The BMA based its decision to keep the abortion time limit at 24 weeks on evidence of the poor survival rates of babies born more prematurely.

A briefing paper prepared for the debate reported the results of the EPICure study, led by Neil Marlow of Queen's Medical Centre, Nottingham.

It looked at 761 babies born up to the 24th week of pregnancy during a 10-month period in 1995 in the UK and Ireland. It found 41% died in the delivery room and 42% in the neonatal intensive care unit. That left 128 (17%) who survived and were discharged, but 2% later died.

The researchers assessed 97 of these children at six years old. They found 28% had a severe disability that was likely to make the child heavily dependent on carers. Another 26% had moderate disability, 33% mild disability and 13% survived without impairment. Survival rates for 379 children born up to the 23rd week were lower, with 60% dying in the delivery room and 33% in intensive care. Of the 28 who survived (7%), one later died. When 24 of these children were assessed at six, 12% were found to be without impairment.

Two of the 138 babies born at 22 weeks survived long enough to be discharged from hospital. One had severe disability and one had mild disability at six.

During the debate, doctors backing a reduction in the abortion time limit said survival rates have improved since the EPICure research.

But BMA leaders said the medical royal colleges recommended that the EPICure results were the most up-to-date measurement of survival and disability.