Risk of Amniocentesis
The first amniocentesis for chromosomal analysis of cultured fetal cells from amniotic fluid was described in a brief article by Steele and Breg in 1966 (Lancet 1966;1:383-5). Ten years earlier, Fuchs and Riis (Nature 1956;177:330) had reported doing the procedure just for fetal sex determination. In both instances, these procedures were done ‘blindly’ inserting a needle without any method of direct visualization into the uterine cavity.
When I started my OB/GYN residency in 1979, our technique was barely better. We sent the pregnant woman to ultrasound (static), a fluid ‘pocket’ was identified, an “X” was placed on her abdomen at the “best place to stick a needle,” she was sent back to our clinic, and we BLINDLY placed an 18 gauge (large bore) needle into the uterus at the site of the “X.” It was not at all unusual to meet some resistance (some part of the baby – never felt real comfortable about that) or draw back blood (from the placenta, umbilical cord, or uterine blood vessels) before finally getting the fluid we needed to do the fetal chromosomal studies.
Despite the relative crudeness of this approach, the chance of losing the baby after this procedure was only about 1 in 200, and that has been the number routinely quoted to women for the “risk” of amniocentesis over the past 30 years, even though we now do these procedures using direct, real-time ultrasound guidance and a thin, usually, 22 gauge needle. Most of us who have done lots of these over the years (I put my numbers somewhere between 5 and 10,000) have known the risk is much lower. In fact, I only know one woman who lost a baby following an amniocentesis I have performed in the past 25 years (KNOCK ON WOOD), and she was probably well on the way to losing that pregnancy due to other complications. Anyway, it has been my routine for awhile to tell most women who come for amniocentesis that the “quoted risk is 1 in 200” but my “undocumented risk is less than 1 in 1000. Besides, I stayed at a Holiday Inn Express last night.”
Finally, as the result of a very large study that included documentation of outcomes following midtrimester amniocentesis, we have data to support the relative ‘safety’ of this procedure. In this study, the FASTER (First And Second Trimester Evaluation of Risk for aneuploidy) trial, the 3,096 women who underwent midtrimester amniocentesis (at many different institutions) were compared to 31,907 who did not (Eddleman, et al., Obstet Gynecol 2006;108:1067-72). The spontaneous fetal loss rate at less than 24 weeks following amniocentesis was 1.0% compared to the background loss rate in the control group of 0.94%. Therefore, the added risk of the amniocentesis was calculated to be only 0.06% (1.0% minus 0.94%) making the risk of the amniocentesis approximately 1 in 1600 – far less than the 1 in 200 (0.5%) risk that is usually quoted.
This information was presented to our patient today, but she was still very anxious about the prospects of the procedure. I told her that we saw no abnormalities of the baby at this time and that alone reduced her “age alone midtrimester risks” of 1 in 100 for Down syndrome (trisomy 21) and 1 in 51 risk for all chromosomal abnormalities by at least 50%. So, I offered her another option. “Why don’t you have a ‘quad screen’ (maternal serum screen) done today as a more reliable estimate of risk than that based on age alone. We will have the results back within 10 days, and you can use that information to further guide your decision regarding amniocentesis.” I then told her, that “because the odds were still in your favor that the baby is chromosomally ‘normal’, I am just going to schedule you for a ‘genetic sonogram’ at 18-19 weeks, and if the ‘quad screen’ results are scary, or if you just change your mind about the amniocentesis, we can see you back sooner and on very short notice.” She seemed happy with that plan…and it was the right decision for her at this time...so, I have a new friend!