The case of Sally Clark - convicted of murdering her two sons - has been one of the most controversial in modern criminal history. However, newly disclosed results of microbiology tests carried out on her second child, the eight-week-old Harry, indicate he died of natural causes.
This crucial information was not disclosed at her trial. Clark's family and lawyers say the jury were left with the impression that hospital tests had excluded all possibility of death from natural causes.
Clark, now 37, has protested her innocence throughout. Christopher died aged 12 weeks in the evening of December 13, 1996, while she was at home alone with him, after she had put him in a Moses basket.
Harry was born on November 29, 1997. Although the family took many precautions, like using an apnoea monitor, the baby died in his bouncy chair on the evening of January 26, 1998. In July of that year, Clark was charged with the murders of both children, and she was convicted the following year.
A key defence witness at the trial, Tim David, professor of child health and paediatrics at the University of Manchester, analysed blood tests carried out on Christopher. He concluded that the child could have died of a rare lung disease, idiopathic pulmonary haemosiderosis.
During Professor David's evidence in chief, and again during cross-examination, the jury passed notes to the judge asking if blood tests had also been carried out on Harry.
The prosecution told the jury: "A sample was not collected from Harry to analyse sodium in his blood".
The Clark family, like the jury, were left with the impression that no tests had been done on Harry. However, Sally's husband Steve spent two years trying to get all medical records relating to the deaths of his children from Macclesfield hospital. Finally, in April, to his astonishment, he was sent a vast number of documents, among which was a copy of a previously undisclosed microbiology report.
It now seemed that Alan Williams, the pathologist who carried out the postmortems, had carried out tests.
"When the postmortem was carried out," explained John Batt, a friend of the Clarks who is also a solicitor, "Dr Williams took cultures from the stomach wall, both bronchi, both lungs, the trachea, the throat, and, most importantly, the cerebro-spinal fluid. He sent the cultures to the lab for microbiology tests. The results were available within 48 hours. Staphylococcus aureus was present in all of the sites."
Such results can sometimes be misleading, as it is difficult to avoid contamination at autopsy and the infection could have arisen after death. So in order to obtain an expert opinion on the significance of this finding, the newly revealed report was sent to consultant pathologist James Morris, honorary professor of biological sciences at the University of Lancaster and one of the country's leading experts in bacterial toxins in cot death cases.
Having read the evidence, he concluded that the possibility of infection arising after death could be excluded. "Cerebrospinal fluid (CSF) is normally sterile," he wrote. "The presence of a pathogen is therefore highly significant. There is no evidence that staphylococcal organisms could get into the CSF after death. The fluid contained a number of polymorphs [white blood cells], showing there had been a reaction to the organism and therefore that it was present prior to death."
Professor Morris noted that "overwhelming staphylococcal infection ... is a very serious and potentially lethal disease. It can progress so quickly that death occurs suddenly without obvious pre-existing illness. It is a recognised cause of sudden and unexpected death in infancy. This is natural disease." On the basis of the evidence now available, he reported that this was the the most likely explanation of Harry's death. He emphasised that "no other conclusion could be sustained."
The Clark family sent the new material to Glyn Walters, a retired consultant in chemical pathology. He, too, concluded that Harry's death was due to "a disseminated staphylococcal infection". In fact, in his postmortem report on Harry, Dr Williams categorically stated: "There is no evidence of acute infection ... there is no evidence that this child died as a result of natural disease." In the light of the new material, this is now in doubt.
"The jury were aware Dr Williams's findings were the subject of much criticism," said Michael Mackey, Clark's lawyer.
"They never knew he had commissioned tests which showed, at the least, a potential natural cause of death to which he made no reference [in his evidence].
How would that knowledge have affected their view of his competence?"
Leading paediatrician Sir Roy Meadow told the jury there was a one in 73 million chance of two cot deaths occurring naturally in the same family - something he estimated would happen less than once a century.
In fact, actual incidence of a second cot death, as others have since pointed out, could be as low as 1 in 64. [presumably as high as 1 in 64 - Ed.]
The referral of the case to the criminal cases review commission [by the CCRC to the Court of Appeal - Ed.] will give renewed hope to mothers convicted of murdering their children in similar circumstances, who have protested their innocence.
These cases include those of Maxine Leggatt, whose two children died in their beds on the same evening near Chester-le-Street, Co Durham, and of Angela Cannings, convicted in Winchester in April of murdering two of her children.
Steve Clark last night acknowledged these other cases: "Others in this situation do not have the resources or support we have been so fortunate to enjoy. But it is frightening to think we might never have discovered those test results."