Barnett v Chelsea & Kensington HMC

Barnett v Chelsea & Kensington Hospital Management Committee [1969] 1 QB 428 is a foundational case in the English law of tort, decided by Nield J in the Queen's Bench Division. The case arose from a tragic set of circumstances involving three nightwatchmen who had been on duty on New Year's Eve and who, in the early hours of New Year's Day, began to experience severe vomiting after drinking tea. The men attended the casualty department of Chelsea and Kensington Hospital, plainly unwell and seeking medical attention.

Upon arrival at the casualty department, the three men were seen by a nurse, who relayed their complaints to the casualty officer on duty. The casualty officer, who was himself unwell at the time, did not attend to examine the men in person. Instead, he sent a message instructing the men to go home and to consult their own general practitioners if their symptoms persisted. No physical examination was carried out, no diagnosis was attempted, and no treatment was offered or administered.

Within a matter of hours, all three men died. Post-mortem examination revealed that the cause of death in each case was arsenical poisoning. It transpired that the tea the men had consumed during the night had been contaminated with arsenic. The question of how the arsenic came to be present in the tea was not in dispute for the purposes of the litigation; what mattered was the response of the hospital to the men's presentation at the casualty department.

An action was brought by the widow of one of the deceased men, Mr Barnett, against the Hospital Management Committee as the body responsible for the hospital and its staff. It was common ground between the parties — and indeed was conceded by the defendant — that the casualty officer had owed a duty of care to the men who had attended the casualty department. It was further conceded that the casualty officer had acted in breach of that duty by failing to examine or treat the patients in accordance with the standard of care required of a reasonably competent medical practitioner.

The sole live issue in the case, therefore, was not whether there had been negligence — that was admitted — but whether that negligence had caused the death of Mr Barnett. The claimant needed to demonstrate not only that there was a duty of care and a breach of that duty, but also that the breach was the factual cause of the harm suffered. It was on this third element of the tort of negligence that the claim ultimately foundered.

Medical evidence was adduced at trial addressing the state of knowledge concerning arsenical poisoning and the treatment options that would have been available to clinicians in the early 1960s. Expert evidence indicated that by the time the men had presented at the casualty department, the arsenic they had absorbed had already reached a lethal concentration within their bodies, and that no treatment then available could have reversed the poisoning or saved their lives. This expert evidence proved decisive to the court's analysis of causation.

Given that both duty of care and breach of duty were conceded by the defendant, the sole issue for the court's determination was whether the admitted breach of duty by the casualty officer had caused, in law and in fact, the death of the claimant's husband. More precisely, the court was required to determine whether the claimant had discharged the burden of proving that, but for the defendant's negligent failure to examine and treat Mr Barnett, he would not have died from arsenical poisoning.

Subsidiary to this central question was the related inquiry into the appropriate legal test for factual causation in medical negligence cases, and in particular whether the standard "but for" formulation was capable of satisfying the requirements of justice in a case where a defendant's negligence had indisputably occurred but where the counterfactual outcome remained uncertain. The case thus invited the court to consider the relationship between moral culpability and legal liability in the context of professional negligence.

Nield J commenced his analysis by setting out the well-established three-part structure of the tort of negligence: the claimant must prove the existence of a duty of care, a breach of that duty, and damage caused by the breach. Authority for the fundamental proposition that a duty of care arises where harm to the claimant is reasonably foreseeable had long been established by the House of Lords in Donoghue v Stevenson [1932] AC 562, and the existence of a duty of care owed by a hospital's casualty department to a patient presenting for treatment was not in doubt. The court accordingly treated the duty question as settled and turned its attention to causation.

In addressing the standard of care, Nield J noted the framework articulated in Bolam v Friern Hospital Management Committee [1957] 1 WLR 582, which established that a medical practitioner is not negligent if he acts in accordance with a responsible body of medical opinion. However, since breach was conceded in the present case, the Bolam standard did not require extended analysis. Its relevance lay rather in providing the backdrop against which the casualty officer's conduct was measured, and the concession of breach meant that the officer's failure to attend and examine the patients fell below even the most permissive formulation of the standard of care applicable to casualty officers.

The principal focus of the court's reasoning concerned factual causation. Nield J adopted the classic "but for" test, which requires a claimant to establish that the damage complained of would not have occurred but for the defendant's negligence. This formulation had received clear endorsement in the Court of Appeal in Cork v Kirby MacLean Ltd [1952] 2 All ER 402, where Denning LJ articulated the test in direct and accessible terms: if the damage would have occurred in the same way whether or not the defendant had been negligent, the negligence is not a cause of the damage. Nield J applied this test with rigour to the facts before him.

The court considered the medical evidence with care. Expert testimony established that arsenical poisoning of the severity suffered by the three men was, by the time they presented at the hospital, irreversible. The dosage of arsenic that each man had absorbed through the contaminated tea was so great, and the time that had elapsed since ingestion so significant, that no treatment then available in a hospital casualty setting could have altered the outcome. Chelation therapy and other potential interventions were either unavailable or, on the expert evidence, could not have been administered with sufficient speed or in sufficient quantity to prevent death.

Applying the but-for test to these facts, the court found that the claimant could not satisfy the causal requirement. Even if the casualty officer had examined the men immediately, even if he had correctly diagnosed arsenical poisoning, and even if he had administered whatever treatment was then available, the men would still have died. The negligent failure to act had not, therefore, made any difference to the outcome. The death of Mr Barnett was not caused by the breach of duty; it was caused by the arsenic poisoning, which had by then progressed to a fatal and irreversible stage.

The court also considered the approach taken in Bonnington Castings Ltd v Wardlaw [1956] AC 613, a decision of the House of Lords in which it had been held that a defendant's breach of duty need only be shown to have materially contributed to the claimant's injury, rather than being the sole cause. However, Nield J found that the Bonnington Castings principle did not assist the claimant in the present case. That principle operates where a disease or injury is the cumulative product of both tortious and non-tortious causes, and where the defendant's contribution, even if not the exclusive cause, has meaningfully added to the harm. Here, however, the position was categorically different: no treatment that the hospital could have provided would have contributed to the men's survival. The defendant's breach had contributed nothing to the progression of the poisoning; it had simply failed to interrupt a process that was already beyond interruption.

Nield J acknowledged the profound moral dimension of the case. It was clear that the casualty officer's conduct had been seriously deficient: a doctor who declines to see and examine patients presenting at a casualty department with acute symptoms is failing in an elementary professional obligation. The court entertained no doubt that the officer's conduct was morally blameworthy. Yet the court was equally clear that moral culpability does not, of itself, generate legal liability. Liability in negligence requires proof of all three elements, and causation is not a technicality to be waived where the defendant's conduct is particularly egregious.

The court further observed that the burden of proving causation rests on the claimant, not on the defendant. It is for the party asserting that the breach caused the damage to demonstrate that proposition on the balance of probabilities. In a medical negligence case, this typically requires the claimant to show that proper treatment would, on the balance of probabilities, have prevented or materially reduced the harm suffered. On the facts of Barnett, the expert medical evidence did not support that conclusion. The evidence went no further than establishing that there was a possibility that treatment might have prolonged life briefly; it did not establish on the balance of probabilities that proper treatment would have prevented death.

The court was careful to distinguish between cases in which a defendant's negligence deprives a claimant of a chance of recovery, and cases in which the evidence establishes that no such chance existed. In the former category, questions about the recovery of damages for loss of a chance might arise; in the latter, where no realistic prospect of a better outcome existed, the but-for test is simply not satisfied and no liability arises. The facts of Barnett fell within the latter category: the medical evidence demonstrated not merely that recovery was uncertain but that it was effectively impossible given the state of the poisoning at the relevant time.

Nield J was also alert to the systemic implications of the claimant's argument. If liability were to be imposed wherever a defendant had acted in breach of duty, regardless of whether that breach had any causal connection to the harm suffered, the tort of negligence would be transformed in a way that would impose liability without limit and would undermine the principled distinction between tortious responsibility and mere moral fault. The requirement of causation serves a legitimate function in confining liability to cases where the defendant's conduct has genuinely contributed to the harm complained of.

Nield J held that, although the defendant hospital had owed the deceased a duty of care and the casualty officer had acted in breach of that duty by failing to examine or treat the men, the claimant had failed to establish that the breach of duty had caused the death of Mr Barnett. Applying the but-for test, the court found that Mr Barnett would have died from arsenical poisoning even if he had received the best treatment available at the time, since the poisoning was already irreversible when he presented at the casualty department. Judgment was accordingly entered for the defendant.

The court confirmed that all three elements of the tort of negligence — duty, breach, and causation — must be independently established by a claimant. The concession of duty and breach was insufficient to ground liability without proof of causation. The claimant's action failed not because the defendant had done nothing wrong, but because the wrong done had not been the legal cause of the damage complained of.

Barnett v Chelsea & Kensington HMC occupies a central position in the academic and practitioner treatment of causation in the law of tort. It is frequently cited as the paradigm illustration of the but-for test in operation and as a demonstration that the test can produce outcomes that are morally uncomfortable but legally correct. The case is routinely used in legal education to emphasise that negligence is not synonymous with liability: a defendant may behave badly and yet escape legal responsibility if the claimant's harm would have occurred regardless.

The decision has been consistently applied in subsequent medical negligence litigation as authority for the proposition that a claimant must prove on the balance of probabilities that the defendant's breach caused or materially contributed to the harm. It has been considered alongside later House of Lords authority, including Wilsher v Essex Area Health Authority [1988] AC 1074, which reaffirmed that where a claimant cannot establish which of several potential causes actually produced the harm, the but-for test is not satisfied. Barnett thus stands as part of a coherent line of authority insisting on rigorous causal proof in medical negligence cases.

The case also provides important context for understanding the subsequent development of exceptional causal doctrines, such as the material increase in risk approach endorsed by the House of Lords in McGhee v National Coal Board [1972] 3 All ER 1008 and the modified but-for approach applied in Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22. These later authorities develop special rules for cases where the causal process is scientifically indeterminate and where strict application of the but-for test would leave all claimants without a remedy. Barnett represents the baseline from which those exceptions depart: a case in which the causal evidence is determinate and the but-for test yields a clear, if harsh, answer.

For students of tort law, Barnett v Chelsea & Kensington HMC serves as an essential analytical reference point. It illustrates that the structure of the tort of negligence is not merely formal but substantive: each element performs a distinct gatekeeping function. The case demonstrates with particular clarity that the causal element is not a rubber stamp on a finding of breach but