“Not exactly a great scientific achievement”: prenatal testing in low risk populations

NEWBERGER D. Down Syndrome: Prenatal Risk Assessment and Diagnosis. Am Fam Physician. 2000 Aug 15;62(4):825-832.

NEWBERGER D. Down Syndrome: Prenatal Risk Assessment and Diagnosis. Am Fam Physician. 2000 Aug 15;62(4):825-832.

Yesterday’s post generated a lot of discussion on Facebook and off-line about the role of maternal age and the incidence of Down syndrome. Today, we consider the value of prenatal screening for Down syndrome in relation to the mother’s age.

Most women accepting prenatal testing for Down syndrome are not doing so in order to find out they are carrying a child with Down syndrome. Rather, most are seeking reassurance that they are not carrying a child with Down syndrome. And, as more and more mothers accept prenatal testing, most of them will get what they are looking for: a negative test result.

This is because the overwhelming majority of expectant mothers are not carrying a child with Down syndrome. Figure 1 takes the age-incidence rate in 5 year increments from the table included in the American Congress of Obstetricians & Gynecologists’ (ACOG) practice bulletin recommending that all women be offered both prenatal screening and diagnostic testing (ACOG Practice Bulletin No. 88):

Age table

Figure 1: ACOG PB 88 maternal-incidence rate

Another way of showing the incidence rate based on the mother’s age is in the chart at the top from the American Academy of Family Physicians. As you can see, in younger mothers, the chance is close to 0% and even for mothers at age 40, the chance is only then nearing 1%. Because of this low-incidence rate, that is why even receiving a “positive” from the new blood test for Down syndrome, with its stated greater-than-99% accuracy, still results in mothers under the age of 30 having a greater chance that their test result is a false positive.  Figure 2 shows this positive predicative value rate based on the ages in Figure 1:

Figure 2: NIPS PPV

Figure 2: NIPS PPV

To further explain, for a 15 year old, if she receives a positive NIPS result at a 99.5% sensitivity rate, she has a 4-in-10 chance of actually having a true positive, and a 6-in-10 chance of her result being a false positive. Similarly, at age 30, the mother has a 51/49 chance for  true positive/false positive.  Only when the mother’s age nears 45 does a “positive” NIPS result with a 99.5% sensitivity actually mean she has a 99.5% chance of a true positive.

But, these figures deal with positive results, when most women are considering prenatal testing in order to be reassured by receiving a negative result. How informative is NIPS with a negative result?

This question was prompted by a comment  by Robert Resta on the blog post that I discussed on Tuesday. Resta is the same Genetic Counselor who noted that in this century, most Down syndrome pregnancies are occurring in women 35 and older, due to the increase in the number of women in that age group being pregnant coupled with the higher incidence rate once a mother turns 35.

Regarding a negative NIPS result, even at the stated 99.9% specificity rate (i.e. the ability to rule out those pregnancies not carrying a child with Down syndrome), Resta offered the following observation:

A 35 [year old] woman has less than a 0.5% chance of having a baby with Down [s]yndrome, so age alone as a screening tool has a negative predictive value of 99.5%. After a normal NIPS, she has a 99.9% chance of not having a baby with Down syndrome. Not exactly a great scientific achievement. Cognitively, most of us cannot distinguish between 99.5% and 99.9%.

This point is made even clearer when considering the negative predictive value of age in mothers under 35 years old.

Table 3 shows the negative predictive value (NPV) based on the incidence rate associated with maternal age:

Figure 3 NPV maternal age

Figure 3 NPV maternal age

So, what is the utility of offering a 15 year old, or a 20 year old, or a 25 year old, seeking reassurance from a negative NIPS result with a 99.9% specificity if, by their age alone, they already have a 99.9% chance of not having a child with Down syndrome? Similarly, how much reassurance is to be gained by the 30 year old by having learned she has a 0.01% greater chance of not having a child with Down syndrome based on a negative NIPS result as compared to her baseline 99.89% chance of not having a child with Down syndrome?

Resta’s opinion:

The very act of offering screening engenders anxiety for a low risk condition, and then we expect pregnant women to be grateful when a test tells them they are at low risk for the condition?

Would knowing that a negative NIPS result means a 0.5% further chance of not having a child with Down syndrome be worth paying $2,700 for the test? 

Comments

  1. I admit to getting a little fuzzy-headed when it comes to statistics, but for me, as a 35-year-old expectant mother, I very much liked the MaterniT21 (and am frankly in awe of the technology–things sure have changed since I had my first child 9 years ago!). It’s not that I didn’t know I was low-risk to begin with, but the reassurance that I am still in the low-risk pool was very welcome. As for the price tag… my insurance covered it (I’ll be meeting my deductible this year no matter what), but if it hadn’t, I probably would have waited for the anatomical ultrasound at 22 weeks just like I always have, since my husband and I wouldn’t terminate anyway and so would only consider an amino in the case of a positive screening result, and then only so we could know & be prepared. I know there are profound ethical issues with this technology (my healthy first child had choroid plexus cysts in utero, which was another exercise in cost/benefit analysis in prenatal screening and diagnosis), but I still think it’s absolutely fascinating.

  2. Daniel Martin says:

    Is Figure 2 based on the assumption that false positives are evenly distributed across age? If you assume instead that false positives are more likely as a mother gets older (which is possible if placental abnormalities increase with age), then you wouldn’t get this trend. Or is this figure based on actual results? If so, what is the source? Thanks!

    • Daniel Martin says:

      Sorry, I should have said “Is Figure 2 based on the assumption that false positives are equally likely at all ages?” Assuming that false positives are evenly distributed across age would imply that false positives are more likely as a mother gets older (as there are fewer old mothers).

      • Figure 2 applies the Bayesian analysis covered in this post to the incidence rates stated in Figure 1. It presumes that the NIPS tests have the same sensitivity/specificity rate regardless of age. I welcome a link to research that shows those rates may change based on the age of the mother (I’m unaware of any such finding being published). Thank you for your interest in the analysis.

  3. My second trimester testing came back high risk for downs 1:18. Went on to have cell free dna screening by Verifi it came back positive but my PPV for my age 31 is 64% we refused an amnio not worth the risk I’m 21 wks along currently and would not terminate. No markers for Ds on any ultrasounds only thing was a few measurements were small but still in the margin of error. I don’t know what to think my husband and I are being optimistic. Anyone have something like this happen and went on to have a “normal” delivery?
    Also when I got my paper work back for my blood work that showed 1:18 my weight was 20 heavier than I am so someone put it in wrong would that effect my 1:18 ratio?

    • I’m not sure if entering your weight incorrectly would affect the algorithm results, but it might’ve. Regardless, if you’re referring to the quad test or another blood based test that is not a cell free DNA screen, I’m not sure how any of those tests could report a PPV as high as 1:18 given your relative young age. And, regardless, again, the verifi result would take precedence over the quad test. If that PPV is correctly stated, then you have a 36% chance of the result being a false positive, or greater than 1-out-3 chance that the test result is wrong. So, based on those odds, out of 3 pregnancies with that PPV, one of them would be a false positive, and that may be you. Since you have decided to continue following a prenatal test result for Down syndrome, the resource Diagnosis to Delivery is available for purchase or free as a .pdf download. It was written by moms for expectant moms in your situation. I hope that is helpful.

  4. New worried mother for 2017 says:

    Im really upset and scared. Im 37 years old. I have been trying to conceive for 2 years, miscarried in May of 2016 and now tested positive through Materni21 for Trisomy21. I was told by my OBGYN, that the baby is surely to have Downs and that the amino test would only clear up the other 1% chance of the 99% chance of accuracy of the Materni21. How true is this?

    • This is not true and your OB is wrong to the point of malpractice (as addressed at this post). Click through the links in the post and you’ll find fact sheets to share with your OB about what your “positive predictive value” or “PPV” is. Based on your age and the MaterniT21 result your actual chance of having a child with Down syndrome is 86%, not 99%, and your chance your result is a false positive is 14%, not 1%. Only an amnio can provide a diagnosis. If you did not receive information about a life with Down syndrome, as recommended by professional guidelines (and as I presume your OB is unaware of), you can find out more at this link to the recommended resource for expectant moms receiving a prenatal test result.

      • New worried mother for 2017 says says:

        Then I guess I will go for the invasive test to determine the results. Before I receive anymore inaccurate information, how true are the amino tests then? And is the finding the materni21 blood test positive because of my age or were there actual findings? There is no history on either side of our family or any birth defects so this is really hard to understand. And now that I have one positive test, and if it remains true, will all my pregnancies be like this? Im so upset that im sick and unable to keep food down due to the high levels of emotions Im going through since finding this out yesterday. I really do appreciate your quick response as I am getting nowhere with my OB and that they are pretty much telling me it is what it is and that they are referring me to genetic counseling.

        • An amnio is considered medically diagnostic. While there are still false test results, those are extremely rare, so much so that an amnio result is considered diagnostic, but only the full amnio result. You may be offered a FISH test which provides a preliminary result of the amnio; FISH is considered screening and the full amnio results are diagnostic. There is no known cause for Down syndrome, so the absence of any family history is more the norm for pregnancies positive for Down syndrome. The MaterniT21 screen tests cell free DNA from the pregnancy, but that DNA can come from the placenta and why there are false positives and false negatives. If your amnio is positive for Down syndrome, then your genetic counselor can tell you your revised chances for having future pregnancies positive for Down syndrome, but a general rule of thumb is that your baseline chance becomes 1%, i.e. that you have a 1-in-100 chance for Down syndrome with future pregnancies. Those odds change if the form of Down syndrome is what is called “translocation,” but that rarely occurs. All of this should be discussed with your genetic counselor and hopefully they will provide better, more accurate information than what you have received so far.

  5. Being tested tomorrow through CVS and ultrasound due to having a MaterniT21 test come back positive for Trisomy 21. Are the charts listed above for people to see their chances in down syndrome without a blood screen before, or is it based off those who have had the blood screen, and still recieve false positives? I spoke to my OBGYN nurse, and she keeps apologizing to me for what Im facing and keeps telling me its not my fault or that I did anything wrong, and that im not alone. Shes talking to me as if the screening test is a true positive. So I guess my question is this: obviously SOMETHING was found within the blood test to show DS for my baby. No family history of DS in my family, etc. So im really in shock and very upset. Im not sure how to understand my chances of anything being normal at this point. So instead of my knowing my chances of having a child with DS, what are my actually chances that this test will be a false positive and are false positives common with this type of testing?

    • Your chances depend on your age, which you can plug into the calculator linked at this post. You may also want to review the linked fact sheets about tests like MaterniT21 and Down syndrome and share those with your OB since it sounds like she or her staff could benefit from reviewing them as well.

      • So basically the positive test resulted mostly from my age and not just their findings in my blood that they took out? I called Sequenom myself and asked questions just to see what they would say and the gentleman told me that the test does indeed tests the DNA from the baby and by me receiving a positive test, it means that my baby does have DS. That the test is used in place of the amino and cvs due to it not being risky. Im not sure how they are able to sell this product when they give false information to people that are trying to find out about their unborn child. He reassured me that they have been in business for years and that the tests are extremely accurate and its rare for them to be wrong. This is by far the most horrific hair pulling experience Ive ever had. Im so unbelievably scared…….

        • Mothers have a baseline chance depending on their age for having a child with Down syndrome, with that chance increasing with the mother’s age. The MaterniT21 test and tests like it (as explained in the fact sheets) test cell free DNA from the pregnancy. However, that DNA could be from cells that are from the placenta or cells that became the fetus. Because it is not always DNA from the fetus, tests like MaterniT21 have false positives and false negatives. A mother’s chance, or positive predictive value, is then a calculation based off her baseline chance from her age as recalculated using the MaterniT21 result. If the representative from Sequenom said that a positive MaterniT21 test means your baby has Down syndrome, then that is misinformation which you could report to your state attorney general’s office of consumer affairs.

          • Thank you so much! After my appointment today and what I find out, I will be contacting my attorney if things are not adding up. I appreciate you filling in the blanks for me, as Im sure these types of things worry so many expecting parents.

  6. does this information mean anything? I had my CVS test yesterday and just awaiting the results but this is what they were sent as far as blood work went. Im not sure if it means something, but they didn’t go over it with me.
    Abnormal Maternal Serum Screening test ICD-9-cm:796.5
    ICD-10-cm: O28.0
    Advanced Maternal Age in multigravida, 1st trimester ICD-9-cm: 659.63
    ICD-10-CM: O09.521

    in the other corner is said MAT21 positive for Trisomy 21.

    Also Id like to add that when I went in and did the ultrasound, afterwards the doctor came in asking me if my OB seen anything on the ultrasound. It made me feel as if she didn’t see anything and was trying to compare knowledge. I asked about the fluid behind the baby, and she said she could see some, but that even normal babies have a little. I did see a nasal bone in the ultrasound pics also. My baby also measured right to the exact day! Not a week smaller, or a week bigger. Is that a good indication? Something just seems off…….im praying my blood tests were wrong. Im not sure these are even things you can speculate on, but something just feels like MAYBE my baby might be okay….. or Im simply brain washing myself.

    • I am not aware of the numbers you share meaning anything regarding the likelihood of a diagnosis or a false positive; something you may wish to discuss with your provider. Regarding ultrasound findings, including fluid on the neck, those are considered “soft markers” and a cfDNA result trumps soft markers.

  7. reading some of this gives me hope. I am 29 years old…I have a 5yr old son and a 3 year old son, had a blighted ovum in may 2016 and am currently 12 weeks pregnant. My OB called me yesterday and told me that my baby had an extra chromosome 21 which would mean positive for Down syndrome. I see the genetic counselor next week to talk and have the ultrasound. I’ve read so many stories of false positives and am trying so hard to remain optimistic, but that was definitely not what I was expecting, especially over the phone. I was also aware of the fact that its not suppose to be a “positive or negative” test but when I asked my OB about the chances and percentages she just said I was positive. Any insight you can provide to help ease my mind while I wait?

    • See this post here and the links therein to fact sheets and the tab at the bottom for resources. The calculator linked therein would calculate your chances of a false positive being 42%.

  8. What about sonogram soft markers and risk with normal NIPT results? I’m 30. My 20 week sono showed a choroid plexus cyst. A week later, a new sonogram showed “very mild” hydronephrosis measuring between 4-5mm on one side that appeared during that sonogram. 2 weeks later, the hydronephrosis was gone and the cyst was still there, unchanged. I can’t seem to find any info on negative NIPT test results and risk when it comes to sono markers.

    • For the conditions that NIPT tests for, a screen negative is even more reliable and therefore trumps any other screening result, including UT soft markers. The markers seen on ultrasound may deserve monitoring for other reasons, but they should not be considered indicative of a condition tested-for by NIPT if you have a screen-negative result. Of course, diagnostic testing remains the only way to know for sure.

      • I’m 42, 20 weeks and had a positive NIPT. My ultrasound showed normal anatomy ( although it wasn’t discussed in any detail). Heart showed possible VSD and kidney showed slight fluid. Both maternal fetal specialist and OB feel that the baby definitely has DS given positive and the heart. I’m reading that VSD can be common regardless of age. I’ve been asked to highly consider the amnio, but it seems that it would not be necessary since they said he definitely has DS. Anyone ever deal with this? Where do I go for info?

        • According to this online calculator, based on your age, the likelihood that your NIPT is a false positive is 4% and the chances that it is a true positive are 96%. Add to the NIPT result a commonly occurring condition with Down syndrome (VSD) and it is highly likely that the NIPT is correct. However, there remains a chance that the VSD is separately occurring and the NIPT is a false positive. The only way to know for sure is with an amnio or wait until delivery.

Trackbacks

  1. […] incorrectly advising their patients that NIPS is “99+%” accurate–no NIPS test provides that much certainty. Women, though, are relying on this misinformation and terminating their […]

  2. […] I’d link the commenter to this post for the simple chart in the middle that shows the PPV associated with the age of the expectant […]

Leave a Reply

%d bloggers like this: