The latest test for Down's is up to 98 per cent accurate and costs the NHS just £25. So, asks Annalisa Barbieri, why isn't it offered to all pregnant women?

When I was pregnant with my first child, in 2003, I had little clue about the efficacy of the screening tests one is offered. When I was booked into University College Hospital, my midwife mentioned something called the integrated test for Down syndrome (UCH introduced it in February 2003).

At 12 weeks, I duly went along for a scan where the nuchal fold of my developing baby was measured; I gave some blood. At 15 weeks, I went back for more blood tests. The one thing I remember clearly was being told that the integrated test gave you no results until at least week 16 – that was the disadvantage – but that it was more accurate than any other non-invasive test offered.

At about week 17, I got a letter: my odds for having a baby with Down syndrome were less than one in 50,000. I'd never heard of such odds, and neither had anyone I talked to. "Wow," everyone said, "but you're 36." (I was 37 at the time of giving birth, the age they take into account).

Soon after, I joined the world of internet chat-rooms and began to realise that very few women had heard of – or been offered – the integrated test. Most were offered a blood test, maybe a nuchal fold scan: the triple test or the quadruple test, occasionally something called the combined test. Some, horrifyingly, were given odds based only on age (this is only 32 per cent accurate). They were given odds that confused them, and gave them days of worry: should they have an amniocentesis or chorionic villus sampling (CVS), the conclusive, but invasive, tests that carry the risk of miscarriage?

This month, reseachers at Stamford University in California said they had devised a blood test for Down syndrome, which could be available in five years. But last month came news that too many healthy foetuses are dying after unnecessary amniocentesis and CVS: in preventing the births of 660 babies with trisomy-21, 400 foetuses that didn't have it are miscarried. The Royal College of Obstetricians and Gynaecologists stands by the accepted risk of between 1 and 2 per cent in having such procedures. Whatever the reality, it raises a question: why isn't every woman who wants it offered the integrated test?

At this juncture, I must make an important point. This isn't about finding a test so accurate that no children with Down syndrome, or trisomy-21, will ever be born. It's about giving parents who want it as accurate a picture as possible so they can make an informed choice and not frightening them into having invasive and perhaps unnecessary procedures such as amnios and CVS.

Let's take the tests in turn. The first was the double test. This took maternal age into account as well as looking for two blood markers: alpha fetoprotein (AFP) and human chorionic gonadotrophic (hCG). AFP tends to be decreased in Down foetuses, hCG increased. It has a 59 per cent accuracy rate: that is, it picks up 59 per cent of babies affected. The triple test added another marker to the blood test, uE3, which tends to be decreased in affected babies. Its detection rate is 63 per cent. The quadruple test added a further marker, inhibin A, which tends to be increased in foetuses with trisomy-21; the quad test is picks up 72 per cent of affected babies.

After this came the combined test, which added to the blood tests by combining data from a nuchal fold scan (which must be done between 11 and 13.6 weeks). Research had found that the thicker the nuchal fold the more likelihood of a baby having Down syndrome. (New markers now also look at other markers such as the presence of the nasal bone – the sooner this is visible, the less likelihood of the syndrome – and the tricuspid valve flow through the heart.) The combined test is said to be 76 per cent accurate. The nuchal test (NT) on its own is only 69 per cent accurate, but relies heavily on the skill of the sonographer.

Then came what should be the industry standard: the integrated test. This combines maternal age, NT result, one blood test at about 12 weeks that looks for PAPP-A (pregnancy associated plasma protein A) and another blood test at 15 weeks that looks at further biochemistry: AFP, hCG, uE3 and inhibin-A. At worst, it has an 89 per cent detection rate, at best 98 per cent (it is more accurate the older the mother is). Put personally, at 37 I had a risk of 1:200 vs the less than 1:50,000 the integrated test gave me. At 42, I had a risk of 1:55 vs 1:1,900.

Crucially, the integrated test cuts the number of false positives right down, and the cut-off point for screen-positive odds (after which women are advised to have an amnio) is 1:150 for the integrated, compared to 1:250 for the others. The bottom line is that if women have screening tests with a low rate of accuracy, the reality is that they are more likely to opt for invasive tests that put their babies at risk.

Only two hospitals currently offer the integrated test on the NHS, St Mary's and UCH (my investigations tell me that UCH will cease to offer the integrated test routinely next month, but it had not confirmed or denied this at the time of going to press). Women can opt to have it done privately at a cost of about £170 (it would cost the NHS much, much less), but even then it often involves travelling large distances.

Why isn't it industry standard? Pat O'Brien, obstetrician at UCH and spokesperson for the Royal College of Obstetricians and Gynaecologists, says it's largely to do with "expense [of the scans; not all hospitals offer scans at 11-13 weeks] and training". It takes an extra day to train a sonographer to do an NT, but there is a huge shortage of sonographers. A further setback was that, in March, NICE published guidelines recommending the combined test, partly, one can imagine, due to funding.

Many hospitals already offer nuchal scans, and the quad test. Adding the extra blood test at 15 weeks (to make the combined test into the integrated test), according to one screening insider, would cost the NHS "about £25". This isn't an official figure, but for about that, you could have a test that would substantially increase the detection rate and cut down false positives. We need to weigh this up against the cost of an amniocentesis, which has to be performed by an obstetrician further trained in fetal medicine. Oh, and against the potential loss of a healthy baby – but I guess that doesn't figure on anyone's spreadsheet.