Participation in maternal serum screening for Down syndrome, neural tube defects, and trisomy 18 following screen-positive results in a previous pregnancy

Abstract

Objective To determine whether women who have had a positive serum screening result for Down syndrome or neural tube defect in 1 pregnancy have a lower rate of participation in screening in their next pregnancy. Setting A triple-marker screening program at a university hospital. Methods Pregnancy and screening information was collected from laboratory and hospital databases to compare subsequent screening participation of women who were screen-negative and screen-positive for the risk of a fetus with Down syndrome or a neural tube defect. Results In an age-matched comparison, 108 women who had a previous screen-positive result were significantly less likely than 108 women who were screen-negative to participate in maternal serum screening in their next pregnancy. When examined according to the type of screen-positive result, the effect was significant for both those who were screen-positive for Down syndrome and those who were screen-positive for neural tube defect. The degree of risk in screen-positive women did not significantly affect their participation in screening in the next pregnancy. Conclusions Anxiety related to a screen-positive result probably causes decreased participation in maternal serum screening in the next pregnancy. Reducing the screen-positive rate in prenatal serum screening would alleviate maternal anxiety and would probably lead to more stable participation.

Maternal serum screening, performed at 15 to 20 weeks of pregnancy, is commonly used in the prenatal detection of open neural tube defects, Down syndrome, and trisomy 18. By its nature as a screening test, maternal serum screening cannot determine with certainty if the baby is affected with a fetal anomaly. Most women who receive screening results indicating a high risk (screen-positive) have unaffected babies (false-positive), and some receiving results indicating a low risk (screen-negative) have affected babies (false-negative). Risk cutoffs are chosen to optimize the detection rate while maintaining a relatively low (5%-10%) false-positive rate.

Farrant first reported that false-positive results were associated with higher levels of maternal anxiety.¹ All subsequent studies of women who received false-positive results, indicating an increased risk of Down syndrome or neural tube defects, have confirmed this observation.^{2,3,4,5,6,7,8,9} Heightened anxiety was found to persist in screen-positive patients even after their initial positive results were shown to be false.^{2, 10} It is not surprising, then, that women with false-positive results, compared with those with true-negative results, were more opposed to maternal serum screening^{11, 12} and were less likely to say that they would choose maternal serum screening in a future pregnancy.^{11, 13} However, whether this negative attitude translates into reduced screening participation in a future pregnancy has not yet been reported.

The purpose of this study was to determine whether women who have had a positive serum screening result in 1 pregnancy have a lower rate of participation in screening in their next pregnancy than women who were screen-negative in their previous pregnancy. We also examined whether it made a difference if women were screen-positive for Down syndrome or neural tube defect or whether the degree of increased risk made a difference in subsequent screening participation. Finally, we asked if screen-positive women who declined serum screening in their subsequent pregnancy chose other forms of prenatal testing instead, such as amniocentesis or ultrasound examination.

PATIENTS AND METHODS

All women who underwent maternal serum triple-marker screening at the Women and Infants Hospital, Brown University School of Medicine, Providence, RI, in 1994 and who had a subsequent pregnancy at Women and Infants Hospital before May 1998 were eligible for inclusion in this study. Women who were screen-positive for trisomy 18, for both Down syndrome and neural tube defect, who had a previous pregnancy with fetal neural tube defect or Down syndrome, and those who were carrying twins during their 1994 pregnancy were excluded.

Our study population was identified using the 1994 comprehensive laboratory screening log book. Of the 9,250 screenings performed in that year, 14% of women had initial screen-positive results—9% for Down syndrome, 4.5% for neural tube defect, and 0.5% for trisomy 18. The screening test result was considered positive if the second-trimester risk for Down syndrome was 1 or more in 270 pregnancies.¹⁴ A screen-positive result for the risk of neural tube defect was defined as a maternal serum α-fetoprotein (AFP) level 2 or more multiples of the median (MoM).¹⁵ A screen-positive result for risk of trisomy 18 was defined as a maternal serum AFP level of 0.75 or less (MoM), a human chorionic gonadotropin level of 0.55 or less (MoM), and an unconjugated estriol level of 0.60 or less (MoM).¹⁶ Of the 1,295 women with positive screening results in 1994, 108 were identified as being screen-positive for either Down syndrome or neural tube defect and as having a second pregnancy before May 1998.

A screen-negative group was identified in a similar manner. Women who had negative results on screening tests performed in 1994 were reviewed in the hospital database. Women who were screen-negative in 1994 and who had a subsequent pregnancy between 1994 and May 1998 were selected as a control group, with each control matched with 1 screen-positive case for maternal age at second delivery (±1 year).

Serum screening information on the study population was obtained using a laboratory reporting system (Alpha; Logical Medical Systems, London). For women receiving screen-positive results, medical records and the laboratory database were examined to determine whether maternal serum screening or alternative prenatal testing (chorionic villus sampling, amniocentesis, or targeted ultrasonography performed between 16 and 24 weeks) was chosen in the subsequent pregnancy.

DATA ANALYSIS

To assess differences in screening participation in the subsequent pregnancy according to previous screening results and type of screen-positive result, χ² analysis was performed (StatView; Abacus Concepts, Inc, Berkeley, CA). The odds ratios and 95% confidence intervals of those having screening in a subsequent pregnancy for women having a positive screening test result in the previous pregnancy, compared with those having a negative screening test result, were also calculated. χ² Tests for trend with 1 df were used to compare the participation in maternal serum screening in the subsequent pregnancy according to the degree of risk increase for the previous screen-positive pregnancy. A significant difference was accepted at P<0.05.

RESULTS

The participation of women in screening in their subsequent pregnancy according to their previous screening results is shown in table 1. In the 108 previously screen-negative and -positive women matched by maternal age, a significantly smaller percentage of those who were screen-positive (57% vs 79%, χ² = 11.27; P = 0.001) had screening in their subsequent pregnancy. Overall, the effect of a screen-positive result was to reduce the odds of a woman having serum screening in a subsequent pregnancy by 64%.

Table 1.

Participation in subsequent screening according to the result of the screening test in the previous pregnancy

Result of screen in previous pregnancy	Total No.	No. (%) having screening in subsequent pregnancy	OR (95% CI)
Screen-negative	108	85 (79)	0.36 (0.20-0.66)
All screen-positive	108	62 (57)*
DS screen-positive only	60	31 (52)†	0.42 (0.20-0.90)
NTD screen-positive only	48	31 (65)‡	0.26 (0.09-0.74)
OR = odds ratio; CI = confidence interval; DS = Down syndrome; NTD = neural tube defect.

^*P = 0.001 relative to matched screen-negative.

^†P = 0.02 relative to matched screen-negative.

^‡P = 0.009 relative to matched screen-negative.

Of the 108 previously screen-positive women, 60 were screen-positive for risk of a child with Down syndrome, and 48 were screen-positive for risk of a child with neural tube defect. As shown in table 1, women who were screen-positive for risk of Down syndrome participated in subsequent maternal serum screening significantly less often than women who were screen-negative (52% vs 72%, χ² = 5.08; P = 0.02). Women who were screen-positive for risk of neural tube defect also participated in subsequent screening less often than their matched screen-negative controls (65% vs 88%, χ² = 6.92; P = 0.009). The effect of a Down syndrome or a neural tube defect screen-positive result was to reduce the odds that a women would have screening in a subsequent pregnancy by 58% and 74%, respectively. This reduced participation was not dependent on the degree of increased risk for either Down syndrome (P = 0.50) or neural tube defect (P = 0.79; table 2).

Table 2.

Participation in subsequent screening given a screen-positive result for either Down syndrome or neural tube defect in the previous pregnancy, stratified by degree of increased risk of previous result

Down syndrome			Neural tube defect
Degree of increased risk in previous pregnancy	Total No.	No. (%) having screening	Degree of AFP elevation in previous pregnancy, MoM	Total No.	No. (%) having screening
1/201 to 1/270	17	10 (59)	2.00-2.49	31	20 (65)
1/101 to 1/200	24	12 (50)	2.50-3.49	13	9 (69)
1/100 to >1/2	19	9 (47)	≥3.50	4	2 (50)
AFP = α-fetoprotein; MoM = multiple of the means.

Forty-six of the previously screen-positive women declined serum screening in their subsequent pregnancy. Of the 41 of these women for whom follow-up information could be obtained for their second pregnancy, 18 (44%) had no prenatal testing in the second trimester, 14 (34%) had a second-trimester ultrasound examination, 8 (20%) had amniocentesis, and 1 had chorionic villus sampling. All of the 9 women having amniocentesis and chorionic villus sampling were aged 35 or older.

DISCUSSION

The present study shows that women receiving screen-positive test results in 1 pregnancy are less likely than women receiving negative results to participate in screening in their subsequent pregnancy and that the degree of risk of a screen-positive woman does not appear to influence her decision about screening in her next pregnancy. These findings are not surprising, given that women receiving positive screening results experience anxiety^{2,3,4,5,6,7,8,9} and exhibit negative feelings toward serum screening.^{11, 12} Both those who were screen-positive for having a child with Down syndrome and those who were screen-positive for having a child with neural tube defects were less likely than screen-negative women to have screening in the future.

Marteau has shown that anxiety associated with screen-positive results would be reduced if before they are tested women knew more about the purpose of screening, the possibility of receiving positive results, and the sequence of decisions and procedures that may follow.¹⁷ Other studies have shown that some women who had screening would not have chosen to undergo testing if they had been given more information before being tested,^{18, 19} and conversely, some women who were not screened would have chosen to participate.²⁰ In practice, however, comprehensive and even accurate information about prenatal testing is not always provided by obstetricians and midwives,²¹ perhaps because they do not fully understand maternal serum screening.²² In addition, evidence shows that women are poorly informed about the screening tests²³ and the implications of test results.²⁴ Improved patient education may increase the likelihood that previously screen-positive women would participate in screening in future pregnancies. We can be assured, however, that many women who receive positive results and decline screening in their subsequent pregnancy are accepting other prenatal tests in later pregnancies.

Although improved patient education on prenatal screening and better understanding of the consequences of a screen-positive result are valuable, the fact remains that current screening for Down syndrome and neural tube defects is associated with many false-positive results. In triple-marker screening, for example, about 5% of all screened women are offered amniocentesis because of their risk of a Down syndrome pregnancy, and another 2% are offered either targeted ultrasonography or amniocentesis because of their risk of a neural tube defect. In women who are screen-positive for Down syndrome, by the triple-marker test, the odds of having an affected fetus are of the order of 1 in 60, meaning that for every true-positive result, there will be 60 false-positive results. As the years go by, the number of women with false-positive screening results will grow substantially, and based on our results, participation in screening will decline.

Now that detection rates for Down syndrome are approaching and may surpass 80%, through the use of second-trimester 4-marker testing and first-trimester combined serum specimen-ultrasonography testing, efforts should focus on reducing the false-positive rate. This would decrease the number of women experiencing anxiety because of screen-positive results and would improve the safety of the diagnostic process by decreasing the number of women having to be offered amniocentesis. As an example of this approach, a recent study by Wald and colleagues in which markers measured in the first and second trimesters were integrated into 1 risk assessment showed that a detection rate of 85% can be attained at a false-positive rate of less than 1%.²⁵ With this approach, not only would the number of amniocentesis procedures be reduced by more than four fifths, but the odds of having an affected fetus if a woman is screen-positive would be increased to about 1 in 7. Having fewer screen-positive results may translate into more stable participation in the prenatal screening process.