A sick child or a sick mind?

In 1977 a paediatrician called Roy Meadow was called to see a six-year-old girl. Her name was Kay, and she had been referred to Seacroft Hospital in Leeds with foul-smelling, bloody urine.

In 1977 a paediatrician called Roy Meadow was called to see a six-year-old girl. Her name was Kay, and she had been referred to Seacroft Hospital in Leeds with foul-smelling, bloody urine.

Something about the girl's symptoms didn't add up and, after investigations, Meadow found out why. It turned out that Kay's mother had been mixing her own urine, contaminated with menstrual blood, with her daughter's, and over a period of years had involved the medical profession in a cat-and-mouse game to see if they could diagnose Kay's condition. Each time they completed one series of investigations, Kay would mysteriously develop a new symptom that required further inquiry.

By the time she arrived at Seacroft Hospital, Kay had already been admitted to hospital 12 times, had had seven major X-ray investigations, six examinations under anaesthetic and been treated with eight antibiotics, vaginal pessaries and creams. Sixteen consultants were involved in her care and her urine was cultured in laboratories 150 times before the doctors realised it was her mother who was ill.

Kay's mother became the first patient in the world to be identified with Munchausen's by proxy. The syndrome - and the recognition that parents can harm their children under cover of appearing to care for them - has since entered the national consciousness.

Today, Professor Sir Roy Meadow is a household name, known not so much for this discovery, as for his work as an expert witness in a number of controversial cot-death cases. To some, he is a hate figure, seen to be responsible for a string of wrongful convictions of innocent parents.

There are few crimes to equal abusing a child, but one of them is to accuse a parent falsely of committing such a crime. This is the dilemma faced by paediatricians working in child protection. How can they steer a course between wrongly diagnosing abuse and missing it altogether, when the consequences of either error are equally devastating?

Their role will be highlighted today when the General Medical Council opens its case against Professor David Southall, a consultant paediatrician at the North Staffordshire Royal Infirmary in Stoke-on-Trent. Professor Southall is an expert in child protection, who has been targeted for a decade by campaigners seeking to expose false allegations of abuse. He is accused by the GMC of intervening in the case of Sally Clark - the solicitor convicted of killing her babies in 1999, and who was cleared by the Court of Appeal last year. The GMC maintains that Professor Southall made allegations against Clark's husband, Stephen, purely on the basis of information he gleaned from a television programme.

Depending on the outcome of that case, he will also face eight further charges before the GMC in relation to his use of the controversial diagnosis of Munchausen's syndrome by proxy, to be heard at a later date.

But the big fish, due to appear before the GMC later this year, is Professor Meadow. It was his evidence that helped secure convictions in a series of high-profile cases, including those of Clark and Angela Cannings, who was also later freed. The collapse of the cases and criticisms of Sir Roy's evidence by the Court of Appeal led the Attorney General to order a review of 258 child-protection cases, the biggest in legal history. That outcome has been hailed as a victory by campaigners, and has provoked a bout of soul-searching among paediatricians. It has led many to conclude that the diagnosis of Munchausen's by proxy, the condition whereby parents induce symptoms in their children in order to gain medical attention for them, is now discredited and that doctors have been seeing child abuse where none in fact exists.

The paediatricians have, however, been heartened by the first stage of the review ordered by the attorney general. Of about 100 cases examined so far, where mothers had been convicted of murdering their children, only five were considered potential candidates for referring back to the Court of Appeal. Although even one wrongful conviction is unacceptable, the fact that 95 per cent of the convictions have been found to be safe does suggest that the notion that mothers sometimes harm their children is not one invented by evil doctors.

One victim of the harm that mothers can do is Julie Gregory. She describes in her book Sickened, published earlier this year, how, as a young girl growing up in north America, an outing with her mother usually meant an appointment with a specialist for another needless and unpleasant test. For her, the problem was not doctors seeing child abuse where it did not exist, but the opposite. They refused to listen to her pleas that there was nothing wrong with her, and she has suggested that the investigations being carried out amounted to assault.

Her life became one, long round of visits to doctors and hospitals, culminating in a week of invasive heart tests when she was 13, as her mother used her as a vehicle to gain the professional attention she craved. "They [the doctors] tried everything they could to find out what was wrong with me. My mum told them I had a mysterious heart condition and described all sorts of symptoms. But the truth is that my so-called illness was all in her head. She robbed me of my childhood."

Julie was lucky. She survived, left home at 17 and now, at the age 34, has told the tale of her bizarre upbringing, based on the 200 pages of her medical records that she has managed to collect. When she confronted the surgeon who conducted the heart catheterisation test on her, and told him she had been a victim of Munchausen's by proxy, he broke down in tears. "It was an important moment for me," Julie writes. "It was very emotional, but I felt vindicated."

Julie grew up in the 1970s, before Munchausen's by proxy was recognised. It was some years before it fully entered the public consciousness when it helped to secure the conviction of Beverly Allitt, the nurse-turned-serial killer, jailed for killing four children and harming nine others in 1993.

But, by the late 1990s, there was concern that the condition was being diagnosed too widely. In some cases in which mothers suffocated their children, they did not seem to be doing it to attract medical attention, but in order to exert control over their babies. These mothers were often disturbed and were frequently unaware of what they were doing, and were driven instead by a sense that their new baby was taking over and that they must get control of their life back.

Aside from these were the cases in which mothers fed their children salt or medicines to induce symptoms in them that would then lead to a round of medical investigations - the classic Munchausen's by proxy case.

In a report published in 2002, the Royal College of Paediatrics said that the term Munchausen's by proxy should be abandoned because it made unwarranted assumptions about the parent's mental state and motivation. Fabricated or Induced Illness (FII) is now the preferred term, because it makes no assumption about why the mother is doing it, but shifts the focus onto the child.

So how common is it? This is the most contentious question. Campaigners such as the Mothers Against Munchausen by proxy Allegations (MAMA) group, a militant collective of aggrieved parents and their supporters, say it is over-diagnosed. Professor Southall told The Independent in April that it was "grossly under-diagnosed and under-reported".

Writing in the British Medical Journal two weeks ago, Professor Alan Craft, president of the Royal College of Paediatrics and head of child health at the Royal Victoria Infirmary, Newcastle, and Professor David Hall, of the Northern General Hospital, Sheffield, cited a UK-wide survey on Munchausen's syndrome by proxy in 1992-94 that identified 128 cases in two years. That, they acknowledged, made it rare. But, they added, in support of Professor Southall's view: "A growing number of case reports suggest that many manifestations of Munchausen's syndrome by proxy are under-identified because of lack of professional expertise and public awareness."

Of those 128 cases, 32 involved suffocation. The most recent estimate, the authors say, is that one in 10 otherwise unexplained cot deaths may be caused by deliberate suffocation, amounting to about one a fortnight.

These are the cases that have sparked the recent controversy and lie at the centre of the court cases involving Clark and Cannings. Where two or three cot deaths occur in the same family there are only two possibilities, Professors Craft and Hall say - homicide or inherited conditions such as heart defects or metabolic disorders. Although these conditions probably account for a "small fraction" of cases, their existence and the possibility that other conditions may yet be discovered adds a dimension of uncertainty to any death, they say.

Any certainty in cases of child abuse is hard to come by and that makes them inappropriate for the criminal courts. Far better, as Professors Craft and Hall suggest, that they be dealt with in the civil courts, where experts can review the evidence in a non-adversarial meeting focused first of all on minimising risk to all the children in the family, and second of all on meeting the parents' needs, too.

As the professors themselves put it: "Although the reasons why parents kill their children are poorly understood, perhaps many deserve sympathy and treatment rather than imprisonment."