Gray v Southampton & S.W. Hampshire Health Authority

In a clinical-negligence claim, proof that injury involved an episode of hypoxia or hypotension does not by itself prove negligent management or causation. The claimant must establish, on the balance of probabilities, sufficient evidence of the episode’s nature, degree and duration, and that competent monitoring or treatment would probably have prevented the injury. A clear breach in failing to take an advised blood-gas sample does not establish liability without the necessary causal inferences. Poor records may justify criticism but cannot supply missing positive evidence. An appellate court must respect unchallenged primary evidence from a witness whom the trial judge accepted as honest, especially where the appellate court did not hear the witness.

Michelle Gray suffered severe permanent brain damage following surgery and seizures in hospital in 1984. She claimed that the respondent health authority was liable for negligent post-operative management, principally failure to detect and correct hypoxia or hypotension while she was ventilated on the E Neuro ward. Toulson J accepted that the damage involved relative hypoxia and/or hypotension but dismissed the claim because negligence had not been proved. On appeal, Gray argued that inadequate monitoring and records justified an inference of causative negligence. The health authority challenged the evidential basis for the eventual case and the findings on causation. The central questions were whether breach was established in relation to blood-pressure and blood-gas monitoring, and whether breach and causation could properly be inferred.

Appeal dismissed. May LJ delivered the judgment of the court. Rix LJ and Aldous LJ agreed.

1. The health authority’s submission that there was no evidence supporting the claimant’s eventual case was rejected. Dr Bates’s opinion had from the outset supported the case that continuing seizure activity, together with relative hypoxia and hypotension, caused the damage and that negligent management was an available evidential case.
2. The appellate court could not reject Dr Read’s evidence about taking additional blood-pressure readings. The trial judge had not rejected that evidence, it had not been challenged in cross-examination, and the appellate court had not seen or heard the witness. The court therefore proceeded on the basis that additional readings, including readings after the fentanyl administration, had been taken and were normal.
3. Although blood-pressure and heart-rate observations should have been recorded at intervals of no more than 10 minutes, the evidence did not establish that additional readings would have detected the relevant hypotension. There was no outward sign of relative hypotension, and no proper basis for determining when it occurred, how long it lasted, or whether treatment would have prevented the injury. A negligence case based on failure to detect hypotension therefore failed.
4. There was a clear breach in failing to take blood-gas samples by 18.00. However, liability still required the court to infer that earlier samples would probably have shown serious acidosis, that the damage had not already occurred, that corrective steps were then available, and that those steps would probably have prevented the damage. The evidence could not properly support all those inferences. The causes of the severe acidosis remained unexplained, and later treatment had not apparently resolved the problem.
5. Poor record keeping justified criticism but could not supply positive evidence absent from the case. The court agreed with the judge that the nature, degree and duration of the hypoxia or hypotension were insufficiently established to prove negligent causation. The remaining contentions in the respondent’s notice did not require decision. The appeal was dismissed with costs, subject to detailed assessment.

• Court of Appeal (Civil Division): In [2001] EWCA Civ 855, the court dismissed the appeal and upheld the dismissal of the claim.
• High Court: Toulson J, on 13 June 2000, found that the brain damage involved relative hypoxia and/or hypotension but dismissed the claim because negligent management had not been established.