New study casts doubt on Down's syndrome test

A blood test for Down's syndrome, hailed as a breakthrough when it was introduced eight years ago and costing the NHS up to £9m a year, is useless, says new research.

A blood test for Down's syndrome, hailed as a breakthrough when it was introduced eight years ago and costing the NHS up to £9m a year, is useless, says new research.

The "triple test" developed at St Bartholomew's Hospital and also known as the "Barts test" was launched in 1992. By 1994, more than half of all obstetricians were offering it to all the women under their care. The test is claimed to double the proportion of babies with Down's detected to over 60 per cent when compared with relying on the mother's age and ultrasound scans. But researchers in Southampton have shown that using the traditional method alone can also detect more than 60 per cent of Down's cases.

The finding undermines the case for the triple test, which costs £15 to £20 a time in laboratory costs alone and provides only an indication of the risk of Down's. The counselling time increases the cost to an estimated £20 to £30 a test. On the basis of 300,000 births a year, half the national total, that implies a total cost to the NHS of £6m to £9m a year.

Down's syndrome is the most common cause of mental handicap, affecting one in 750 births, and its incidence rises steeply among older mothers. Doctors have traditionally offered mothers over the age of 35 the option of amniocentesis, which provides a definite diagnosis of Down's. However, it is an invasive test, involving inserting a needle into the womb to extract a sample of fluid, and carries a small risk of miscarriage. Obstetricians saw the triple test, based on a blood sample, as a safe way of telling which women are at high risk.

David Howe and colleagues at Princess Anne Hospital, Southampton, which does not offer the triple test, studied all 31,000 babies born at the hospital between 1993 and 1998.

All women over 35 were offered amniocentesis. In addition, all women of any age were offered an ultrasound scan at 19 weeks followed by amniocentesis if any abnormality was detected. The results, published in the British Medical Journal, show they detected 57 Down's babies before birth (68 per cent), more than twice the 30 per cent generally thought to be achievable using the mother's age and scans alone.

Of the 17 not detected antenatally, seven had refused amniocentesis and three were twin pregnancies in which the triple test would not have worked.

Mr Howe, a consultant obstetrician, said the triple test has never been submitted to the rigours of a comparative trial.