Gouldsmith v Mid Staffordshire General Hospitals NHS Trust

In a clinical-negligence claim based on an omission, causation first requires a factual finding, on the balance of probabilities, of what would have happened had the omitted act been performed. The Bolam standard does not govern that counterfactual inquiry. The separate question whether a failure to take a particular step would itself have been negligent arises only if the factual inquiry does not establish causation.

Where the breach was a failure to refer, accepted evidence that most appropriate specialist units would probably have performed the beneficial procedure could establish that it would probably have occurred. The claimant was not required to identify the particular specialist who would have treated her where the defendant had provided no evidence to displace that inference.

The claimant suffered progressive ischaemia in her dominant left hand and ultimately lost all its digits. She alleged that the respondent hospital negligently failed to refer her promptly to a specialist vascular unit. The trial judge found that referral should have occurred in the first week of October 1999 and that surgery by 13 October would probably have saved three digits. He nevertheless dismissed the claim because he was not satisfied that every competent specialist would have operated.

The claimant appealed on causation. The central issue was whether, had the required referral been made, surgery would probably have been undertaken and would have prevented the relevant amputations.

1. Appeal allowed by a majority. Pill and Wilson LJJ held that the judge had erred in his causation analysis. The finding of negligent failure to refer was upheld. The respondent's challenges to the findings of continuing embolisation and to the critical date for surgery were rejected.

2. The majority applied the causation analysis in Bolitho v City and Hackney Health Authority [1998] AC 232. The first question was factual: what would probably have happened upon referral to a specialist unit? The Bolam test was irrelevant to that question. Only if surgery would not in fact probably have occurred would it be necessary to ask whether a failure to operate would itself have been negligent.

3. The judge had impermissibly merged those questions by asking whether every specialist acting to the proper standard would have operated. His acceptance of Professor McCollum's evidence that most specialist vascular units would have operated answered the factual question in the claimant's favour. Professor Bell's contrary view did not defeat causation because it rested on the diagnosis which the judge had rejected, namely that the subclavian lesion was irrelevant.

4. The majority held that it would be unjust to order a retrial. The respondent had directed its case to denying the need for referral and had not directly contested the likely consequences of a referral on the factual basis ultimately found. Damages were remitted to the High Court.

5. Maurice Kay LJ dissented. He considered that the claimant had not proved the likely destination or treatment of the hypothetical referral, and that evidence that most units would operate did not establish what the unidentified treating specialist would probably have done.

6. Gregg v Scott [2005] 2 AC 176 concerned a different issue. Once factual causation was proved on the balance of probabilities, no question of loss of a chance arose.

• Court of Appeal (Civil Division): By a majority, allowed the claimant's appeal on causation, upheld the finding of breach of duty, and remitted damages to the High Court: [2007] EWCA Civ 397.
• Queen's Bench Division, Liverpool District Registry: His Honour Judge Gilmour QC, on 19 December 2005, found negligent failure to refer but dismissed the claim because causation had not been proved.