Conceptually, prenatal screening is simple. The goal is to find some characteristic of the fetus that is more common for babies who have Down syndrome, which can then be used to provide information to parents about the chance that their baby is affected.
It may be easier to understand the basic idea with a nonpregnancy-related example. Consider the much less emotionally fraught process of shopping for fruit. At the store you’re faced with a selection of cantaloupes, and you want to make sure the one you choose is ripe. In order to know for sure whether a particular one is ripe you’d have to cut into it and taste it. This, of course, isn’t possible to do before you buy.
What you likely do instead is try to figure out whether the fruit is ripe by looking at things you can see on the outside. What color is it? How does it smell? People have all sorts of tricks for doing this. Someone once told me you can figure this out based on whether the fruit is especially heavy. Whatever your personal system is, it’s all the same theory. Take the color. On average, melons that have some green rind are less likely to be ripe. If you see a melon with no green rind, therefore, you think it’s more likely to be ripe. A statistician would say that you are trying to infer the truth (whether or not it is ripe) based on a signal (in this case, whether or not the rind is green).
Using these techniques, you pick the cantaloupe you think is most likely to be ripe and you buy it. But you know that no matter how good your tricks are, there is still some risk. There is some chance that when you get home and cut into the melon, you will find it is not ripe. Some of the cantaloupes that look ready are not. On the flip side, there are some cantaloupes in the bin that get left behind because they don’t look ripe—they are green, or they don’t smell much—and yet they actually are ripe. These are two different kinds of “mistakes.” In the first case, you think everything is fine but it is not. In the second, you think there is a problem but there isn’t.
This example may seem completely unrelated—and in terms of emotional valence and importance, it no doubt is!—but in terms of the statistics, this is very similar to how the first trimester screening works. Doctors want to identify babies who are healthy (the ripe melons in the example above). They have found some features that are more common among healthy babies (in the example, no green on the cantaloupe rind). If they see that the fetus has one of these good signs, this makes it more likely that the baby is healthy.
In the case of Down syndrome, there are three markers the doctor is looking for in the first trimester screen. The most useful of these is a measurement of the amount of fluid behind the baby’s neck (called the nuchal translucency, or NT). Fetuses who have Down syndrome are much more likely to have a lot of fluid behind the neck. Doctors also measure two hormones in Mom’s blood (PAPP‑A and HCG). Women who are carrying fetuses with Down syndrome also tend to have different hormone levels than those whose babies have normal chromosomes. By comparing your baby’s measurements and your hormone levels to those of mothers and fetuses with and without Down syndrome, your doctor can learn about your baby’s health.
Just like with the fruit, these procedures cannot tell you for sure whether or not the baby has Down syndrome. Sometimes you think the baby is healthy based on these signs, but she is not. In other cases, the signs are worrisome, but in fact the baby has normal chromosomes. The key is that even though it is not perfect, the test can provide a lot of information.
Why use both hormones and the fluid scan rather than just one or the other? Simple: if you have more signals, you can learn more. Go back to the ripe fruit example. Most people would use both color and smell to pick the best cantaloupe. You do this because you think that by using both pieces of information you are more likely to get the right melon. You would do worse if you were allowed only to look, or if you were allowed only to smell. In fact, some doctors go further and also look for a nasal bone on the ultrasound scan. Fetuses with Down syndrome are less likely to have a visible nasal bone, so if doctors see one it’s another signal that the baby has normal chromosomes.
After doing all these measurements, your doctor will tell you a new level of risk. This risk will be based on these new measurements, and also on your age. Your old risk was based only on your age, so, depending on what the measurements look like, your new risk might be higher or lower. If your screening was normal, your risk will go down relative to what it was based on age alone. If your screening was abnormal—for example, if your fetus has a lot of fluid behind the neck—your risk will go up.
This new risk is a number. The doctor will say something like: based on your age and the blood test and the ultrasound, your fetus has a one in 500 (or 1,000, or 10,000) risk of having a chromosomal problem.
This number—this final risk—is an important piece of information. But, in fact, you’re not quite done. Because at this point, after the screening, your doctor will make some recommendation about more testing. Remember: this screen cannot tell you for sure if your baby has a chromosomal problem. If the risk is high after the screening, your doctor will typically recommend that you go on to do more testing—either the CVS [Chorionic villus sampling] or the amniocentesis.
Doctors base this recommendation on a cutoff. A common one is a risk of one in 300. If your final risk is higher than one in 300 (say, one in 200) you’ll be told that you should undergo more extensive testing. If it is lower (say, one in 400) you’ll be told there is no need to do more.
At the risk of belaboring the point, return for a final time to the cantaloupe selection example. This cutoff approach is equivalent to dividing the bin of fruit into “ripe” and “not ripe.” Customers are told the ripe fruit is fine and the not-ripe fruit should be further investigated. But because the selection isn’t perfect, there are mistakes in both bins; the same is true here.
Everyone who has a final risk above one in 300 is told they are positive. In fact, a lot of these people have perfectly healthy babies. It’s easy to see why this is. Think about someone with a risk of one in 200. This person is told that her screening is positive. However, there is still less than a 1% chance that her baby has Down syndrome! Of 200 women like this, one will have an affected baby and 199 will turn out to be false positives.
On the flip side, think about someone with a risk of one in 400. She is told her screen is negative, and it’s true that there is an overwhelming chance that her baby is healthy. However, there is still a chance that her baby has Down syndrome: of 400 women like this, we expect that one will have a baby with Down syndrome. This person is a false negative.
Reprinted with permission from Expecting Better: Why the Conventional Pregnancy Wisdom is Wrong—and What You Really Need to Know, by Emily Oster, published by the Penguin Group. Copyright 2013 Emily Oster.