A chance remark by a doctor at another hospital alerted Carl and Debbie Henshall to the fact that two of their daughters had been involved as guinea-pigs in a research trial.

Sofie, now aged eight, suffered brain damage after being born prematurely at the North Staffordshire Hospital in Stoke-on-Trent in December 1992. She was treated on a new type of ventilator imported from the United States, which was being tested at the hospital.

According to the Henshalls, staff at the hospital emphasised the virtues of the new treatment in stark terms. The CNEP ventilator - the initials stand for continuous negative extrathoracic pressure - consisted of a tank placed over the body, which used suction to expand the lungs instead of the conventional method of inserting a tube down the windpipe and blowing air into the lungs, which could damage delicate tissues.

"We were offered it as the newest, best treatment and the safest gentlest option and we were told of the horrors of the conventional option. We felt damn lucky - it was literally sold to us," Mrs Henshall said.

However, Sofie did not do well. A neurologist later confirmed she had suffered brain damage after birth but was unable to say if it was caused by the ventilator. The couple consulted a solicitor and started legal action against the hospital in 1994. In 1996 the solicitor sent them to a medical expert at St James's Hospital, Leeds.

"He [the expert] said to us, 'What do you expect from an experimental treatment?' We said 'What do you mean experimental?' Until then we believed our daughter had been given the best treatment available," Mrs Henshall said.

It was the second tragedy to strike the family. Stacey, born in February 1992, also prematurely and with breathing problems, had died after being treated on the CNEP ventilator. The North Staffordshire Hospital NHS Trust stresses that many babies on ventilators are seriously ill and have a high chance of dying anyway. But the Henshalls were furious that their children had been included in research without their knowledge. "We were duped twice and we didn't understand how. That is what has driven us on," Mrs Henshall, who has seven children, said.

Experts say it is extremely difficult to obtain meaningful consent from mothers immediately after birth, when they may be suffering the aftereffects of drugs. Studies show that even in the most carefully conducted trials, parents have no recollection of what they were told a year later.

But the Griffiths report notes that the parents' stories about the CNEP trial and their failure to grasp that it was an experimental treatment show a remarkable consistency. The report says many of the parents used the words "a kinder, gentler treatment" in relation to the ventilator, which were the same words the research team used.

"Members of the research team were very ready to tell us that parents were often poor at... recalling what they were told. In effect they were asking us to believe that these parents could not remember consenting to [the research] but did remember that CNEP was a 'kinder, gentler treatment'," the report says.

More than 20 families have complained to the General Medical Council (GMC) that they were misled. Sharon Bradley, 31, whose son Stephen, now seven, was treated on the CNEP ventilator and has severe learning difficulties and is autistic, said: "Stephen had mild respiratory distress after he was born and I was told they had two treatments.

"There was positive pressure [the conventional method], which could cause blow-out to the lungs and sounded horrific, and there was negative pressure [the CNEP technique] which would teach him to breathe for himself and he would not have to be ventilated for so long. One was sold to me and one was given to me like a horror story. [The CNEP ventilator] was the safest, best treatment, I was told. I was thrilled."

The Griffiths inquiry has been unable to decide whether the new ventilator was better or worse than the standard treatment. There is particular concern that the period of follow-up - eight weeks - was too short to pick up all cases of brain damage. Sofie Henshall, who was not diagnosed with brain damage until she was two, is not included among the 10 offical cases of brain damage associated with CNEP.

When an experimental treatment is offered it is crucial to find out what effect it has in the long term, the report says. It recommends that all children treated with the CNEP device should be followed up

It was the Henshalls' complaint to their local MP, Llin Golding, that triggered the inquiry. Their cool approach to emotive issues has won praise locally, in contrast to the campaigning tactics of the mothers accused of Munchausen by proxy, another aspect of the hospital's work examined by the Griffiths inquiry. A working party set up by the hospital trust has sought their advice on tightening ethical guidelines for the conduct of research.

A key issue still to be resolved is whether signatures on consent forms for the ventilator trial were forged. The hospital claimed that it held consent forms signed by the parents for all the babies involved in the trial but the inquiry found some were missing, some were not signed by parents or doctors and some "appeared to have signatures that appeared not to be those of the parents". It does not reach a judgement on these "very serious" allegations but recommends that they be passed to the GMC.

Mrs Henshall said: "When Stacey was born, Carl was asked to sign a form but he was not told it was for a trial... When I had Sofie they came up to me after the birth and said the baby was struggling to breathe. I had had a baby in CNEP treatment before and would I like to do it again. I assumed they wanted to do the best for us because we had lost a baby before. I never signed anything."

The hospital says it did have a consent form signed by Mrs Henshall. To prove its case, it sent copies to Channel 4 and other media organisations; the medical director of the trust, Keith Prowse, is now underinvestigation by the GMC in relation to unauthorised disclosure of patient records. Mrs Henshall says the signature on the form is not hers.

In a statement yesterday Staffordshire police said that they had interviewed parents about allegations of forged consent forms and were waiting to study the Griffiths report before deciding how to proceed.