Mugweni v NHS London

In a negligence claim involving competing medical explanations, causation must be assessed globally on the balance of probabilities. The court should weigh the evidence as a whole, recognising uncertainty in individual steps of a reconstructed sequence.

A defendant may be liable where a reasonably competent clinician should have detected warning signs of a developing condition. However, breach alone does not establish causation. The claimant must show that the breach probably caused or materially contributed to the damage. Where the period attributable to negligent delay is too short to cause the alleged injury, and the evidence supports a possible non-negligent cause, the claim fails.

Grace Mugweni claimed damages for severe neurological disabilities allegedly caused by the negligence of an anaesthetist during open-heart surgery at Guy’s Hospital in August 1983. She contended that the anaesthetist failed to detect a developing tension pneumothorax after closure of the sternum, causing cardiac arrest and hypoperfusion-related brain damage.

NHS London, as successor to the relevant strategic health authority, accepted that a tension pneumothorax had occurred but contended that the brain damage was a consequence of the operation and time spent on cardiopulmonary bypass. The central issues were whether there had been a breach of duty and, if so, whether that breach caused or materially contributed to the claimant’s brain damage.

1. Breach of duty. The developing tension pneumothorax would have produced changes in arterial and central venous pressure before cardiac arrest. Given the ever-present risk of postoperative tamponade, the anaesthetist’s responsibility for ventilation and perfusion, and the evidence that detectable warning signs should have been noticed, Dr Hasbury fell below the required standard by failing to act shortly after those signs appeared.
2. Assessment of causation. The court assessed the claimant’s case globally on the balance of probabilities. It did not treat each disputed step in the reconstructed medical sequence as separately established merely because it was more likely than not on the assumption that an earlier step had occurred.
3. Time available for injury. The pneumothorax could not have begun until the sternum was being brought together. The time before cardiac arrest was at most about 25 minutes, and the period of materially detectable hypoperfusion was probably only 2–3 minutes. Circulation was restored promptly after the chest was reopened. The evidence therefore did not provide the approximately 30-minute period of damaging hypoperfusion required on the medical case advanced.
4. Alternative cause. The evidence supported the conclusion that apparently uneventful cardiopulmonary bypass could cause an appreciable risk of neurological injury of the watershed type, particularly in an unusually sick infant whose cerebral autoregulation was compromised. The absence of an observed perfusion incident did not exclude that mechanism.
5. Disposition. The cardiac arrest did not produce an acute anoxic injury, and there was insufficient evidence that the negligent delay materially contributed to the watershed damage. The claimant therefore failed to establish causation on the balance of probabilities. The claim was dismissed.

First-instance decision. No prior appellate decision is stated in the judgment.