In cases of alleged medical negligence, it is often more difficult to establish that the failure of care affected the outcome rather than whether there was sub-standard medical care.

It is necessary to show that, on the balance of probabilities, the patient would have fared appreciably better if he or she had received proper care.

It is not enough to show that the patient lost the chance of benefiting from proper care; instead it has to be proved that more likely than not (ie at least 51%) he or she would have fared better had he or she received proper care (Hotson v Fitzgerald [1987] AC 750).Many cases have failed on causation, either at the outset or at trial.

There are the loci classici such as Barnett v Chelsea and Kensington Hospital Management Committee [1969] 1 QB 428 where a failure to administer an antidote to an arsenic victim did not lead to compensation for the widow because the judge concluded that the antidote could not have been administered in any event soon enough to avoid death; and Robinson v Post Office [1974] 1 WLR 1176, where a failure to administer a test dose of an anti-tetanus serum before the full dose did not attract compensation for injury suffered from the full dose because the judge held that in the time available to the doctor the test dose would not have revealed any contra-indications to the injection of the full dose.More recently, a failure to offer an amniocentesis to a pregnant mother did not lead to compensation for the birth of a handicapped child because the judge held that, even if the possibility of the test had been mooted, the woman would have accepted her consultant's advice not to go ahead with it (Gregory v Pembrokeshire Health Authority [1989] 1 Med LR 81).An unusual, and important, aspect of causation recently went to the Court of Appeal in the case of Bolitho v City and Hackney Health Authority [1993] PIQR P334.

In this case a child was ill in hospital; it was agreed that it was negligent that during the night no doctor had responded to a call made by the night sister; it was agreed that if a doctor had come and had intubated the child the cardiac arrest and brain damage that he went on to suffer would have been avoided.

One might think that the defendants would have paid up, but they argued (successfully) that the plaintiff could not prove that if the doctor had come she would probably have intubated.

Each side provided a leading expert on paediatric respiratory problems, both of whom were accepted by the judge as representing a responsible body of medical opinion (ie so as to discharge the Bolam test of professional negligence).

The plaintiff's expert said it would have been mandatory to intubate; the defendants' expert said he would not have intubated.

The doctor who should have responded to the call from the nurse - Dr Horn - said that she would not have intubated (the judges were alive to the 'self-serving' nature of this evidence).Faced with this conflict of medical opinion the judge held that the plaintiff had not proved that the outcome would probably have been different if the doctor had responded to the nurse's call.The Court of Appeal upheld this decision by a majority.

Dillon LJ said that if one accepted the evidence of Dr Horn that she would not have intubat ed the child, the plaintiff could not succeed unless it could be shown that it would have been negligent not to intubate.

Clearly, upon the evidence of the defendants' expert there was a responsible body of medical opinion that would find the decision not to intubate to be acceptable.If one felt impelled to ignore the evidence of Dr Horn as possibly self-serving, there remained the two schools of thought about the need to intubate and, said Dillon LJ, there was no material which entitled the court to say that on the balance of probabilities the child would have been intubated if the nurse's call had been responded to.Farquharson LJ said that the issue between the parties which had to be resolved was what Dr Horn would have done had she come: 'Whether Dr Horn's failure to appear would have made any difference in the event depended upon what she would have done had she been present.' This is clearly correct, but the issue was muddied to an extent by the apparent failure of the trial judge to decide on what basis he was approaching the question of causation.

At one point he was saying, rightly, that the issue was what the doctor would have done had she come.

At another point, he appeared to be saying that if it would have been within the bounds of acceptable medical treatment not to intubate the child, the plaintiff could not succeed.This is clearly wrong.

The fact that it would have been acceptable either to intubate or not to intubate may be useful evidence on the question of what the doctor would have done, or may make it impossible for the court to decide that question one way or another (in which case the plaintiff fails).

However, the issue is not whether it was acceptable not to intubate but whether the child would in fact have been intubated (or, if not intubated, whether that would have been negligent).

Farquharson LJ was led astray in this context when he said: 'It was for the plaintiff to prove that [Dr Horn] would probably have intubated, and further that if she did not do so [that] failure was contrary to accepted medical practice.' The words 'and further' must be replaced by the word 'or'.Simon Brown LJ disagreed.

He said that the trial judge had approached the matter on the wrong basis, in that he had rejected the claim on the ground that there was a body of medical opinion that would not have intubated.

As the trial judge put it: 'It has not been proved that any competent doctor in the position of Dr Horn would have intubated Patrick before his collapse.' Quite rightly, Simon Brown LJ said that was not the issue.

Causation issues are not, he said, to be decided according to some sort of application of the Bolam test, but by deciding what would probably have taken place if proper care had been given.In place of the trial judge's flawed conclusions Simon Brown LJ substituted his own conclusion that on the evidence, whatever the defendants' expert, and Dr Horn, might say in retrospect about what would, should or could properly have been done at the time, the only reasonable inference was that any doctor who had responded to the nurse's call would, in the circumstances prevailing in the ward at the time, have intubated.I find this last an extremely bold conclusion, given that the defendants' expert's opinion that he would not have intubated was based on the medical history and the clinical condition of the patient prevailing at the time.

One may perhaps be forgiven for thinking that Simon Brown LJ was swayed by sympathy for the child and perhaps also by antipathy towards a defence based on a last-ditch stand o n causation when not only was negligence admitted but also it was admitted that intubation would have avoided the injury.From the legal point of view it is difficult to fault the decision of the two courts, given the evidence of the defendants' expert and (less importantly though not negligibly) that of Dr Horn.

I have to conclude that Dillon LJ was right when he said that there was no evidence entitling the court to conclude that a doctor responding to the nurse's call would have intubated (or, I would add for completeness, would have been negligent not to intubate).