Human or Statistic? Do we know what percentage of Down syndrome kids are killed by abortion?

How many children diagnosed with Down syndrome in the womb are subsequently killed by abortion?  (This is usually called a “termination rate.”)

J.L. Natoli and colleagues published in 2012 the best estimate available on this question.  It’s a systematic literature review, rather than a study with original data.  [J.L. Natoli, et. al. Prenatal diagnosis of Down syndrome: a systematic review of termination rates (1995-2011); Prenatal Diagnosis, 2012, 32, 142-153.]

This literature review does not make a claim about the overall termination rate in the US.  The researchers note that the percentage of kids killed (the study doesn’t use that language, of course, but that language is accurate) varies by region in the US and by age of the child (or “pregnancy”), and then they give a “weighted mean” in three different categories.

• The range of population-based studies was 61-93%, and the weighted mean was 67%.
• The range of hospital-based studies was 60-90%, and the weighted mean was 85%.
• The range of the anomaly-based studies was 0-100%, and the weighted mean was 50%.

The weighted mean in the third category is not as helpful, since the sample sizes of these studies were so low.  It appears, then, that about two out of three (67%), and possibly as high as nine out of ten (85-90%), children diagnosed with Down syndrome in the womb are subsequently killed by abortion.  (Take care, though, when citing even this statistic, especially keeping in mind Note 2 below.  I am not claiming here that 67-90% of children diagnosed in the womb with Down Syndrome were killed because they had Down syndrome.)

Here are a few thoughts about what we can learn from this literature review:

1. In terms of relevance to the question of the US termination rate, this study replaces the oft-cited Mansfield literature review (1999) for two reasons.  First, the Mansfield review includes only a few studies from the US, and they are small studies.  So, when it reports in its abstract that 92% is the termination rate for Down syndrome, that number really must be read within the context of the study itself.  Second, the review is now 16 years old, so we must take care in projecting its findings onto the present day since various factors might be different now.  For more on the liabilities of citing the Mansfield study, see the introduction to the Natoli review (2012).  See also my note here.  So, articles like Amy Harmon’s “Prenatal Test Puts Down syndrome in Hard Focus” (New York Times, May 9, 2007) which cite Mansfield’s study as applicable to the US (and probably utilize the 92% number from the abstract) should now be read in light of Natoli’s more comprehensive review of more recent US stats.
2. Keep in mind that we can’t know for sure from this study whether or not Down syndrome was the primary reason for termination.  This is because this study didn’t separate out situations in which Down syndrome was the only prenatal diagnosis of deformity or disease that was made. For example, if a woman underwent prenatal diagnosis which found Down syndrome and hypoplastic left-heart syndrome, we cannot know if she terminated for Down syndrome or HLHS or something else. If we looked at the studies themselves, rather than this systematic review, we might be able to determine that Down syndrome was selected for in the particular study, but we cannot ascertain that from this published systematic review.  (I am reminded, though, as I am re-reading this before posting, that it is hardly a comfort to the compassionate heart if more of the parents in these studies were motivated to kill their kids due to the presence of some other disability and less due to the presence Down syndrome.  While it’s important to be accurate, it’s also important to see the forest for the trees.)
3. I was particularly happy to see that Natoli et al. excluded studies which “failed to provide the total number of pregnancies with prenatal diagnosis” (144).  I think this is critical to ascertaining an accurate termination rate.  If I’m not mistaken, this mistake is the source of Leroi’s flawed assessment that there’s a 29% abortion rate for Down syndrome in the US (in “The Future of Neo-Eugenics”).  The study Leroi is referring to is Cragen, et al., “Including Prenatal Diagnosis in Birth Defects Monitoring: Experience of the Metro Atlanta Congenital Defects Program,” Birth Defects Research (Part A), 85, 20-29 (2009).
4. I cited the Mansfield study in Common Ground Without Compromise (Question 10).  See Endnote 63 here.  This blog post updates the work I did at that time to try to determine an accurate termination rate, as reflected in that portion of my book.

Related Post: What percentage of Spina Bifida kids are aborted? (June 23, 2011)