Wardlaw v Farrar

In clinical negligence claims, causation must be determined on the balance of probabilities by considering all relevant evidence, including evidence particular to the patient and evidence arising after the negligent delay. Statistical evidence may be given appropriate weight, but it cannot displace individual evidence or safely predict an outcome where the underlying study lacks an agreed definition of a critical category. An appellate court should be slow to interfere with a trial judge’s assessment of expert evidence. A new legal argument may be raised on appeal only where it would not affect the factual basis on which the evidence was given. A transcript-based argument that should have been clearly advanced below cannot be entertained.

Peter Wardlaw appealed against Judge Singer’s order in the Manchester County Court on 14 August 2002. The judge awarded him £1,000 in his claim, as widower and personal representative of his wife, under the Fatal Accidents Act 1976 and the Law Reform (Miscellaneous Provisions) Act 1934.

Dr Farrar had negligently failed to consider pulmonary embolism or arrange hospital admission on 14 September 1997. The judge found that the delay caused a week’s pain and suffering but did not materially contribute to Mrs Wardlaw’s death. The appeal challenged the judge’s assessment of expert evidence, his treatment of statistical evidence, a proposed new causation argument based on the transcripts, and the refusal of fresh evidence. The central issue was whether the negligent delay caused or materially contributed to the death.

Brooke LJ delivered the judgment, with Latham LJ and the President agreeing. The appeal was dismissed.

1. Appellate restraint. The court reiterated that an appellate court should be very slow to interfere with a trial judge’s assessment of expert evidence, because the trial judge had seen and heard the witnesses. The principle stated in Wilsher v Essex Area Health Authority [1988] AC 1074 at 1091 applied. There was no basis for interfering with Judge Singer’s preference for Dr Hill’s evidence.
2. Causation and statistical evidence. In deciding causation on the balance of probabilities, the judge had to consider all relevant evidence and the rival cases advanced at trial. The failure of anticoagulation therapy to prevent a subsequent massive pulmonary embolism was material evidence. Statistical evidence could be given appropriate weight, but the court could not disregard evidence placing the individual patient in a particular category. The judge was entitled to reject the ICOPER statistics as an unsafe means of predicting the outcome for a particular patient, since the study lacked an agreed definition of haemodynamic instability.
3. New argument on appeal. The court could entertain a new point of law where it would not have affected the factual basis on which evidence was given. It could not entertain an entirely new transcript-based argument where the evidence might have been different had the point been clearly taken below. The attempted reliance on Hotson v Berkshire Area Health Authority [1987] AC 750 did not establish that probabilities had to be assessed only when negligent treatment began.
4. Fresh evidence and case management. The refusal of further medical literature as fresh evidence was justified because it could have been obtained with reasonable diligence before trial and added nothing material. The court also emphasised, as wider guidance, the importance of directions requiring timely disclosure of expert literature and agreement of the trial bundle in multi-track clinical negligence cases.
5. The judge was entitled to conclude that the delay caused pain and suffering but did not increase the risk of death or materially contribute to it. Appeal dismissed.

• Court of Appeal (Civil Division): On 27 November 2003, the appeal was dismissed. The court upheld Judge Singer’s causation findings and award of £1,000.
• Manchester County Court: On 14 August 2002, Judge Singer found negligence causing a week’s pain and suffering, but no material contribution to the death, and awarded £1,000 damages.