- Top of page
Objective The objective of this study was to follow up and evaluate the statewide first-trimester combined screening programme for Down syndrome and trisomy 18 at Genetic Health Services Victoria, Australia.
Design Retrospective population cohort.
Setting Maternal Serum Screening Laboratory records.
Sample All women screened between February 2000 and June 2002 (16 153 pregnancies).
Methods Screening results were matched to Victorian perinatal and birth defect data via record linkage, with an ascertainment of 96.8% of pregnancy outcomes. Manual follow up with health professionals increased ascertainment to more than 99%.
Main outcome measures Fetal Down syndrome or trisomy 18, and combined screen results, to calculate test characteristics.
Results Using a risk threshold of 1 in 300 at time of ultrasound, the sensitivities for standard first-trimester combined screening and augmented 13-week combined screening for Down syndrome were 87.3 and 90.5% and the false-positive rates (FPR) were 4.1 and 3.9%, respectively. The sensitivity for trisomy 18 was 66.7% (10/15, 95% CI 42.8–90.5%) with a 0.4% FPR and 15.2% positive predictive value (1 in 250 risk threshold).
Conclusions The combined use of record linkage and manual follow-up techniques was effective in ascertaining more than 99% of pregnancy outcomes for calculations of accurate test characteristics of the combined screen. The sensitivity for Down syndrome at Genetic Health is comparable to similar populations. However, the sensitivity for trisomy 18 is lower than that elsewhere, which may reflect the overall low birth prevalence of trisomy 18 and associated small numbers in this particular cohort.
- Top of page
Prenatal screening for Down syndrome and other chromosomal abnormalities has become increasingly available in Victoria, Australia, where approximately 63 000 women give birth each year.1 The Maternal Serum Screening Laboratory at Genetic Health Services Victoria introduced a combined first-trimester screening pilot programme in February 2000, and this became a full service in July 2002. This test combines the nuchal translucency (NT) ultrasound measurement with first-trimester maternal serum, together with maternal age, weight and gestation to produce a risk estimate of the fetus having Down syndrome or trisomy 18. There are various reported sensitivities of the combined screen for Down syndrome (73–93%)2–13 and trisomy 18 (91–100%).9,14,15 Variations between test characteristics can be a reflection of different maternal age distributions, methods of calculating the risk estimates and risk thresholds.
Having a newly established service in Victoria, the aim of this study was to follow up and evaluate data from the pilot programme to determine the test characteristics for the combined screen for the detection of Down syndrome and trisomy 18. This would allow for comparisons with well-established services elsewhere and provide local information to those receiving testing and to those planning future services.
- Top of page
Of the 16 153 screening records, pregnancy outcome was determined for 16 002 (99.1%). Record linkage found the pregnancy outcome for 15 630 of the screened pregnancies (96.8%) and a further 372 (2.3%) were ascertained through manual follow up with health professionals, leaving 151 lost to follow up. Live births from manual follow up included 198 interstate, 33 Victorian (including homebirths), 14 overseas and 14 in unknown locations. Manual follow up also ascertained 17 cases of Down syndrome (all terminations), 3 cases of trisomy 18 (two terminations and one unknown pregnancy outcome), 68 miscarriages (<20 weeks of gestation), 41 terminations of pregnancy (36 for birth defects), 2 neonatal deaths and 1 stillbirth (≥20 weeks of gestation). The trisomy 18 case with an unknown pregnancy outcome is reported in the ‘lost to follow up’ group but included in the calculations of test characteristics, as the karyotype was known. Eighty-one of the women lost to follow up had moved overseas (47) or interstate (34). In some cases, doctors did not respond to the contact letter (30), the woman was unknown to them (23) or the doctor was unable to be contacted (11). Participation was declined for five cases.
Women lost to follow up were not more likely to be at increased risk for Down syndrome (RR = 1.5, 95% CI 0.8–2.8, P= 0.2) or to be aged 37 years or older (RR = 1.2, 95% CI 0.9–1.6, P= 0.3) when compared with women who were able to be followed up. Women living in rural Victoria were more likely to be lost to follow up compared with those living in metropolitan areas (RR = 2.1, 95% CI 1.6–2.7, P < 0.01).
Maternal age ranged from 16 to 51 years, with a mean and median age of 33 years, which is higher than the mean age of 30 years in the general Victorian population of women who had babies in 2001–02.1 As shown in Figure 1, the maternal age distribution of screened women was statistically different (χ2[6df] = 4696, P < 0.001) than that of the general population.1
Figure 1. Age distribution of screened women and the general victorian population of women who had a baby in 2001–02.
Download figure to PowerPoint
Screening for down syndrome
Six hundred and ninety-two (4.3%) women had an increased risk for Down syndrome. Thirteen percent (376/2985) of women aged 37 years and older had an increased risk result compared with 2% (316/13 168) of younger women (RR = 5.3, 95% CI 4.5–6.1, P < 0.001). There were 63 cases of Down syndrome ascertained in this population, which was comparable to the 66 expected cases calculated from our maternal age distribution.16
The sensitivity of the standard combined screen for the detection of Down syndrome (i.e. all pregnancies screened using standard first-trimester markers, between 10 and 13 weeks of gestation) was 87.3% (95% CI 79.1–95.5%), with a FPR of 4.1%. For the screening test augmented with second-trimester markers for those screened at 13 weeks (and standard markers for those screened <13 weeks), the sensitivity was 90.5% (95% CI 83.2–97.7%), with a 3.9% FPR (Table 1). There were three cases of Down syndrome that were screened at 13 weeks, all of which were at increased risk using the 13-week screening method, with only one being at increased risk from the standard combined screen.
Table 1. Results of prenatal screening for Down syndrome and trisomy 18 with combined first-trimester screening, including the sensitivity (DR), FPR and PPV
|Screened condition||Number of screened pregnancies||Number of increased risks||Test characteristics (%)|
|Risk ≥1 in 300**||16 003||63||15 940||683||57||626||90.5||3.9||8.4|
|Risk ≥1 in 300***||16 003||63||15 940||714||55||659||87.3||4.1||7.7|
|Bloods <13 weeks of gestation***||15 243||60||15 183||605||55||550||91.7||3.6||9.1|
|Fixed 5% FPR, risk ≥1 in 351**||16 003||63||15 940||799||58||741||92.1||5.0||7.3|
|Women aged <37 years**||13 050||35||13 015||314||30||284||85.7||2.2||9.6|
|Women aged ≥37 years**||2953||28||2925||369||27||342||96.4||11.7||7.3|
|Risk ≥1 in 250**||16 003||15||15 988||66||10||56||66.7||0.4||15.2|
|Risk ≥1 in 25**||16 003||15||15 988||29||10||19||66.7||0.1||34.5|
|Women aged <37 years**||13 050||6||13 044||40||4||36||66.7||0.3||10.0|
|Women aged ≥37 years**||2953||9||2944||26||6||20||66.7||0.7||23.1|
|Bloods <13 weeks of gestation***||15 243||15||15 228||66||10||56||66.7||0.4||15.2|
Women younger than 37 years who had augmented screening had lower sensitivity and FPR (85.7% [95% CI 74.1–97.3%] and 2.2%, respectively) than women aged 37 years and older (96.4% [95% CI 89.6–100%] and 11.7%, respectively).
If a 5% fixed FPR had been used in the augmented screen (risk threshold = 1 in 351), 117 extra pregnancies, including one with Down syndrome, would have been at increased risk, changing the sensitivity to 92.1% (95% CI 85.4–98.7%). When women screened at 13 weeks (n= 760) were excluded from the 5% fixed FPR analysis, the sensitivity and FPR remained similar to the overall sensitivity of 91.7% (95% CI 84.7–98.7%) and 3.6%, respectively.
Screening for trisomy 18
Sixty-six women (0.4%) were screen positive for trisomy 18, with a higher proportion of women aged ≥37 years at increased risk (0.9%) than younger women (0.3%) (RR = 2.2, 95% CI 1.6–2.9, P < 0.001). Of those at increased risk, 42 (64%) were also at increased risk for Down syndrome. The overall sensitivity for trisomy 18 was 66.7% (95% CI 42.8–90.5), and this did not vary between the two age groups (Table 1). There were no cases of trisomy 18 screened during the 13th week. Detailed information on the screening results and analyte measurements for the five low-risk cases of trisomy 18 are shown in Table 2.
Table 2. Detailed information of fetuses with T18 that had low-risk combined screening results for T18 and DS and MoMs for NT, PAPP-A and beta hCG
|Maternal age (years)||Gestation at time of test (weeks + days)||Pregnancy outcome||Gestation of pregnancy outcome (weeks)||Combined screen T18 risk (1 in X)||Combined screen DS risk (1 in X)||NT (MoMs)||PAPP-A (MoMs)||Beta-hCG (MoMs)|
|36||10 + 6||Stillbirth||20||349||2980||0.71||0.33||0.19|
|37||11 + 1||Unknown||Unknown||1030||3640||1.07||0.59||0.23|
|37||9 + 2||Live birth||39||10 600||1840||0.97||0.49||0.45|
|38||11 + 2||Neonatal death||33||13 000||5500||0.87||1.05||0.31|
|34||10 + 1||Stillbirth||23||45 300||16 200||0.84||0.91||0.71|
The FPR for trisomy 18 was lower in younger (0.3%) than older women (0.7%), as was the PPV (10 and 23%, respectively). Nine of the 15 (60%) fetuses with trisomy 18 were categorised as at increased risk for both Down syndrome and trisomy 18. Only 40% (6/15) of the fetuses with trisomy 18 were at increased risk from NT alone (in combination with CRL and maternal age). The lowest true positive risk was 1 in 24, and therefore, setting the risk threshold at ≥1 in 25 would have produced the same sensitivity but with a lower FPR (0.1%) and higher PPV (35.5% or 1 in 3). However, of the 37 women who would have changed from an increased to a low risk for trisomy 18 (i.e. those with a risk between 1 in 26 and 1 in 250 for trisomy 18), 15 (41%) were also at increased risk for Down syndrome.