In medical negligence litigation, the causation enquiry often proves to be the most complex aspect of a case. The recent judgment of the Supreme Court of Appeal (the SCA) in
N K obo U K v MEC for Health Eastern Cape (805/2024) [2026] ZASCA 105 provides a striking illustration, producing a split bench with two sharply divergent approaches to expert medical evidence on causation.
The case concerned a child, U K, who suffered severe brain damage manifesting as cerebral palsy. His mother, Ms N K, claimed ZAR 28.2 million from the MEC for the Department of Health, Eastern Cape. She was admitted to Dora Nginza Provincial Hospital on 3 March 2019 with severe cramps and pre-eclampsia. Over five days she was repeatedly given a labour-inducing substance but did not deliver. On 7 March a doctor ruptured her membranes and indicated a caesarean section would be necessary, yet none was performed. During the night she was left to push unassisted; vacuum extraction and fundal pressure both failed. U K was born on 8 March 2019 and was confirmed to have sustained severe brain damage. Ms N K's medical records had been lost by the hospital. The only documents available were a histology report diagnosing "severe acute chorioamnionitis with maternal and foetal inflammatory response", a blood gas analysis, an MRI and a CTG.
Both the High Court and the Full Court dismissed the claim. Although the High Court found, on Ms N K's unchallenged evidence, that the staff had been negligent, it held that the negligence was not causally connected to U K's injury. Both courts accepted the MEC's expert evidence that chorioamnionitis had set in before labour and was not attributable to the staff's conduct.
On further appeal, the SCA was divided. The minority would have upheld the appeal, holding that the lower courts erred in uncritically accepting the MEC's experts without regard to Ms N K's unchallenged testimony. The majority, however, dismissed the appeal. It held that the objective documentary evidence, principally the histology report and blood gas analysis, was pivotal. The MEC's expert evidence established that the histology revealed severe acute chorioamnionitis with vasculitis and funisitis, confirming a foetal inflammatory response. The presence of funisitis in particular indicated that the infection predated the onset of labour. Because acute chorioamnionitis is asymptomatic, it could not have been detected by CTG monitoring or routine clinical observation and was therefore not preventable. The blood gas analysis, taken from the umbilical cord immediately after birth, corroborated this finding. Accordingly, the majority concluded that the brain injury was caused by placental insufficiency arising from pre-existing chorioamnionitis, not the conduct of the medical staff during labour. The negligent conduct, however egregious, was not the factual cause of U K's injury. The appeal was dismissed with costs.
The judgment offers important reminders. First, negligence alone does not establish liability; the plaintiff must still prove causation on a balance of probabilities. Second, objective medical evidence such as histology and blood gas reports carries decisive weight and expert opinions that fail to engage with it will be found wanting. Third, courts must assess whether expert opinions rest on logical reasoning, rather than merely accepted practice. Fourth, the absence of medical records is not in itself proof of negligence – this aspect continues to prejudice plaintiffs, a situation the SCA described as disconcerting. Fifth, an adverse inference may be drawn from the failure to call witnesses who could have testified.
The split bench underscores the inherent difficulty of causation where competing experts disagree on the timing and mechanism of injury. In medical negligence, proving what went wrong is often far easier than proving that it made the difference.
