence the appellant had not recognized as counterfactual evidence was exactly that. That is my reading of the pair rather than anything either panel said, and neither decision purports to change the law. Together, though, they suggest that the bare assertion of an evidentiary void is not going far on appeal in Ontario at the moment.
Mitigation: a conjunctive burden, and who carries it
The third ground was that the trial judge ignored mitigation. The word does not appear in her reasons, which the Court of Appeal acknowledged, saying it would have been preferable if she had used it. Substance prevailed over vocabulary.
The analysis is worth setting out precisely, because it is the most practically useful passage in the decision. Under Southcott Estates Inc. v Toronto Catholic District School Board, 2012 SCC 51, [2012] 2 S.C.R. 675, at para. 24, the burden of proving a failure to mitigate rests on the defendant, and it has two limbs: that the plaintiff failed to make reasonable efforts to mitigate, and that mitigation was possible. Both must be established.
The trial judge never found that Ms. Bradley made reasonable efforts. On the first limb, in other words, she had no finding in her favour. But she had dealt with the second. She accepted Dr. Ruggles’ evidence that Ms. Bradley could not return to emergency department nursing, and rejected Dr. Wang’s evidence that other suitable positions were open to her, because all of them lay beyond the 45 km commute her injury allowed her to tolerate. That was enough. Because the defendant’s burden is conjunctive, failing the availability limb ended the argument regardless of what the plaintiff had or had not done.
The distance problem existed largely because Ms. Bradley had moved from Courtice to Marmora in 2017. The trial judge found the move was not unreasonable: it was made for financial reasons, after she lost the ability to work as an emergency department nurse. Those findings attracted deference, and the appeal failed on the simple ground that Dr. Wang had not proved suitable jobs existed where she actually lives.
There is a real objection to be made here, and the reasons do not answer it. On this analysis a plaintiff’s own post-injury relocation can shrink the pool of alternative employment, and the defendant absorbs the consequence. The reply in the decision is that the relocation was itself a downstream effect of the income loss the negligence caused, and was found reasonable, which is a factual finding an appellate court will not revisit lightly. But where the limit sits is unaddressed. A relocation unconnected to the injury, or one the trial judge found unreasonable, would presumably be treated differently, and nothing here tells us how differently. Defence counsel should read the passage as an evidentiary instruction rather than a doctrinal defeat: the alternative positions had to be ones the plaintiff could actually take, in the place she actually lived, and that evidence was not led.
How much this decision carries
Calibration matters. This is a short-form set of reasons for decision, sixteen paragraphs, per curiam, resolving three grounds almost entirely on deference. It creates no new law and does not claim to. Its value is illustrative, and on three points the illustration is a clear one.
A physician’s evidence of usual practice is weak medicine against a patient’s specific account of the act in question, particularly where the chart records nothing about it. Counterfactual proof can be built from evidence excluding the alternative explanation rather than from a positive narrative of the alternative history. And a mitigation defence that cannot establish the availability of suitable alternatives fails, whatever the plaintiff did or did not do to look for them.
The appeal was dismissed with costs of $50,000, all inclusive, in the amount the parties had agreed.
Bradley v Wang, 2026 ONCA 570, on appeal from Bradley v Wang, 2025 ONSC 2267.
Bradley v Wang: A Misplaced Injection, Usual Practice Evidence, and the Mitigation Burden
The Court of Appeal for Ontario dismisses Dr. Wang's appeal in Bradley v Wang, upholding findings that a tetanus vaccine was injected in the wrong part of the arm, that the misplacement caused an adhesive capsulitis, and that no mitigation defence was made out. The decision confirms that counterfactual proof can be assembled from evidence excluding the competing explanation, and that a defendant who cannot show suitable alternative work was available loses the mitigation argument regardless of the plaintiff's own efforts.
Most medical negligence trials turn into a contest between experts. Occasionally one turns on whether the court believes the patient about something she saw. Bradley v Wang, 2026 ONCA 570, released on August 7, 2026, is the second kind, and the Court of Appeal for Ontario’s short reasons dismissing the appeal are a useful reminder of how little room there is to attack that kind of finding.
Bonnie Bradley attended her family doctor’s office on December 1, 2015 for a tetanus vaccination. The injection was given by Dr. Richard Wang, then a first-year family medicine resident being supervised by her family doctor. She felt severe pain on injection, looked at the needle while it was still in her arm, and saw that it was too high and too far forward. She was a registered nurse who had given many vaccinations herself, so she knew where the needle was supposed to be.
The pain did not resolve. She tried to continue in her job as an emergency department nurse, stopped going in a little over a week after the vaccination, and has never returned to any employment. Specialists who assessed her concluded that she had suffered an adverse inflammatory response to the vaccination and developed adhesive capsulitis, or frozen shoulder. Justice Vallee found Dr. Wang negligent, found that the negligence caused the injury, and awarded $931,585 for loss of income and $361,600 for pension loss. I wrote about the trial decision and the injection landmarking evidence behind it when it was released in 2025. This appeal leaves all of it standing.
A word on framing, because this case will attract readers looking for something it is not. The finding was not that a vaccine caused an injury. It was that this injection was placed in the wrong part of the arm, and that a correctly placed one would very likely have caused nothing at all. Ms. Bradley’s own expert accepted that adverse reactions to properly administered vaccines occur, and testified that they are rare. The injury pattern is described in the immunization literature as shoulder injury related to vaccine administration, or SIRVA, although the Court of Appeal does not use that term in these reasons.
A concession that made the case a credibility contest
The single most consequential thing that happened at trial was an agreement. The parties agreed that if Dr. Wang had administered the vaccine where Ms. Bradley said he did, he had breached the standard of care. That agreement removed the clinical question from the case and replaced it with a purely factual one: where did the needle go in.
Dr. Wang could not recall the specifics of the appointment, which after nearly a decade is unremarkable. More significantly, his notes did not record the precise location of the injection. What he was left with was a denial grounded in his usual practice: he would not inject in the area Ms. Bradley described. The trial judge preferred her direct eye-witness account over his evidence of habit, and rejected the challenges to her credibility and reliability, which included arguments that she had no real opportunity to observe the needle, that her evidence was internally inconsistent, and that she was an unreliable historian.
On appeal, the argument was that the trial judge had not properly analyzed the competing evidence, had left material inconsistencies in Ms. Bradley’s testimony unaddressed, and had given reasons too thin to permit meaningful review. The panel of Huscroft, Dawe and Wilson JJ.A. rejected all of it, applying settled principles: reasons are read functionally and as a whole, in light of the evidence and submissions (Farej v Fellows, 2022 ONCA 254; Monk v Farmers’ Mutual Insurance Company (Lindsay), 2019 ONCA 616, 92 B.L.R. (5th) 1), and a trial judge need not canvass every piece of evidence, answer every argument, or resolve every alleged inconsistency in a witness’s testimony (Liu v Huang, 2020 ONCA 450). Where the judge has made a reasoned assessment and dealt with the key credibility challenges actually raised, the decision to believe or disbelieve attracts significant deference: Housen v Nikolaisen, [2002] 2 S.C.R. 235, at para. 72.
None of that is new. What makes it worth noting is the structural point the concession created. Once both sides accepted that liability turned entirely on a fact about needle placement, and the defence had no contemporaneous record of that fact, the case reduced to whether an experienced nurse’s account of watching a needle enter her own arm would be believed. That is a difficult position to defend, and an almost impossible one to appeal from. The same dynamic appears whenever a defendant’s answer to a specific allegation is general habit rather than a note, which is one reason the contents of the chart matter so much in these cases.
It is worth adding that the appeal never reached the more interesting question the trial raised. The trial reasons addressed the standard applicable to a first-year resident performing a routine procedure. The concession meant none of that was in issue on appeal, so the resident standard-of-care analysis survives untested. In a different case it may not be conceded, and this decision says nothing about it. On the standard of care being established without a duelling-experts contest, compare Drain v Ziesmann.
What counts as evidence of the but-for world
The causation ground is the part of the decision with the widest application. Dr. Wang argued that Ms. Bradley had led no evidence at all of what would have happened in the but-for world, and that the trial judge had therefore filled the void with common-sense inference and reasoned backwards from the fact of injury to its cause.
The Court of Appeal disagreed, and the reason it gave is the useful part. Dr. David Ruggles, the plaintiff’s expert, testified that it would be extremely unlikely for a patient to experience Ms. Bradley’s symptoms after a properly administered vaccination. He accepted that abnormal immune responses happen, but described them as rare, and testified that immediate, severe and unremitting pain of the kind Ms. Bradley had was not consistent with that mechanism. There was also evidence that she had received tetanus vaccinations before without incident. Contrary to the appellant’s submission, the court held, “this was evidence of what might have occurred in the ‘but for’ world.”
That sentence repays attention. The appellant’s premise was that counterfactual evidence has to be a positive account of the alternative history. The court’s answer is that evidence which excludes the competing explanation, combined with the plaintiff’s own uneventful history with the same intervention, does the same work. The counterfactual was assembled from a differential, not narrated.
The point becomes sharper when this decision is read alongside Ibrahimova v Cavanagh, 2026 ONCA 573, which the Court of Appeal released the same day. Different panel, different facts, and a far longer set of reasons, but the same appellate argument: the plaintiff proved no counterfactual, so the trial judge must have speculated. In both cases the court answered that evidence the appellant had not recognized as counterfactual evidence was exactly that. That is my reading of the pair rather than anything either panel said, and neither decision purports to change the law. Together, though, they suggest that the bare assertion of an evidentiary void is not going far on appeal in Ontario at the moment.
Mitigation: a conjunctive burden, and who carries it
The third ground was that the trial judge ignored mitigation. The word does not appear in her reasons, which the Court of Appeal acknowledged, saying it would have been preferable if she had used it. Substance prevailed over vocabulary.
The analysis is worth setting out precisely, because it is the most practically useful passage in the decision. Under Southcott Estates Inc. v Toronto Catholic District School Board, 2012 SCC 51, [2012] 2 S.C.R. 675, at para. 24, the burden of proving a failure to mitigate rests on the defendant, and it has two limbs: that the plaintiff failed to make reasonable efforts to mitigate, and that mitigation was possible. Both must be established.
The trial judge never found that Ms. Bradley made reasonable efforts. On the first limb, in other words, she had no finding in her favour. But she had dealt with the second. She accepted Dr. Ruggles’ evidence that Ms. Bradley could not return to emergency department nursing, and rejected Dr. Wang’s evidence that other suitable positions were open to her, because all of them lay beyond the 45 km commute her injury allowed her to tolerate. That was enough. Because the defendant’s burden is conjunctive, failing the availability limb ended the argument regardless of what the plaintiff had or had not done.
The distance problem existed largely because Ms. Bradley had moved from Courtice to Marmora in 2017. The trial judge found the move was not unreasonable: it was made for financial reasons, after she lost the ability to work as an emergency department nurse. Those findings attracted deference, and the appeal failed on the simple ground that Dr. Wang had not proved suitable jobs existed where she actually lives.
There is a real objection to be made here, and the reasons do not answer it. On this analysis a plaintiff’s own post-injury relocation can shrink the pool of alternative employment, and the defendant absorbs the consequence. The reply in the decision is that the relocation was itself a downstream effect of the income loss the negligence caused, and was found reasonable, which is a factual finding an appellate court will not revisit lightly. But where the limit sits is unaddressed. A relocation unconnected to the injury, or one the trial judge found unreasonable, would presumably be treated differently, and nothing here tells us how differently. Defence counsel should read the passage as an evidentiary instruction rather than a doctrinal defeat: the alternative positions had to be ones the plaintiff could actually take, in the place she actually lived, and that evidence was not led.
How much this decision carries
Calibration matters. This is a short-form set of reasons for decision, sixteen paragraphs, per curiam, resolving three grounds almost entirely on deference. It creates no new law and does not claim to. Its value is illustrative, and on three points the illustration is a clear one.
A physician’s evidence of usual practice is weak medicine against a patient’s specific account of the act in question, particularly where the chart records nothing about it. Counterfactual proof can be built from evidence excluding the alternative explanation rather than from a positive narrative of the alternative history. And a mitigation defence that cannot establish the availability of suitable alternatives fails, whatever the plaintiff did or did not do to look for them.
The appeal was dismissed with costs of $50,000, all inclusive, in the amount the parties had agreed.
Bradley v Wang, 2026 ONCA 570, on appeal from Bradley v Wang, 2025 ONSC 2267.
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.
A Bimalleolar Ankle Fracture, Lost to Follow-up, and a Below-Knee Amputation
A settlement involving a 65-year-old man whose ankle fracture went 40 days without orthopedic follow-up, leading to joint infection and below-knee amputation.
Szeto v Kives: Laparoscopic Hysterectomy, Bowel Injury, and the Surgical Duty to Inspect
Ontario Superior Court finds gynecologic surgeon liable for unrecognized intraoperative bowel injury. The duty to inspect framework in laparoscopic surgery.
Brown v Meaney: Pyridoxine, Informed Consent, and the Low-Risk Alternative
Ontario Superior Court finds pediatric neurologists negligent for delayed pyridoxine treatment in PDE. The low-risk alternative framework and twin sister causation.