Noninvasive prenatal testing (NIPT) using cell-free (cf) DNA in maternal blood has the potential to dramatically alter the way prenatal screening and diagnosis for Down’s syndrome is delivered. In many countries, before NIPT can be implemented into routine practice, information is required on its costs and benefits. There are differences in cost-effectiveness between cfDNA-based NIPT used in a contingent testing strategy and cfDNA-based NIPT used as a primary test. The latter strategy is more effective than current screening practice, but also more costly, with high incremental cost-effectiveness ratios. Cell-free DNA-based NIPT is associated with lower procedure-related miscarriages compared with usual screening. Several factors need to be considered when evaluating the cost-effectiveness of cfDNA NIPT, including the choice of outcome measure, impact on uptake, study perspective, and transferability of findings.
KeywordsNoninvasive prenatal testing, Cell-free DNA, Cost, Cost-effectiveness, Health economics, Economic appraisal, Prenatal screening, Prenatal diagnosis, Down’s syndrome
Systematic reviews have demonstrated that noninvasive prenatal testing (NIPT) based on sequencing of cell-free DNA in maternal plasma is a highly effective screening test for Down’s syndrome (trisomy 21) . It can also be used to screen for Patau syndrome (trisomy 13), Edwards’ syndrome (trisomy 18), and sex chromosome aneuploidies, though with less accuracy . For example, a recent meta-analysis showed that the pooled detection rate (DR) of cfDNA testing for Down’s syndrome was 99.7% (95% confidence interval (CI) 99.1%–99.9%) and the false positive rate (FPR) was 0.04% (0.03%–0.07%); for Patau’s syndrome the pooled DR and FPR were 97.9% (94.9%–99.1%) and 0.04% (0.03%–0.07%), respectively; for Edwards’ syndrome they were 99.0% (65.8%–100.0%) and 0.04% (0.02%–0.07%), respectively .
In most developed countries, pregnant women are offered some form of prenatal testing, but the delivery of these services varies between countries. “Standard” screening comprises either first-trimester combined (serum parameters and ultrasound) screening, second-trimester serum screening or integrated screening (a two-stage test where risk is estimated after all first- and second-trimester screening tests). Risk is based on the computation of a series of factors such as maternal age, maternal serum markers, and measurement of nuchal translucency with ultrasound. Those who screen positive, with a predictive risk higher than a predetermined value, are offered a definitive invasive diagnostic test, typically amniocentesis or chorionic villus sampling (CVS) . Current evidence suggests that the procedure-related risks of miscarriage for amniocentesis and CVS are 0.11% (95% CI − 0.04% to 0.26%) and 0.22% (− 0.71% to 1.16%), respectively .
Cell-free DNA-based NIPT has been proposed as an addition to the current prenatal testing pathway to decrease the number of invasive diagnostic tests, and consequently reduce the number of procedure-related miscarriages .
There are two main ways in which cfDNA NIPT may be introduced into a prenatal screening pathway. The first is as a contingent strategy to lower the false positive rate of testing prior to invasive diagnostic procedures. In this case, all pregnant women are offered standard prenatal screening and those with a risk above a prespecified threshold are offered cfDNA NIPT rather than an invasive test. Invasive diagnostic testing is then offered to those with an abnormal cfDNA NIPT result. The second alternative is to use cfDNA NIPT as the principal screening test—a first-line or universal testing strategy. In this case cfDNA NIPT replaces standard prenatal screening and it is offered to all women instead of the combined testing or serum screening; invasive diagnostic testing is offered to those with abnormal cfDNA NIPT results. Note that in both testing strategies cfDNA NIPT is not a replacement for invasive procedures because it can produce false positive results, as discussed in Chapter 5 —a diagnosis can only be made by invasive diagnostic testing . However, while cfDNA NIPT does not completely remove the need for invasive testing, it can potentially reduce the number of women needing it, thereby also lowering the number of procedure-related miscarriages.
Cell-free DNA-based NIPT is now becoming increasingly available worldwide, but before it can be implemented into routine practice, information is required to identify whether and how it fits best in the prenatal testing pathway, based on the predicted costs and benefits. The benefits of NIPT when used as a contingent test or as first-line testing are expected to be different. With contingent testing, the main benefit is expected to be fewer invasive tests and procedure-related miscarriages. With universal testing, in addition to a reduction in false positives and number of invasive tests and procedure-related miscarriages, the number of cases detected is expected to rise compared with standard prenatal screening. The impact of the introduction of cfDNA NIPT on costs depends on several factors, such as the cost of the test and the range of costs included, which will depend on the perspective taken to measure costs.
The aims of this chapter are twofold. First, we consider current evidence on the cost-effectiveness of cell-free DNA-based NIPT for prenatal testing both as a contingent test and as a universal test. Second, we explore challenges in evaluating the cost-effectiveness of cell-free DNA-based NIPT, which need to be borne in mind when considering the extant evidence. In the following section we provide an overview of economic appraisal, introducing the concepts and methodological framework utilized in the rest of this chapter.
Overview of Economic Appraisal
Organizing and planning health care services to meet the needs of a population has to be considered in the context of the finite resources available to do this. Decisions are needed about whether or not to implement a health care program (e.g., should cfDNA NIPT for prenatal testing be made available?), in what ways should it be made available (e.g., as a contingent or universal test?), and to whom should it be made available (e.g., to all pregnant women or a subset?). Economic appraisal is a systematic way of analyzing such choices. The term “economic appraisal” describes a set of techniques that weigh up the costs of an action, such as providing cfDNA NIPT, against the benefits that the action may elicit. The term “economic evaluation” is also used. The distinction between the two terms is that appraisal is undertaken before the action is taken, to help in deciding whether or not and how the action is to be done, and evaluation is undertaken after the action, to judge its effects.
The principles of economic appraisal can be applied to all kinds of health programs, from interventions at the individual level (e.g., whether cfDNA NIPT should be offered or what type of surgery should be used to treat a specific type of cancer) to the population level (e.g., whether there ought to be a nationally organized prenatal testing program for Down’s syndrome or whether acute stroke services should be centralized).
The methods and principles of economic appraisal are well articulated and described . Several of these methods and principles are particularly relevant when undertaking an economic appraisal of cfDNA NIPT, including scarcity, opportunity, cost, incremental analysis, perspective, cost-effectiveness analysis, cost-effectiveness plane, and cost-effectiveness acceptability.
Scarcity exists because the claims on health care resources outstrip the resources that are available. This provides the rationale for economic appraisal, because without scarcity there would be no need to make difficult resource allocation decisions. For example, the decision about whether or not to implement a cfDNA NIPT testing strategy would not need to depend on the cost of that strategy, but could focus instead purely on the benefits to pregnant women and families.
The existence of scarcity leads directly to an important economics concept, namely, opportunity cost. If scarce resources are used to produce a particular health care program or service, those resources cannot be used to produce other goods or services. Opportunity costs reflect the benefits that are forgone by not producing those other services. The opportunity cost of using resources in a particular way is defined as the benefits that would have resulted from their best alternative use. When economists refer to costs, they usually mean opportunity costs. This is different from the financial costs—the costs of goods and services in terms of money—though very often financial costs are used to measure opportunity costs. As an example, from the perspective of a national committee responsible for all screening programs in a public health system, the opportunity costs of additional investment in cfDNA NIPT could be the potential health benefits of investing instead in new screening programs for cancer.
Economic appraisal involves a comparison between two or more alternatives in terms of their costs and consequences. An economic appraisal ought to be comparative, because a health care program can only be thought to be worth doing or not relative to some alternative—even if that alternative is a do-nothing strategy. An incremental analysis is carried out to express the results of an economic appraisal. This is an assessment of the extra costs and benefits of one option compared with another—the additional costs incurred by one health care program are compared with another, in relation to the additional benefits of that health program compared with the other. Note that if the comparator is a do-nothing strategy it is unlikely to have zero costs and consequences, and these should also be accounted for.
To give an illustration, Torgerson and Spencer showed the importance of incremental analysis by analyzing data on the cost-effectiveness of biochemical screening for Down’s syndrome as a replacement for screening by maternal age. An estimate had been published showing that introducing biochemical screening in a population of 20,000 pregnant women would cost £413,500 (including costs of screening, invasive testing, and terminations of pregnancy) and might prevent eleven affected births, giving a cost-effectiveness ratio of £37,591 per case identified. Torgerson and Spencer pointed out that for the same number of women existing antenatal screening procedures based on maternal age already cost £79,000 (including invasive testing and terminations of pregnancy) with four affected births potentially prevented. So, the true cost-effectiveness of biochemical screening when compared to preexisting services was (£413,500 − £79,000)/(11 − 4) = £334,500/7 = £47,786, a higher cost per extra affected case identified than for the estimate based on average cost alone.
The perspective of an economic appraisal is important because among other things it determines what constitutes a cost, that is, what costs are included in the appraisal. Perspectives commonly taken in economic appraisals are those of the health service (or individual components within it, e.g., the hospital), the purchaser or payer of health care, and society as a whole. A societal perspective, which is the broadest perspective, might include health services costs, costs borne by patients and families, and costs in other sectors, for example, costs arising due to production losses from time off work. Researchers should be explicit about the viewpoint they adopt so an assessment can be made as to whether the costs included in an economic appraisal are complete and appropriate. The choice of perspective is important for an economic appraisal of cfDNA NIPT, as it affects the costs included. For example, one perspective is that of the prenatal test provider, who might only be interested in the costs of providing standard prenatal screening, cfDNA NIPT, and invasive diagnostic testing. A broader health service perspective might also include health care costs of pregnancy and childbirth, and termination of pregnancy. A societal perspective additionally includes lifetime costs of caring for individuals born with Down’s syndrome, such as lost earnings from parents taking time off work.
Cost-effectiveness analysis is one form of economic appraisal; others are cost-benefit analysis, cost-utility analysis, cost-minimization analysis, and cost-consequences analysis. The main form of economic appraisal that is encountered in the context of cfDNA NIPT is cost-effectiveness analysis, so we focus on that here. As with other forms of economic appraisal, cost-effectiveness analysis seeks to evaluate the costs and consequences of different options. One alternative is preferred to another if it provides greater benefit at the same or lower cost, or has lower cost for the same or greater benefit. Once data on the costs and benefits of the options being compared has been calculated, the first decision rule for this kind of appraisal is to reject any intervention dominated by another alternative or combination of alternatives; a dominated alternative has a greater cost with fewer or the same benefits, or lower benefits with the same or greater costs. For example, Garfield et al. calculated that a contingent cfDNA NIPT test strategy incurred a cost of US$59,228,142 in 100,000 pregnant women to detect 170 cases of Down’s syndrome. This compared favorably with standard prenatal screening, which costed US$59,748,721 and detected 148 cases of Down’s syndrome. Therefore in this case cfDNA NIPT dominated standard prenatal screening, and the latter strategy was therefore rejected.
The choice between nondominated alternatives is more complex because it leaves open the question which of two alternatives is preferred if one provides greater benefit than the other, but at a higher cost. A cost-effectiveness ratio (CER), defined as costs divided by benefits, can be calculated to inform this. The CER most often used in health economics is called an incremental CER, or ICER, which measures the incremental cost of an activity relative to its best alternative divided by its incremental effect.
In cost-effectiveness analysis, costs are regarded as opportunity costs measured in monetary terms, and benefits are measured in units other than money. For example, in the case of cfDNA NIPT, benefits might be measured in terms of the number of cases of Down’s syndrome diagnosed or the number of procedure-related miscarriages. Costs and benefits are therefore measured in different units and are not comparable with each other; directly weighing the value of effects against the value of costs is not possible. For decision-making, the ICER is measured against a ceiling ratio or threshold; to be regarded as cost effective any intervention must have an ICER which is on, or below, this threshold. To illustrate, Morris et al. calculated the cost-effectiveness of cfDNA NIPT as a universal testing strategy compared with standard prenatal testing in a population of 10,000 pregnant women. The costs of the cfDNA NIPT universal strategy and standard screening were estimated to be UK£1,825,000 and UK£279,000, respectively. The two strategies were estimated to produce 16.49 and 13.24 diagnoses of Down’s syndrome, respectively. Neither option was dominant. The incremental cost and incremental effect of cfDNA NIPT as a universal testing strategy versus standard prenatal testing strategy were UK£1,546,000 and 3.25 cases, respectively, with an associated ICER of UK£475,692 per extra case detected. By using the aforementioned ceiling ratio it is possible to judge whether or not cfDNA NIPT is cost effective in this case. For example, if the ceiling ratio was UK£500,000 per extra case identified, that is, we are prepared to pay up to UK£500,000 to diagnose an additional case of Down’s syndrome, then cfDNA NIPT is good value for money because the ICER is below the threshold. Alternatively, if we are only prepared to pay UK£100,000 per extra case identified then cfDNA NIPT would not represent good value for money. A potential problem with cost-effectiveness analysis, which affects its usefulness for decision-making, is that the ceiling ratio may not be known (see later).
Incremental analyses in economic appraisal are often illustrated using the cost-effectiveness plane ( Fig. 1 ). ICERs are presented graphically as a combination of the costs and the effects of a health intervention, described in the figure as cfDNA NIPT compared to an alternative. Incremental costs are conventionally placed on the north-south axis and incremental effects on the east-west axis. The origin is the point where costs and benefits of both options are equal. Incremental costs and effects can be positive, negative, or zero.
An intervention can be positioned anywhere on this diagram according to its incremental costs and benefits versus the comparator. For example, suppose we were comparing a contingent cfDNA NIPT testing strategy to standard prenatal screening. If cfDNA NIPT lies in the northwest quadrant, such as point A, the costs of cfDNA NIPT are higher than standard prenatal screening, and its benefits are lower. It is therefore dominated by the alternative. Conversely, in the southeast quadrant, at a point such as B, costs are lower, and benefits are higher, so cfDNA NIPT dominates standard prenatal screening. In the northeast quadrant, at a point such as C, higher benefits of cfDNA NIPT are gained at a higher cost over standard prenatal screening. So, we can calculate an ICER, the incremental cost per unit of effect gained, measured as the slope of the line from the origin (the point where costs and effects are equal between both strategies) to point C. In the southwest quadrant, at a point such as D, the cfDNA NIPT costs are lower, but also the benefits are lower than the alternative strategy. Again, we can calculate an ICER, although this is now a cost saving per unit of effect lost, which is again measured as the slope of the line from the origin to the point.
The ceiling ratio can also be visualized in the cost-effectiveness plane diagram ( Fig. 2 ). In this diagram, the dotted diagonal line marked Rc represents the ceiling ratio. If cfDNA NIPT lies above the line, it will not be acceptable on cost-effectiveness grounds versus standard prenatal screening. This is either because cfDNA NIPT is dominated by standard prenatal screening, whatever the value of the ceiling ratio, as in point A (or any other point in the northwest quadrant), or its ICER does not satisfy the ceiling ratio, as in points B and C (or any other point above the Rc line in the southwest and northeast quadrants). It should be noted that for points in the southwest quadrant above Rc, the value of the ICER (i.e., the incremental cost per unit of effect gained or the slope of the line from the origin to the point) is smaller than the ceiling ratio, as in point B, while for points in the northeast quadrant above Rc it is larger, as in point C. Below the Rc line cfDNA NIPT is acceptable on cost-effectiveness grounds. This is either because it dominates the alternative, as in point D, or its ICER satisfies the ceiling ratio, as in points E and F. In this case, the ICER is then either larger than the ceiling ratio in the southwest quadrant, as in point E, or smaller than it in the northeast quadrant, as in point F.
Evidence on the Cost-Effectiveness of Cell-Free DNA NIPT
Systematic Review: García-Pérez et al.
In this section we review empirical evidence on the cost-effectiveness of cfDNA NIPT for prenatal screening and subsequent diagnosis. We focus on economic studies identified by a recent, up-to-date systematic review of the cost-effectiveness of cfDNA NIPT for trisomies 21, 18, and 13 based on papers published in English or Spanish between January 2006 and April 2017 . The review included universal and contingent strategies, compared with screening strategies that do not include cfDNA NIPT or with no screening. To be included in the review the study had to present the number of cases of trisomy 21, 18, and/or 13 detected for every comparator, and their costs or ICERs.
Fourteen papers related to 12 studies, all published after 2012, were identified and included in the review (one study was reported three times ). The review authors assessed the methodological quality of the economic appraisals and found it to be acceptable in most cases, but noted that “the lack of transparency and details on sources prevented a more accurate assessment of the bias” . The sensitivity, specificity, and false positive rate of the strategies compared were consistent with wider published evidence .
Five studies were undertaken in the United States; two in Australia; two in the United Kingdom; and one each in Canada, the Netherlands, and Belgium. All studies were model based, combining data from several sources. All studies compared a prenatal screening strategy (usually first-trimester screening or a combination of first- and second-trimester screening or integrated screening) and some form of screening with cfDNA-based NIPT. Nine studies evaluated cfDNA-based NIPT in a contingent testing strategy; nine evaluated it in a universal testing strategy. Two studies included detection of trisomies 13, 18, and 21 ; the remainder focused on trisomy 21. The perspective was not always stated but a health care system perspective was most common (see later). Some form of health care costs were included in every study. Two studies used multiple perspectives, including educational costs and lost productivity costs as well as costs to the health care system . The time horizon was the duration of pregnancy in all studies; three studies included longer time horizons as well (5 years , lifetime ).
There are several published studies that have assessed the cost-effectiveness of cfDNA NIPT but which were not included in García-Pérez et al.’s systematic review because they did not meet the inclusion criteria .
NIPT and Procedure-Related Miscarriages
Ten of the 12 studies included procedure-related miscarriages as an outcome. In all of these studies the number of procedure-related miscarriages was lower for cfDNA NIPT, whether in a contingent or a universal testing strategy, compared with the strategy without cfDNA NIPT.
Contingent CELL-FREE DNA NIPT Versus Standard Prenatal Screening
Nine studies evaluated cfDNA NIPT in a contingent testing strategy compared with the usual screening strategy . In terms of the incremental costs and benefits, the studies were represented in all four quadrants of the cost-effectiveness plane, with contingent cfDNA NIPT being shown to be more or less costly than standard prenatal screening, and more or less effective at identifying true cases of Down’s syndrome, across different studies ( Table 1 ).