It is hard to think of a more terrible wrong that Professor Sir Roy Meadow could have inflicted upon Sally Clark, short of raping or murdering her. Mrs Clark is the solicitor from Wilmslow, Cheshire, who was sentenced to six years' imprisonment in 1999 on false charges of having smothered her baby sons, Christopher and Harry. She stayed behind bars until her conviction was quashed by the Court of Appeal in 2003. Sir Roy is the eminent paediatrician whose "erroneous and misleading" evidence, as an expert witness at her trial, helped to convict her.

Nobody is claiming that Sir Roy set out deliberately to mislead the court, or that his evidence was motivated by any feeling of malice towards Mrs Clark. What he told the jury at her trial, he firmly believed to be true.

But there comes a point when a man, however well meaning, must take responsibility for his conduct and beliefs, and he must be brought to account when his beliefs are shown to be stupid and false, and his conduct causes dreadful suffering to others.

That point came at Mrs Clark's trial when Sir Roy announced, with all the authority of his knighthood and his professorial chair, that the odds against two natural cot deaths occurring in a single middle-class family were 73 million to one. This week, the General Medical Council heard that the true figure was something like 77 to one. So Sir Roy's evidence was out by a factor of getting on for a million.

The GMC also heard, from Sir Roy's own mouth, how he arrived at his astronomically inaccurate figure. He had read an article in The Lancet, in which a physiologist at Bristol University, Professor Peter Fleming, had claimed that there was a one-in-8,543 chance of a child in a middle-class family falling victim to a cot death. Sir Roy had then whipped out his pocket calculator and squared that figure, thinking that the answer represented the odds against two cot deaths happening in one middle-class family.

You only have to think about that for 20 seconds to realise that it cannot be true. I didn't even have to think about it for that long. This is because I once knew a doctor's wife - a thoroughly nice, ordinary, devoted mother - two of whose children had died in their infancy, for no reason that the pathologists could discover. I remember her telling me that when her second baby died, she had been held overnight at the local police station. Having just suffered the unimaginable trauma of losing a second child, she was now being suspected of having killed it. It was the most awful thing that had ever happened to her. But that was in the days, she said (and she was talking about the 1950s), when nobody knew about sudden infant death syndrome (SIDS).

A couple of years ago, I mentioned this woman's terrible experience to a friend, who said that he, too, knew a patently sane couple (not the same people), who had lost two babies to cot deaths. If even I have come across two instances of multiple cot deaths, in a circle of acquaintance numbering hundreds rather than thousands, then it doesn't take a genius to work out how wildly improbable it is that the odds against these double-tragedies happening are anything like as high as 73 million to one.

But, as I say, nobody needs this sort of anecdotal evidence to realise that Sir Roy's evidence at Mrs Clark's trial was questionable to the point of incredibility. The calculation that he made, when he squared that figure of 8,543, took no account at all of the possibility that there may be genetic or environmental factors at work in SIDS. Simple common sense should surely have told Sir Roy that a couple who have already suffered one cot death are considerably more likely than any other couple, picked at random from the telephone directory, to suffer a second.

We still don't know exactly what causes SIDS. But whatever it is - whether it has something to do with diet, genes, "passive smoking", sleeping positions, living near electricity pylons or any combination of a thousand other possibilities - it is quite clear that siblings share genes and tend to live in similar conditions. If one is at risk, there is a fair likelihood that his brothers and sisters may be too.

But this doesn't seem to have occurred to Sir Roy, when he pulled out his calculator and condemned Mrs Clark to years of torment, on the strength of something that he had read in The Lancet. What infuriates me most of all about his behaviour is the contrast between the sloppiness of his evidence and the extreme gravity of Mrs Clark's plight. You would think that any man, no matter how eminent in his field or how sure of himself, would take the very greatest care before saying anything under oath that might send a bereaved mother to prison on a false charge of murdering her child.

If he had bothered to ask around, he would quickly have heard of several instances of multiple cot deaths in the United Kingdom alone. Last year (admittedly long after the Clark case) a team led by Professor Robert Carpenter, of the London School of Hygiene and Tropical Medicine, identified no fewer than 18 families who had suffered more than one cot death, in circumstances beyond suspicion. Prof Carpenter said: "Our data suggest that second deaths are not rare and that the majority, 80 to 90 per cent, are natural." So much for Sir Roy's one-in-73 million and his famous "Meadow's Law": "One in a family is a tragedy; two is suspicious; and three is murder."

Mrs Clark is by no means the only woman to have suffered at Sir Roy's well-meaning hands. Two other mothers, convicted partly on his "expert" evidence, have had their convictions quashed on appeal. One of them, poor Donna Anthony, served more than six years of a life sentence before justice was done by her. My heart goes out, too, to the 5,000-odd children and their parents who have been separated because of Sir Roy's belief in "Munchausen's Syndrome by Proxy" - a smart-sounding name for the theory that some parents harm their children in order to draw attention to themselves. We will never know how many of those who have had their children taken into care are innocent.

Experts who have pet theories like this are a menace. I don't much care what happens to Prof Sir Roy Meadow, when the GMC decides his fate. He is retired now and has done his worst. I will only observe that the kindest thing to be said about him is that he is intensely stupid.

Information appearing on telegraph.co.uk is the copyright of Telegraph Group Limited and must not be reproduced in any medium without licence. For the full copyright statement see Copyright