Meadow’s Law and the bad assumptions that put innocent women behind bars

Content Warning: This article refers to sudden death of children in infancy. Red Nose Australia provides 24/7 Bereavement Services on 1300 308 307.

When expert witness Dr Calum MacRae was giving evidence to the Folbigg Inquiry relating to the deaths of Kathleen Folbigg’s four children, he remarked it was hard for him to entertain the G114R calmodulin variant as a cause of her girls’ deaths.

That variant, possessed by Sarah and Laura Folbigg is suggested as the cause of their deaths.

Their mother Kathleen was sentenced to 30 years in prison for killing them and their two brothers. Research into their genetic variant was the trigger for NSW attorney-general Mark Speakman recommending this judicial review into her convictions.

But because their brothers Caleb and Patrick died before them without carrying the same variant, MacRae said it was more likely a shared genetic variant between the four (no plausible one has been yet identified) or other shared external reason caused their deaths.

Counsel assisting the inquiry Julia Roy asked MacRae whether he knew of so-called Meadow’s Law: an assumption that more than two sudden infant deaths in a family is “murder until proven otherwise”.

Can the law keep up with science?

MacRae acknowledged there was no difference in the reasoning between Meadow’s Law and the way he assessed the probabilities that two separate clusters of deaths – one caused by the genetic variant in the girls, and another by unknown means in the boys – would occur.

But using probabilities to explain Sudden Infant Death Syndrome (SIDS) and other unexpected deaths in children, particularly amid negative autopsies where no definitive cause of death is found, is fraught.

The use of Meadow’s Law in the nineties shows why.

Sally Clark’s two children died suddenly, “Palpably False” Meadow’s Law was used to convict her

Meadow’s law is named for the British paediatrician Roy Meadow, who adopted the opinions of American pathologists Vincent and Dominick Di Maio when giving expert evidence in prosecutions of British women over then-named ‘cot deaths’.

Meadow’s Law was also referenced in the original trial of Australian woman Kathleen Folbigg in 2003.

It has since been rejected as a valid form of evidence.

That’s because three British mothers convicted of killing their children in trials where Meadow’s Law was used — Sally Clark, Angela Cannings and Donna Anthony — had their sentences overturned. In June 2003, it took 90 minutes for a jury in the trial of another woman, Trupti Patel, to acquit her of similar charges.

Clark’s was the most high-profile of these trials. She was originally convicted of murdering her two infant sons after Meadows suggested the chances of two children in the same family dying from SIDS was 1 in 73 million.

This figure was reached by squaring the risk – a 1 in 8543 chance – of a sudden infant death in a family with no recognised environmental danger: a basic calculation trying to determine the chance of two independent events – such as the unexplained deaths of two children – happening.

But the chance of two (or more) SIDS deaths being entirely independent of each other is incredibly remote.

Professor Peter Donnelly, a mathematician at Oxford University, referenced Meadow’s erroneous reasoning in his TED Talk on how statistics can fool juries back in 2007.

“It’s palpably false,” Donnelly said.

“There are lots and lots of things that we don’t know about sudden infant deaths, it might well be that there are environmental factors that we’re not aware of, and it’s pretty likely to be the case that there are genetic factors that we’re not aware of.

“So if a family suffers from one cot death, you’d put them in a high-risk group: they’ve probably got these environmental factors and/or these genetic risk factors we don’t know about.

“And to argue then that the chance of a second death is as if you didn’t know that information, is really silly. It’s worse than silly, it’s really bad science.”

Several observers of the Folbigg case that Cosmos spoke to commented that the reasoning of the Di Maios and Meadow discards the presumption of innocence.

In analysing the data used to support Meadow’s original claim, Salford University mathematician Professor Ray Hill noted it’s less likely that a child would be the victim of a homicide than cot death.

In the aftermath of Clark’s release from prison, he wrote: “There are fewer than 30 infant homicides per year among the 650,000 births each year in England and Wales. So the chances of an infant being a homicide victim are about 1 in 21,700.”

“If these cases are to teach us anything, it is that simplistic rules of thumb should have no place in our thinking, and that all sudden deaths, whether first, second or subsequent, should be thoroughly, but sympathetically, investigated.”

Professor Ray Hill

Compared to the 1 in 1300 chance of a cot death during the time of the study referenced by Meadows, that would mean an infant is 17 times more likely to be a SIDS case, rather than infanticide victim.

Why proteins matter

Further, a child born into a family that has already suffered one sudden infant death was considered 10 times more likely to also be a victim of SIDS.

Finally, Meadow’s calculation was strongly rebuked by Britain’s Royal Statistical Society, which emphasised the environmental and genetic risk factors associated with multiple SIDS cases that would later be scrutinised by Donnelly.

So concerned about Meadow’s use of the statistic, that the society urged the Lord Chancellor (who heads Britain’s justice ministry) to implement changes so statistics are “presented only by appropriately qualified statistical experts, as would be the case for any other form of expert evidence”.

Using statistics to explain sensitive issues like this can be risky

In correspondence to the British Medical Journal in 2004, Hill noted infant deaths deserve sensitive and methodical investigation.

“Surely, if these cases are to teach us anything, it is that simplistic rules of thumb should have no place in our thinking, and that all sudden deaths, whether first, second or subsequent, should be thoroughly, but sympathetically, investigated.”

Professor Roger Byard is a world-leading forensic pathologist and SIDS researcher who provided expert evidence at the original Folbigg trial and co-founded the National Scientific Advisory Group of Red Nose Australia, the charity for safe sleep education and bereavement support for the loss of infants and children.

How scientists work out which genetic variants are deadly

He observes that simply relying on statistics is not necessarily helpful to illustrate the nuance of unexplained infant deaths.

“Just because something is vanishingly rare, to me, actually is an admission that [event] can happen,” Byard says.

“I think we are understanding that there are a number different causes of unexpected infant deaths that leave no marks, as in all these cardiac conduction disorders and there may be genetic backgrounds to those.

Professor Roger Byard

SIDS, as decades of research has shown, can take place multiple times in the same family.

Sometimes that’s because of shared genetic and environmental factors, as proposed in the current Folbigg inquiry.

SIDS was listed as the cause of death for both Caleb and Sarah Folbigg. Patrick’s death was asphyxia resulting from epileptic fits from an encephalopathic disorder. Laura’s death was listed as “undetermined.”

Subsequent forensic pathology evaluations were largely the same, though several considered Sarah to have died from myocarditis.

These diagnoses will undoubtedly be explored in expert evidence from Stephen Cordner, Matthew Orde and Alan Cala, the latter of whom performed the analysis on Laura Folbigg.

While SIDS has been dramatically reduced thanks to ongoing sleep safety campaigns (a more than 80% decrease in unexplained deaths from SIDS in Britain since the Back to Sleep campaign began in the nineties), there is still much we don’t know about the pathology.

As Byard observes, researchers and pathologists are today far more open minded about the possibilities of SIDS triggers.

“The way we’ve evolved is to understand the complexity of infant death more,” Byard says.

“I think we are understanding that there are a number different causes of unexpected infant deaths that leave no marks, as in all these cardiac conduction disorders and there may be genetic backgrounds to those.

“Certainly back in ’89 when you say the case is a homicide based on negative autopsy findings, I don’t think there’s be many pathologists around the world who would say that nowadays.”

More on the Folbigg inquiries

• Expert: It would be extraordinary for someone to suffocate Folbigg children without leaving a mark
• Video: Inquiry hears that new evidence puts the conviction of Kathleen Folbigg in doubt
• Folbigg inquiry head must decide on evidence now, “can’t wait for science to fully develop”
• Researchers affirm science behind G114R study
• Scientists grilled on advocacy and accuracy
• Calmodulin variants “not benign” say Danish experts