Disclaimer: No medical training.

Amniocentesis is a sample of the amniotic fluid to test fetal DNA at 15-20 weeks of pregnancy. It involves inserting a long needle into the amniotic sack, which is potentially risky. But it could catch serious defects.

Arguably the main cost of the procedure comes from an increased risk of miscarriage, which increases by about 1 in 1,000. The best-estimated risks (see this meta-analysis) may still be confounded by selection into the procedure, but experts agree the risk is non-zero. Depending on the person, you might also consider the stress entailed in the process, the feeling of not being done testing, and the cost of a false positive.

What are the benefits? This is harder, because pregnancy involves a lot of correlated tests. For instance, Down syndrome can be detected in "cell free" DNA testing (also known as NIPT) at week 8. It's also detected with much higher accuracy in the amnio. But a negative in the cell free DNA test will drastically reduce the chance that an amnio comes back positive for Down, so it certainly reduces the benefit of amnio--enough that current convention is to not recommend amnio after a negative NIPT result. 

There are also ultrasounds. These can detect issues that are also caught in genetic tests, like Down. But they can also detect issues that are not currently detectable in DNA. For instance, it's estimated that roughly half of all cases of Noonan syndrome are novel, meaning a genetic test for them wouldn't come back positive based on known mutations. 

So the amnio can help catch things that are (i) genetic and (ii) invisible to the ultrasound and other tests. 

To complicate things further, DNA testing of the amniotic fluid is currently performed at three levels of detail. From lowest to highest resolution:

  1. Karotyping or cytogenetic. The lowest resolution, this essentially counts chromosomes. It detects the same things as the cell free DNA test, but with greater accuracy. 
  2. Microarray. Can catch missing or deleted chromosomal segments. 
  3. Prenatal exome sequencing: Genotyping the fetus. Explainer here

Getting an increased risk of Down in the cell free DNA test will greatly increase the benefits of the test. Most medical advice treats the amnio as an obvious choice if there's a finding of increased risk. 

What about for the majority of people, who have not had any indications of elevated risk?  I consider the case of the microarray, which is becoming more available.

Because everyone agrees that the procedure is at least a tiny bit dangerous, and that the risk of a serious syndrome is small, the question hinges on exactly what information is added by the test. The benefit from the amnio is zero if it's only catching things that the NIPT, blood, nuchal translucency, or ultrasound screens would catch (assuming the timing is roughly similar). What is the added information?

Most of the medical papers on the topic were not suited to this question, but Srebniak et al (2018) is almost a perfect fit. They study the results from microarrays and restrict to fetuses that were “karyotypically normal” (normal number of chromosomes)--which is where you would be after a negative NIPT.

The most important outcome seemed to be an "early-onset syndromic disorder." In their meta-analysis of 10,000 mothers, 0.37% of fetuses had this kind of issue. These were serious issues. They write: "most were deletions of various chromosomal regions causing loss and disruption of many genes and leading to intellectual disability, developmental delay, dysmorphic features and variable structural anomalies...A significant number of these syndromes are actually more severe than Down syndrome."

Would these have been detected otherwise? They say that “detection of these disorders by routine ultrasound was assessed as generally unlikely” and suggest around a 50% chance. Not very exact, but still useful. 

Taking their findings literally, amnio reduces your chance of a surprise serious issue by 1/540 ( = 0.37% * 50%), while increasing your chance of a miscarriage by 1/1,000. Serious sub-chromosomal issues are surprisingly common, and about half of them need an amnio (with microarray) to be detected. This might be enough to justify the amnio for some people, although that goes against the standard recommendations.

The main reservation I have is that these numbers seem too high: most people do not get these tests, so does this mean that the population rate of early-onset syndromic disorder is around 0.37%? It does not seem right that 1/270 live births have a genetic abnormality often worse than Down syndrome, but I couldn’t find a source speaking directly to this issue.

One possibility is that the mothers in the meta-analysis had unobserved factors that put them at higher risk. This is pure speculation, but you might want to check if your cost-benefit calculation changes if you further decrease the prevalence estimates by a factor of 2.


New to LessWrong?

New Comment
4 comments, sorted by Click to highlight new comments since: Today at 11:06 PM

A couple of thoughts, having been through this.

  1. Not all of the issues caught would have made it to birth, so it's less than 1/270 live births -- a bunch of chromosomal abnormalities end in miscarriage. It's a little hard to say exactly how this affects the benefit of amnio. On the one hand, maybe you didn't need to do anything for the cases that would miscarry anyway. On the other hand, there's probably some value in knowing that this isn't going to be a good outcome, and being able to choose a more defined outcome if you want.
  2. It's worth thinking a bit about what you would do if you found something. For instance, if you're not going to abort if you find Down's, then the value of finding goes down a lot. ~90% of people do abort in that situation so maybe that's not the mainstream case, but that could cut down on 10% of the amnios done.
  3. It's a big-ass needle that they use for amnio. Possibly this will freak out some people more than others -- i.e. the cost is higher if you don't enjoy being stabbed by a giant syringe. I thought it was interesting to watch, but I wasn't the one being stabbed, so YMMV.

Really agree with all of these, thanks. Curious, in your decision-making process, did you ever have a way to calculate “the chance of a really disabling (as bad as Down syndrome) disorder”?

I looked up some stats, but unfortunately this was ~15 years ago and I have no idea where I found them or what they are.

(Also not medically trained.)

Something missing from this analysis is that the expected probability of these conditions for any given pregnancy is not the same as the incidence in the population at large. The factor that I've most often heard about is increasing age being highly associated with increasing incidence of Down syndrome, though there may be others, and I'm not sure whether there are known correlates with the other conditions you mention.

That might also relate to the last point about incidence of these conditions in the wider population and the incidences that study reported. It could be that older pregnant people are more likely to opt for the test, knowing that they are at elevated probability.

You finish by suggesting people think about prevalence differences shifting by a factor of two, but from a quick Google, it looks like age can shift prevalences by orders of magnitude. The first table that popped up suggested 1 in 2000 at age 20, increasing to 1 in 100 at age 40 and 1in 10 at age 49.