In December 2003, Angela Cannings walked free from the Court of Appeal, her conviction for murdering her two infant sons overturned as 'unsafe.' She had spent nearly two years in prison for crimes she did not commit. Her case, along with those of Sally Clark and Trupti Patel, stands as a damning indictment of how junk science can corrupt the criminal justice system when prosecutors and courts fail to scrutinise expert testimony.

At the heart of these miscarriages of justice was Sir Roy Meadow, a paediatrician whose fatally flawed statistical evidence helped convict multiple mothers of infanticide. His legacy is not one of protecting children, but of destroying families through pseudoscientific testimony that any competent statistician would have recognised as worthless.

The Making of a Flawed Expert

Meadow's reputation was built on his 1977 paper describing 'Munchausen Syndrome by Proxy' — a condition where caregivers harm children to gain medical attention. While this phenomenon exists, Meadow's approach to identifying it was dangerously simplistic. His infamous dictum — 'one cot death is a tragedy, two is suspicious and three is murder, until proved otherwise' — became known as Meadow's Law and influenced child protection agencies across Britain.

This reductive thinking ignored the complexity of sudden infant death syndrome (SIDS) and the possibility of genetic or environmental factors that might make some families more vulnerable. It also betrayed a fundamental misunderstanding of probability and statistical inference that would prove catastrophic in court.

The Sally Clark Case — Statistics in the Dock

The most notorious example of Meadow's flawed methodology came in the 1999 trial of Sally Clark, a solicitor accused of murdering her two baby sons. Meadow testified that the odds against two cot deaths occurring in the same affluent, non-smoking family were 73 million to one. He claimed such an event would occur only once every hundred years in England, Scotland and Wales — equivalent to backing an 80-1 Grand National winner four years running.

This figure was statistical nonsense on multiple levels. The Royal Statistical Society issued an unprecedented press release calling it groundless, with no statistical basis. Professional statisticians identified three fundamental errors:

First, Meadow committed the prosecutor's fallacy — confusing the probability of innocence given the evidence with the probability of the evidence given innocence. These are not the same unless murder itself is almost certain a priori, which it manifestly is not.

Second, he fell into the ecological fallacy, assuming that the aggregate population probability applied equally to all families. This ignored the possibility that genetic or environmental factors might cluster within families, making some more susceptible to SIDS than others.

Third, and most fundamentally, he assumed SIDS cases within families were statistically independent. In reality, one SIDS death suggests the family may have predisposing factors, making subsequent deaths more likely than the population average — potentially as much as 1 in 100 rather than 1 in 8,500.

When statisticians corrected these errors, the odds shifted dramatically — potentially 2:1 in favour of Clark's innocence rather than 73 million to one against it.

The Angela Cannings Tragedy

By the time of Angela Cannings' 2002 trial, Meadow's statistical approach was well-established. Cannings had lost three children: Gemma in 1989, Jason in 1991, and Matthew in 1999. While never charged over Gemma's death, she faced murder charges for Jason and Matthew based largely on Meadow's testimony.

The conviction rested on the same faulty logic that had condemned Sally Clark — the assumption that multiple SIDS deaths in one family were so improbable as to indicate murder. Yet a BBC investigation later revealed a crucial piece of evidence that the prosecution had overlooked or ignored: Cannings' family history showed a pattern of sudden infant deaths spanning generations.

Her paternal great-grandmother had suffered one sudden infant death, and her paternal grandmother two. Clinical geneticist Michael Patton told the BBC that genetic inheritance was the most likely explanation for the cluster of deaths. This was precisely the kind of family-specific risk factor that rendered Meadow's population-level statistics meaningless.

The Unravelling

The cracks in Meadow's edifice became visible when new evidence emerged in the Sally Clark case. It transpired that Home Office pathologist Dr Alan Williams had failed to disclose test results showing Clark's second child had died from Staphylococcus aureus infection, not smothering. When Clark's case returned to the Court of Appeal in 2003, Lord Justice Kay described Meadow's statistics as 'grossly misleading' and 'manifestly wrong.'

The judges stated that had the statistical evidence been fully argued, 'we would, in all probability, have considered that the statistical evidence provided a quite distinct basis upon which the appeal had to be allowed.' Sally Clark's conviction was overturned, but the damage was irreversible. She died in 2007 from acute alcohol poisoning, never having recovered from the trauma of losing her children and then being imprisoned for their murders.

Trupti Patel, tried in 2003 for allegedly killing three babies, was acquitted in just 90 minutes after Meadow's credibility had been fatally undermined. Angela Cannings was freed that December, her conviction deemed unsafe.

The System's Failure

These cases expose a cascade of institutional failures. Prosecutors relied uncritically on Meadow's testimony without seeking independent statistical verification. Trial judges allowed complex statistical arguments to proceed without ensuring the jury understood their limitations. The first Clark appeal dismissed the statistical evidence as a 'sideshow,' demonstrating a troubling lack of appreciation for how fundamentally flawed statistics can contaminate an entire case.

Perhaps most disturbing was the prosecution's continued faith in Meadow even as his methods crumbled. After Patel's acquittal, a Crown Prosecution Service spokesperson said they would still be 'very happy' to use Meadow's evidence in future trials. It took the intervention of Solicitor General Harriet Harman to effectively bar Meadow from court work by requiring defence lawyers to be warned of court criticisms of his evidence.

The Human Cost

Behind the statistical arguments lie shattered lives. Sally Clark, a successful lawyer and mother, died broken at 42. Angela Cannings endured nearly two years in prison while grieving the loss of her children, branded a murderer by the state that should have supported her. Their families faced not only the tragedy of infant death but the added trauma of false accusation and wrongful imprisonment.

The General Medical Council struck Meadow from the medical register, though he was later reinstated on appeal when judges ruled his courtroom errors did not constitute serious professional misconduct affecting patient care. This narrow interpretation of professional responsibility fails to acknowledge the broader duty medical experts owe to the justice system.

Lessons Unlearned?

The Meadow cases prompted a review of 297 other convictions based on similar expert evidence, with the Attorney General finding three cases requiring court reconsideration. But the fundamental problem remains: how can the justice system protect against expert witnesses who exceed their competence or present flawed science as established fact?

The answer requires humility from experts about the limits of their knowledge, scepticism from lawyers about claims that seem too precise to be credible, and judicial recognition that statistical evidence, however impressive it sounds, can be profoundly misleading when applied incorrectly.

Angela Cannings' case stands as a monument to the dangers of prosecutorial certainty built on scientific uncertainty. Her suffering, like that of Sally Clark and others caught in Meadow's web, reminds us that the price of forensic overreach is measured not just in wrongful convictions, but in lives destroyed by a justice system that confused statistical sophistication with scientific rigour. In the end, it was not the complexity of the science that failed these women — it was the simplicity of minds that could not grasp the difference between correlation and causation, between improbable and impossible, between expertise and arrogance.