LXLP v St George's University Hospitals NHS Foundation Trust: causation fails despite admitted negligence in cerebral palsy claim

The High Court has dismissed a cerebral palsy claim against St George's University Hospitals NHS Foundation Trust, finding that although the defendant admitted negligent failures to prescribe antibiotics following preterm prelabour rupture of membranes (PPROM), the claimant could not establish on the balance of probabilities that those failures caused her injury.

The claimant, born at 29 weeks and 3 days after her mother presented with PPROM at 27 weeks and 5 days' gestation, developed bilateral periventricular leukomalacia (PVL) resulting in asymmetric four-limb cerebral palsy requiring full-time support. The defendant admitted that a ten-day course of erythromycin should have been commenced no later than 26 April 2016 and that there were further missed opportunities to prescribe. It was also established that the claimant's mother had tested positive for group B streptococcus (GBS) resistant to erythromycin but sensitive to penicillin, and that local hospital guidance — drafted by the very consultant overseeing her care — specifically called for the addition of oral penicillin in such circumstances.

Penicillin: no breach found

Despite the apparent force of the local guidance, Kimblin J found that departing from it in favour of national RCOG and NICE guidance did not constitute a breach of duty. National guidance explicitly advised against administering penicillin for GBS colonisation prior to labour, reserving its use for intrapartum prophylaxis. The court accepted the evidence of the defendant's obstetrics expert, Professor Gupta, and microbiologist Dr Gray, that prescribing oral penicillin carried its own risks: disruption to the lactobacilli-dominated vaginal flora, raising pH and potentially creating conditions more favourable to other virulent organisms including E. coli. The Bolam/Bolitho test was applied, with the court satisfied that reliance on national guidance had a sound logical basis notwithstanding the local protocol's contrary position.

Causation: the central obstacle

On causation, the claimant advanced both "but for" and material contribution arguments. Both failed. The court found that erythromycin reduces the risk of chorioamnionitis across populations but does not prevent it, and that the literature does not support prolongation of pregnancy beyond approximately one week. The neonatal medicine expert, Dr Rennie, gave unchallenged evidence that prolongation of one week would not have avoided PVL in this claimant's case, given the severity and speed of her cytokine response. Kimblin J accepted that evidence and found accordingly that the injury would have occurred in any event.

The material contribution argument was also rejected. The court held that the claimant's case was, in fact, the opposite of an indivisible injury: her case depended on reducing bacterial load to reduce severity, which is characteristic of a divisible condition. Extrapolating from population-level risk reduction data to a finding of material contribution to injury in a specific case was described as tenuous, particularly given the variability between individual microbiomes and the unknown polymicrobial aetiology of the chorioamnionitis. The court applied Williams v Bermuda Hospitals Board [2016] AC 888 and Holmes v Poeton Holdings Ltd [2024] KB 521, declining to modify the "but for" test where the evidence pointed to multiple potential causes and science could not isolate the defendant's contribution.

The case is a significant illustration of the limits of admitted negligence in birth injury litigation: an admission of breach, even a broad one, does not shift the burden of proof on causation, and statistical evidence of reduced risk at a population level will not substitute for proof of probable causation in the individual case.

Copyright & permissions