Causation: the last refuge in a weak case?

It used to be thought by clinical negligence practitioners that to take a causation point at trial smacked of desperation. The real business of clinical negligence litigation, the thinking ran, was arguing the issue of breach of duty which would in and of itself determine the outcome of the trial. But if that were ever true it is no longer. Many contested clinical negligence trials have causation disputes at their heart and in recent years such trials have generated some of the most interesting and important jurisprudence in this area of practice.

Whilst determining whether or not there has been a breach of duty often involves finely balanced argument as to clinical judgement, the legal framework underpinning such argument is usually uncontroversial. The principles enunciated in the landmark rulings in Bolam, ¹ Sidaway, ² Maynard ³ and Bolitho ⁴ are well understood now not only by lawyers but by many of the medical practitioners engaged in medicolegal work. ⁵

Yet such cannot be said of causation. The over-riding principle is easy to state: the claimant must establish a causal link between the breach and his loss or damage. In many cases this will be an explanatory factual inquiry: how, in fact, did the damage occur? ⁶ And the starting point in such inquiry will be ‘common sense’. ⁷ But the difficulty in understanding the concept of causation stems from the fact that there can be more than one ‘common sense’ answer to the factual inquiry in any given case. In others, answering this question will not clearly and straightforwardly establish causation in law. The factual inquiry may yield a number of different contributory causes of the loss or damage. The factual cause of the Claimant's loss or damage may be unknowable (a) at all (because medical science cannot tell us the answer) or (b) as between a negligent and non-negligent cause.

Thus a number of difficulties in understanding issues of causation persist and continue to interest practicing and academic lawyers alike. And the case-law has produced decisions which, depending upon one's viewpoint, may appear counter-intuitive or plainly wrong.

From the perspective of a Claimant it may seem to defy common sense to deny recovery to a person whose obvious symptoms of cancer were negligently overlooked by their GP with the result that their chance of being cured was substantially reduced (albeit by less than 50%): Gregg v Scott. ⁸ Equally it may seem contrary to justice and common sense to anyone looking at matters from a corporate perspective that a company could be liable for someone's injury and death when what they actually exposed their employee to may very well have caused no injury at all: Fairchild. ⁹

This article seeks to elucidate three principles by reference to recent cases cognizance of which, the authors suggest, will increase prospects of success at trial in a difficult causation case: burden of proof, material contribution and multiple causes giving rise to a compound injury.

Burden of proof

It perhaps should not need stating that the burden of proof in establishing causation lies with the Claimant. But this principle may sometimes be overlooked in a complex clinical negligence case where the issue of breach of duty is difficult, the weight of evidence obtained by the parties goes substantially to that issue, and therefore in consequence all the creative energy of the barristers and solicitors involved in the case is focused upon succeeding on that issue.

An example is the case of Naraji. ¹⁰ The claim was brought by an ex-professional footballer who, whilst on Sheffield United's books, sustained an injury to his anterior cruciate ligament (ACL). He underwent an initial reconstruction operation. It failed and he underwent further reconstruction surgery performed by the first Defendant. ¹¹ The case was heard by Popplewell J over seven days in October 2011. His judgment devotes much time to analysis of the issue of breach of duty and relatively little to the issue of causation. His analysis of breach of duty involved consideration of the testimony of six orthopaedic surgeons, four of whom were called as independent experts by the parties, two expert radiologists, and nine other professional witnesses whose opinions were either reported to the trial judge by the Claimant or put before him in written form (and from which he gained little assistance, with one notable exception). Such evidence went to the minutiae of ACL reconstruction surgery.

In his judgment on that evidence the judge found that the first defendant had acted negligently in placing the femoral tunnel too anteriorly “beyond the range which would have been regarded as proper and acceptable by a respectable and responsible body of surgical opinion”.

He then turned to the issue of causation, where the issue was whether or not the Claimant would have been able to achieve the same level of playing ability as before the ACL injury had the operation been performed without negligence. On this issue the judge found for the Defendant. He began his analysis by remarking: “I have not found this an easy question, and very little attention was paid to it in evidence by comparison with the issues of liability”.

That might seem an extraordinary statement given that causation is the second of the twin hurdles to be surmounted by a Claimant in winning his case on liability. As to the evidence of the Claimant's orthopaedic experts the judge said: “Professor Fairclough … opined that had Dr Shelbourne … not provided substandard care Mr Naraji would have been able to return to professional football… There was nothing more put forward to support this opinion than its simple articulation, although … Mr Strover's report expressed his opinion that if further revision surgery had been undertaken, Mr Naraji would on the balance of probabilities have been able to return to a career as a professional footballer… He too put forward nothing more to support this opinion than its simple articulation.”

Thus the judge found that the burden of proving the issue of causation was not discharged by the Claimant. He found that there was a paucity of evidence on this issue and that such evidence as was available to him was superficial, this at the conclusion of a seven-day trial involving oral expert evidence on the issue from eminent experts and consideration of numerous opinions from other sources.

In the rare case judges may be prepared to ‘finesse’ the issue of causation in the Claimant's favour. Perhaps the high point of such an approach is Chester v Afshar. In that case the Claimant could not prove that had she been properly advised by the Defendant of the risks of her spinal surgery she would not have had it. She failed the ‘but for’ test and any other test of causation you care to mention. Yet she succeeded. The decision is widely seen as an exercise in judicial policy and distributive justice. Lord Steyn said that the Defendant's negligence had denied the Claimant the right to have given her informed consent to the surgery and “her right of autonomy and dignity can and ought to be vindicated by a narrow and modest departure from traditional causation principles”. Lord Steyn followed an academic lawyer, Professor Honore. He said: “in my view Professor Honore was right to face up to the fact that Chappel v Hart – and therefore the present case – cannot neatly be accommodated within conventional causation principles. But he was also right to say that policy and corrective justice pull powerfully in favour of vindicating the patient's right to know.”

Lord Hope was also influenced by Professor Honore's views and concluded “I would hold that justice requires that Miss Chester be afforded the remedy which she seeks, as the injury which she suffered at the hands of Mr Afshar was within the scope of the very risk which he should have warned her about when he was obtaining her consent to the operation which resulted in that injury”.

In reaching this decision their Lordships were not put off by the fact that “…there is no direct English authority permitting a modification of the approach to the proof of causation in a case such as the present,” ¹² illustrating, if nothing else, that this is a dynamic area of law. The editors of ‘Medical Negligence’ wrote in an edition which pre-dated Chester: ‘…it is important to appreciate that there is an element of judicial policy at work in attributing causal connections’.

But if the Claimant's legal team in Naraji were hoping that the judge would follow policy or distributive justice to find for the Claimant they were disappointed. In the case of a momentary slip by a skilled surgeon, as Popplewell J clearly categorized Dr Shelbourne's too anterior placement of the femoral tunnel, there were no public policy considerations upon which the Claimant could rely and distributive justice did not call for success for the Claimant. The onus in that case (as in the majority) was on the Claimant to prove causation in fact (i.e. that, in fact, the surgical negligence caused his retirement from professional football, or to put it another way, that but for the negligence he would have been able to continue playing professional football). In that case (as in the majority) causation in fact would be identical with causation in law, and determining the former would determine the latter. And the judge found that the Claimant had not discharged his duty in establishing causation in fact, so the claim failed.

It may be thought by some who read the entirety of the judgment in Naraji that the judge was harsh on the Claimant on this issue – how can a Claimant prove a hypothetical outcome from different medical treatment from that which he received other than by inviting eminent experts to opine on the matter? But be that as it may, the message is clear: Claimant's lawyers should underestimate at their peril the attention which needs to be paid to establishing causation in fact in a clinical negligence trial. Whilst the Claimant's legal team may be thought to have done fantastically well in establishing breach of duty in such a complex case, the case was lost despite such victory. One is reminded of the skilled surgeon performing the complex operation who proudly announces that the operation “went extremely well, yet the patient died”.

Material contribution

There are a number of cases where Claimants have succeeded in recovering for the whole of their injuries despite other causes of the injury predominating, in reliance on the doctrine of material contribution. Two relatively recent examples are Bailey v MOD ¹³ and Canning-Kishver. ¹⁴ In each case the Claimants would have failed on a strict application of the ‘but for’ principle but succeed on an application of the ‘material contribution’ test.

This test has a long lineage. The House of Lords held in Bonnington Castings ¹⁵ that the Claimant does not need to prove that the Defendant's breach of duty was the main cause of the damage provided that it materially contributed to it. The Claimant sued his employers for contracting pneumoconiosis from inhaling air laden with silica dust. The evidence in the case established that there were two types of dust from different sources: ‘innocent dust’ from pneumatic hammers for which the employers were not liable and ‘guilty dust’ from swing grinders for which they were liable. It could not be proved which dust had caused the Claimant's condition. There was no evidence as to the proportions of each dust he inhaled but there was evidence that it was likely that the greater proportion of inhaled dust came from the innocent source. ‘But for’ causation could not be proved, i.e. it could not be proved that but for the guilty dust he would not have contracted the disease. However, the House of Lords drew an inference of fact that the guilty dust was a contributory cause, holding the employers liable for the full extent of the loss. They found that the Claimant did not have to prove that the guilty dust was the sole or even the main cause of his disease, just that the guilty dust had materially contributed to the disease if such contribution could be categorized as anything other than a de minimis cause.

The editors of ‘Medical Negligence’ ¹⁶ conclude that the case was significant for two reasons, firstly for marking a departure from the ‘but for’ principle of causation – the Claimant did not have to prove that he would not have suffered the damage i.e. injury or illness but for the breach of duty – and secondly for the fact that the Court was willing to draw an inference of fact that there had been a material contribution, when in reality it was impossible to say whether or not there had been any such contribution or even to make a statistical guess.

The material contribution test was applied to the Claimant's benefit in the case of Bailey. The Claimant, when an inpatient at St Mary's Hospital in Portsmouth, aspirated her vomit leading to cardiac arrest that caused her to suffer hypoxic brain damage. She had undergone an unsuccessful procedure to remove a gallstone, the view of which had been obscured by considerable bleeding during the procedure. Following the procedure there was a period of lack of care, in particular a failure to resuscitate. The Claimant was diagnosed as possibly developing pancreatitis. Her condition continued to deteriorate and she was subsequently transferred to an intensive care unit where she underwent additional procedures, and it was confirmed that she had pancreatitis. She later aspirated her vomit which led to cardiac arrest that caused her to suffer hypoxic brain damage.

The judge found that the physical cause of the Claimant's aspiration and subsequent cardiac arrest was her weakness and inability to react to her vomit. He found that such had several contributory causes: the negligent postoperative care at the defendant's hospital, and the pancreatitis, a non-negligent cause. He found that each factor had made a material contribution to her overall weakness.

He found for the Claimant, applying a modified ‘but for’ test. He found that the Defendant's negligence had made a more than de minimis contribution to the damage. The Defendant appealed, submitting (inter alia) that unless the Claimant could establish that but for the negligence of the Defendant the brain damage would not have happened she could not succeed, and the pancreatitis was the mostly likely cause of the vomit and could have been a cause on its own of her inability to protect her airway.

The Court of Appeal held that it was enough for the Claimant to establish that, on the balance of probabilities, a lack of care made a material contribution, namely something greater than negligible, to the weakness of her condition, although that was not an application of the ‘but for’ test.

The judgment of the Court was given by Waller LJ, who said:

“I would summarize the position in relation to cumulative cause cases as follows. If the evidence demonstrates on a balance of probabilities that the injury would have occurred as a result of the non-tortious cause or causes in any event, the Claimant will have failed to establish that the tortuous cause contributed … If the evidence demonstrates that ‘but for’ the contribution of the tortuous cause the injury would probably not have occurred, the Claimant will (obviously) have discharged the burden. In a case where medical science cannot establish the probability that ‘but for’ an act of negligence the injury would not have happened but can establish that the contribution of the negligent cause was more than negligible, the ‘but for’ test is modified, and the Claimant will succeed.”

The judge found on the evidence that there were a number of possibilities for causation of the Claimant's disabilities, including her prematurity. Imaging had revealed that the Claimant's cerebellum was atrophied. This was thought by the Claimant's experts to be the cause of her disabilities. However, they conceded that the atrophied cerebellum may have arisen from her premature state. The cause of the atrophy may have been congenital, metabolic or by way of insult. The Claimant's neonatologist posited that the most likely cause was a cerebellar haemorrhage occurring as a result of insult. Absent the Claimant's collapse caused by the Defendant's negligence, the Claimant's experts reasoned, absent the insult causing the haemorrhage causing the cerebellar atrophy.

The judge accepted evidence that without negligent treatment a baby born at 25 weeks gestation has a 50% chance of disability of some sort, assuming survival. Of those 50% with disability, 50% will have a major functional impairment such as arising from cerebral palsy. The balance will have lesser, more subtle disorders.

In his judgment the judge relied upon the case of Bailey, citing the passage already quoted in full in this article and found for the Claimant notwithstanding that she would have lost on a classic application of the ‘but for’ test. He found:

“The evidence simply does not establish on balance of probabilities that Ayesha's brain injury arose simply from her immaturity. That cannot be excluded as a possibility but there is nothing that suggests that that non-tortious cause was probable … the fact of residual possibilities militates against success for Ayesha by reference to a ‘but for’ test. However, I am entitled to find – and do find – that on balance of probabilities the contribution of the collapse occasioned by the breach of duty constituted a contribution to the atrophy of the cerebellum that was more than merely negligible so that the claim succeeds. As to this, the Claimant's experts have satisfied me that, notwithstanding the necessary speculation as to cause and effect in this ‘cutting edge’ sphere, the probabilities are that it was the impact of events leading up to and following the collapse that resulted in the atrophy of the cerebellum.”

Multiple causes giving rise to a compound injury

In a recent case in the Court of Appeal in Wright v Cambridge Medical Group [2011] EWCA Civ 669 an 11-month-old child was admitted to hospital with chicken pox where she contracted an undiagnosed super-infection with Streptococcus pyogenes. She was taken to the GP who, it was admitted, was negligent in failing to refer her to hospital. Thereafter the infection caused an effusion in the hip joint which led to circulatory difficulties and the child was taken to hospital two days after the day when she ought to have been admitted. Although the hospital recognized the infection, and treated it with antibiotics, the hospital staff did not make a proper assessment in relation to the lack of movement at the hip, and an ultrasound scan was not performed until four days after admission. When it was carried out it showed widespread septic arthritis. The infected collection was surgically drained but by this stage the infection had been in situ for such a length of time that it had caused significant damage to the structures within the hip. The Claimant was inevitably going to suffer from lifelong difficulties as a result.

Had the Claimant been referred promptly, however, and treated with appropriate antibiotics, the evidence was that she would have recovered without a need for surgery or any permanent damage.

Had both the GP practice and the hospital been sued then it is probable that the case would have been resolved prior to trial or at the very least without the Court of Appeal being troubled. However, what led to the issues of successive causes being so intensely debated was the fact that the only Defendant was the GP practice and the practice had not sought to join the hospital as Third Party.

This set up a position in which the Defence put forward by the GP practice was (i) that given the response of the hospital when the Claimant was in fact taken, it would not be appropriate to conclude that they would have acted any more swiftly in detecting the infected hip had she been admitted two days earlier. The delay by the GP therefore was of no causative significance as a matter of fact, and (ii) that although they had been negligent in not referring the Claimant, that negligence had no causal relationship to the outcome in law, as the scope of the GP's duty was limited to ensuring referral at a time when effective treatment could be given. This had been done, as had the Claimant been treated promptly and properly at the hospital when she in fact went two days later, she would have made a full recovery and not suffered permanent damage, albeit that slightly more invasive treatment might have been required. The cause of the Claimant's damage was thus whatever was done or not done by the hospital, and not the delay in referral.

The trial judge, Mackay J, accepted the first element of this argument, in finding that "the compelling criticisms of the performance of the paediatric unit make it impossible for me to find as a fact that it is more likely than not that if the Claimant had been placed in their hands during the evening of 15 April she would have been so treated as to avoid permanent damage to her hip." He accordingly dismissed the claim on the basis that the Claimant had failed to establish that the GP's negligence had caused the damage she had suffered.

The judge made no finding as to whether or not the hospital was negligent on the grounds that the hospital was not party to the proceedings and so not in a position to answer the criticisms made. But as the Master of the Rolls observed the findings and evidence were such that no other conclusion was reasonably possible for the purposes of the case before the Court.

The trial judge's finding was overturned by the majority of the Court of Appeal. However, whilst the Master of the Rolls and Dame Janet Smith agreed in that outcome (Elias LJ dissenting) their reasoning was different.

The view of the Master of the Rolls was straightforward: “in a case where a doctor has negligently failed to refer his patient to a hospital, and, as a consequence, she has lost the opportunity to be treated as she should have been by a hospital, the doctor cannot escape liability by establishing that the hospital would have negligently failed to treat the patient appropriately, even if he had promptly referred her … because by negligently failing to refer the patient promptly, he deprived her of the opportunity to be treated properly by the hospital, and, if they had not treated her properly, that opportunity would be reflected by the fact that she would have been able to recover damages from them.” (Paragraph 61)

Dame Janet Smith said “where the condition in question is not difficult to diagnose, and the correct treatment is simple and uncontentious, there must be a presumption that a hospital will provide a correct diagnosis and treatment.”

The Defendant argued, however, that notwithstanding these conclusions the GP practice should not be held liable on the second ground, namely, that the actual cause of the Claimant's injury was outside the scope of the GP's duty, and so was solely the responsibility of the hospital.

The relevant governing principle where there are successive tortfeasors are well established but was set out by the Court: the Court must look at the causative potency of the negligence of each, i.e. it must look at how common sense answers the question: what did each piece of negligence cause? The negligence of the first tortfeasor will still be an operative cause of the damage sustained following the negligence of the second unless there is negligence so “gross and egregious as to break the chain of causation and make it unjust for that reason to impose liability on the doctor” ¹⁷ i.e. unless common sense dictates that it should not be so regarded.

Thus, in Rahman v Arearose Ltd [2001] QB 351, an employer who failed to take reasonable care for the security of its employee, which led to an assault, was liable for the loss of sight consequent upon subsequent negligent medical treatment, because the cause of the original damage was still operative in creating the risk of the subsequent injury.

Similarly, in Webb v Barclays Bank plc and Portsmouth Hospitals [2001] Lloyds Rep Med 500 the employer was responsible for an unsafe path at the side of its bank. An employee tripped and fell. Subsequently the employee was found to have been negligently advised to undergo amputation of her leg. It was held that whilst the predominant cause of the loss of the limb was the negligent medical advice, the cause of the original fall was still a material operating cause and liability for the amputation was apportioned between the parties.

The Master of the Rolls held that whilst the negligence of the hospital was substantial it did not amount to a new cause which broke the chain of causation. There was composite delay, to which both the GP practice and the hospital contributed, which contributed to the adverse outcome. Had both been before the Court he was of the clear view that the damage would have been apportioned between them, the larger element falling upon the hospital.

Dame Janet Smith accepted that there was a two-stage test, as set out in Kuwait Airways Corporation v Iraqi Airways Co [2002] 2 AC 883, but held that the delay by the GP was causally linked to the outcome because “it had the effect of reducing the period available to the hospital to provide correct and effective treatment”. She added that she had found the case difficult, and regarded it as being “very close to the line”.

Elias LJ dissented. He suggested that the principle set out in the cases of Rahman and Webb had been established because, as a matter of policy, there were good reasons why the original tortfeasor who had caused the initial injury should remain liable despite negligence in the medical treatment required thereafter. His difficulty was in seeing why the GP practice, which had not caused the original condition, should effectively be liable for a delay in referral which, but for the negligence of the hospital, would have made no difference to the outcome. He did not see the additional two-day delay as being relevant where there was still ample time for the hospital, acting properly, to treat the Claimant and prevent permanent damage.

It is perhaps relevant to note that a claim could still be made against the hospital in that case and it was the expectation of the Court that a very substantial contribution would be made by the hospital in due course to the Claimant's damages. One suspects that this also affected the trial judge's decision at first instance: he may well have been more reluctant to exonerate the GP practice had there been some bar to obtaining recovery against the hospital.