The Marshall v Schembri appeal: “Counter-factual” causation and the precise mechanism of harm

Shortly before the lockdown caused by the Coronavirus pandemic, the Court of Appeal gave their judgement in an interesting and complex clinical negligence case on causation, Mario Schembri v Ian Marshall [2020] EWCA Civ 358. The appeal was by the Dr Mario Schembri, the defendant general practitioner (“D”), against the judgement of Mr Justice Stewart a year beforehand, who had held that he had been negligent in failing to suspect a pulmonary embolism (“PE”) and refer the late Mrs Doreen Marshall (“DM”) to hospital. The claim was brought by DM’s husband (“C”).

The facts

The facts are straightforward. DM had attended D’s surgery at about 1600 hours on 25 April 2014 complaining of chest pain and breathlessness. She had a past history of PE in 2008. D considered that DM had a muscular strain affecting a hiatus hernia and recommended ibuprofen and rest. DM remained at home. She deteriorated suddenly the following morning at about 0800 hours, suffering a collapse and subsequent cardiac arrest from which she could not be resuscitated by the attending ambulance crew.

The trial

It had been agreed between the parties that, had D referred DM to hospital at the proper time the previous day, she would have been diagnosed as having a PE and she would have received either anticoagulation (heparin) or thrombolysis (alteplase). Stewart J. found at trial that DM would have received heparin by 2100 hours which would have prevented further clotting by midnight at the latest. C had asserted that, absent D’s breach, the massive PE which had caused DM’s death would have been avoided; or that, even if a massive or sub-massive PE had occurred in hospital, thrombolysis would have been administered and on balance DM would have survived. D asserted that DM would have had less than 12 hours of heparin before her massive PE at 0800 and on balance she would still have suffered this even with anticoagulation by 2100 the night before and would have died.

The judgement

The judge had formulated 4 key questions the court had to answer, as follows (§41):

Q1 Has C proven on the balance of probabilities that there were progressive PEs during the night of 25/26 April 2014? [An indicator for prescribing thrombolysis.] The judge answered this as a “No”.

Q2 If so: (a) would progressive pulmonary emboli have been picked up on monitoring had DM been in hospital? (b) if so, would thrombolysis have been prescribed and with what effect? Although this question did not demand an answer because of the negative response to Q1, it still needed a response because it remained relevant. Accordingly, the judge answered it, as a (a) Yes and (b) No.

Q3 If the answer to Q1 and/or Q2 is negative, had DM been in hospital, would thrombolysis have saved her? In other words, had there not been progressive PEs, can C prove that thrombolysis would have saved her had she gone into cardiogenic shock or arrested in hospital? The judge addressed this question by reference to the medical literature.

1. First, he concluded that, had DM been given alteplase at least about 3 hours before her arrest at about 0830 hours, she would probably have survived. However, he also concluded that DM had not reached the threshold for thrombolysis prior to her arrest.
2. He then looked at the literature on thrombolysis in cardiogenic shock and cardiac arrest, which was favourable to C. With respect to the shock, he concluded that C could not prove on balance that DM would have been in the 64-75% category who would have survived with alteplase, but her chances of survival would have been significantly increased if she had been in hospital.
3. With respect to the arrest, DM had no co-morbidity and so possibly would have been in the 35%-37% category of patients who survive with alteplase.

Q4 If the answer to Q1-Q3 are negative in that C cannot prove a specific train of events or mechanism which would, absent D’s negligence, have saved her [then] looking at the evidence as a whole, is it nevertheless more likely than not that DM would have survived had she been referred to Southend Hospital?” The judge referred to a passage by Toulson LJ in Drake v Harbour [2008] EWCA Civ 25:

“… where a claimant proves both that a defendant was negligent and that loss ensued which was of a kind likely to have resulted from such negligence, this will ordinarily be enough to enable a court to infer that it was probably so caused, even if the claimant is unable to prove positively the precise mechanism. That is not a principle of law nor does it involve an alteration in the burden of proof; rather, it is a matter of applying common sense. The court must consider any alternative theories of causation advanced by the defendant before reaching its conclusion about where the probability lies. If it concludes that the only alternative suggestions put forward by the defendant are on balance improbable, that is likely to fortify the court’s conclusion that it is legitimate to infer that the loss was caused by the proven negligence.”

At §128, the judge noted:

“As is accepted, the Claimant has the burden of proving causation. Yet the Claimant needs to prove no more than that Mrs Marshall would probably have survived had she been admitted to hospital. The Claimant does not need to prove the precise mechanism by which her survival would have been achieved.”

The judge had noted from C’s medical expert evidence and the literature that patients very rarely die from PE in hospital, in particular those without any co-morbidities and risk factors as in this case. He concluded that DM’s chances of survival had she been referred to hospital were very high.

The appeal

D appealed on the basis that the judge was wrong for the following reasons:

(i) Having found C had not proved on balance that DM would have survived had she been admitted to hospital (Q1-3), he should have concluded the claim failed.

(ii) He should not have posed a separate overriding Q4 based on “general” survival rates.

(iii) He should not have found for C on the basis of a “general” analysis that most people do not die from PEs in hospital.

(iv) In answering Q4, (a) he did not give sufficient weight to D’s experts’ evidence (whose evidence he had preferred in answering Q1-3); and, (b) he wrongly applied general propositions to DM’s individual case when he should have treated it individually.

In response, C relied on Drake v Harbour, referred to above, that, where a breach of duty is established and the injury is of a kind likely to have resulted from that breach, that is usually enough for the court to find the injury resulted from the breach.

D accepted that C did not need to prove the precise mechanism by which the injury would have been prevented absent the breach, but argued that, where the judge had found that C had failed to establish the necessary standard that DM would have survived with receipt of either or both treatments, the claim must fail. This was on the basis of the answers to the earlier questions, namely that C had not proved that: (a) there were continuing PEs overnight; (b) thrombolysis would have been administered overnight; and, (c) thrombolysis during shock or the arrest would have led to survival; accordingly (d) causation had not been made out.

D argued that the judge had fallen into error by the incorrect use of statistics on a “general” basis, applying Wardlaw v Farrar [2003] EWCA Civ 1719. However, the appeal court did not consider that Wardlaw helped, as in that case the patient’s response to treatment was a matter of fact, whereas the present case concerned a “counter-factual” hypothesis, i.e. a matter of hypothetical fact. The court referenced Clerk & Lindsell on Torts (22nd Edition (2018), at para. 2-30, p.75), as follows:

“On the other hand, care should be taken not to take the logic of this reasoning too far in the opposite direction. If the evidence is that, say, 80 per cent of patients survive with prompt treatment, but 20 per cent die even with prompt treatment, the fact that the patient died following delayed treatment does not establish that he probably fell into the 20 per cent category at the outset and therefore the delay did not contribute to the death. The assessment of causation would turn upon the detailed medical evidence, both as to the overall statistical chances of survival and the particular condition and circumstances of the patient” [my emphasis].

And, after quoting Lord Nicholls in Gregg v Scott [2005] 2 AC 176 at [27]-[28] and [32], and his observati0ns that statistics can be useful and their use appropriate, the appeal court further relied on the last sentence of the above paragraph:

“Proof of causation is almost inevitably about a burden of persuasion and sometimes statistics can be highly persuasive.”

The appeal court (McCombe LJ giving the sole judgment) concluded that the judge had considered in Q1-3 whether C could establish a specific mechanism which would have saved DM and, having not been able to do so, had gone on to look at all the evidence to assess whether it was more likely than not that DM would have survived. D’s appeal was heavily dependent upon what had actually happened at home whereas the judge was attempting on the basis of all the available evidence to identify the “counter-factual” case of what would have happened in hospital, on a balance of probabilities. The judge had concluded that, on a “clear” balance of probabilities, DM would have survived. The appeal court considered that the judge was entitled to reach such a conclusion on all the evidence and he had not been wrong to pose Q4 or answer it the way he had. There was a legitimate place for statistics in cases of this nature and the judge’s employment of that evidence had been closely linked to DM’s own particular condition, in which her prospects of survival had been very good indeed; cases like this one were intensely “fact-specific”.  Holroyde and Phillips LJJ agreed. The appeal was dismissed.

It is perhaps of passing interest to note that C’s cross-appeal asserted that the judge should have concluded that DM would have survived cardiogenic shock if she had been in hospital because the weight of the evidence was favourable. McCombe LJ said he could “see the force of the argument raised and it might have persuaded some judges to answer [Q3] in the positive sense”. Although the point did not need determination, it seems likely that the appeal court felt that Stewart J. could have answered the question in C’s favour (and they would have).

Conclusion

The case is interesting for a couple of reasons.

First, a claimant does not need to prove the precise mechanism of harm, or of avoiding harm, to be able to succeed in proving causation (see also the recent case of AXO v Salisbury NHS Foundation Trust [2019] EWHC 1454 (QB)). This is trite law but the case is a helpful reminder.

Secondly, in assessing a counter-factual or hypothetical factual causation scenario, i.e. what on a balance of probabilities would have happened as a matter of hypothetical fact absent the breach, it is right to look at both statistics and an individual patient’s condition or characteristics and consider the evidence in its totality to assist in reaching a conclusion; in a case (such as this) where statistics are not determinative, all the evidence is required to be reviewed and considered.