Prenatal screening for trisomy 21 (Down syndrome), trisomy 18 (Edwards syndrome), and trisomy 13 (Patau syndrome) using plasma (cell-free) DNA analysis detects nearly all affected pregnancies (pregnancies with trisomy 21, 18 or 13) with a much lower positive rate (proportion of unaffected pregnancies with a positive screening result) compared with conventional screening methods based on the measurement of ultrasound and serum markers.[1,2] DNA testing has however not generally been implemented as a method of routine screening, because it is more complex and costly than conventional screening methods, and has a failure rate due to technical reasons or biological reasons, for example when the fetal fraction (percentage of cell-free DNA from the placenta) is low.[3] Two screening methods to overcome this have been proposed.

One method, “recall DNA screening” recommended by the UK National Screening Committee [4], involves women having a first trimester Combined test (based on the measurement of nuchal translucency, pregnancy associated plasma protein A, free beta human chorionic gonadotrophin, and maternal age). The 2–3% of women with positive Combined test results (risk of an affected pregnancy ≥1 in 150) are recalled for counselling and offered an invasive diagnostic test (amniocentesis or chorionic villus sampling) or having another blood sample collected for a DNA test. Women with a positive DNA test are then offered a diagnostic test. The recall method has a detection rate (proportion of affected pregnancies with a positive screening result) similar to that of conventional screening. [5,6] In one study [5] about 18% of women with a positive Combined test were sufficiently worried by their Combined test risk to choose an invasive diagnostic test without having a DNA test first. The recall DNA screening method is illustrated in Fig 1A.

Table 3 shows a comparison of the screening efficacy and costs between recall and reflex DNA screening according to Combined test markers risk cut-off levels and the cost of a DNA analysis. (S1 Table shows screening performance and costs for a greater range of Combined test risk markers cut-offs and DNA analysis costs). Reflex and recall DNA screening leads to the diagnosis of a similar number of affected pregnancies but with reflex DNA screening far fewer women with unaffected pregnancies are given a positive result and far fewer invasive diagnostic tests are performed in unaffected pregnancies. For example, in 100,000 women screened, using a Combined test markers risk cut-off of 1 in 150 more than 100-fold fewer women with unaffected pregnancies are given a positive test result (2389 v 19) and about 20-fold fewer women have an invasive diagnostic test (17 v 332). The cost per woman screened and the cost per affected pregnancy diagnosed are lower with reflex DNA screening than with recall DNA screening regardless of the Combined test markers risk cut-off and DNA analysis cost shown in Table 3 and S1 Table. For example, with a DNA analysis cost of £250 reflex DNA screening costs about £1 less per woman screened than recall DNA screening for all the specified Combined test markers risk cut-offs.

Our results show that using published UK unit cost estimates, reflex DNA screening is less expensive and more cost-effective than the recall method and has over 100-fold lower false-positive results (see Fig 2). As the cost of the DNA analysis continues to decline, as is likely, a greater proportion of women can be reflexed to a DNA analysis with the associated increase in the detection rate while maintaining a low number of false-positive results and cost effectiveness (see Table 3), an advantage not achievable with the recall method. Costs vary from place to place and over time. For this reason, we have produced an online screening cost and efficacy calculator (www.screening-cost-calculator.com) that allows local unit costs to be entered (in £, $ or €) and local versions of Table 3 are generated (S2 Fig) and local programme costs determined for the two screening methods.

Our cost estimate of screening using the recall method with a 1 in 150 Combined test risk cut-off (£3.8 million per 100,000 women screened) is similar to the estimate in the study we used to obtain the component unit costs [5], taking into account that that estimate was based on women offered screening with 66% accepting (£3.7 million per 100,000 women screened)

The main contribution to the total cost for both the recall and reflex methods is from measuring the Combined test markers and calculating the risk of an affected pregnancy based on them, which all women receive. Although the DNA analysis is substantially more expensive, only a small proportion have a DNA analysis so it contributes less to the total cost. There may be additional costs for the recall method should women with a high Combined test risk but a negative DNA test result be sufficiently anxious about the Combined test risk to request and amniocentesis. However, this would only apply to a small number of women, and Chitty et al did not report any such women. [5].

Costs are just one element in determining policy. It cannot be ignored that the number of false-positives with the recall method is over 100 times greater than with the reflex method; a considerable burden of anxiety that the reflex method avoids (Table 3). The anxiety caused is well illustrated in an implementation study of the recall method that observed that 18% of the 2–3% of women recalled due to a high Combined test risk estimate requested an invasive diagnostic test without having a DNA analysis first [5], even though many of these pregnancies would be unaffected, the proportion being higher in affected pregnancies than unaffected pregnancies because they have, on average, higher Combined test risk estimates and are therefore more anxious (see Table 2). These unnecessary diagnostic procedures would be avoided with the reflex method, making it the safer screening method. The reflex method also avoids the time, inconvenience and costs to the women returning to the hospital or clinic for counselling that arise with the recall method.

Our estimates of costs were based on the use of EDTA blood collection to provide plasma for a DNA analysis from all women as is current practice in the Wolfson Institute antenatal screening service. The use of EDTA tubes is acceptable as there is evidence that the separation of plasma from cells up to a least 48 hours after blood collection does not significantly degrade the sample needed for DNA analysis [10,11] and unpublished data from the Wolfson Institute screening service indicate that this is the case for up to 60 hours. The use of DNA stability blood collection tubes would add significantly to the cost of reflex DNA screening. We have included a cost input item for this as an option to the screening cost and efficacy calculator. S2 Table, provides in a similar way to Table 3, cost estimates using DNA stability tubes. The added cost of DNA stability blood collection tubes when EDTA tubes can be used emphasises the importance of ensuring that blood samples are delivered to the DNA laboratory within 60 hours of collection.

In summary a reflex DNA screening programme is no more expensive than the recall method, is more effective, safe and more cost-effective than a programme based on the recall method.