Medical causation


Medical causation

This is the third hurdle a claimant needs to jump. 

This is factual causation. This is judged on the 'but for' test on the balance of probabilities. 

There are difficulties with factual causation in medical negligence because the patient is already ill- the unfortunate event may have happened anyway. 

Complex causation in clinical cases- 

  • The majority of patients who see a Dr and subsequently consider bringing a claim were not in good health anyway, or they would not have needed the medical attention.
  • Most have a pre-existing illness or injury which means they are more vulnerable to further injury, it is quite possible that even with perfect care they would not have been 100% healthy after medical treatment. 
  • The aim of tort is to put them back in the position before the negligence occured. 
  • More so than with other personal injury cases, therefore it is necessary to produce evidence that the injury for which compensation is sought was a result of the breach of duty and would not have occurred but for the breach of duty. 
1 of 16

Factual causation

The but for rule protects the medical profession. Prevents a flood of claims. 

The main case for simple application of the rule is Barnett v Chelsea Hospital 1969- Nightwatch man on duty. He had an alcoholic drink and some intruders harmed one of them. He went to hospital, was let out and went back to work. At 5am all three had tea and started to vomit. 8am they went to hospital. Left in A+E as was NYE so the Dr didnt want to see them, thinking they were just drunk. 1pm Barnett is readmitted to hospital but dies on arrival. 

The poison antidote wouldn't have worked before he died at 8am. The 'but for' test was not satisfied. 

Hindsight is relevant in causation cases- unlike in breach (Roe v MoH). After the event scientific knowledge can put a figure on the chances of the wrongdoing causing the harm and experts are called to give evidence in this respect. Has a lot of relevance. Have to seek expert opinion to know what could have happened. 

Multiple causes- where there are competing causes of potential injury, the patient may recover compensation if he/she can prove that the breach of duty 'materially contributed' to the injury. 

2 of 16


Material contribution is a contribution which is more than negligible. Concurrent cases= at the same time. Consecutively= after each other. Concurrent makes it hard to determine the but for rule. 

Bonnington Castings v Wardlaw 1956- man worked at steel works for 8 years. Exposed to dust. Suffered from scilicosis. Guilty dust came from sawing grinders. Legally had to clear this up. Other dust was naturally in the factory. Couldnt extract which dust caused the harm, this was indivisible harm. The but for test wouldnt work. He won his case and the full claim against the factory was allowed. Defendants didnt argue that it was divisible harm. Based on the whole amount. It was not possible to identify or apportion the extent to which the negligent exposure contributed to the disease. Material contribution is not de minimus. Contribution doesnt have to satisfy the balance of probabilities. 

With technology this harm may no longer be indivisible. 

Lord Reid- cant distinguish which type of silica caused the disease. 'The disease is caused by the whole of the noxious material inhaled, and if that material comes from two sources, it cannot be wholly attributed to one source or the other. It appears to me, that the source of his disease was the dust from both sources, and the real question is whether the dust from the...

3 of 16


swing grinders (the guilty dust) materially contributed to the disease. What is a material contribution must be a question of degree.' 

Bailey v MoD 2008- where the injury caused by the breach of duty is indivisible from the non negligently caused injury, the claimant can recover for the entirety of his injury. In both these cases the harm was indivisible and so you may get one defendant to pay for all the damage.

If the injury caused by the breach of duty can be distinguished from injuries that would have occurred in any event, then the claimant can only recover the proportion of the injury that the defendant caused. The burden is on the claimant to prove that the additional injury was, on the balance of probabilities, caused by the defendant. 

Wilsher v Essex AHA 1988- premature baby looked after by junior and senior Dr. Gave too much oxygen and baby goes blind. Could you prove breach caused the damage? There were 5 potential causes each on their own, could've caused it but none caused it on the balance of probabilities. 

The judges didnt want to apply an industrial case to medical negligence. They didnt talk about material contribution, and distinguished Bonnington. 

4 of 16


The HoL adopted the persuasive precedent of Lord Browne Wilkinson in CoA of this case.

Lord Hoffmann considers policy- with medical negligence claims 'political and economic arguments involved in a massive increase of liabilty for the NHS... are more complicated than the reasons for imposing liability on an employer who has failed to take a simple precaution.' 

The HoL said it should be retried using the test of causation outlined. By the time the case came to the HoL Martin was 9 and now they were back to square one. Finally a settlement was reached, so a retrial never occurred. Martin was paid a sum of money originally order at trial but without interest (this would have doubled the amount). Ironically in later years research into his condition showed that excessive oxygen is the only likely cause of Martin's blindness. If this had been known at the time of trial, he almost certainly would have won his case as he'd have been able to prove this was the cause on the balance of probabilities. 

Special rules in Fairchild- it wasn't known how mesothelioma developed (ie from a single strand or otherwise). Here there were several tortious exposures, but each one individually may not pass the but for test. HoL said claimant could recover if D had materially increased the risk of harm but scientific knowledge about how it is caused must be impossible. It wouldn't apply to Wilsher as they were distinctly different causes. 

5 of 16


They adopted Bonnington. Allowed for further risk of harm but only for mesothelioma. Only if you dont know where the harm came from. 

Barker v Corus- proportioned the damages as he was self employed for a bit. 

Cumulative cases- if it can be established that the final harm was on the balance of probability caused by a non tortious reason, then the claimant cannot recover in full. (Wilsher).

However if it is difficult to distinguish how the final harm was caused and the causes operate in combination, it will not be necessary to prove causation on the balance of probabilities. A claimant simply needs to prove that D materially contributed to that harm in breaching the duty of care. 

Bailey v MoD is significant as the material contribution to damage test had been ruled out in Wilsher, Gregg and Hotson. May be able to use Bonnington harm?

Bailey v MoD 2008- B went to hospital because she had gall stones. Routine investigative operation to try to extract the stones. Unsuccessful as had internal bleeding. Negligently treated in a series of events. Not resuscitated properly, not checked on etc. 

6 of 16


She did have complications but there was clear negligent treatment also. She also succummed to pancreatitus, which happened naturally. Went for operation, and 17 days later she vomits and chokes. Shes in such a weak state from the treatment that she suffers cardiac arrest and brain damage (in the second hospital she was moved to). They werent negligent. She alleged the first hospital was. She said they caused her final harm. Pancreatitus caused weakened state. But her bad care also contributed. Which rule should be used as tortious and non tortious conduct? 

Bonnington can now be applied to medical negligence. If they don't know through science, the judge may be more claimant friendly. Evidential uncertainty may always be a negative to defendants. 

Waller LJ- 'if the evidence demonstrates on the 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 tortious causes contributed.' 'If the evidence demonstrates that but for the contribution of the tortious cause, the injury would probably not have occurred, the claimant will have obviously 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 to the negligent cause was more than negligible...

7 of 16

Canning- Kishner

the but for test is modified and the claimant will succeed.' So if the pancreatitius caused the weakened state she would fail. This approach separates natural from torts. Accepts Bonnington can apply to medical negligence.

There has been critique of the case. Some said normal but for should have been used. May open flood gates as many patients will have weakened states. 

Canning- Kishner v Sandwell NHS 2008- Claimant is premature and goes to specialist unit. Monitored hourly. Many complications and 6 days after birth, she is extubated (they take the tubes out). Next 36 hours are touch and go. Allegation that 2 days after extubation her heart rate falls, she is in a weak state, has a cardiac arrest and suffers brain damage. The nurses were negligent in not getting a Dr to arrive on time. They admitted negligence. Couldn't be shown the brain damage was caused by the delay though. 

They applied Bailey reasoning. On balance the court said they couldnt prove they'd have brain damage but couldn't prove the delay caused it. Delay was a possible cause and she won the case as it was a material contribution. The non tortious cause was not a probable cause, but there was a material contribution (more than negligible) as a result of the delay. 

8 of 16


Williams v Bermuda Hospital 2016- claimant has stomach ache and Dr suggests a CT scan. Suspected apendicitus. It ruptures at 3.15pm. Such a long delay between ordering the scan and results that his sepcis gets worse. Appendicitus is diagnosed at 7pm. If operation had taken place when it should have he would've been cured. Not operated on until 9.30pm. 2 hours and 20 minutes of negligent delay. Harm is the sepcis. The sepsis developed incrementally over a period of approx 6 hours, progressively causing another problem. It was a single continuous process that affected the heart and lungs. Was it all from the tort or nature as well? Indivisible harm. Kind of consecutive case. 

The trial judge used the 'but for' test to refute the claim. 

The CoA said the proper causation test was not 'whether the negligent delay and inadequate system caused the injury, but whether the breaches of duty contributed materially to the injury. It was considered that Bailey v MoD was a case where the traditional but for test is modified. An academic said the CoA could have halted the relaxation of the but for rule. He had started to had rupture anyway and so had to pay for whole delay. 

Privy council- The defendant tried to distinguish Bonnington Castings by suggesting that in Bonnington the inhalation of dust from the two sources was simultaneous (concurrent)

9 of 16


whereas the development of sepsis in the claimant was not. The way it was submitted was that sepsis had begun already as a consequence of the claimants underlying condition and that the subsequent delay only caused additional/more sepsis to develop, as such D should not be liable for the whole amount of loss. They wanted apportioned blame as the chain of events was a series. Sarah Green agrees with the defendants. The Privy Council did not accept this argument. 

Privy Council- found that the injury sustained by the claimant was indivisible. It was not possible to distinguish the extent of the injury Mr Williams would have sustained in any event from that which resulted as a direct consequence of the delay in performance of the procedure. Where there is delay and the claimant has underlying issue, if the court thinks that delay contributed even in a very small way, you can win. 

Although not said explicitly, the PC found the injury sustained by the claimant was indivisible. It was not possible to distinguish the extent of the injury Mr Williams woudl have sustained in any event from that which resulted as a direct consequence of delay in performance of the procedure. As a matter of principle, successive events are capable of each making a material contribution to the subsequent outcome. 

10 of 16


The decision will potentially have implications for many kinds of clinical negligence claims including cases involving birth injuries and delayed diagnosis of an illness, such as cancer, where the injury is indivisible. If it can be established that a negligent period of delay made a material contribution to the outcome then the claimant can succeed in full. They didnt care if consecutive or concurrent. 

Where it can be established that the injury is divisible, it is arguable that the claimant may not be entitled to recover the full value of the claim, and can only be compensated for the additional injury caused by the negligence. But for will still apply in this scenario and damages can be apportioned. 

Critiques of the case- Sarah Green thought that the court could have worked out when the sepsis occurred and the damages could be apportioned.

Laura Sutherland- 'there appears to have developed in the cases a difference in approach depending on whether injuries are divisible or indivisible. This may ultimately be extremely important in terms of what can be recovered for a claimant. However, the courts have not given clear guidance on what constitutes divisible or indivisible injury.' 

11 of 16

Where are we now?

Divisableness of harm is the first thing considered. With technological advances, things may change to divisble. 

Are there similarities between Williams and Barnett?

  • Couldnt the question have been 'on attending hospital was the patient more likely than not to have developed complications from appendicitis?'
  • If the answer is yes, then breach would be irrelevant.
  • Here there was chronological space between factors.
  • Is Williams effectively making the medical profession effective insurers of their patients? 
  • Cant say but for the delay he wouldnt have got the disease. 

Where are we now? It is a mess.

  • The but for test remains the standard causation test to be applied. The test is but for the defendants breach of duty, would the injury on a balance of probabilities have been avoided. 
  • Where the cumulative causes and the injury is a divisible one (eg lung cancer caused by smoking or exposure to asbestos) it is correct to seek to apportion the extent to which the...
12 of 16

Where are we now?

  • the injury is caused by the breach, essentially an application of the but for test in order to make that apportionment. 
  • Where the injury is indivisible and there are cumulative causes of which at least one is attributable to the defedants breach of duty, the but for test can be applied more broadly and provided the defendants breach of duty contributed materially to the injury, causation will be established. For these purposes, material means more than minimal. 
  • This is easier to prove. It is not on the balance of probability, just more than minimal. Can get all damages even if they have a pre-existing condition. 
  • Where there are cumulative causes it does not matter whether they are concurrent or successive. (Williams). 

Where there are a number of disparate factors causing the harm, as in Wilsher, the but for rule still applies. 

After Williams, where there are separate but cumulative causes, it may be possible to categorise the non negligent cause as effectively being an application of the egg shell rule and of no causative effect. 

Everyone at a hospital has a pre-existing condition. 

13 of 16

Legal causation

Remoteness- in addition to proving factual causation, a claimant must prove that the type of harm they have suffered is reasonably foreseeable. This is usually easy to prove in medical negligence cases and easier than factual causation. 

The Wagonmound- main rule. The type of harm must be foreseeable. 

R v Croydon HA 1997- claimant had medical examination as applied for a new job. Given X ray and discovered that she suffers from PPH. This is a condition that could cause sudden death if you become pregnant. Purpose is for her job. Radiographer doesnt tell her but tells employer. She gets pregnant and gives birth to a healthy child. She has serious complication has to have histerectomy, depression and loses her job. Sues radiographer for failure to tell her. She sues for wrongful birth and pain and suffering from pregnancy and bringing up a child. The radiologist hadnt met the claimant. They were simply examining the X ray.

The type of damage was to warn of risks associated with fitness to work. There was no duty to warn about things that would affect domestic damage. Acknowledged award for pain, but not for having a child. 

14 of 16

Legal causation

Page v Smith 1996- psychiatric shock. Man involved in low speed car crash. Not his fault. He didnt really suffer physical injury but he had his ME triggered by the accident. Claimed for this. Can claim for psychiatric injury if physical injury was foreseeable. 

Group B case 2000- children who had dwarfism were involved in clinical trials in 1977. Injected with growth hormone. Later discovered this could infect the children with mad cow disease. Children claimed for psychiatric damage for fear of contracting disease. 

This was a fear that most would suffer, particularly given the media coverage. They were allowed their claim as this was reasonable. 

Novus Actus Interveniens- If the chain of causation breaks as a result of an unforeseeable act by the claimant or a third party, D will not be liable for the full extent of the injuries. 

Hyde v Tameside 1981- goes to hospital with a shoulder injury. He starts to get depression, thinking the injury is really bad. Attempts suicide by jumping out of window and is seriously injured. Trial judge said should be liable as they should've been on notice and didnt break chain. CoA said suicide was too remote and NAI. They weren't aware of the extent of depression. Denning was against recognising suicide as an NAI but in this case he broke the chain. 

15 of 16


Wright v Cambridge Medical Group also. 


Contributory negligence- St George v Home Office 2008- prisoner was a drug addict and alcoholic. Suffers from withdrawal and has a lot of seizures. Placed in a top bunk and has seizure, falls out and suffers severe head injuries and brain damage. Trial judge said he was contributory negligent as he was a drug addict. He reduced the damages but only by 15% (the same as not wearing a bicycle helmet.) The CoA said they shouldnt take into account the alcholism, and drugs. The prison guards were there to help. Not contributory negligence. 

Illegality- Clunis v Camden HA 1998- he had a mental problem and they let him out of hospital. He killed someone and claims he shouldnt have been allowed out. There is a statutory duty under Mental Health Act saying they dont owe a duty to every patient. Defendant could have a defence of illegality. He cannot profit from his crime. 

Remoteness/legal causation needs to be mentioned but not lots of detail needed. 

Jones v Royal Wolverhampton 2015- similar case to the seminar question. 

16 of 16


No comments have yet been made

Similar Law resources:

See all Law resources »See all Tort resources »