Support, not suspicion, should be the first reaction to the tragedy of a cot death

It will be the role of the expert witness in court that will grab the headlines in reports of Baroness Helena Kennedy's recommendations on how the law should deal with cases of sudden unexpected infant death. Perhaps understandably. For there has been something nightmarish about the idea that mothers and fathers, already traumatised by the death of their child, can be jailed on the say-so of a single expert, whose acquaintance with them may, at best, be brief and, at worst, second-hand.

It will be the role of the expert witness in court that will grab the headlines in reports of Baroness Helena Kennedy's recommendations on how the law should deal with cases of sudden unexpected infant death. Perhaps understandably. For there has been something nightmarish about the idea that mothers and fathers, already traumatised by the death of their child, can be jailed on the say-so of a single expert, whose acquaintance with them may, at best, be brief and, at worst, second-hand.

Professor Sir Roy Meadow is the man who comes to most people's minds here. At the trial of the solicitor Sally Clark, who was accused of murdering her two sons, he famously said that the chances of two members of the same family dying of cot death were "73 million to one". What became known as Meadow's Law - that one cot death in a family was a tragedy, two is grounds for suspicion and three, unless proved otherwise, is murder - also helped convict Angela Cannings and Trupti Patel. All three women were later cleared after his notorious statistic was cast into doubt.

The Kennedy report looks far wider than one man, and indeed beyond the reliability of expert statements in general. It suggests that, like lawyers, expert witnesses may become susceptible to "case hardening" which leads them to see their role as securing a "win" for their side. No doubt our adversarial legal system encourages this hawkishness, pushing expressions of probability into assertions of certainty. The report is right to remind doctors, lawyers and judges that in cases of child protection they have a foremost duty to the truth. It makes a number of sensible recommendations to strengthen this - training witnesses in what is permissible and judges in how to establish the expertise of a witness.

But it is the report's less dramatic proposals that will make a bigger difference in preventing miscarriages of justice. Every cot death, it suggests, should be investigated by a paediatric pathologist. This is sensible for a number of reasons. It would aid the family with grieving; address anxiety about genetic problems that might affect future pregnancies, and assist research that might spare other families the same anguish.

The problem is that there are only 40 paediatric pathologists in the country. And there are - thanks to the hate mail and hostility increasingly levelled at the specialism - few trainees in the pipeline. The Government needs to address this issue, but so does the media whose irresponsible reporting has encouraged a fivefold increase in official complaints against paediatricians in the past year.

But the most profound change the report proposes will cost nothing at all. It is a change in attitude among professionals for whom cot deaths have in the past aroused automatic suspicion. Bereaved couples have found the accusatory word "unascertained" written on the death certificate as the cause of death. Other distraught parents have found yellow police incident tape wound around their house by officers who insist on referring to their home as a "crime scene".

Children's charities who feared that the Cannings judgement would become a "murderer's charter" - making it difficult to prosecute mothers suspected of killing their children - will no doubt express concern at the Kennedy report. But we need to keep a sense of proportion. There are some disturbed parents who kill their children. Earlier this year, the anti-Roy Meadow campaign suffered a setback when another potential victim of a miscarriage of justice, Maxine Robinson, after years of denial, admitted that she had murdered her three children.

But 600 babies a year die suddenly - children are four times more likely to die in the first year of life than at any other time - and in the vast majority of cases nothing unlawful has taken place. What is needed, where a child dies suddenly for no apparent reason, is the presumption that the parents are innocent. In such cases, our first response should not be suspicion, but support.