Clarey v Gillis: A Near-Fatal Surgical Complication and No Breach of the Standard of Care

Norman Clarey went into hospital for planned surgery to remove a pre-cancerous growth from his bowel. He came close to dying. Four days after the operation, the reconnected section of his colon lost its blood supply and became necrotic, forcing emergency surgery, a permanent stoma, multi-organ failure, and months of hospitalization and rehabilitation. On any human measure, the outcome was devastating.

Yet in Clarey v Gillis, 2026 PESC 15, the Supreme Court of Prince Edward Island dismissed his claim in full. Justice Gordon Campbell found that the surgeon, Dr. Alex Gillis, met the standard of care at every stage: in the colonoscopy, in obtaining informed consent, and in performing the surgery itself. The decision is a clear illustration of a principle that patients often find hard to accept, and defendants rely on heavily: in a medical negligence action, a bad, even catastrophic, outcome is not by itself proof that anyone did anything wrong.

Clarey is a Prince Edward Island trial decision, so it is persuasive rather than binding in Ontario. On the law, though, that distinction matters less than usual here. Justice Campbell decided the case almost entirely on Ontario and Supreme Court of Canada authority, including Crits v Sylvester, Ter Neuzen v Korn, Reibl v Hughes, Clements v Clements, Snell v Farrell, and the Court of Appeal for Ontario’s reasons in Armstrong v Royal Victoria Hospital (endorsed by the Supreme Court in Armstrong v Ward). Those authorities are binding in Ontario, and the way a PEI court applied them is instructive for how a similar fact pattern would likely be analysed here.

What happened

Two routine colorectal cancer screening tests in early 2015 were positive for blood in the stool, and Mr. Clarey’s family physician referred him for a diagnostic colonoscopy. Dr. Gillis performed the colonoscopy on March 4, 2015, assisted by a second general surgeon. They found and removed one small polyp, and identified a second, irregular lesion near the end of the ascending colon. According to the two surgeons, they made repeated attempts to remove that lesion but could not do so safely; they took multiple biopsies instead. Pathology confirmed the lesion to be a tubular adenoma, a pre-cancerous growth.

Dr. Gillis met with Mr. Clarey on April 8, 2015. He advised that the polyp could not be removed endoscopically, reviewed the risks and benefits of either leaving it in place or removing it surgically, and recommended a laparoscopic right hemicolectomy, removal of the right side of the colon. The recognized risks he described included a leak at the anastomosis (the surgical reconnection of the two ends of the bowel), bleeding, the possibility of converting to open surgery, and the potential need for an ostomy. Mr. Clarey consented, and the family physician reviewed the plan with him again before surgery.

The hemicolectomy took place on May 8, 2015, with a second general surgeon assisting. Both surgeons testified that the operation was uncomplicated and that the anastomosis was well perfused, meaning it had a good blood supply, when they completed the surgery. Over the following weekend Mr. Clarey was cared for by nursing staff and an on-call surgeon, and no one identified anything beyond the expected course of recovery. When Dr. Gillis resumed care on May 11 and found worsening pain and no bowel movement, he ordered imaging. Mr. Clarey deteriorated overnight, and on May 12 Dr. Gillis performed an emergency laparotomy. He found that a portion of the transverse colon beyond the anastomosis had become ischemic and necrotic, although the anastomosis itself was intact and had not leaked. The dead tissue was removed and an ileostomy was created. Mr. Clarey’s recovery was long and dangerous, and he was discharged in August 2015.

The colonoscopy: a genuinely non-resectable polyp

Mr. Clarey’s experts argued that Dr. Gillis fell below the standard of care during the colonoscopy: that he should have removed or more fully characterized the polyp, that poor bowel preparation obstructed the view, and that a second colonoscopy or a referral should have followed. Justice Campbell rejected each of these submissions on the evidence.

He preferred the evidence of the defence experts, Dr. Attard and Dr. Redstone, and accepted Dr. Gillis’s contemporaneous operative note over the reconstructions offered by the plaintiff’s experts. He found there was no problem with the bowel preparation that interfered with the assessment, that the difficulty identifying the polyp’s margins reflected the nature of the lesion rather than any deficiency in technique, and that two qualified surgeons had independently concluded the polyp could not be safely resected endoscopically. He also found that a further colonoscopy would not have changed anything, because the polyp still could not have been removed that way.

Notably, Justice Campbell held that the involvement of two independent surgeons in the single colonoscopy was, in substance, equivalent to one surgeon performing a colonoscopy and then referring the patient to a colleague for a second look. That finding neutralized much of the plaintiff’s “should have referred” theory.

Informed consent and the alternatives that were not offered

The informed consent claim was the plaintiff’s strongest line of attack. His experts said Dr. Gillis should have offered further options before recommending surgery: a repeat colonoscopy, a referral for endoscopic mucosal resection (EMR), an advanced technique for removing larger polyps, or non-surgical management. Because those alternatives were not put to him, the argument went, his consent to the hemicolectomy was not truly informed.

Justice Campbell set the claim against the established framework. A physician must disclose the nature of a proposed treatment, its material risks, and any special or unusual risks, and must disclose reasonable alternatives that a patient in the same position would want to know about. But, drawing on Denman v Radovanovic and the cases it cites, he emphasized two limits. First, a physician is not obliged to offer alternatives that are not clinically indicated or that would not benefit the patient. Second, a plaintiff generally needs expert evidence to establish that a proposed alternative was in fact a medically reasonable option. On the evidence he accepted, neither a repeat colonoscopy nor a referral was clinically indicated in 2015, and the only medically reasonable choices were to leave the polyp in place, knowing it could eventually become cancerous, or to remove it surgically. Both were disclosed. The court described an alternative that offers no benefit and adds unnecessary risk as not a medically appropriate option at all.

The claim also failed on causation, and this is where informed consent cases so often turn. Under the subjective and modified objective tests confirmed in Denman v Radovanovic, a plaintiff must prove both that he personally would have declined the surgery had he been fully informed, and that a reasonable person in his position would have declined it. Justice Campbell was not satisfied that a reasonable person, told of the risks and benefits of the realistic options, would have disregarded the surgeon’s advice and refused the recommended surgery. He went further and found that even if further options had been discussed, the outcome would have been the same: the polyp would not have been endoscopically resected in any event, because EMR in Atlantic Canada was in its infancy in 2015 and no one was known to be accepting referrals for polyps like this one. The same structure defeated the recent bowel-surgery claim in Reistad v La Berge, where a British Columbia court held that a plaintiff needs expert evidence to show that an undisclosed alternative was a genuine, reasonable option before a consent claim can succeed.

The anastomosis and the operative note

The surgical negligence claim narrowed to a single, technical point: whether Dr. Gillis properly assessed the blood supply to the anastomosis, and whether his operative note recorded that he had. The plaintiff’s expert, Dr. Dixon, testified that failing to assess perfusion falls below the standard of care, and that failing to record the assessment in the operative note was itself a breach. He relied on the well-known line of authority, from Kolesar v Jeffries through Sozonchuk v Polych, that the absence of a contemporaneous record can support an inference that the event simply never happened.

Justice Campbell accepted the first proposition as a matter of principle: failing to assess the adequacy of the blood supply would fall below the standard. But he found, as a matter of fact, that Dr. Gillis did assess perfusion and did record it. Dr. Gillis’s note stated that a “suitable location” for the transection “was selected,” and he explained in his testimony that selecting a suitable location necessarily involves confirming that the bowel is adequately perfused and free of tension at that point. That explanation was corroborated by the assisting surgeon, who testified that he saw nothing to suggest the anastomosis was poorly perfused and would have raised it if he had. Applying the “invariable practice” doctrine from cases such as Boutcher v Cha and Bafaro v Dowd, the court held that the absence of a more explicit phrase in the operative note did not defeat credible evidence that the assessment was performed.

The court also made a point worth underlining for practitioners on both sides. Even if the note had been silent on perfusion, and even if that silence were a breach of the documentation standard, no damage flowed from the documentation gap itself. A failure to record a step does not cause an anastomosis to fail. On this record, Justice Campbell was satisfied that the plaintiff’s experts had overstated the “he did not document it” argument, because the note did in fact record the relevant step.

Causation and the danger of reasoning backwards

The plaintiff’s core causation theory was that the transverse colon was marginally perfused at the moment the anastomosis was created, which, he argued, meant there must have been a technical failure by the surgeon. Justice Campbell was not prepared to draw that inference, for reasons that repay attention.

The defence experts testified, and the court accepted, that a marginal blood supply can look no different from an adequate one at the time of surgery, and that perfusion can deteriorate afterward for reasons unrelated to technique: patient factors such as age, smoking history, diabetes, and vascular disease, and ordinary post-operative swelling that can compress the blood supply for days. Critically, all of the experts, including the plaintiff’s, agreed that an anastomosis can fail even when everything is done correctly, and that the recognized failure rate is in the range of one to five percent. Against that background, the court declined to infer negligence from the bad result alone.

Justice Campbell drew explicitly on the warning in Armstrong that it is a logical error to treat an improbable adverse outcome as proof that the defendant must have been negligent: rare events do sometimes occur, and correlation is not causation. He also invoked the long-standing caution in Crits v Sylvester, quoting Denning L.J.’s admonition that a court “must not condemn as negligence that which is only a misadventure.” Dr. Redstone put the same idea in plainer terms, attributing Mr. Clarey’s outcome to “a giant dose of bad luck, and misfortune.” The court concluded that the anastomosis most likely failed because the blood supply was, or became, marginal and could not sustain perfusion during healing, not because of any negligent act. That reasoning mirrors the recent Ontario decision in Dallner v Gladwell, which examined when a court may, and may not, infer negligence from a rare but recognized surgical complication.

What the decision means for Ontario patients

For patients and families, Clarey is a hard but honest lesson. A terrible outcome, standing alone, does not win a medical malpractice case. To succeed, a plaintiff must prove, with expert evidence, that the physician fell below the standard of a reasonable practitioner of the same experience and standing, and that the breach caused the injury on a balance of probabilities. A known complication that can occur despite competent care, without more, will not get a claim across either threshold. That is the same reality that defeated the surgical claim in Forget v Gibb, where a real and serious complication was found not to be a breach of the standard of care.

Three practical themes stand out. First, expert evidence is decisive: this case, like most, was won and lost on which experts the judge found more persuasive, and the plaintiff’s experts were found to have reconstructed events and overstated the documentation point. Second, informed consent claims usually fail on causation, not disclosure, because of the modified objective test and the difficulty of showing that a reasonable person would have refused recommended surgery. Third, the standard of care is a legal standard proven by expert testimony, not a description a commentator can supply; the summary above of what the surgeon should or should not have done reflects the evidence the court accepted in this case, not a fixed clinical rule.

One further point is worth flagging for Ontario readers. Justice Campbell accepted, drawing on Dumesnil v Jacobs and Gumbley v Vasiliou, that while the standard of care is uniform across Canada, applying it fairly means considering the facilities and resources realistically available where the physician practises. A general surgeon in Charlottetown is measured against reasonable practitioners in similar circumstances, not against a subspecialist at a large urban teaching hospital. That is a persuasive analysis rather than binding law in Ontario, but it is consistent with how Ontario courts have approached the same question.

If you or a family member has suffered a serious complication after surgery and you are trying to understand whether it reflects negligence or an unavoidable risk, our overview of suing for medical malpractice in Ontario explains how these cases are assessed and what has to be proven.

Clarey v Gillis, 2026 PESC 15, is a decision of the Supreme Court of Prince Edward Island and is persuasive, not binding, in Ontario.