The Facts Speak for Themselves: Understanding How Courts Infer Negligence in Medical Cases

One of the hardest things for a patient to prove in a medical negligence case is what actually went wrong.

Patients lack the necessary medical expertise and often do not have access to the full clinical records or explanations required to understand what occurred during the incident or treatment. Courts therefore sometimes have to decide whether negligence can be inferred from the surrounding facts, especially when the key information sits with the hospital or doctor.

A phrase that often comes up in this context is res ipsa loquitur, a Latin expression meaning “the thing speaks for itself.” In simple terms, it refers to situations where the facts strongly point towards negligence, even if there is no direct eyewitness or proof of exactly what happened. South African courts have repeatedly said, however, that this is not a legal shortcut, and it does not necessarily mean the healthcare provider was negligent. It is simply a way of reasoning to get from proved facts to the most likely explanation.

The Supreme Court of Appeal explained the point in The MEC for Health of the Gauteng Provincial Government v CBM [2026] ZASCA 80 as follows:

“The maxim is part of inferential reasoning applied when evaluating conflicting or mutually contradictory evidence, to determine whether a particular factual inference can be drawn from the facts, having regard to the probabilities and the credibility of the witnesses. It is not a presumption of law or fact. The question is simply whether, in the case of a civil dispute, the probabilities support a particular inference as the more probable one”

That distinction is especially important in medical cases. From Van Wyk v Lewis (1924 AD 438) to Buthelezi v Ndaba (2013 (5) SA 437 (SCA)), courts have warned that medicine is complex, and that a bad outcome does not by itself prove negligence. As our courts have put it, a doctor is not negligent simply because something went wrong. The question is always whether the proved facts support the conclusion that negligence is the probable explanation.

South African law has developed this approach over many years. Early cases such as Union Government (Minister of Railways) v Sykes (1913 AD 156) recognised that, where important facts are peculiarly within the knowledge of one party, less evidence may be needed from the other side to make out a case. Later decisions, including Macleod v Rens (1997 (3) SA 1039 (E)) and Goliath v MEC for Health, Eastern Cape (2015 (2) SA 97 (SCA)) made it clear that res ipsa loquitur is not a separate doctrine or presumption. It is merely part of the court’s common-sense process of inference by looking at the facts that are proved, weighing probabilities, and deciding which explanation is more convincing.

It does not mean courts are free to speculate. South African law draws a firm line between reasonable inference and guesswork. In civil cases, the claimant does not have to prove that their explanation is the only possible one. They do have to show that it is the more acceptable and likely inference on the evidence.

That is where the idea of an adverse inference comes in. An adverse inference arises when a party who could have produced helpful evidence does not do so. The principle is reflected in Elgin Fireclays Ltd v Webb (1947 (4) SA 744 (A)), where the court said that a party’s failure to place available evidence before the court “leads naturally to the inference that they fear that such evidence will expose facts unfavourable to them”. But that inference is not automatic, it depends on whether the evidence is genuinely available and whether it would have helped explain the facts. If the evidence is not available, the defendant should make that fact known to the court.

In HAL obo MML v MEC for Health, Free State [2022] 1 All SA 28 (SCA), the majority of the Supreme Court of Appeal treated missing records as a neutral fact, where there was a plausible explanation for their loss and no proof that they were deliberately withheld. But the position can change where doctors, nurses, or other staff who were directly involved, are not called to testify, and no proper explanation is given for their absence. A practical example is Ntsele v MEC for Health, Gauteng Provincial Government [2013] 2 All SA 356 (GSJ), where the concern was that the people who knew what had happened during treatment were the hospital staff, yet they were not called to explain it.

The recent Supreme Court of Appeal decision in CBM shows how this works in practice. A patient required a hysterectomy following a caesarean section. The question was whether negligence during the operation caused the complication. The court noted from the evidence that there was a very high probability against a hysterectomy becoming necessary in the ordinary course of a caesarean section unless a culpable mistake had occurred. The court repeated that res ipsa loquitur should not be treated as a rule that automatically proves negligence in medical cases. At the same time, the court accepted that it is still part of inferential reasoning. The doctors who performed the caesarean section were not called to give evidence, and their absence was not properly explained. Because the critical facts were within their knowledge, and because there was no rebuttal evidence from the healthcare provider’s side, the court held that negligence could be inferred from the proved facts and probabilities.

On the issue of whether the delay in treatment amounted to negligence, the court preferred the evidence of the expert called for the healthcare provider, whose opinion was grounded in the proved facts and applicable guidelines, over that of the expert called for the patient, whose evidence lacked proper foundation and whose registration had been suspended. That reflects the caution referred to in Kennedy v Cordia (Services) LLP [2016] UKSC 6, that expert evidence must be reasoned and supported by facts, not simply asserted.

For liability insurers, the practical lesson is clear. The strength of a defence depends not only on what the patient can prove, but on what the insured can explain. It is not enough to rely on weaknesses in a patient’s case if the hospital or doctor has relevant staff or practitioners who can explain what happened but does not call them.

Insurers and healthcare providers should preserve clinical records, identify key witnesses early, and make sure that the treating staff and experts can give a clear, evidence-based account. If records are missing, the reason should be documented. If witnesses are unavailable, that too should be explained.

The phrase res ipsa loquitur is not a magic formula and does not shift the burden of proof. Patients must still prove negligence. But where the key facts are known only to the hospital and its staff, courts will expect the defence to provide answers. Failing to do so can make the court more willing to infer negligence.