Dallner v Gladwell: Brachial Plexus Injury and Inferring Negligence from a Rare Complication

Most surgical operations carry a known catalogue of complications. Patients are warned about them as part of the informed consent process. Some of those complications occur even when the surgery is performed competently; others do not. The legal question that often divides surgical malpractice cases is which side of that line the particular complication falls on. Where the complication is a recognized adverse event that can occur with appropriate care, the defendant typically prevails on the standard of care analysis. Where the complication is extreme and rare to the point that no competent surgeon should be producing it, the trial judge may be entitled to infer that something fell below the standard during the procedure even where the precise mechanism is not directly observed.

Dallner v Gladwell, 2024 ONSC 6557, is a clean Ontario illustration of the second category. The plaintiff underwent a total shoulder arthroplasty (a routine joint replacement procedure) for glenohumeral arthritis. He emerged from surgery with a permanent injury to the medial cord of his brachial plexus, the network of nerves that supplies strength and sensation to the arm. Both orthopedic surgeons who testified at trial — one for the plaintiff and one for the defendant — said they had never seen or heard of an injury of this degree from this procedure. The Ontario Superior Court of Justice (November 25, 2024) found the defendant surgeon negligent on the standard of care, with the negligence consisting of applying an excessive degree or duration of force or stretch to the patient’s arm during the surgery. Judgment was entered for the plaintiff; damages had been settled before trial.

The decision is doctrinally significant for several reasons. It applies the Armstrong v Royal Victoria Hospital, 2019 ONCA 963, framework that allows the trier of fact to consider the nature of the injury in resolving whether the standard of care has been breached, where the injury bears on “what happened” in a surgical procedure that no one else witnessed. It illustrates the careful judicial reasoning required to distinguish “bad outcome alone” from “outcome plus surrounding evidence that supports an inference of breach.” It demonstrates the role of the surgeon’s own contemporaneous operative documentation in the post-event reconstruction. And it adds a substantive orthopedic surgery case to the cluster — a practice area that had not previously been well represented.

The case in summary

The procedure. A total shoulder arthroplasty (TSA) is a joint replacement procedure in which both the ball (humeral head) and socket (glenoid) of the shoulder are replaced with implants. The surgery is performed for end-stage glenohumeral arthritis where conservative measures have failed. It is a standard operation in modern orthopedic practice; in Ontario, it is performed in many community and academic centres.

A necessary feature of the procedure is repositioning of the upper limb. The arm must be moved into non-anatomic positions to give the surgeon access to the glenoid (the socket) and to the top of the humerus (the bone of the upper arm). Tissue release is required. Force is applied to dislocate and to position the shoulder. The arm is repositioned several times during a single procedure. None of this is itself the negligence; the manipulation is unavoidable. The question is whether the force and duration applied stayed within the bounds of what a competent shoulder surgeon would apply.

The patient. Mr. Dallner presented with right shoulder pain and was diagnosed with glenohumeral arthritis. He was a good candidate for the surgery. There was no dispute that he needed it. He was not a complicated patient.

The surgeon. Dr. Gladwell was an orthopedic surgeon practising in Owen Sound, Ontario, in his early years of independent practice. He had performed total shoulder arthroplasty many times during residency and on his own (he estimated 40 to 50 times) since completing his training approximately two and a half years before this surgery.

The injury. During the surgery, Mr. Dallner sustained a permanent injury to the medial cord of his brachial plexus. The medial cord is one of three cords (medial, lateral, and posterior) of the brachial plexus, the nerve network that runs from the cervical spine through the axilla (armpit) to supply the upper limb. Injury to the medial cord produces loss of strength and sensation in territories supplied by that cord. In Mr. Dallner’s case, the injury was permanent and irreversible.

The defendant’s position. Dr. Gladwell took all reasonable precautions to protect the brachial plexus during the surgery. Nerve injuries can and do happen during this kind of procedure. No one knows exactly how the injury occurred. At worst, he may have erred about the amount of force or traction applied, but error in judgment of this kind is not negligence.

The plaintiff’s position. The injury could only have been caused by force or duration of stretch outside the bounds of competent practice. The extreme rarity of the complication supports the inference of breach. The mechanism is consistent with the surgeon’s own contemporaneous documentation, which identified intraoperative stretch and retraction during glenoid exposure as the most likely cause.

The surgeon’s own admission

A doctrinally important feature of the case is the surgeon’s contemporaneous operative addendum note. After the surgery, Dr. Gladwell documented his own assessment of what had likely caused the injury. The trial judge characterized the note as identifying the most likely cause of the injury within hours of the surgery: “Most likely, this is due to intraoperative stretch.” And: “Overall, I presume this is from retraction of the humerus during glenoid exposure, but again I am surprised by this.”

The note operates as evidence on two levels. First, it identifies the mechanism that the trier of fact ultimately accepted (intraoperative stretch). The surgeon’s own real-time assessment substantially aligned with the plaintiff’s case on what had happened. Second, the note’s reference to surprise — “but again I am surprised by this” — supports the rarity finding. The surgeon himself characterized the outcome as unusual, consistent with the expert evidence that the complication is extreme and rare.

Contemporaneous documentation of a surgeon’s own assessment of what went wrong is a recurring feature of plaintiff-success surgical malpractice cases. Where a surgeon documents an honest assessment of cause shortly after the event, the documentation can become decisive evidence on the mechanism analysis. The principle is bidirectional: documentation that supports the defence position is equally probative. The point is that contemporaneous documentation is durable in a way that retrospective trial testimony is not.

The expert evidence

The trial proceeded with expert evidence from orthopedic surgeons on both sides. Each was qualified in the field of orthopedic surgery with relevant experience in shoulder arthroplasty.

The plaintiff’s expert testified that a permanent and irrevocable injury to the medial cord of the brachial plexus, in a patient having an anatomic shoulder replacement for primary glenohumeral osteoarthritis with no prior surgeries or injuries to the shoulder, was something he had not seen in his practice. His evidence, distilled: “I haven’t seen it, it doesn’t happen. It shouldn’t happen.” The injury, in his view, fell outside the catalogue of complications that occur with appropriately performed total shoulder arthroplasty.

The defendant’s expert agreed that the complication was exceptionally rare. He had never seen or heard of an injury of this degree from this procedure either. His evidence supported the rarity finding while preserving the defence position that the rarity did not in itself establish negligence — that some complications occur even with competent care and the trier of fact should not work backwards from the outcome.

The agreement on rarity is unusual in malpractice litigation. Most cases involve expert disagreement on the underlying scientific or clinical questions. Where both experts converge on rarity, the trial judge is in an unusually strong evidentiary position to reason from the rarity to the inference of breach — particularly where the surgeon’s own contemporaneous documentation also points to the mechanism.

The legal framework

The trial judge worked through several legal principles in arriving at the conclusion of breach.

The bidirectional approach to standard of care and causation. The framework was articulated in Bafaro v Dowd, 2010 ONCA 188, which says it is “typically best” to determine breach before causation. The principle is that the trial judge should not work backwards from the injury to find a breach that may not have actually occurred — the “post hoc ergo propter hoc” fallacy.

However, the Ontario Court of Appeal in Armstrong v Royal Victoria Hospital, 2019 ONCA 963, refined the principle. In that case, both Paciocco JA and van Rensburg JA agreed that where the nature of the injury is relevant to “what happened,” it is not an error to consider the injury in resolving whether the standard of care has been breached. The framework is not absolute. Where the injury itself is evidence of what occurred — where the mechanism of injury bears on the standard of care analysis — the trier of fact may properly consider the injury as part of the SOC analysis.

The principle is supported by several earlier Ontario cases: Meringolo (Committee of) v Oshawa General Hospital, [1991] OJ No 91 (CA), leave to appeal refused [1991] SCCA No 115; Grass (Litigation guardian of) v Women’s College Hospital, 2001 CanLII 8526 (ON CA), leave to appeal refused [2001] SCCA No 372; and Kennedy v Jackiewicz, [2003] OJ No 1854 (SCJ), affirmed [2004] OJ No 4816 (CA).

The application to surgical cases is particularly important. In a surgical malpractice case, the trier of fact typically does not have direct observation of what occurred during the procedure. The operative record may be incomplete. The witnesses are often the surgeon and the assistants, all aligned in interest with the defendant. The injury itself is often the most accessible evidence of what happened. The Armstrong framework allows the trier of fact to engage with the injury as evidence rather than artificially excluding it from the SOC analysis.

The “bad outcome alone” principle. A separate principle, equally important, is that a bad outcome does not by itself establish negligence. The principle was articulated by the trial judge as: “If an event can occur, one must expect that it will sometimes occur.” The proposition is foundational in Canadian malpractice law. Some complications occur with competent care. The fact of a complication does not establish that the care was substandard.

The principle is sometimes confused with the Armstrong framework but the two are distinct. The “bad outcome alone” principle says the injury cannot by itself prove negligence. The Armstrong framework says the nature of the injury can inform the SOC analysis where it is relevant to what happened. The two operate together: a bad outcome of a type that can occur with competent care does not establish breach; a bad outcome of a type that cannot occur with competent care can support an inference of breach. The Dallner case turned on which side of that line the medial cord brachial plexus injury fell.

The plaintiff does not need to rule out all non-negligent causes. The trial judge addressed an argument sometimes raised by defendants: that the plaintiff must rule out all non-negligent explanations before the trier of fact can find breach. The court rejected the argument. The plaintiff’s burden is balance of probabilities — to show that breach is more likely than not. The plaintiff is not required to eliminate every conceivable non-negligent explanation. The right approach is to assess the relative likelihoods of each possibility. Where the negligent explanation is more probable than not, the plaintiff has met the burden.

The principle is operationally important. In any rare-complication case, there will almost always be some non-negligent explanation that is theoretically possible. If the plaintiff had to disprove every such possibility, the burden would exceed balance of probabilities. The right framework is comparative: which is more likely, on the evidence, the negligent or the non-negligent explanation?

Inferring what happened from the outcome. Where the trier of fact has limited direct evidence of what occurred during a procedure, the outcome can support inferences about the mechanism. In Dallner, the trial judge inferred from the extreme rarity of the injury, from the surgeon’s own contemporaneous identification of intraoperative stretch as the mechanism, and from the expert evidence on how stretch injuries occur, that the breach was the application of excessive force or duration of stretch on the patient’s arm during the surgery.

The inference is not pure speculation. It is grounded in specific evidence: the rarity finding from both experts, the surgeon’s own note, the anatomical analysis of how the medial cord could be injured during glenoid exposure, the expert evidence on the mechanics of brachial plexus injury. The trial judge worked through the evidence methodically and arrived at the conclusion through reasoned analysis rather than from outcome alone. The reasoning is the model for plaintiff counsel in similar surgical cases.

The brachial plexus and the shoulder

A brief clinical aside on the anatomy and the mechanism of injury, since the underlying biology is foundational to the legal analysis.

The brachial plexus is a network of nerves arising from the cervical spinal nerves C5 through T1. The roots combine to form trunks (upper, middle, lower), the trunks divide into divisions (anterior and posterior), and the divisions reassemble into cords (medial, lateral, posterior). The cords pass under the clavicle and into the axilla, where they branch into the major nerves of the upper limb: the median, ulnar, radial, musculocutaneous, and axillary nerves.

The medial cord specifically gives rise to the ulnar nerve and contributes to the median nerve. Injury to the medial cord produces weakness and sensory loss in the territories of these nerves: weakness of intrinsic hand muscles, sensory loss along the ulnar (little-finger) side of the hand, and various functional impairments depending on the severity.

The brachial plexus is anatomically close to the shoulder joint and the operative field for shoulder arthroplasty. During total shoulder arthroplasty, the surgical exposure of the glenoid (socket) involves retraction of the humerus and surrounding structures away from the operative field. The retraction creates the mechanical conditions under which a stretch injury to the brachial plexus is theoretically possible. Modern shoulder arthroplasty technique includes specific measures to minimize the risk: careful positioning, controlled retraction, awareness of the duration that the arm is held in non-anatomic positions, and attention to the patient’s specific anatomy.

A stretch injury to a nerve occurs when the nerve is held under tension beyond its tolerance. The injury can be temporary (neuropraxia, where the axon is preserved and recovery occurs over weeks to months) or permanent (axonotmesis or neurotmesis, where axons are damaged and recovery is incomplete or absent). The severity correlates with the magnitude and duration of the tension. Modern neurophysiology research supports the framework that nerves can tolerate moderate stretch for short periods but suffer permanent injury when the stretch is excessive in magnitude, duration, or both.

The clinical and legal point converges: in Dallner, the trier of fact concluded that the magnitude or duration of stretch in this surgery exceeded what a competent shoulder surgeon would apply. The conclusion was supported by the extreme rarity of the resulting injury, by the surgeon’s own characterization of the mechanism, and by the expert evidence on the biology.

Why this case matters

For prospective clients. Surgical malpractice cases are typically harder for plaintiffs than other categories of medical malpractice. The trier of fact does not directly observe what happened during the procedure. The defendant’s expert evidence often includes the proposition that “complications happen even with appropriate care.” Plaintiffs in surgical cases face structural disadvantages.

Dallner illustrates that the structural disadvantages are not insurmountable. Where the complication is extreme and rare to the point that competent surgical practice does not produce it, the trier of fact can find breach on inference from the outcome combined with surrounding evidence. The case is a useful reference for any family considering whether to pursue a claim arising from a surgical complication that everyone agrees is unusual.

The key questions to assess at the outset:

• How rare is the complication? Truly rare (the plaintiff’s expert in Dallner had never seen it) is a different category from uncommon-but-recognized.
• What does the surgeon’s own contemporaneous documentation say? A surgeon who documented an honest assessment of the mechanism in the immediate aftermath is providing potentially decisive evidence.
• What does the mechanism analysis support? The expert evidence on how the injury could have occurred is part of the foundation for any breach analysis.

For more on how I evaluate surgical malpractice cases generally, see Suing for Medical Malpractice in Ontario: What You Need to Know and the firm’s Surgical Error practice page.

For surgeons and clinical leaders. Several observations from the other side of the file:

• Document your own assessment of the mechanism. A surgeon’s contemporaneous operative addendum that honestly identifies the most likely cause of an adverse event is the right clinical practice. It supports patient care, supports professional integrity, and is legally consistent with the disclosure framework. Dallner illustrates that such documentation can also become evidence in subsequent litigation; that is a feature of the framework, not a bug.
• The “complications happen” defence is bounded. It is a legitimate part of the surgical malpractice framework, and it succeeds in many cases. But it does not succeed where the complication is extreme and rare to the point of being outside the catalogue of recognized adverse events. Where the complication is one your expert has never seen, the defence position becomes much harder.
• Calibrate the force and duration. Modern shoulder arthroplasty technique includes deliberate attention to the magnitude and duration of stretch. The technique exists because the brachial plexus is anatomically vulnerable and because stretch injuries are mechanically possible. Adhering to the technique is the operational expression of the standard of care.
• Operative documentation matters. The operative note is your contemporaneous account of what happened. It should reflect the actual events: the positioning, the duration of retraction, any unusual features of the anatomy, any adjustments made during the procedure. Subsequent legal analysis will turn substantially on what the operative documentation shows.

Cluster integration

The plaintiff-success cluster now includes:

• Kotorashvili v Lee (orthopedic, administrative-clinical workflow)
• Henry v Zaitlen (delayed referral, Court of Appeal affirmance)
• Denman v Sabapathy (informed consent)
• Hemmings v Peng (obstetric anaesthetic, $12M, Court of Appeal affirmance)
• Hasan v Trillium Health Centre (stroke, Court of Appeal affirmance, Snell framework)
• Gumbley v Vasiliou (asthma intubation delay)
• Dallner v Gladwell (orthopedic brachial plexus injury, “inferring negligence from rare complication” framework)

The bidirectional SOC-causation framework cluster:

• Armstrong v Royal Victoria Hospital, 2019 ONCA 963 (foundational appellate authority)
• Bafaro v Dowd, 2010 ONCA 188 (the prior framework)
• Dallner v Gladwell (application of the Armstrong refinement)

The “inference from outcome” framework:

• Dallner v Gladwell (principal cluster authority on inferring breach from a rare complication)
• Connects to the Snell v Farrell “robust and pragmatic” causation framework in Lorencz v Talukdar and Hasan v Trillium Health Centre
• Distinct from the Snell framework in that Dallner applies the inferential framework on the SOC question rather than the causation question

The “bad outcome alone” doctrine:

• A foundational principle in Canadian malpractice law
• Dallner v Gladwell illustrates the bounded operation of the principle — bad outcome alone does not establish negligence, but a bad outcome of a type that does not occur with competent care can support the inference

Orthopedic / surgical error practice area:

• Dallner v Gladwell is now the principal cluster authority on shoulder arthroplasty
• Connects to Kotorashvili v Lee (clavicle revision)
• Anchors the Surgical Error practice page with substantive case content

---

Decision Date: November 25, 2024

Jurisdiction: Ontario Superior Court of Justice

Citation: Dallner v Gladwell, 2024 ONSC 6557 (CanLII)

Outcome: Judgment for plaintiff on liability. Damages had been settled prior to trial. The trial judge found that Dr. Gladwell breached the standard of care by applying an excessive degree or duration of force or stretch on the patient’s arm during the surgery, resulting in a permanent injury to the medial cord of the brachial plexus. The conclusion on breach was supported by the extreme rarity of the complication, the surgeon’s contemporaneous operative addendum identifying intraoperative stretch as the most likely cause, and the expert evidence on the mechanism of brachial plexus stretch injury.

Key authorities: Armstrong v Royal Victoria Hospital, 2019 ONCA 963 (bidirectional SOC-causation framework; injury as evidence of “what happened”); Bafaro v Dowd, 2010 ONCA 188 (the typical order of analysis); Meringolo (Committee of) v Oshawa General Hospital, [1991] OJ No 91 (CA); Grass (Litigation guardian of) v Women’s College Hospital, 2001 CanLII 8526 (ON CA); Kennedy v Jackiewicz, [2003] OJ No 1854 (SCJ), affirmed [2004] OJ No 4816 (CA).