Judgment given on 29th August by the Court of Appeal in Manzi v King’s College Hospital NHS Foundation Trust gives useful guidance on whether adverse inference should be drawn from an absent witness. The full judgment can be found here.

Initial Trial

The case involved a claimant who had given birth at the defendant's hospital, but retained placenta following discharge. She suffered significant pain and collapsed, prompting a return to hospital for an ultrasound which revealed an 8cm retained mass. The claimant underwent surgery to remove the mass and suffered haemorrhage and psychiatric injury.

The makeup of the 8cm mass was the issue of contention at the first trial. The claimant alleged it was all placenta and if that was correct the experts agreed it was negligent to miss this following the delivery. The Trust argued it was a mixture of blood clot and placenta (which look very similar) and if that was right the experts agreed that there had been no negligence. A doctor had made a contemporaneous note that the mass was all placenta but the court drew no adverse inference from the fact that the Trust had not called this doctor to give evidence due to proportionality reasons. The Trust's argument was accepted by the judge.


The claimant appealed the decision claiming that the judge had erred in his evaluation of the evidence and failing to draw an adverse inference against the Trust.

In its judgment the Court of Appeal dismissed the appeal and held:

  • It is up to the discretion of the court whether an adverse inference should be drawn based on all the facts of the case. There was no obligation to do so.
  • The significance of the role of the particular witness in the index events is a relevant factor – in this case the role was described as "tangential".
  • A proper explanation for an absence is relevant – in this case a decision on proportionality grounds was taken by the Trust.
  • It is also relevant that the claimant could have called the witness to give evidence, or asked for a direction which contained a specific warning as to adverse inference. The claimant did neither.

Finally, the court also held that the proposition that a contemporaneous record was inherently likely to be accurate did not create a presumption in law that had to be rebutted in the manner submitted by the claimant.


The judgment provides a useful guide as to how the courts should approach the relevance of an absent witness. If a witness is significant to the index event then clearly it is important to try and get evidence from him/her. However, if a witness is less important then it may be reasonable not to call him/her, particularly in lower value cases.