Causation

What does causation mean?

Causation asks what difference the breach of duty actually made to the patient’s outcome.

A clinical negligence claim is not established simply by showing that care fell below the required standard. The claimant must also establish the necessary connection between that failure and the injury or loss for which compensation is sought.

This is why breach of duty and causation are investigated separately.

What is the ‘but for’ test?

The usual starting point is known as the ‘but for’ test.

The central question

What would probably have happened but for the negligent care?

If the evidence shows that the same injury would probably have occurred even if appropriate care had been provided, the breach will ordinarily not have caused that injury.

If the evidence shows that appropriate care would probably have avoided the injury, reduced it or prevented it from occurring when it did, causation may be established.

What does ‘on the balance of probabilities’ mean?

Clinical negligence is a civil claim. The ordinary standard of proof is the balance of probabilities.

In simple terms, the claimant generally has to establish that the relevant proposition is more likely than not.

For causation, that may mean showing that appropriate treatment would probably have prevented the injury, that earlier diagnosis would probably have produced a better outcome, or that a negligent act probably contributed to the harm in a legally sufficient way.

Why does causation involve a hypothetical question?

Causation often requires comparing two histories.

• What actually happened.
• What would probably have happened if appropriate care had been provided.

The second history is sometimes called the counterfactual.

Because nobody can observe a version of events in which the breach never occurred, the court has to determine that hypothetical outcome from evidence.

That evidence can include medical records, test results, scans, clinical guidelines where relevant, witness evidence, the natural history of the condition and independent expert opinion.

How does causation work in delayed diagnosis cases?

Delayed diagnosis claims illustrate the difference between breach and causation particularly clearly.

Suppose a condition should reasonably have been diagnosed three months earlier. Establishing that delay may prove breach, but it does not by itself establish what damage the delay caused.

The causation investigation may need to ask:

• what stage the condition was probably at when it should have been diagnosed;
• what treatment would probably have been offered;
• when that treatment would probably have started;
• how the condition would probably have responded;
• whether the eventual injury would probably have been avoided;
• whether the outcome would probably have been less severe; or
• whether the negligence caused an additional period of symptoms, treatment or disability even if the ultimate outcome could not have been avoided.

What if the patient would have suffered some harm anyway?

Clinical negligence often occurs against a background of illness or injury that already existed independently of the breach.

The defendant is not normally responsible for harm that would have occurred in any event. But that does not mean a claim necessarily fails if the negligence made the position worse.

Depending on the evidence, causation may concern:

• additional injury caused by the negligence;
• worsening of an existing condition;
• an avoidable complication;
• an additional period of pain or disability;
• treatment that became necessary because of the breach; or
• an outcome occurring sooner than it otherwise would have done.

What is material contribution?

Some injuries do not result from one isolated cause. They develop through a cumulative process in which negligent and non-negligent factors may both contribute.

In an appropriate case, a claimant can establish causation by showing that the defendant’s breach materially contributed to the injury or to the process that produced it.

The Court of Appeal applied this principle in Bailey v Ministry of Defence, where negligent and non-negligent factors both contributed to the claimant’s weakened condition, which led to aspiration and serious brain injury.

A later Privy Council decision, Williams v The Bermuda Hospitals Board, provides a useful explanation of the distinction. The patient’s injury resulted from a continuous process of worsening sepsis. The negligent delay prolonged that process, and the Board held that the negligence materially contributed to the injury.

Is increasing the risk of injury enough?

Usually, no.

There is an important difference between showing that negligence caused or contributed to an injury and showing only that it increased the risk that the injury might occur.

In Wilsher v Essex Area Health Authority, a premature baby developed serious eye damage. Negligent excess oxygen was one possible cause, but several different non-negligent conditions could independently have caused the same injury.

The House of Lords rejected the suggestion that it was enough merely to show that the negligence had increased one possible risk among several distinct possible causes.

What is a ‘loss of chance’ in clinical negligence?

Sometimes negligent treatment reduces a patient’s statistical chance of recovery or survival, but the evidence cannot establish on the balance of probabilities that the negligence caused the adverse outcome itself.

The leading authority is Gregg v Scott.

The claimant’s cancer diagnosis was negligently delayed. His statistical prospect of long-term survival had already been below 50% before the negligent delay and was reduced further by the time treatment began.

By a majority, the House of Lords did not recognise the reduction in that chance as a separate personal injury for which damages could be recovered on the facts of the case.

How does causation work in informed-consent cases?

Where the breach concerns failure to give adequate information about risks or treatment alternatives, causation has an additional dimension.

The investigation may need to consider what the patient would probably have done if properly informed.

For example, the evidence may concern whether the patient would have declined the treatment, chosen a reasonable alternative or deferred the procedure while considering the options.

Chester v Afshar concerned unusual facts. The surgeon negligently failed to warn of a small but serious inherent risk of surgery. That very risk then occurred. The majority treated the causal requirement as satisfied even though the claimant could not establish that she would never have undergone the surgery.

Later Court of Appeal decisions have emphasised that Chester is exceptional and narrow. In Duce v Worcestershire Acute Hospitals NHS Trust, the Court confirmed that Chester does not generally remove the need to prove a causative effect from the failure to warn.

What if there are several possible causes of the injury?

A patient’s condition can sometimes have several medically plausible explanations.

The fact that negligence is one possible cause does not automatically prove that it was the legal cause.

The evidence may need to distinguish between:

• several factors that acted together to produce one injury;
• several separate conditions, any one of which could have caused the injury;
• a pre-existing condition and additional negligent injury; and
• uncertainty about the medical mechanism itself.

Wilsher shows why this distinction matters. Where several distinct agents could independently have produced the same injury, it was not enough simply to identify the negligent agent as one possibility.

By contrast, Bailey and Williams involved findings that negligent and non-negligent factors had actually contributed to a cumulative process producing the harm.

What role do medical experts play in causation?

Independent medical expert evidence is often central to causation.

A causation expert may be asked to consider the patient’s condition, the relevant chronology and what probably would have happened if appropriate care had been provided.

Depending on the case, the expert may address:

• the natural progression of the underlying condition;
• when a diagnosis probably could and should have been made;
• what treatment would probably have followed;
• likely response to earlier or different treatment;
• whether the breach caused or worsened an injury;
• whether several factors contributed to the outcome; and
• the patient’s likely condition in the absence of the breach.

Different specialties may be needed for breach and causation. The clinician best placed to say whether earlier care was required is not necessarily the expert best placed to explain the effect that earlier care would probably have had.

Can breach be admitted while causation is disputed?

Yes.

A healthcare provider may accept that part of the care fell below the required standard while denying that the failure caused the injury being claimed.

For example, it may be admitted that a scan should have been performed earlier, while remaining disputed whether an earlier scan and resulting treatment would probably have altered the outcome.

This is not an unusual contradiction. It reflects the fact that breach and causation are separate elements of negligence.

Why can causation be the hardest part of a clinical negligence claim?

Clinical negligence often concerns patients who were already ill or injured before the negligent event.

The law must distinguish the consequences of the underlying medical condition from the consequences of the negligence, often by reconstructing a hypothetical course of events that can never be directly observed.

That can require detailed chronology, multiple medical specialties, statistical evidence and careful consideration of competing explanations.

A case can therefore involve an obvious failure in care but a genuinely difficult causation question.