PROFESSOR Sir Roy Meadow misused statistics from the world’s biggest survey of cot deaths when he gave evidence in a murder trial, the General Medical Council was told yesterday.
The distinguished paediatrician, whose evidence helped to convict Sally Clark of killing two baby sons, was given exclusive access to the unpublished report because he had been invited to write the foreword. The author of the study told the GMC that he would have refused permission for Professor Meadow to cite the unchecked draft in court, but the expert witness never asked.
The author was so alarmed by the possible misuse of numbers that he wrote to Mrs Clark’s lawyers before the trial explaining their correct meaning.
Advertisement
Professor Meadow, 72, has been charged by the GMC with serious professional misconduct over his use of statistics suggesting that the chances of two natural cot deaths for an affluent, non-smoking, mature couple like the Clarks was vanishingly small. He told the murder trial that the odds were 72 million to one.
The disciplinary panel has the power to strike the retired paediatrician off the Medical Register.
Professor Peter Fleming, another cot-death expert, was the author of the study of cot deaths among 472,000 British babies during the 1990s. Professor Fleming, a specialist in infant health and physiology at the Institute of Child Health at Bristol University, had already made history by discovering that cot deaths were much more common among babies who slept on their fronts.
He led the “Back to Sleep” campaign, which urged parents to put infants to bed on their backs. This was a spectacular success, reducing cot deaths by as much as 70 per cent. Professor Fleming told the GMC that he had invited Professor Meadow, a former president of the Royal College of Paediatrics and Child Health, to write the foreword to his study.
He had had “great respect for Professor Meadow’s work” in highlighting the importance of child abuse. Making the understanding of abuse an accepted aspect of understanding child care was “of enormous significance”, Professor Fleming said.
Advertisement
Before Mrs Clark’s trial her lawyers had contacted Professor Fleming to say that the prosecution was claiming that unpublished research gave a 72 million-to-one chance of two natural cot deaths in a family like hers. “I was quite concerned because I felt this was not in the context that it had been used in the report,” Professor Fleming told the GMC. “The statistics were never meant as an estimate of risk. They were a form of mathematical modelling to try and find families for whom we should provide practical forms of intervention.”
He wrote to Mrs Clark’s lawyers immediately. “I felt very worried,” he said. “I felt very concerned that a sentence out of context would be used in a way which didn’t reflect the real content of the report.”
The figure appears in an analysis showing the hugely increased risk of cot deaths in families who smoke, are unemployed, or where the mother is young. In a family with none of those dangers, the risk of a cot death was 8,543 to one. In such families, the chance of two cot deaths was 72 million to one, the report said. The figure was reached by multiplying 8,543 by itself; but it was a statistical exercise that excluded known causes of cot death such as genetics, familial conditions, sleeping positions and heavy bedding, Professor Fleming said.
In spite of his letter, Mrs Clark’s lawyers omitted to challenge Professor Meadow about the 72 million-to-one statistic at her trial. She was jailed for life but cleared on appeal and freed after three years.
Under cross-examination by Nicola Davies, QC, for Professor Meadow, Professor Fleming accepted that the figure was a “statistically invalid assumption”.
Advertisement
Miss Davies: “Is it possible to misinterpret it because the text itself is not clear?”
Professor Fleming: “Clearly, it was not. People were misinterpreting it.”
The hearing continues.