I happened to get my hands on some interesting literature on pre-natal genetic screening, literature that amply reinforces my impression that clinical genetic testing is still in the dark ages.

Let's say you (or your wife/fiance/girlfriend) are pregnant, and you're interested in taking a blood test to see if your baby is going to develop a neural tube defect, like spina bifida. Should you take the test?

In this case, it's a no-risk test (unlike amniocentesis) that involves measuring a blood protein called AFP. Here's what the pamphlet I've got says:

- There is a 1:1000 chance that your baby will have spina bifida.
- The blood test can identify 80% of spina bifida cases.
- The false positive rate for the test is 1%-3% (let's be generous and call it 1%).

Given these numbers, you're much, much more likely to have a healthy baby, but get a positive test result than you are to have an unhealthy baby and get a positive result:

Probability that your baby is healthy and the test comes up positive:

99.9% x 1% = ~0.99% - roughly 1 out of 100.

Probability that your baby has spina bifida and the test comes up positive:

0.1% x 80% = ~ 0.08% - less than 1 out of 1000

In other words, you're about 12 times more likely to get a false positive result than you are to get a true positive result. Now if there is a more reliable, non-invasive follow-up of initial false positive test, this might not be such a bad idea, but otherwise, this test doesn't appear to be that helpful.

Here's another example from the physician's office brochures I've got here: Trisomy 18, which is when a baby has an extra copy of chromosome 18.

- There is a 1 in 7500 chance (~0.013%) that your baby will have Trisomy 18
- The blood test detects 90% of trisomy cases, and has a 3.3% false positive rate.

Your chance of having a healthy baby but a positive test result is about 3%. Let's say that, after a positive test result you decide to follow-up with amniocentesis, which carries (according the the pamphlet I've got) a 1:500 risk of miscarriage.

So your risk of having a healthy baby, a false positive test result, and an amniocentesis-induced miscarriage is about 0.006%

What's your risk of having a baby with trisomy, and getting a positive test result (recall that the test picks up 90% of trisomy cases)? About 0.012%.

In other words, your risk of amniocentesis-induced miscarriage after a false positive test is fully half of your risk of having a trisomy fetus and a positive test.

People can have many reasons for taking the test - like peace of mind. If you do get a negative result, you can rest easy, but your chances of getting a false positive result are high - much higher than the chances that your baby actually has any of these conditions.

What's needed is a non-invasive way of collecting fetal DNA that doesn't put you at risk for miscarriage, followed by truly modern genetic testing - the kind that involves DNA sequencing, and which, for many Mendelian diseases (diseases caused by mutations in single genes), are much more reliable than the typical blood protein tests or quick assays for major chromosomal abnormalities.

I don't give medical advice here, but it's worth getting some before you take these tests - preferably from someone who doesn't stand to benefit financially from your decision to test.