A Court of Appeal in Northern Ireland has ruled that a mother convicted of murdering her baby son did not receive a fair trial, finding that prosecutorial misconduct, unreliable expert witness evidence, and inadequate judicial oversight collectively undermined the integrity of the 2023 proceedings. The original convictions — for murder and attempted murder — were quashed in 2025 and a retrial ordered. The woman subsequently pleaded guilty to infanticide and attempted murder, and was sentenced to ten years’ imprisonment.
The appeal court’s judgment, made public only after the conclusion of the retrial proceedings, identified three compounding failures: the conduct of the prosecution, the evidence of a psychiatric expert witness, and the handling of the trial by the presiding judge. “The serious errors in the conduct of this trial which began with the prosecution’s misconduct were compounded by the expert witness and the judge,” the court stated.
A Theory Introduced Outside the Record
At the centre of the appeal court’s concerns was Dr Richard Bunn, a psychiatrist who gave evidence for the prosecution at the 2023 trial. During his testimony, Dr Bunn introduced the theory that the most likely motive for the stabbings was “spousal revenge” — a claim that appeared nowhere in his written reports submitted to the court.
It subsequently emerged that Crown counsel Richard Weir KC had provided Dr Bunn with an article on spousal revenge prior to his taking the stand. The two men also held a telephone conversation lasting nearly an hour before Dr Bunn gave evidence, a call that was not recorded.
The problems with Dr Bunn’s evidence did not end there. The appeal court found that a significant quotation attributed to the defendant in his report — “he destroyed my life, so I destroyed his” — did not appear in his original clinical notes. His notes recorded only the phrase “he destroyed my life.” When challenged on this discrepancy during the appeal proceedings, Dr Bunn initially accepted that he had lied about what the woman said to him during his examination, before withdrawing from that position.
The Court of Appeal concluded that “Dr Bunn’s evidence was not evidence that a jury could rely on in a case of this nature.”
Prosecution Conduct Below Recognised Standards
The appeal court’s assessment of Crown counsel Richard Weir KC was equally severe. The court found that “the prosecutorial conduct in this case fell below the recognised standards of good practice,” citing both the provision of the spousal revenge article to the expert witness and the unrecorded pre-testimony phone call as conduct incompatible with the obligations of a Crown prosecutor.
The contrast with the defence’s psychiatric evidence was stark. Dr Murphy, an expert in infanticide, testified that the woman had been suffering from moderate to severe depression that substantially impaired her rational decision-making and self-control. Dr Murphy further told the court that the defendant appeared to have been a victim of coercive control by her partner, the father of her children, and that the defence of infanticide was properly available to her.
Judge’s Failure to Discharge the Jury
The defence team raised its concerns about Dr Bunn’s evidence — and about the circumstances in which the spousal revenge theory had surfaced — with Judge Donna McColgan during the trial itself. They applied twice for the jury to be discharged. Judge McColgan refused both applications.
The Court of Appeal found that the judge “did not pay adequate regard to this issue and appeared to minimise the import of the prosecution conduct and Dr Bunn’s evidence.” It ruled that the jury should have been discharged, describing the case as “a rare case where the judge has made an error in failing to discharge the jury.”
The woman, who cannot be named for legal reasons, pleaded guilty at retrial to infanticide — defined in law as the unlawful killing of a child under twelve months by its biological mother when the balance of her mind was disturbed — and to the attempted murder of her toddler. She was sentenced to ten years’ imprisonment earlier this week.
The case raises unresolved questions about the mechanisms through which expert witnesses are briefed by prosecutors, the absence of recording requirements for pre-trial communications, and the threshold at which trial judges must intervene when misconduct is identified before them in real time.

