Wilsher v Essex AHA [1987] UKHL 11

A very premature baby suffered severe visual impairment from retrolental fibroplasia and sued the health authority, alleging negligent oxygen management. The House of Lords held that McGhee did not alter the ordinary burden of proving causation, set aside earlier judgments, and ordered a retrial on causation.

Facts

The infant plaintiff, Martin Graham Wilsher, was born nearly three months prematurely on 15 December 1978, weighing only 1200 grammes. He suffered many of the complications associated with extreme prematurity and passed through a series of crises, nearly dying. Advances in medical science and the treatment he received in the special baby care unit of the Princess Alexandra Hospital, Harlow, meant that he survived with unimpaired brain function.

However, he developed retrolental fibroplasia (RLF), an incurable retinal condition that caused total blindness in one eye and severely impaired vision in the other. He sued the Essex Area Health Authority, responsible for the hospital, alleging that his RLF was caused by excessive oxygen tension in his bloodstream in the early weeks of life due to negligent management of his oxygen supply.

There were two main phases of alleged negligence. First, during the initial 38 hours after birth, an umbilical catheter intended to monitor arterial oxygen tension (PO₂) was mistakenly inserted into a vein so that the sensor and sampling aperture lay in the heart rather than the aorta. This led to sampling of mixed arterial and venous blood and therefore misleading PO₂ readings. Both the house officer and registrar failed to detect the misplacement on X-ray. The trial judge held this failure to be negligent, for which the authority was liable. The plaintiff’s case was that misleadingly low monitor readings led to excessive oxygen administration, raising true arterial PO₂ to an unsafe level until the error was discovered on the morning of 17 December 1978.

Secondly, between 20 December 1978 and 23 January 1979, it was alleged that on five occasions medical and nursing staff allowed arterial PO₂ to remain above an accepted safe level. The judge found four of these five periods resulted from the authority’s negligence. The Court of Appeal, however, later reversed those particular negligence findings (while upholding negligence regarding the mis-sited catheter).

Issues

Negligence

The Court of Appeal unanimously upheld the finding that the health authority was vicariously liable for the registrar’s failure to recognise from the X-ray that the umbilical catheter lay in a vein rather than an artery. They unanimously overturned the trial judge’s findings of negligence concerning the later high-oxygen periods, holding that he had misdirected himself as to the burden of proof by treating it as reversed.

No issue on those liability findings arose before the House of Lords. The critical outstanding negligence finding, accepted by all appellate courts, was the negligent failure to detect the misplaced catheter in the first 38 hours.

Causation

The central issue before the House of Lords was whether the negligence relating to the misplaced catheter caused or materially contributed to Martin’s RLF. There was extensive and conflicting expert evidence from paediatricians and ophthalmologists for both sides, as well as numerous medical journal articles about RLF.

It was common ground that a sufficiently high level of arterial PO₂ in a very premature baby, if maintained long enough, can have a toxic effect on immature retinal blood vessels and lead to RLF. It was equally common ground that RLF can also occur in premature babies who receive no artificial oxygen, and that various other conditions commonly affecting premature infants (such as apnoea, hypercarbia, intraventricular haemorrhage and patent ductus arteriosus), all of which afflicted Martin, had been correlated with RLF, though causal mechanisms remained unclear.

The causation dispute had two dimensions: (1) whether the oxygen levels actually experienced as a result of the misplacement were, in degree and duration, capable of causing RLF; and (2) more fundamentally, whether any such excess oxygen administration in fact caused or materially contributed to Martin’s RLF in circumstances where several other potential causal conditions were also present.

The case further raised an important legal issue: the proper interpretation and scope of the House of Lords’ earlier decision in McGhee v National Coal Board, particularly whether that case justified a reversal or modification of the ordinary burden of proving causation and whether it applied where there were multiple independent potential causes.

Judgment

Procedural outcome

The House of Lords ordered that the orders of the Court of Appeal and of Peter Pain J be set aside (save as to costs) and that there be a retrial, before a different judge, of the single issue whether the negligence of the appellants, as found by the Court of Appeal, caused or materially contributed to the respondent’s RLF.

“It is Ordered and Adjudged, by the Lords Spiritual and Temporal in the Court of Parliament of Her Majesty the Queen assembled, That the said Order of Her Majesty’s Court of Appeal (Civil Division) of the 24th day of July 1986 and the Order of Mr. Justice Peter Pain of the 21st day of December 1984 complained of in the said Appeal be, and the same are hereby, Set Aside, save as to costs, and that the Cause be, and the same is hereby, remitted back to the Queen’s Bench Division of the High Court of Justice with a Direction that there be a retrial before a different judge of the issue whether the negligence of the Appellants, as found by the Court of Appeal, caused or materially contributed to the Respondent’s retrolental fibroplasia”

The House agreed that it was not open to it, on the written record, to resolve the complex and conflicting expert evidence on causation, and that neither side invited it to do so. Accordingly, a retrial of the causation issue was inevitable.

Burden of proof and causation in tort

Lord Bridge, with whom the other Law Lords agreed, reaffirmed the orthodox position that a plaintiff must prove, on the balance of probabilities, both breach of duty and causation. He traced the starting point to Bonnington Castings Ltd v Wardlaw, in which the House of Lords had rejected the notion that breach of statutory duty in itself shifts the burden of proving causation onto the defendant.

“It would seem obvious in principle that a pursuer or plaintiff must prove not only negligence or breach of duty but also that such fault caused or materially contributed to his injury, and there is ample authority for that proposition both in Scotland and in England… the employee must in all cases prove his case by the ordinary standard of proof in civil actions; he must make it appear at least that on a balance of probabilities the breach of duty caused or materially contributed to his injury.”

Lord Bridge emphasised that neither common-law negligence nor breach of statutory duty ordinarily reverses the onus of proof of causation:

“No distinction can be drawn between actions for common law negligence and actions for breach of statutory duty in this respect. In both the plaintiff or pursuer must prove (a) breach of duty and (b) that such breach caused the injury complained of”

Interpretation of McGhee v National Coal Board

The trial judge and the Court of Appeal majority had relied on McGhee v National Coal Board as supporting a shift in the burden of proof where a defendant’s breach materially increased the risk of the very harm that eventuated. Peter Pain J had previously stated a principle in Clark v MacLennan that:

“It seems to me that it follows from McGhee that where there is a situation in which a general duty of care arises and there is a failure to take a precaution, and that very damage occurs against which the precaution is designed to be a protection, then the burden lies on the defendant to show that he was not in breach of duty as well as to show that the damage did not result from his breach of duty.”

Lord Bridge rejected such a reading of McGhee, holding that it did not lay down any new principle reversing the burden of proof of causation. Instead, it was simply an example of robust inferential reasoning on the particular facts.

He noted that some passages in Lord Wilberforce’s speech suggested a reversal of the burden in that case, but considered those views to be a minority opinion inconsistent with the emphatic statements in Bonnington and unsupported by the other Law Lords in McGhee:

“My Lords, it seems to me that both these paragraphs, particularly in the words I have emphasised, amount to saying that, in the circumstances, the burden of proof of causation is reversed and thereby to run counter to the unanimous and emphatic opinions expressed in Bonnington Castings Ltd, v. Wardlaw … I find no support in any of the other speeches for the view that the burden of proof is reversed and, in this respect, I think Lord Wilberforce’s reasoning must be regarded as expressing a minority opinion.”

He then surveyed the other speeches in McGhee (Lords Reid, Simon, Kilbrandon and Salmon), concluding that the majority had treated the issue as one of inference from the facts, not of shifting burdens or creating a special rule of law. In particular, they regarded “material increase in risk” and “material contribution” as, in the circumstances of a single causal agent (brick dust), functionally indistinguishable from a common-sense perspective.

Lord Bridge summarised his conclusion:

“The conclusion I draw from these passages is that McGhee v. National Coal Board … laid down no new principle of law whatever. On the contrary, it affirmed the principle that the onus of proving causation lies on the pursuer or plaintiff… the attempt to extract from it some esoteric principle which in some way modifies, as a matter of law, the nature of the burden of proof of causation which a plaintiff or pursuer must discharge once he has established a relevant breach of duty is a fruitless one.”

Multiple potential causes versus a single causal agent

Lord Bridge distinguished the factual pattern in McGhee from that in the present case. In McGhee, there was only one causal agent (brick dust) capable of causing the dermatitis, and the employer’s negligence merely extended the period of exposure to that agent. In those circumstances, it was reasonable to infer that the negligent extension of exposure probably made a material contribution to the disease.

By contrast, in the present case there were several different potential causative agents for RLF (including excess oxygen, hypercarbia, intraventricular haemorrhage, apnoea and patent ductus arteriosus), all of which Martin had suffered from. The Vice-Chancellor’s dissent in the Court of Appeal, quoted and endorsed by Lord Bridge, stressed that this factual difference made application of the McGhee reasoning inappropriate:

“In the present case the question is different. There are a number of different agents which could have caused the RLF. Excess oxygen was one of them. The defendants failed to take reasonable precautions to prevent one of the possible causative agents (e.g. excess oxygen) from causing RLF. But no one can tell in this case whether excess oxygen did or did not cause or contribute to the RLF suffered by the plaintiff… A failure to take preventative measures against one out of six possible causes is no evidence as to which of those six caused the injury.”

The House accepted that, on this analysis, McGhee did not justify treating the health authority’s negligence (concerning oxygen) as causative merely because it increased one of several risks where no evidence showed that this risk was more likely than the others to have materialised.

Critique of the trial judge’s reasoning on causation

The trial judge had treated the plaintiff as having established a “prima facie” case of causation and then placed on the defendants the burden of proving that their negligence had not caused or materially contributed to the RLF. His reasons included a mistaken understanding that there was no dispute that the periods of raised PO₂ “materially increased the risk of RLF”.

“This statement, it is now accepted, was a misunderstanding of the evidence… it was certainly not accepted by the defence that any of the levels to which Martin was subjected were sufficient in degree or duration to have involved any material increase in that risk.”

Lord Bridge noted that, in assessing causation, the judge repeatedly emphasised that the onus lay on the authority:

“For the purpose of this action I need go no further than to consider whether the breaches have probably made no substantial contribution to the plaintiff’s condition.”

“So I have to consider whether the exposure that occurred probably did no harm.”

He then concluded that the defendants had failed to show that certain periods of exposure did not cause or materially contribute to Martin’s RLF, adding in respect of the first and third periods:

“On the basis of this evidence I find that the defendants fail to show that the first and third periods of exposure did not do any damage; indeed the probability is that they did.”

Lord Bridge held that these passages demonstrated a misdirection as to the burden of proof. While the highlighted phrase suggested that the judge might have regarded the plaintiff’s case as probable, he had not indicated any clear preference for the plaintiff’s experts or undertaken the detailed analysis that would normally accompany an orthodox finding that the plaintiff had discharged the burden of proof.

Accordingly, the appellate courts could not treat the judgment as though causation had been properly found in the plaintiff’s favour on a balance of probabilities with the burden correctly allocated.

Limits of appellate fact-finding and the Court of Appeal’s approach

The Court of Appeal, although unable to resolve the primary conflict on causation, had gone on to find that the PO₂ levels resulting from the misplacement of the catheter were of a kind capable of causing RLF. Mustill LJ acknowledged concern about whether the court was going too far in an effort to avoid a retrial, but considered that it was legitimate, on the “weight of the expert evidence”, to make this limited finding. Treating excess oxygen as “one of the possible causes” of Martin’s RLF, he then applied a principle derived from McGhee (as formulated in his own judgment) to attribute legal causation to the authority’s negligence.

Lord Bridge doubted whether the Court of Appeal had been entitled to make even this limited finding in the face of unresolved and sharp technical disagreements among experts, stressing the importance of the trial judge’s advantage in seeing and hearing such witnesses:

“Where expert witnesses are radically at issue about complex technical questions within their own field and are examined and cross-examined at length about their conflicting theories, I believe that the judge’s advantage in seeing them and hearing them is scarcely less important than when he has to resolve some conflict of primary fact between lay witnesses in purely mundane matters.”

He concluded that, given the absence of clear findings by the trial judge on the relevant factual and scientific issues, there was “really no alternative to a retrial” and that the judge conducting the retrial should approach causation with an entirely open mind, uninfluenced by the Court of Appeal’s factual views.

Implications

The decision reaffirms the orthodox rule that the plaintiff bears the burden of proving, on the balance of probabilities, that the defendant’s negligence caused or materially contributed to their injury, even in complex medical causation cases. It clarifies that McGhee did not modify this rule as a matter of law, but exemplified a permissible inference of fact in a case involving a single causal agent and an extension of exposure to that agent.

The House emphasised that increasing the risk of harm is not, in itself, sufficient to establish causation where there are multiple independent potential causes and no evidence that the negligent factor was more likely to have operated than the non-negligent ones. A failure to guard against one of several possible causes does not, without more, justify an inference that that particular cause materialised.

The case is particularly important in medical negligence, where scientific uncertainty and multiple potential aetiologies are common. It signals that courts should resist relaxing causation standards to accommodate evidential difficulty and should not treat risk-creation as equivalent to causation in multi-cause scenarios.

Finally, the judgment illustrates the limits on appellate courts’ ability to resolve complex technical conflicts on paper alone and underscores the importance of trial-level evaluation of expert evidence. It also highlights the systemic tension between the need for clear proof of fault-based causation and the hardship this may cause to severely injured claimants in cases of scientific uncertainty.

Verdict: The House of Lords allowed the appeal, set aside the orders of the Court of Appeal and of Peter Pain J (save as to costs), and remitted the case to the Queen’s Bench Division for a retrial before a different judge limited to the issue whether the appellants’ negligence caused or materially contributed to the respondent’s retrolental fibroplasia. Monies paid into court were to remain pending retrial, and the respondent’s costs of the appeal to the House were to be taxed under Schedule 2 to the Legal Aid Act 1974.

Source: Wilsher v Essex AHA [1987] UKHL 11