Graham v Bridgepoint Health: Osteomyelitis, Hospitalist Standard of Care, and Patient Treatment Decisions

Some malpractice trials are defendant-favourable on a single dispositive issue. The standard of care is met. Or causation fails. Or damages can be limited. Graham v Bridgepoint Health, 2025 CanLII 380 (ON SC), released by the Ontario Superior Court of Justice on January 2, 2025, is the rarer case in which the defendant prevails on every contested issue. The standard of care for the defendant’s role as a hospitalist was met. Causation failed on three independent grounds: the timing did not work because the bone infection likely originated before the defendant’s involvement in the patient’s care; the medical treatment would have been the same regardless of timing; and the patient’s own decision to decline recommended surgery contributed to the prolongation of recovery. The plaintiff also achieved a full recovery, with the tibial fracture united and the plaintiff returning to demanding professional roles, which bore on damages.

For prospective clients, the case is doctrinally useful as an illustration of how a malpractice action can fail despite a serious underlying injury. Mr. Graham was the victim of a catastrophic motorcycle accident with open fractures in three locations. He developed an infection. The infection produced osteomyelitis (bone infection). He had a complicated and prolonged recovery. By the metric of “bad outcome,” this is a case where one might expect significant damages exposure. But the legal framework requires more than a bad outcome. It requires breach of a specific standard of care, causation by that breach, and damages that flow from the breach rather than from the underlying injury. The trial judge found that the defendant met his SOC, that the breach (if any) did not cause the harm, and that the plaintiff’s eventual recovery was not legally attributable to the defendant’s conduct. The case was dismissed.

The case is also a useful introduction to the hospitalist medicine and post-operative infection contexts, both of which are recurring areas in modern Ontario malpractice litigation. It introduces a new substantive practice area to the cluster (osteomyelitis / orthopedic post-operative infection) and a new clinical context (hospitalist medicine / post-acute rehabilitation).

The facts

The accident. Scott Graham suffered a serious motorcycle accident on July 1, 2012. He sustained:

• Open fractures of the left tibia (the larger of the two lower leg bones; “open” means the bone broke through the skin, creating an immediate risk of contamination and infection)
• A fractured right forearm
• A burst fracture of the T12 vertebra (the lowest of the thoracic vertebrae; a “burst” fracture involves multiple fragments of the vertebral body)

This was a multi-trauma presentation requiring urgent care across orthopedic, spinal, and (potentially) neurological dimensions.

The initial surgical care. Mr. Graham received initial treatment at St. Michael’s Hospital in Toronto. An urgent orthopedic operation was performed by the treating surgeon to address the open tibial fracture. The procedure included fixation of the bone and management of the open wound. Open fractures carry a high baseline risk of infection because the bone has been directly exposed to environmental contamination at the time of injury; the standard surgical approach involves debridement (removal of devitalized tissue), thorough irrigation, and prophylactic antibiotics.

The bloodstream infection. After the initial surgery, Mr. Graham developed fever and a confirmed bloodstream infection. A bloodstream infection in the post-operative period is a recognized risk for any patient with an open fracture; the bacteria can enter the bloodstream directly through the wound or indirectly through the surgical site. This bloodstream infection became central to the subsequent causation analysis.

The transfer to Bridgepoint Hospital. Once Mr. Graham was clinically stable, he was transferred to Bridgepoint Hospital in Toronto for orthopedic rehabilitation. Bridgepoint is a post-acute rehabilitation hospital that provides extended care for patients recovering from major trauma, surgery, and acute illness. It is part of the Sinai Health system. Patients at Bridgepoint are typically past the acute phase of their illness or injury but require continued inpatient care for rehabilitation, mobilization, and ongoing medical management.

Mr. Graham’s care at Bridgepoint was provided by Dr. Berger, a family physician practising as a hospitalist. (A hospitalist is a physician whose practice is dedicated to the inpatient care of hospitalized patients; in rehabilitation hospitals, hospitalists provide the general medical oversight while subspecialist care is provided on a consultation basis.) The treating orthopedic team at St. Michael’s remained involved in directing the orthopedic plan; Dr. Berger’s role was to implement that plan and manage the day-to-day care.

The emerging lump on the tibia. During Mr. Graham’s stay at Bridgepoint, a lump developed over the surgical wound on his tibia. Dr. Berger monitored the lump. Initially the lump appeared to respond to conservative measures including warm compresses. The clinical picture at this point did not include the classic signs of infection (redness, heat, drainage, systemic features). Dr. Berger sought the advice of the orthopedic surgeon from St. Michael’s, who visited the patient at Bridgepoint to assess the lump.

The osteomyelitis diagnosis. The visiting orthopedic surgeon diagnosed the lump as an infection (osteomyelitis — bone infection). Mr. Graham was transferred back to St. Michael’s Hospital for management.

The declined surgery and the second opinion. At St. Michael’s, the treating orthopedic surgeon recommended surgical irrigation and debridement to clear the infection. Mr. Graham declined this recommendation initially. He opted instead for local irrigation at bedside (a less invasive approach that involves washing out the wound at the bedside without taking the patient to the operating theatre). He returned to Bridgepoint, where the treatment continued under the orthopedic team’s orders.

A piece of blue suture was found and removed from the wound. The antibiotic regimen was adjusted to Ampicillin based on sensitivity testing of the identified bacteria.

The treating orthopedic surgeon repeatedly recommended surgical intervention. Mr. Graham continued to decline until, after a second opinion, he accepted surgery.

The recovery. Following surgery, Mr. Graham was managed with intravenous and oral antibiotics over several months. He ultimately recovered at home. His tibial fracture healed completely. He returned to his work as a Crown Attorney in January 2013, approximately six months after the accident. He gradually resumed his military duties.

The action. Mr. Graham brought a malpractice action against multiple defendants including the various physicians and hospitals involved in his care. By the time of trial, all other defendants had been released. Dr. Berger remained as the sole defendant. The trial proceeded on all three issues: standard of care, causation, and damages.

Hospitalist medicine and the clinical context

A few clinical and structural observations on hospitalist medicine, because the framework is central to the case.

The hospitalist role. Hospitalists are physicians who practise primarily in the inpatient setting. The role is now well-established in Canadian hospital medicine, particularly in academic and tertiary care centres. Hospitalists typically come from family medicine backgrounds and provide general medical care for patients admitted to hospital, while subspecialist physicians are consulted as needed for specific clinical questions in their fields.

In a post-acute rehabilitation hospital like Bridgepoint, hospitalists are typically responsible for:

• The day-to-day medical management of inpatients
• Monitoring clinical progress and adjusting treatment plans
• Coordinating with consulting specialists (orthopedic surgery, infectious diseases, internal medicine, and others)
• Managing acute clinical changes
• Discharge planning

The standard of care for a hospitalist. The standard of care for a hospitalist is the standard of a reasonable family physician practising hospital medicine in the relevant setting. It is not the standard of an orthopedic surgeon, an infectious diseases specialist, or any other subspecialty. The hospitalist is expected to recognize when subspecialty input is needed and to obtain it appropriately, but is not expected to substitute their own judgment for the subspecialty input on questions within the subspecialist’s domain.

The Supreme Court of Canada’s articulation in Wilson v Swanson, [1956] SCR 804, of the standard for specialists (“the average of competent practitioners of that branch of the profession”) applies analogously to hospitalists. The standard is the standard of the role, not the standard of the consulting specialists.

The consultative framework. Where a hospitalist faces a clinical question outside their general medical scope, the appropriate response is to consult the relevant subspecialist. The consultation framework operates as follows:

• The hospitalist identifies the clinical question
• The hospitalist arranges for the subspecialist to assess the patient or provide written input
• The subspecialist provides their recommendations
• The hospitalist implements the recommendations as part of the patient’s overall care plan
• The subspecialist remains available for further input as the clinical picture evolves

The framework is fundamental to modern hospital medicine. It allows the hospitalist to provide comprehensive care across multiple clinical dimensions while ensuring that subspecialty expertise is available where needed. The framework was operating in Mr. Graham’s case: Dr. Berger consulted the orthopedic surgeon when the lump appeared; the orthopedic surgeon assessed the patient; the assessment led to the diagnosis of osteomyelitis and the transfer back to acute care.

Osteomyelitis and the post-operative infection context

A brief clinical overview of osteomyelitis is useful for understanding the causation analysis.

The disease. Osteomyelitis is an infection of bone. It can develop through several mechanisms:

• Direct inoculation — bacteria are introduced directly into the bone during trauma or surgery (the typical mechanism in open fractures)
• Hematogenous spread — bacteria circulating in the bloodstream seed the bone (particularly common in pediatric patients and in patients with bloodstream infections)
• Contiguous spread — bacteria spread from adjacent soft tissue infection into the bone

The clinical course. Acute osteomyelitis typically presents with localized pain, swelling, erythema, and sometimes systemic features (fever, elevated white blood cell count, elevated inflammatory markers). Chronic osteomyelitis may present more subtly with a draining sinus or non-healing wound. Imaging (MRI, bone scan, X-ray) and microbiological sampling are typically required for definitive diagnosis.

The treatment. Treatment of osteomyelitis typically requires:

• Prolonged antibiotic therapy (often six weeks or more of intravenous antibiotics, followed by oral antibiotics)
• Surgical debridement to remove devitalized bone and infected hardware where present
• Stabilization of the bone if not already addressed
• Management of any contributing factors (diabetes, vascular insufficiency, foreign bodies)

The combination of surgical and medical management is typical; antibiotic therapy alone is generally inadequate for established osteomyelitis because the infected bone tissue has limited blood supply and the antibiotics cannot reliably penetrate.

The post-operative infection risk. Patients with open fractures have a substantial baseline risk of osteomyelitis because the bone has been directly exposed to contamination. The standard surgical approach (debridement, irrigation, prophylactic antibiotics) reduces but does not eliminate the risk. Bloodstream infections in the post-operative period add a separate pathway: bacteria circulating in the blood can seed the freshly operated bone and produce hematogenous osteomyelitis on top of any direct contamination.

The clinical context is important for the Graham causation analysis. Mr. Graham had multiple risk factors for osteomyelitis: open tibial fracture (direct contamination risk), surgical fixation hardware (foreign body that bacteria can colonize), post-operative bloodstream infection (hematogenous seeding pathway). The infection could have originated at the time of the initial trauma, at the time of the surgery, or as a downstream consequence of the bloodstream infection. These are all etiologies that pre-date the patient’s transfer to Bridgepoint and Dr. Berger’s involvement in the care.

The legal framework — standard of care

The trial judge found that Dr. Berger met the standard of care. The reasoning, distilled.

Adherence to the treatment plan. Dr. Berger had implemented the treatment plan as directed by the treating orthopedic team. The plan involved continued inpatient care, monitoring of the surgical wound, mobilization as tolerated, and standard post-operative medical management. Dr. Berger’s adherence to the plan was not in dispute.

Monitoring of clinical progress. Dr. Berger had monitored the patient’s clinical progress throughout the admission, including the surgical wound on the tibia. The medical records reflected ongoing assessments. When the lump appeared, Dr. Berger documented it and continued to monitor.

Appropriate consultation when the clinical picture changed. When the lump did not resolve and the clinical picture suggested potential infection, Dr. Berger sought consultation from the orthopedic surgeon from St. Michael’s. The consultation produced the diagnosis of osteomyelitis and the transfer back to acute care. The consultation framework operated as expected.

Expert evidence supporting the SOC analysis. The defence retained two experts: a hospitalist (a retired family physician who had practised for over 40 years as a hospitalist in a rehabilitation hospital similar to Bridgepoint) and an infectious diseases specialist. Both experts supported the conclusion that Dr. Berger’s management of the patient met the standard of care for a family physician acting as a hospitalist.

The court’s credibility findings. The plaintiff’s allegations included challenges to Dr. Berger’s account of events. The trial judge rejected those challenges. The court found that the evidence supported Dr. Berger’s narrative. The credibility finding is itself an important feature of the case: in any malpractice trial where the plaintiff alleges that the defendant’s account of the clinical decisions is unreliable, the trial judge’s assessment of credibility is dispositive.

The trial judge accordingly found that the plaintiff had not established a breach of the standard of care. The case could have been dismissed on this finding alone. But the trial judge also addressed causation.

The legal framework — causation

The trial judge addressed causation on three independent grounds.

Ground 1: The osteomyelitis likely originated before the defendant’s involvement. The plaintiff’s theory was that the osteomyelitis was caused by an undetected suture abscess (a localized infection forming around the surgical sutures). The defence theory was that the osteomyelitis was the result of pre-existing bacteremia (the documented post-operative bloodstream infection) that had spread to the bone before Mr. Graham was transferred to Bridgepoint and came under Dr. Berger’s care.

The defence theory was supported by the defence infectious diseases expert, who testified that:

• The bone infection most likely originated through hematogenous spread from the post-operative bloodstream infection
• The timing supported this etiology (the bloodstream infection pre-dated the Bridgepoint admission)
• The clinical course was consistent with hematogenous osteomyelitis becoming clinically apparent during the Bridgepoint stay rather than originating during the Bridgepoint stay

If the osteomyelitis originated before Dr. Berger’s involvement, then no action or inaction by Dr. Berger during the Bridgepoint admission could have caused the infection. The causation analysis fails at the threshold.

Ground 2: The treatment would have been the same regardless of the timing of the diagnosis. Even on the plaintiff-favourable assumption that the infection had developed during the Bridgepoint admission, the defence argued that the medical treatment of the established osteomyelitis would have been the same regardless of when it was diagnosed. The standard treatment (prolonged antibiotics with eventual surgical debridement) was the treatment that was ultimately provided. The defence infectious diseases expert testified that earlier diagnosis would not have altered the course of treatment or the outcome.

The framework is a variation on the “counterfactual would not have helped” principle now well-established in the cluster. In Yang v Freed, the counterfactual failed because the relevant pathology was not yet present at the time of the breach. In Lorencz v Talukdar, the counterfactual failed because the wait times made the resulting intervention speculative. In Graham v Bridgepoint Health, the counterfactual failed because the treatment that would have followed from earlier diagnosis was the same treatment that was ultimately provided. The framework operates wherever the plaintiff cannot establish that a different timing would have produced a different outcome.

Ground 3: The plaintiff’s own conduct contributed to the prolongation of recovery. The plaintiff had declined the recommended surgical irrigation and debridement on multiple occasions. The surgery was eventually accepted after a second opinion. The defence position was that the period of declined surgery represented the patient’s own contribution to the prolongation of recovery rather than any conduct by Dr. Berger.

The principle that a plaintiff’s own conduct can defeat a malpractice claim is well-established in Canadian tort law. The framework operates through several distinct doctrinal mechanisms:

• Failure to mitigate — a plaintiff who fails to take reasonable steps to mitigate damages cannot recover for harm that the mitigation would have prevented
• Patient autonomy and informed refusal — a patient who, after being properly informed, declines a recommended treatment cannot subsequently complain that the treatment was not provided
• Contributory conduct affecting causation — where the plaintiff’s own conduct is part of the causal chain leading to the harm, that conduct affects the causation analysis

In Graham v Bridgepoint Health, the plaintiff’s decision to decline surgery was the patient’s own informed choice. The choice slowed his recovery. The defence successfully argued that the consequences of that choice could not be transferred to Dr. Berger.

The cumulative effect. The trial judge accepted all three grounds. Even taking the plaintiff’s case at its highest, causation was not established. The case was accordingly dismissed.

The full recovery factor

A final observation on the damages dimension of the case.

Mr. Graham achieved a full recovery. His tibial fracture united. The infection resolved. He returned to his work as a Crown Attorney in January 2013 (approximately six months after the accident) and gradually resumed his military duties. Both of those professional roles are demanding: a Crown Attorney’s work involves complex litigation, court appearances, and substantial cognitive demands; military service involves physical capacity, mobility, and operational responsibilities.

The full recovery is a feature of the case that bears on damages. While the trial judge did not need to assess damages because liability was not established, the broader principle is relevant for prospective clients evaluating their own cases. Damages in a malpractice case are tied to the consequences of the breach, not to the underlying injury. Where the patient achieves a full recovery, the damages available are limited to the costs and losses sustained during the recovery period. Significant ongoing damages (loss of future earning capacity, future care costs, ongoing pain and suffering) require a permanent or substantially prolonged impairment.

The framework was relevant to the strategic posture of the case. A plaintiff seeking damages for a prolonged recovery from osteomyelitis who has ultimately fully recovered to demanding professional roles is in a different posture than a plaintiff seeking damages for permanent disability. The damages framework is narrower; the plaintiff must establish that the breach caused a specific period of recovery prolongation, and the costs and losses associated with that specific period.

Doctrinal anchors

Several doctrinal anchors emerge from the case.

The defendant trial win on both SOC and causation. Graham v Bridgepoint Health is a clean example of a case where the defendant prevails on every contested issue. The pattern is distinct from the breach-without-causation cases in the cluster (Williamson, Papineau, Lorencz, Yang), where the SOC was breached but causation failed. In Graham, the SOC was met AND causation failed. The doctrinal point: a malpractice action can fail on either ground, and where both grounds operate in the defendant’s favour, the dismissal is structurally robust.

The hospitalist standard of care framework. Graham v Bridgepoint Health is the first cluster case to articulate the standard of care for hospitalists in post-acute rehabilitation. The framework: the standard is the standard of a reasonable family physician practising hospital medicine in the relevant setting, not the standard of any subspecialist. The hospitalist is expected to recognize when subspecialty input is needed and to obtain it appropriately, but is not expected to substitute their own judgment for the subspecialty input on questions within the subspecialist’s domain.

The “pre-existing condition” causation defence. The defence successfully argued that the osteomyelitis originated before Dr. Berger’s involvement in the care. The framework is generalizable to any case where the plaintiff’s harm has multiple potential etiologies and the alleged breach is only one of them. The plaintiff must establish on the balance of probabilities that the breach was the operative cause; where the harm could equally have originated through a pathway that pre-dates the breach, the causation analysis can fail.

The “treatment would have been the same” framework. This is a third variation on the counterfactual reconstruction principle now well-established across the cluster. Where the treatment that would have followed from earlier diagnosis is the same treatment that was ultimately provided, the breach does not connect to a different outcome. The principle joins the Yang v Freed (temporal mismatch) and Lorencz v Talukdar (wait times) variations as a third anchor in the cluster’s treatment of the counterfactual reconstruction framework.

The “patient’s own conduct” framework. The plaintiff’s decision to decline recommended surgery contributed to the prolongation of recovery. The framework operates through failure-to-mitigate principles, informed refusal of treatment, and contributory conduct affecting causation. The principle is doctrinally important because it places limits on the transfer of consequences from the patient’s own choices to the defendant.

The multi-defendant release pattern. All other defendants had been released prior to trial; Dr. Berger remained as the sole defendant. The pattern is doctrinally common in multi-defendant malpractice cases and reflects strategic decisions about where to focus the claim. The implication for prospective clients: complex malpractice cases often involve initial broad pleadings followed by progressive narrowing of the defendant pool as the evidence develops. Settlement, release, or discontinuance against some defendants is a routine feature of the litigation pathway.

The “full recovery” factor in damages. Mr. Graham’s full recovery to demanding professional roles bears on the damages framework. The principle: damages are tied to the consequences of the breach, not to the underlying injury. A plaintiff who has fully recovered is limited to damages associated with the recovery period itself.

Why this case matters

For prospective clients. Graham v Bridgepoint Health illustrates several important practical points about how malpractice cases operate.

A bad outcome is not the same as a malpractice case. Mr. Graham suffered a serious motorcycle accident with major orthopedic injuries, developed a bloodstream infection, suffered osteomyelitis, had a prolonged recovery, and required eventual surgical intervention. By the metric of “bad outcome,” this could appear to be a major malpractice case. The legal analysis required substantially more than the bad outcome. The plaintiff needed to establish a specific breach by a specific defendant, causation by that breach, and damages flowing from the breach. The plaintiff did not succeed on any of these requirements.

Pre-existing conditions can defeat causation. Where the harm has multiple potential etiologies and only one of them is attributable to the defendant, the causation analysis can fail. The plaintiff must establish on the balance of probabilities that the defendant’s conduct was the operative cause. The framework was central to the Graham defence and is a recurring feature of post-operative infection cases generally.

Patient choices have legal consequences. A patient who is properly informed and who declines a recommended treatment cannot subsequently transfer the consequences of that choice to the treating physician. The principle is doctrinally well-established and operates through multiple mechanisms (failure to mitigate, informed refusal, contributory conduct). It is a significant strategic consideration in any case where the patient’s own decisions were part of the clinical course.

Full recovery limits damages even where liability is established. The damages framework is tied to the consequences of the breach. Where the patient has fully recovered, the available damages are limited to the recovery period itself. This is not a defence to liability but it is a substantial constraint on the case’s strategic value.

For more on the general framework for evaluating these cases, see Suing for Medical Malpractice in Ontario: What You Need to Know and the firm’s hospital negligence practice page.

For hospitalists and clinical teams. A few practical observations:

Document the rationale for clinical decisions. Where a clinical question is referred to a subspecialist, document the question, the referral, the subspecialist’s response, and the implementation of the response. The documentation establishes that the consultative framework was operating as expected. Graham v Bridgepoint Health turned in part on the documentation of Dr. Berger’s consultation with the orthopedic surgeon when the lump appeared.

Maintain clear lines between hospitalist and subspecialist responsibility. The standard of care for a hospitalist is the standard for a family physician practising hospital medicine. It is not the standard for the consulting subspecialists. Where the clinical question is within the subspecialist’s domain, the hospitalist should obtain the subspecialist’s input and implement it rather than substituting independent judgment. The framework protects both the patient and the hospitalist.

Respect patient autonomy and document informed refusals. Where a patient declines a recommended treatment after being properly informed, document the discussion, the information provided, the patient’s reasoning, and the patient’s decision. The documentation is essential to the subsequent legal analysis and to the patient’s own care.

Cluster integration

The defendant-win cluster (full):

• Pellerin v Balfour (BC ER appendicitis)
• Sutherland v Encompass Health (Ontario stroke)
• Williamson v Y (BC anaesthesia)
• Papineau v Romero-Sierra (Ontario delayed Lyme)
• Noel v Hawrylyshyn (Ontario birth injury)
• Lorencz v Talukdar (Saskatchewan failure-to-refer)
• Yang v Freed (Alberta critical care)
• Graham v Bridgepoint Health (Ontario hospitalist post-operative infection)

Eight documented defendant-win cases now anchor the cluster. The pattern is well-developed across multiple jurisdictions and clinical contexts.

The counterfactual reconstruction framework:

• Yang v Freed (temporal mismatch — pathology not yet present)
• Lorencz v Talukdar (wait times speculative)
• Graham v Bridgepoint Health (same treatment would have followed)

Three variations of the counterfactual reconstruction defence now anchored in the cluster.

The “patient’s own conduct” framework (new for the cluster):

• Graham v Bridgepoint Health is the principal cluster authority
• Connects to failure-to-mitigate, informed refusal, and contributory conduct doctrines

The “pre-existing condition” causation defence (new anchor):

• Graham v Bridgepoint Health is the principal cluster authority
• Generalizable to any case with multiple potential etiologies for the alleged harm

The hospitalist medicine practice area (new for the cluster):

• Graham v Bridgepoint Health is the principal cluster authority
• Distinct from the existing cluster on specialist standards of care

The osteomyelitis / post-operative infection practice area (new):

• Graham v Bridgepoint Health is the principal cluster authority
• Distinct from the medication error and surgical error practice areas

The multi-defendant release pattern:

• Graham v Bridgepoint Health illustrates the common pattern of progressive narrowing of the defendant pool through the pre-trial period

---

Decision Date: January 2, 2025

Jurisdiction: Ontario Superior Court of Justice

Citation: Graham et al v Bridgepoint Health et al, 2025 CanLII 380 (ON SC)

Outcome: Action dismissed. The trial judge found that the defendant family physician, practising as a hospitalist at Bridgepoint Hospital, met the standard of care in the management of the plaintiff’s post-operative care following transfer from St. Michael’s Hospital. The trial judge further found that the plaintiff had not established causation on three independent grounds: (a) the osteomyelitis most likely originated through hematogenous spread from a pre-existing post-operative bloodstream infection that pre-dated the defendant’s involvement in the care; (b) the medical treatment of the established osteomyelitis would have been the same regardless of when it was diagnosed; and (c) the plaintiff’s own decision to initially decline recommended surgical intervention contributed to the prolongation of recovery and could not be transferred to the defendant. The plaintiff ultimately achieved a full recovery with healing of the tibial fracture, resolution of the infection, and return to demanding professional roles as a Crown Attorney and military officer.

Key authorities (implicit in the analysis): Wilson v Swanson, [1956] SCR 804 (standard for specialists, applied analogously to hospitalists); Clements v Clements, 2012 SCC 32 (but-for causation framework); Snell v Farrell, [1990] 2 SCR 311 (robust and pragmatic causation); failure-to-mitigate and informed refusal doctrines under general tort law.