Social workers did not know of previous attempt to kill a child: A mother with a conviction for attempted infanticide and history of depression was allowed to take home a newborn child. Stephen Ward and Mike Walsh report

THE LITANY of errors which ended with the death of seven- week-old Mia Gibelli began in December 1991 when her mother's GP phoned the social services office in West Norwood. Christine Gibelli was 22 weeks pregnant. The doctor said he was worried about what might happen after the child was born.

He warned that Gibelli's other two children were in care and she had a history of psychiatric admissions, drugs and heavy drinking. Crucially, he failed to add that four years earlier Gibelli had tried to kill another of her two children by throwing her out of a third-floor window. The local social services office - Gabriel House - flooded with what it thought were more urgent child protection cases, did not give his call priority.

In January 1992, Gibelli - who suffers from a severe form of post- natal depression called Puerpal psychosis - attended the ante- natal clinic at King's College hospital. Some bizarre remarks about her children led to a hospital-based social worker being asked to compile background information on the family. But she failed to find the key files detailing the attempted infanticide. In cases where risk to the child is suspected, there should be a meeting of child protection experts well in advance of the birth, while specially trained teams assess the family and come up with ideas for handling matters in the long term.

But here, after the social worker had missed the key file, the appropriate warning bells were not sounded. Instead, tussles began as the hospital social workers and those in West Norwood apparently tried to palm off the case on each other.

The hospital staff wanted rid of it, fearing the local office would refuse to take over once mother and child were out of hospital. The local office, covering a tough patch known as Area 8, did not want it because it was swamped with 200 new child protection referrals a year.

At the end of April, Area 8 finally accepted responsibility for the case, four months after the GP had expressed his concern. There had been no assessment and no pre-birth meeting.

A meeting was squeezed in a week ahead of the birth. It was so hurried that the social worker charged with preparing a presentation had just half an hour to read the files.

A child protection plan was verbally agreed. Gibelli was to be offered a place in a special mother and baby unit and a new hormonal treatment for Puerpal psychosis. If she refused this, the fallback plan was to seek legal advice - with a view to a care order to protect the baby before she was discharged from King's.

Meanwhile, social workers would be interviewing Gibelli and her family before mother and child left hospital to assess what should happen in the long term.

None of it happened. Before the birth, Christine Gibelli turned down the treatment proposal. She announced that she had a Harley Street specialist with a different treatment for her condition that would allow her to take the baby home. The specialist was Dr Katherina Dalton, a pioneer of hormonal treatment of post-natal depression. Marque Shields, Gibelli's boyfriend, had read about her. After examining Gibelli, Dr Dalton declared that her own treatment - heavy daily doses of progesterone - was the answer to her psychosis.

Dr Dalton's methods are not universally accepted in the medical profession. Some doctors at King's were not enthusiastic. After Mia's birth, while mother and baby were still in hospital, the doctors tried again to persuade Gibelli to let them monitor her in the special unit. However, she insisted on going home. The protection plans fell apart. Talk of taking Mia into care was forgotten. The assessment sessions to decide the long-term future were postponed. Mother and baby were to be allowed home. Channi Kumar, a psychiatrist, warned of a risk to the child and the need to monitor mother and baby.

The progesterone injections started on 7 May, the day after Mia was born, and seemed to be working, with Gibelli totally normal. To confirm it, a psychiatrist examined her before discharge. He pronounced her mentally well.

With mother and child at home, the onus was now on Area 8 to ensure they were monitored effectively and convene a conference that would take final decisions on Mia's future. Area 8, however, was distracted by industrial action and difficulties in finding staff.

Social workers now began to make unscheduled visits, often in pairs, as monitoring and assessment got under way. But most of the visitors had not read the case files.

Even the case worker assigned to Mia could not 'take out a significant block of time to read the file' and only completed reading it two weeks before Mia's death. Health visitors and midwives who called were unaware of the full background.

Meanwhile, Gibelli was fretting over the conference she knew could decide her daughter's future. 'We kept asking 'when's the conference?',' Mr Shields recalls. 'Christine was convinced they'd take Mia away because of what happened before.'

She would have been wrong. The conference, repeatedly postponed, was finally set for 29 June, nearly eight weeks after Mia's birth. Briefs prepared for it show that she would have kept the baby.

To control her depression effectively, the progesterone had to be injected every 24 hours. On 28 June, a Sunday, the district nurse was unable to call to administer the injection. The next morning, shortly after 7am, Gibelli drowned her daughter.

(Photographs omitted)