Bryan Caplan  

Public Opinion About Fetal Testing

More on Taleb... Price Controls on Gas: Is the ...

If you're having a high-risk pregnancy, it's pretty obvious that you'll be more likely to have your fetus tested for birth defects and other problems. But is risk the only factor that predicts demand for fetal testing? Nope. Here's an interesting abstract from a study of 612 women (2/3 high risk, 1/3 low risk):

As expected, traditional risk status predicted desire for screening and also invasive testing following either a negative or positive screen. Yet, controlling for risk status, many psychosocial variables predicted a women's interest in screening and in invasive testing: more knowledge about prenatal testing, concern about fetal health, willingness to terminate a pregnancy, and an internal or medical profession health locus of control.
If you buy the Enlightened Preference approach, this suggests that women aren't doing enough fetal testing. What do you think?

P.S. Here's another thought-provoking abstract:

A survey asked 190 pregnant women their opinions on whether a hypothetical other woman would be justified in having an abortion under ten different circumstances, four of which related to abnormalities of the fetus. They were then asked whether they would be justified in having an abortion under these same circumstances... Respondents were more likely to view abortion as justifiable for a hypothetical other than for themselves. However, there was no evidence that pregnant women's views about abortion are more conservative than those of the general public. The majority of respondents considered abortion of a defective fetus to be justified, both for themselves and for others. The best predictor of abortion attitudes was respondent's views about the ideal number of children in a completed family.

Comments and Sharing

COMMENTS (9 to date)
Mitch Oliver writes:

My wife and I are dealing with this decision currently. We have foregone additional fetal testing because of the additional risk introduced by the tests themselves and the invasiveness of the procedures without providing sufficient benefit. Whether we will have "enough" testing is quite irrelevant to us. The opportunity cost of additional testing is too high.

Jeremy Bettis writes:

Is there any action that can be taken if the fetal screening comes back bad other than abortion? If not (or if hardly ever) it makes zero sense to do any screening if you are unwilling to perform an abortion due to moral convictions.

And "ideal number of children" is probably a proxy for religious belief statistically.

Justin Time writes:

My wife and I were just notified that our unborn child has a CP Cyst. This is one marker for Trisomy 18. Babies with Trisomy 18 are severely handicapped and have low life expectancy. Neither of us are open to abortion, but knowing in advance may affect decisions during delivery if there are complications. In other words, doctors will place more priority on the life of the mother.

Fortunately, there were other markers and a blood test that made amnio unnecessary for our piece of mind.

To answer your question, Jeremy, yes there are conditions that are treated. You can just see this Wikipedia article for a few examples. Whether that meets your standards for "hardly ever" I don't know.

Dr. T writes:

As with almost all aspects of medicine in the United States, fetal testing is excessive. My specialty is clinical pathology (laboratory medicine), and I have much expertise in this area.

Many years ago, obstetricians arbitrarily established a risk cutoff of 1 in 200 for deciding when to recommend amniocentesis. Why 1 in 200? Because that equaled the risk that amniocentesis (sticking a needle into the sac of fluid around the fetus) would cause the death of a healthy fetus. I would have chosen 1 in 500 as the risk cutoff for recommending amniocentesis, because I think a death rate of 1 in 200 is too high. (Note: As amniocentesis techniques improved, the risk of fetal death fell to 1 in 270, which some programs adopted as their cutoff.)

Because of the 1 in 200 (or 270) cutoff, many amniocenteses are performed, and many genetic and chemistry tests are done. Initially, testing was for neural tube defects (NTDs) that affected 1 in 500 pregnancies in the US. Later, it was discovered that the screening test for NTDs could be used to recalculate the risk of Down's syndrome, so testing served two purposes. NTDs mostly have been eliminated after we learned that folate is preventive. Most pregnancy testing now is focused on Down's syndrome and similar chromosomal disorders. Because the incidence rates are low (and the screening tests are imperfect), many women must be tested to uncover one Down's syndrome fetus.

At present, approximately three chromosomally disordered fetuses are aborted for every healthy fetus lost due to amniocentesis. I believe that the costs of this protocol (in dollars, worry, and fetal deaths) exceed the benefits (prevention of some births of some defective babies). I believe the screening and testing program should be redesigned to improve the cost/benefit ratio.

Matt C writes:

At a guess, too much testing. I agree with the premium medicine/overtreated view of how medical care works in the US, and this is no exception.

When my wife was pregnant with our son, we got an abnormal result on the "triple screen". Maybe possibly he had Down's syndrome. We were counseled and got a lot of pressure to get an amniocentesis. We knew it was risky and declined. The counseling doc was quite unhappy with us--if he had been able to, I think he would have forced one on us.

Anyway, after it was done, we realized it didn't make much sense to get these screenings if we weren't willing to abort in the case of verified abnormality (at least for Down's syndrome, we weren't). All the screening did for us was raise our anxiety levels during the last 2/3 of the pregnancy. (Our son was normal BTW.)

The other guy said there were some pregnancy-treatable things to screen for, so perhaps these screenings make sense sometimes. But I agree with the other other guy that before testing, you ought to think through what you're testing for and what you're going to do if the results come back abnormal. If the answer is "nothing different", just skip the test.

Ed Lopez writes:

I found the comment by "Dr. T." to be highly informative. Thanks.

My wife and I are about 7 weeks away from delivery. Thankfully our chromosome screenings went well so we didn't have to consider the testing trade-offs. But the 3-to-1 ratio Dr. T. cites seems amazingly, sadly low. So from a normative view, I'd have a hard time agreeing that there isn't enough fetal testing. But from a positive view, I dont' think Bryan gets the enlightened preferences implication right. The variables listed in the abstract (I couldn't access the paper, even through my university's e-journals) are knowledge of screening/testing, concern for fetus, attitudes toward termination, and health locus of control. I buy the part that if knowledge and willingness to test are directly related, then Bryan's implication is right. But the other variables don't easily fall onto an enlightenment dimension. If you spend more time thinking and reading about it, will you become more concerned for the fetus? Doubt it. I'd say a corner solution driven by unfolding circumstances is more likely, as with the scenario laid out by Justin Time--which I hope does not come about. As for attitudes toward termination, I don't think there is a defensible way to relate abortion attitudes monotonically to enlightenment. And locus of control seems to have more to do with family background than education ( For the most part I buy the enlightened preference application to politics. I just don't think it applies here except on the knowledge variable. Maybe that's enough. That's an empirical question. Wish I could see the tables in the article.

Consider libertarian paternalism. Having read most of the book, Nudge, I haven't yet found Sunstein and Thaler talking about fetal testing. Suppose the locus of control were at the external maximum, so couples felt completely powerless. If you were the enlightened expert establishing a default stance for uninformed couples facing decisions on fetal testing--specifically risky ones like amniocentesis--what would you do?

James writes:

"The best predictor of abortion attitudes was respondent's views about the ideal number of children in a completed family."

Does this sentence stand on its own or is it connected to the defective fetus scenario?

IOW, is the attitude: "The fetus is defective and I already have the ideal number of children. Therefore, combining those two factors, abortion is an acceptable decision."

Or is the attitude: "I already have the ideal number of children. Therefore, based solely on that factor, abortion is an acceptable decision."

Peter St. Onge writes:

Given the massive costs, including nonmonetary, of a handicapped child, compared to the costs of 'keep trying' or another IVF cycle, it seems insane to not test every single pregnancy.

Assuming, of course, one is willing to abort. If not, I suppose there's no point in knowing, and the risks to the foetus dominate.

Comments for this entry have been closed
Return to top