This video discusses the decision with John Coughlan QC, Jodie Kembery, Jamie Gamble and Rebecca Livesey of No5’s Clinical Negligence Group, in which we are joined by Imogen Halstead of No5’s Business & Property Groups for a view from the commercial perspective.

Nothing to see here” *

* Save for a novel and first principles reconfiguration of the tort of negligence


On Friday 18th June 2021, the much-anticipated Judgment in Khan v Meadows was given by the Supreme Court. At the same time, the Court gave Judgment in Manchester Building Society v Grant Thornton LLP, an auditors’ negligence case. The cases were heard by the same panel of seven Justices on different dates in November 2020.

It will not be immediately apparent why a failure to give proper advice in relation to hedge funding and swap transactions should have anything to do with a doctor failing to arrange proper investigation of genetic haemophilia.

In fact, the two cases taken together may just represent the most significant attempt of the highest Court in the land to rationalise (or even reformulate or codify) the common law of negligence in living memory.

This article seeks to explain and explore the decision in Khan and consider its wider ramifications for the clinical negligence practitioner.

Factual Background

For a beautifully pithy summary of the facts, look no further than paragraph 1 of the Judgment of Lord Hodge and Lord Sales, with whom three of the other Justices agreed:

A woman approaches a general medical practice for testing to establish whether she is a carrier of a hereditary disease. Tests which are inappropriate to answer that question are arranged. A general medical practitioner when informing her of the results of those tests negligently fails to advise her that she needs a genetic test to establish whether she is a carrier of the relevant gene. In fact, she is a carrier of the disease. Several years later, she gives birth to a baby boy who sadly not only suffers from the hereditary disease but also has an unrelated disability. Is the medical practitioner liable in negligence for the costs of bringing up the disabled child who has both conditions or only for those costs which are associated with the hereditary disease?

More specifically, in 2006, the Claimant learnt that she may be a carrier of the haemophilia gene and wished to avoid having a child with the hereditary condition. Shortly thereafter she sought the advice of her GP and blood tests were organised. However, these were sufficient only to establish whether the person has haemophilia and not whether they carry the gene, for which genetic testing by a haematologist is required.

The Claimant returned to discuss the results on 25 August 2006 with the Defendant, a different GP within the same practice. The Claimant was negligently advised that the results were “normal”, leading the Claimant to believe that any future child would not have haemophilia.

The Claimant gave birth to her son Adejuwon in 2011. Adejuwon suffers haemophilia and severe autism, with his autism serving to exacerbate the risks and effects of his haemophilia. Adejuwon’s autism is such that he will never live independently or take paid employment.

Subsequent genetic testing confirmed that the Claimant was a carrier of the haemophilia gene. It was common ground that, had proper testing been undertaken in 2006, the Claimant would have undergone foetal testing when she became pregnant and the diagnosis of haemophilia would have been made. The Claimant would have chosen to terminate her pregnancy and her son would not have been born.

It was agreed between the parties that the additional costs in raising Adejuwon on account of his haemophilia are £1.4m, but the additional costs in raising the actual Adejuwon (haemophilia plus autism) are £9m.

The decisions below

The Trial was heard by Yip J, who found that the focus of the Defendant’s duty and the Claimant’s purpose in seeking confirmation was to provide her with the necessary information to allow for termination any pregnancy afflicted by haemophilia (§62); and the Defendant assumed a responsibility which, if properly fulfilled, would have avoided the birth (§63).

The Defendant appealed successfully to the Court of Appeal, which found that the focus of the consultation was whether the Claimant was a carrier of the haemophilia gene and not the wider issue of whether she should become pregnant. Applying the scope of duty test identified in South Australia Asset Management Corpn v York Montague Ltd [1997] AC 191 (“SAAMCO”), damages for autism were not within the scope of the risks against which the Defendant had undertaken to protect the Claimant.

The decision of the Supreme Court

The Claimant’s appeal was rejected unanimously. The Claimant’s submissions in the appeal – effectively that SAAMCO a) does not apply to clinical negligence cases, alternatively b) should not apply to clinical negligence cases – were entirely rejected. SAAMCO applies to all negligence claims, hence the marriage of this case to the Grant Thornton Judgment. On the facts of this case, autism lay outside the scope of duty of the duty owed by the Defendant.

There, however, unanimity ends. The “majority” Judgment (of course, there is no dissent in the case but let’s call them that) penned by Lord Hodge and Lord Sales comes in for ripe criticism of different flavours and intensity in the Judgments of Lord Burrows and Lord Leggatt. That there was tense debate between the Justices is apparent from the Judgment (see Lord Burrows at §78, Lord Leggatt at §96) and made explicit at §3 of the Grant Thornton Judgment.

The tension appears to derive in large part from what Lords Leggatt and Burrows perceive to be a novel / unnecessary / potentially damaging attempt by the majority to revise or reframe the conceptual structure of the tort of negligence.

The rule of six

Whilst expressly declaiming any novelty in their approach, the “majority” set out what they consider to be “a helpful model for analysing the place of the scope of duty principle in the tort of negligence, and the role of the other ingredients” (§28):

1. “Is the harm (loss, injury and damage) which is the subject matter of the claim actionable in negligence? (the actionability question)
2. What are the risks of harm to the claimant against which the law imposes on the defendant a duty to take care? (the scope of duty question)
3. Did the defendant breach his or her duty by his or her act or omission? (the breach question)
4. Is the loss for which the claimant seeks damages the consequence of the defendant’s act or omission? (the factual causation question)
5. Is there a sufficient nexus between a particular element of the harm for which the claimant seeks damages and the subject matter of the defendant’s duty of care as analysed at stage 2 above? (the duty nexus question)
6. Is a particular element of the harm for which the claimant seeks damages irrecoverable because it is too remote, or because there is a different effective cause (including novus actus interveniens) in relation to it or because the claimant has mitigated his or her loss or has failed to avoid loss which he or she could reasonably have been expected to avoid? (the legal responsibility question)”.

Putting it bluntly, this is not the traditional sequence of questions that personal injury and clinical negligence lawyers ask themselves when assessing a case: duty / breach / causation / loss. Questions 2 and 5 (the scope of duty and duty nexus questions) are the focus of the appeal and require special attention below.

What of the “minority” Judgments in the case? In what is presumably as close to a rebuke as one comes in the Supreme Court, Lord Burrows describes the “majority” approach as “novel” and doesn’t think it “necessary” or “helpful”. Lord Leggatt joins in on this point, thinking it “undesirable and unnecessary to engage in such an exercise”.

However, as a result of his criticisms of the efforts of the “majority” to impose the new conceptual framework, Lord Burrows seems to have felt bound to offer a rival set of questions along more orthodox lines (§79):

1. Was there a duty of care owed?
2. Was there a breach of duty?
3. Was the damage factually caused by the breach?
4. Was the damage too remote?
5. Was the damage legally caused by the breach?
6. Was the damage within the scope of the duty of care?
7. Are there any defences?

Lord Leggatt declines to offer any new analysis, holding (twice) that the analysis in the present case is “straightforward” and to add more would be an “excursus”.

Irrespective of those criticisms, the excursus is now firmly out of the bag, and practitioners and students alike will now need to consider the six questions in every case and every exam question.

The (relatively) uncontroversial questions

Question 1 – actionability. Largely uncontroversial, this part of the Judgment makes the point that some damage must be caused in order for the cause of action to be complete. Reference is made to the pleural plaques cases as examples of symptomless damage cases where this hurdle would not be cleared.

Somewhat surprisingly, there is no discussion in this Judgment as to establishing existence of a duty of care. Lord Burrows makes the point that there is no place in the new framework for a discussion of the difficult cases on establishing the existence of a duty of care. Presumably those issues would fall under, or more likely even precede, the actionability question.

Question 3 – breach. The normal principles apply, but one can only breach a duty that is owed and within scope.

Question 4 – factual causation. One would have thought that the factual causation question would be similarly uncontroversial, but the “majority” seem to be at pains to delineate the limits of factual causation and the “but for” test: it is a “threshold but not a sufficient condition for liability”, it is “not of universal utility” and struggles in cases involving multiple wrongdoers and multiple sources of harm. Reading somewhat between the lines, the Supreme Court appear to consider that misapplication of the “but for” test lies at the heart of the problems at first instance in this case, and wish to put the concept firmly back in its box.

Question 6 – legal responsibility. Given that each and all of the six questions go to establishing legal responsibility for the harm caused, this is an odd label. Indeed, the question is not really a question but a catch-all, for all of the other arguments that may serve to prevent a Claimant receiving damages they would otherwise be due: contributory negligence, volenti, novus actus interveniens, failure to mitigate and all the rest of it.

Scope of duty and duty nexus

All of the Justices make the point that scope of duty is not a new question, but to the reader who has practised in personal injury and clinical negligence applying the orthodox framework, it may feel this way. We now know that clinical negligence is not a special case and that all negligence claims can and must be analysed by way of the six questions. The Supreme Court could easily have resolved the scope of duty question by affirming the decision of the Court of Appeal in Khan, but they chose to go wider.

Once the actionability question (question 1) is answered in the affirmative, the Court must address the scope of duty question as a matter of principle. What are the risks that this Defendant assumed responsibility to protect this Claimant from? We are told that the scope of duty question will cover omissions as well as positive acts.

Scope of duty may be very straightforward on the facts of any given case, indeed the majority of cases: the scope of a car driver’s duty is uncontroversial and well known, as is the duty of an operating surgeon to her patient, and so on. But other cases will be more complex. It seems likely that, as in Khan and as in Lord Hoffmann’s famous SAAMCO hypothetical of the mountaineer with the dodgy knee, it will be cases involving advice, the gathering and rendering of information and investigations that the scope of duty questions will serve to limit liability.

Dr. Khan undertook to advise the Claimant on her genetic susceptibility to haemophilia and the likelihood of having a baby with haemophilia (and did it badly), but she did not undertake to advise the Claimant on all the risks of pregnancy and to the baby including that of severe autism. It is not enough simply to assert that “but for” Dr. Khan’s negligence Adejuwon would not have been born. Adejuwon’s autism never lay within the scope for which Dr. Khan could in law be responsible.

In commercial transactions, different actors assume different parts of the responsibility for the agreement, and usually a contract will dictate who is responsible for which risk. In medical cases, it is less clear, but the issue remains one of “risk allocation”. Dr. Khan took the risk on matters haemophilia, but the Claimant took the remainder of the risks of pregnancy including autism.

The scope of the duty undertaken will therefore have to be determined on the facts of every clinical negligence case involving advice and, importantly in medical cases, investigation. This is to be conducted under question 2, the scope of duty question, before any issue of breach or causation is considered: what did this doctor undertake to do to protect this Claimant from harm?

An example: an elderly man suffers a fall at home and attends his GP complaining of chest pain. The GP asks for an x-ray to exclude rib fractures. The radiologist reports the x-ray as clear for rib fractures, but fails (in a way that no responsible radiologist should) to spot the large malignant shadow on the patient’s lung. The authors suspect that every reader, and every radiologist, would expect a claim against the radiologist to succeed: expert radiologists will always say that it is their duty to report everything on the film irrespective of the request. But can the radiologist not now claim that, so far as she was concerned, she undertook to look for rib fractures and was not allocated anything more?

The answer to this problem may lie in a proper analysis, not given by the Supreme Court, as to where the content of the duty fits into its framework, as opposed to its scope. We expect – perfectly understandably – Defendants in advice claims to take many more of these points than hitherto, and for many cases to be fought in the coming years exploring scope of duty in this context.

Once the scope of duty is identified, and questions 3 and 4 are answered in relation to breach and factual causation, we are urged by the Supreme Court to answer the duty nexus question: is there a sufficient nexus between a particular element of the harm for which the claimant seeks damages and the subject matter of the defendant’s duty? If one has conducted the scope of duty exercise accurately, it is difficult to understand the need for a duty nexus question, and there is an element of acceptance in the “majority” framework that this is essentially a cross-check that liability is properly established.

In order to achieve this crosscheck, a further device – “the SAAMCO counterfactual” – is presented to assist. Do not confuse this counterfactual with our standard enquiry in clinical negligence cases (“but for the negligence, what would the decision / course of events / outcome have been?”). Rather, the SAAMCO counterfactual urges the Court to consider what would the position have been if the negligence advice had been correct and all other factors remained the same? Would the same loss have nevertheless been sustained? If so, the loss was outside scope in the first place.

When Lord Burrows applies the SAAMCO counterfactual to the case before him (§77(iii)), he is satisfied that the answer is “yes” as regards the autism losses, and “no” as regards the haemophilia losses. The autism losses were thus outside the scope of the doctor’s duty of care.

It is our view that the SAAMCO counterfactual is here to stay, but that it carries difficulties. Much depends on how the question is put, and many more questions are begged. Is it fair to put the “but for” question so far back in its box that we must assume that Ms. Meadows would ever have had a baby at all in the presence of reasonable care? Had she known she was a carrier of the haemophilia gene, would the autism losses ever have been incurred? Was the balance of probabilities test applied by Yip J at first instance in relation to subsequent pregnancies and the risk of autism not preferable?

Thinking points: there may be trouble ahead

This decision will be subject to much consideration and criticism as the dust settles. As well as the commentary offered above in relation to the six questions, and the prospect of the proper limits of scope of duty being litigated in the coming years, we raise the following as further food for thought:

– How does Khan sit with the eggshell skull principle? Despite the Claimant’s carriage of the haemophilia gene, Khan cannot properly be analysed as an eggshell skull case. But it wouldn’t take too many factual tweaks to bring eggshell skull principles and scope of duty principles into interplay, and perhaps into conflict.

– How does Khan sit with Montgomery and respect for patient autonomy? How does Montgomery sit in the six-question framework? Should the Claimant not have been informed of the extent of the risk allocation process? There is no discussion of Montgomery at all in this case.

– Does Khan sit as comfortably as the Court says with the wrongful birth cases? There is discussion of MacFarlane and Parkinson in this Judgment, and – an interesting aside – Parkinson appears to be considered settled law. It might be said that for failed sterilisation cases the scope of duty is to achieve sterilisation, and so recovery is well grounded, but is this to recognise a new sub-category of wrongful birth cases?

– Does the distinction between pure economic loss cases and those grounded in personal injury survive Khan? The Supreme Court seem to think it an unimportant distinction, which is not the experience of those at the coalface when it comes to issues such as limitation and periodical payments.

– Fundamentally, does Khan achieve its professed aim (§28) of putting the tortious victim as close as back to the position she should have been in had the tort not occurred?