Factors affecting the clinical use of non-invasive prenatal testing: a mixed methods systematic review

Authors


  • Funding sources: None
  • Conflicts of interest: None declared

ABSTRACT

Non-invasive prenatal testing has been in clinical use for a decade; however, there is evidence that this technology will be more widely applied within the next few years. Guidance is therefore required to ensure that the procedure is offered in a way that is evidence based and ethically and clinically acceptable. We conducted a systematic review of the current relevant literature to ascertain the factors that should be considered when offering non-invasive prenatal testing in a clinical setting. We undertook a systematic search of relevant databases, journals and reference lists, and from an initial list of 298 potential papers, identified 11 that were directly relevant to the study. Original data were extracted and presented in a table, and the content of all papers was analysed and presented in narrative form. Four main themes emerged: perceived attributes of the test, regulation and ethical issues, non-invasive prenatal testing in practice and economic considerations. However, there was a basic difference in the approach of actual or potential service users, who were very positive about the benefits of the technology, compared with other research participants, who were concerned with the potential moral and ethical outcomes of using this testing method. Recommendations for the appropriate use of non-invasive prenatal testing are made. © 2013 John Wiley & Sons, Ltd.

INTRODUCTION

Techniques to enable non-invasive prenatal testing (NIPT) of the foetus were developed during the middle to late 1990s[1] and are now available clinically for a limited range of conditions, including some monogenic disorders, and for foetal sexing for sex-specific disorders.[2] Using the technique to identify a foetus affected by a specific monogenic disorder has been referred to as non-invasive prenatal diagnosis;[3] however, for applications such as foetal sexing[4] and screening for Rhesus D status,[5] it is appropriate to use the term testing, rather than diagnosis. For brevity, in this paper, we will use the inclusive term ‘non-invasive prenatal testing’ (NIPT) throughout (except when referring to the original papers in Table 1).

Table 1. Summary of papers included in the review
Authors, title and countryAimsMethodSampleAnalysisFindings/resultsQuality issues
  1. NIPD, non-invasive prenatal diagnosis; RR, response rate; NHS, National Health Service; IPD, invasive prenatal diagnosis; DSS, Down's syndrome screening; PGD, pre-implantation genetic diagnosis.

Hill et al. ‘Incremental cost of non-invasive prenatal diagnosis versus invasive prenatal diagnosis of fetal sex in England’, UK‘To evaluate the incremental cost of NIPD compared with IPD for fetal sex determination in the English NHS’ (p. 268)Cost analysisComparison of NIPD and IPD in two conditions: congenital adrenal hyperplasia and Duchenne muscular dystrophyMean costs of pregnancyNo significant difference in cost between IPD and NIPD. Costs of fetal sexing with NIPD were offset by the smaller proportion of women who required CVS.88%; Limitations: care pathways derived from a study that was primarily concerned with diagnostic accuracy. Costs were calculated to end of pregnancy and did not consider lifetime costs of a diagnosis.
  Care pathways derived from empiric data. Costs established from NHS data. Perspective NHS costs Univariate and probabilistic sensitivity analysis to determine limits of uncertainty  
    Monte Carlo simulation model  
Hill et al. ‘Determination of foetal sex in pregnancies at risk of haemophilia: a qualitative study exploring the clinical practices and attitudes of health professionals in the United Kingdom’, UK‘To determine the current practices of health professionals in offering prenatal care for women who are carriers of haemophilia and explore the introduction and use of NIPD for foetal sex determination’ (p. 576)Qualitative cross-sectional study based on grounded theory32 health professionals who were involved in management of women who were carriers of haemophilia during pregnancy. Of these, 12 worked in haemophilia centres, six in genetics units and six in foetal medicine units.Grounded theory methodOffering NIPD for foetal sexing was felt to enhance prenatal care of carrier women. Use of NIPD was not felt to increase workload for staff. Although respondents reported that most women were relaxed about use of NIPD, the main questions asked by carriers focussed on the accuracy and timing of the test. Respondents felt it was important the test was not seen as routine.90%; No evidence given of researcher reflexivity, otherwise an excellent paper
Kelly and Farrimond ‘Non-invasive prenatal genetic testing: a study of public attitudes’, UK‘To identify the range of viewpoints on NIPD amongst a sample of the UK public with a diversity of experiences and demographic characteristics’ (p. 75)Q-methodology. Qualitative data were collected via postal self-completion questionnaires.Purposive sample of 71 UK individuals aged between 18 and 60 years (73% RR). Sample recruited via media sources.Thematic analysis of participants' first responses to brief factual information about NIPD.Although 63% of first responses were positive, many of those respondents had concerns. Ambivalence about testing focussed on increased safety and utility by individual parents, contrasting with more ethical concerns about increase ease leading to more casual use, changes in attitudes to disabled children and increased termination rates.80%; Good theoretical underpinning for study but lacked sufficient information on sampling procedure and data analysis.
Kooij et al. ‘The attitude of women toward current and future possibilities of diagnostic testing in maternal blood using fetal DNA’, The NetherlandsTo determine ‘women's attitudes toward current and future testing possibilities’ (p. 165) concerning NIPD.Cross-sectional survey. Questions were derived from a questionnaire used in a previous study. Participants used Likert scales to indicate level of agreement with statements on use of NIPD.Women recruited from two groups: (1) pregnant women and (2) female Master's level students. A power calculation indicated 100 women needed in each group, recruitment continued until sample obtained.Descriptive statistics calculated. Chi-square distribution to determine differences in responses between the two groups.Pregnant women were more likely than students to state that NIPD for Down syndrome should be offered to all pregnant women. The majority in both groups supported use of NIPD for prenatal diagnosis of gender-specific condition, but not for family balancing.75%; Study questionnaire is well described. Rationalisation for using only female Master's students rather than broadening recruitment to other young women not given. Does not state if the students had children.
The Netherlands      
Lewis et al. ‘Non-invasive prenatal diagnosis for fetal sex determination: benefits and disadvantages from the service users' perspective’Authors aimed to explore ‘the benefits and disadvantages of NIPD from the service users' perspective’ (p. 1127).Qualitative cross-sectional descriptive study. Data were collected using semi-structured interviews.45 participants who had experience of using NIPD for foetal sex determination were recruited from appropriate health service clinics in the UK.Thematic analysis of transcribed interview data.Participants were overwhelmingly positive about their use of NIPD. Positive attributes cited included safety for foetus and timing of the procedure, whereas increased anxiety and potential misuse of the technology were cited as negative aspects.80%; The sampling strategy and data analysis are well described; however, there is a lack of reflexivity in the account.
UK      
Lewis et al. ‘Fetal sex determination using cell-free fetal DNA: service users' experiences of and preferences for service delivery’‘To determine service users' experiences and preferences for service delivery' regarding fetal sex determination using NIPD’ (p. 735)Qualitative cross-sectional descriptive study. Data were collected using semi-structured interviews.38 women who had used NIPD for foetal sex determination because the foetus was at high risk of a specific condition affecting one sex and six of their partners. All recruited via relevant clinics in England.Thematic analysis of transcribed interview data.Participants believed use of NIPD enhanced reproductive choice. They preferred receiving pre-test information and results from a specialist who understood the procedure and the condition. Accuracy of results considered most important feature.80%; Sampling strategy clear. There is no reflexive account given in the report.
UK      
Sayres et al. ‘Cell-free fetal DNA testing: a pilot study of obstetric healthcare provider attitudes toward clinical implementation’To assess the opinions of healthcare providers on introduction of NIPDQuestionnaire distributed at a conference62 of approximately 180 surveys completed. 34% response rate Majority of respondents had MD degree and majority were practising in the United StatesDescriptive85% did not have a high level of knowledge about cell-free foetal DNA testing. 55% thought it would encourage testing for more people and 45% more conditions. Clinical utility ranked first out of six aspects of prenatal testing followed by risk to mother or foetus and sensitivity. Majority would follow guidelines if issued, felt test should be covered and a minority would support direct-to-consumer testing65%; Small sample
      Low response rate and potential sample bias limit generalisability
United States      
Szczepura et al. ‘A new fetal RHD genotyping test: Costs and benefits of mass testing to target antenatal anti-D prophylaxis in England and Wales’Cost-effectiveness of NIPD targeted prophylaxis in pregnancies of rhesus negative mothers compared with normal practiceCost analysisTwo scenarios compared with normal practice: (1) NIPD used to target antenatal prophylaxis with serology tests directing post-delivery prophylaxis, and (2) in addition to antenatal prophylaxis, NIPD also replaces serology testing postnatally if RhD negative foetus was identified.Financial modelling. Univariate sensitivity analysisNet financial benefit of implementing mass NIPD testing in the antenatal setting while maintaining current strategies for postnatal testing is negligible. NIPD is unlikely to produce any clinical benefit in these scenarios as the number of sensitisations may not fall. Depending on test sensitivity, they may in fact rise88%; Many assumptions in modelling as limited data. Predominantly from a laboratory perspective. Main uncertainty concerns potential royalty fees for tests.
  Costs established from NHS data together with a threshold royalty fee per test.    
  Care pathways derived from literature review and expert opinion    
UK      
  Perspective NHS England and Wales    
Tischler et al. ‘Noninvasive prenatal diagnosis: pregnant women's interest and expected uptake’To assess whether pregnant women are interested in NIPD and what factors would influence their interest and expected uptake’ (p. 1293).Experimental study. Women were allocated to two groups before completing a questionnaire on NIPD. In one group, the women were given brief information on NIPD prior to reading the questionnaire; in the other, more detailed information was given.114 pregnant women (80.9% RR) more than 24 weeks of gestation. Women were recruited in the waiting room of the local antenatal clinic.Descriptive statistics used. Chi-square analysis used to compare responses between those who read the brief and those who read the detailed description of NIPD.The majority of women expressed interest in using NIPD if available. Safety of the foetus and accuracy of the results were most important factors in deciding to use NIPD. Women who read the brief description of NIPD were more likely to use the test.68%; Although randomisation appears to have been possible, it is not mentioned.
United States      
Van den Heuvel et al. ‘Will the introduction of non-invasive prenatal diagnostic testing erode informed choices? An experimental study of health professionals’To assess the extent to which the introduction of NIPD of Down's syndrome may undermine the process of making informed choices (p. 25)A vignette study, between-subjects design. Each participant randomly presented with one vignette: IPD, NIPD and DSS. Outcome variables: topics considered important to communicate, timing of the test and the perceived need for written consent assessed by questionnaire.231 obstetricians and midwives working in the UK. Invitations given to attendees at three conferences on updates in prenatal diagnosis and three maternity units. Majority of respondents were obstetricians (59.3%), female (79.7%) with a mean age of 39.7 years. Age, gender, profession and data source (i.e. the conference or maternity unit) were distributed evenly across intervention groups.Descriptive statistics for responses across the groups. Univariate analyses across test types to compare the three characteristics of prenatal counselling. Logistic regression to identify independent predictors of beliefs about the timing of the test and the perceived need to sign a consent form.Five of the seven topics were considered equally important to communicate across the three vignettes. Procedure itself and miscarriage risk more important for IPD vignette. Respondents receiving IPD vignette more likely to indicate presentation and uptake of tests should occur on different days and that signed consent was necessary as compared with those receiving NIPD and DSS.88%; Randomisation method not clear and sample possibly not representative. However, no significant differences in the characteristics of participants between experimental groups. As discussed in the paper, the main limitation is that the outcome measure may not be a true measure of informed decision-making.
UK      
Yotsumoto et al. ‘Attitudes toward non-invasive prenatal diagnosis among pregnant women and health professionals in Japan’ JapanTo assess attitudes towards non-invasive diagnosis in JapanSelf-completed questionnaire mailed to clinical staff and pregnant women in two hospitals252 respondents 185 health professionals: RR 58% obstetricians, 39% paediatricians, 45% genetic counsellors 67 pregnant women RR unknownDescriptive statistics79% health professionals 97% women agreed with application of PGD p < 0.0175%; Sample selection may be biased. Small number of pregnant women responders may limit generalisability
    Comparison between responses of pregnant women and health professionals  
     Women expected the test to be safer earlier and easier to understand than health professionals who were more concerned about diagnostic accuracy and reliability Health professionals had more concerns about implementation than pregnant women. Suggestion that individuals who had less knowledge were more supportive of NIPD. 

Although there have been numerous papers focussed on the development and quality of laboratory techniques to determine foetal status using cell-free foetal DNA extracted from maternal plasma,[2] these must be placed into a health services context through consideration of the clinical, ethical, legal, social and economic issues related to the introduction of any new technology. The use of NIPT will result in a reduction in the number of invasive tests performed.[2] The major clinical advantages of using NIPT compared with invasive testing have been cited as eliminating the risk of miscarriage due to testing and enabling confirmation of foetal status earlier in the pregnancy.[6, 7] At present, NIPT for aneuploidy is not fully diagnostic, and confirmatory invasive testing should be offered;[8] however, there is anticipation that this will change. Offering one diagnostic test for aneuploidy, as opposed to stepwise screening and invasive testing, has been cited as a factor in simplifying the counselling process, enabling parents to focus on one decision.[7] However, there have been serious ethical concerns aired about the use of NIPT. For example, the potential for the test to be viewed as ‘routine’[9, 10] and the ease with which a blood sample can be taken[3] may reduce the consideration given by parents as to the outcome and implications of the test, placing in jeopardy the need for informed consent[9, 11] to the procedure. Greely[11] makes the point that in other areas of reproductive healthcare, there are demonstrable disparities in access to services between individuals from different ethnic or socio-economic groups, and the use of NIPT might exacerbate these. He, amongst others, also raises the issue that the ease with which NIPT can be used will inevitably result in increases in the number of prenatal tests and subsequent abortions.[11] Concerns that use of NIPT will result in sex selection for social reasons[12] and the fear that pressure to increase the range of conditions for which testing is offered (due to specification creep[6]) will result in eugenic practices have also been voiced.[11]

Although it is evident that both advantages and disadvantages regarding the clinical use of NIPT have been expressed, the evidence base for offering NIPT and the impact on prospective parents and health professionals has not been reviewed. We therefore conducted a systematic review on this topic.

METHODS

Aim

The aim of this review was to investigate factors influencing the clinical use of NIPT. Although we did not approach the topic with an a priori list of themes, we did anticipate that literature focussed on clinical guidelines or protocols, user perspectives, ethical issues, health professional competence and economic considerations might be included. The objectives were to (1) identify studies of relevance to the topic, (2) extract themes from those papers, relevant to factors influencing clinical use of NIPT and (3) make recommendations related to the clinical use of NIPT, on the basis of the existing evidence.

Design

A systematic review can be used to ensure that relevant research-based evidence is identified and the data are synthesised to produce overall findings and enable valid conclusions to be drawn. We followed the method described by the Centre for Reviews and Dissemination[13] for the conduct of systematic reviews and used a set of inclusion and exclusion criteria to guide selection of papers. We also undertook a quality assessment of the studies before conducting a descriptive and narrative synthesis[14] of the results. The search was undertaken during November 2012.

Search methods

Six relevant electronic databases were searched: CINAHL, Medline, SocIndex, Psycharticles, Psyclit and the Web of Science. In accordance with the recommendations of the Centre for Reviews and Dissemination (2008), we hand searched for further publications of the first authors of the initial set of papers and the reference lists of those papers. We also undertook a hand search of the three previous years of publication of key journals: Prenatal Diagnosis, American Journal of Obstetrics and Gynecology, British Journal of Obstetrics and Gynaecology, Journal of Community Genetics, European Journal of Human Genetics and Public Health Genetics. This ensured that we retrieved the maximum number of eligible papers, and this strategy is particularly important when undertaking a review that includes qualitative or mixed method studies, as indexing of those papers may not be accurate.

The combination of key words used in the initial search was ‘non-invasiveornon invasiveor AND ‘prenatal’ or ‘antenatal’ or ‘pregnan*’. There was one paper retrieved through hand searching that had not already been identified via the electronic search. To be certain we had not missed any relevant papers, we also searched the databases using the terms ‘cell-free fetal DNA’ and ‘prenatal’ and did not find any papers that were not identified in the initial search.

Studies were eligible for inclusion if they

  • were published between January 2002 and November 2012;
  • were research studies based on collection of empirical data or reviews involving a meta-analysis of data;
  • focussed on the factors influencing clinical use of NIPT of the foetus using extraction of cell-free foetal DNA from maternal serum;
  • were published in English in a peer-reviewed scientific journal; and
  • were conducted in any country.

Studies were excluded if they focussed on laboratory issues, technical case reports or biochemical or ultrasound methods to test or screen the foetus.

Search outcome

We followed the PRISMA[15] process for reporting systematic review search outcomes. The selection process is summarised in Figure 1. Of 372 papers initially identified through the search strategy, 72 were duplicates, and a further 266 did not fit the inclusion criteria. The full text version of 32 papers was read, resulting in 12 papers for quality appraisal.

Figure 1.

Search flow diagram based on PRISMA[15]

Quality appraisal

We used the quality assessment tool described by Kmet et al.[16] to undertake an evaluation of the quality of all papers. This tool provides two sets of relevant questions, one for qualitative and one for quantitative studies. For each relevant question, the paper is allocated a score of 0 (criterion not met), 1 (criterion partially met) or 2 (criterion fully met). The total score was then converted to a percentage mark. As there were so few papers identified for the review, we used a cut-off point of 55%, which is described by Kmet et al.[16] as liberal. This resulted in exclusion of a further paper, leaving a total of 11 papers for the review.

Data abstraction

Key features of each study were extracted and entered into a table (Table 1)[13] under the following headings: aim, study design, sample, data analysis, main findings and quality issues.

Synthesis

With so great a variation in the studies included in this review, it was not feasible to undertake a meta-analysis. As suggested by the Centre for Reviews and Dissemination,[13] we used the narrative synthesis technique to extract the main themes (Table 2).[14] This involved comparing data across studies that may have been focussed on different uses of NIPT, for example, foetal sexing and identification of single-gene mutations. The findings are presented in the following text under those themes.

Table 2. Example codes and categories under each theme
CodeCategory
  1. NIPT; non-invasive prenatal testing; IPD, invasive prenatal diagnosis; TOP, termination of pregnancy; DTC, direct-to-consumer test; PND, prenatal diagnosis.

Perceived attributes of NIPT
Less risk to foetusPerceived positive attributes
Safer testing 
General ease of testing process 
Enables woman who is going to have IPD (e.g. after foetal sexing) time to prepare 
Decision-making easier in first trimester (before pregnancy is obvious) 
‘Normalisation’ of pregnancy (TOP might take place in period other pregnancies at risk of miscarriage) 
Test will increase anxietyPerceived negative attributes
Increases connection to foetus when pregnancy might be terminated 
Delayed attachment to pregnancy (waiting till after NIPT result) 
Ease of test may reduce thought before taking 
Regulation and ethics
Testing needs to be regulated to avoid abuseEthical concern
NIPT will be too easily used 
Need to avoid offering as ‘routine’ 
Adverse impact by increasing numbers of terminations of pregnancy 
Not raising new social and ethical issues, extension of existing issuesPolicy
Need for wide public debate 
Insufficient support for current tests (so do not introduce new ones) 
Need for health professional input (should not be DTC)Direct-to-consumer tests
DTC would increase use for trivial reasons DTC possible 
NIPT in practice
NIPT will increase demand for PNDPractice issues
Test easy, but results important 
Pre-support and post-support from health professionals needed 
Accuracy important 
List of conditions for which NIPT should be offered 
NIPT should be offered in high-risk casesUses of NIPT
NIPT should be provided to women on request 
NIPT can be used for foetal sexing in gender-specific conditions 
NIPT is useful to detect Down syndrome 
NIPT should not be used for sexing for family balancing 
Decision about termination still hard (even if test easy)Patient experience
Adequate information was given by health professionals pre-test 
Prefer results by phone, followed by in writing 
Pressure from family to have test 
Higher interest in NIPT associated with older age, higher education and white or Asian ethnicityPatient attitude
Would want to use to check from other conditions, for example, aneuploidy if possible 
Economic considerations
Use of NIPT to detect Rhesus positive foetus is cost effective (dependent on royalty fee)Economic analysis
Use of NIPT similar cost to invasive test when used initially for sexing of a foetus at risk of Duchenne muscular dystrophy or congenital adrenal hyperplasia 
Minimal extra expense to family (from travel only)Patient experience
Patients prepared to pay for test 
Cost of test important when considering introduction to health servicePolicy

RESULTS

Description of the studies

Of the 11 studies included, seven[17-23] were based on quantitative methods and four[24-27] on qualitative methods. The majority of the studies were conducted in the UK,[17, 20, 22, 24-27] whereas two were undertaken in the United States,[19, 21] one in Japan[23] and one in the Netherlands.[18] Two studies focussed on the economic issues involved in introducing NIPT,[17, 20] three on the views of health professionals,[19, 22, 26] only one of which examined health professionals actively offering NIPT.[26] Four focussed on service users,[18, 21, 24, 27] including one on women using NIPT[24] and one on pregnant women.[21] Yotsumoto et al.[23] investigated attitudes of both pregnant women and health professionals. The final paper reported results of a study of public attitudes.[25] Original data (and the quality scores) on the included studies are presented in Table 1.

Themes

We identified four major themes under which we present the analysis. These were perceived attributes of the test, regulation and ethical issues, NIPT in practice and economic considerations. One of the striking aspects of the findings in general was the difference in emphasis between the reported attitudes of those who had or anticipated using NIPT and others including health professionals and members of the public. Where the users understandably focussed on benefits and disadvantages to the individual, including enhanced decision-making and support for the management of a pregnancy,[24, 26, 27] other respondents were more likely to voice their concerns about the impact of the use of NIPT on attitudes towards the foetus and disabled members of society.[19, 25, 27] This central issue of individual versus societal considerations is reflected strongly in the first two themes.

Perceived attributes of the test

Positive aspects of NIPT were perceived by service users to be greater safety for the foetus,[21, 25-27] earlier information about foetal status[24, 27] and general ease with which the sample could be taken.[24, 26, 27] In particular, those who had used NIPT were positive that benefits outweighed the slight disadvantages.[24, 27] There was also a sense that NIPT empowered women in several ways; having a test that could be performed in the first trimester made decision-making about possible termination of pregnancy easier and gave women more time to prepare for either further testing or birth of an affected foetus.[24, 27] In addition, women interviewed for one study spoke of the availability of a test before 12 weeks as ‘normalising’ the pregnancy, as all pregnancies are considered to be at risk until 12 weeks of gestation,[27] and if termination took place, this would be consistent with the timing of other pregnancy losses.

Views on the psychological impact of using NIPT were expressed. Some participants spoke of the test as promoting peace of mind,[24, 26, 27] although it was also thought that the use of NIPT could provoke additional anxiety in parents.[25] Some women indicated concern that having a test so early in pregnancy could enhance bonding with the foetus at a time when miscarriage was still a significant risk,[27] thus potentially increasing distress, whereas others in the same study felt that testing at that time could interfere with bonding with the foetus.[27]

Regulation and ethical issues

There were a number of ethical concerns raised by respondents in seven of the included studies:[19, 21, 23-27] these concerns could be summarised as a fear that use of NIPT could diminish the value of foetal life and the lives of those who are disabled. This was expressed by participants in studies by Kelly et al.[25] and Lewis et al.[27] as increasing the possibility of parental rejection of any child who was not perfect, with concern shown about a perceived concurrent increase in the numbers of pregnancies that would be terminated.[25]

Two authors[25, 26] cited concerns about the ease of NIPT leading to ‘routinisation’ of the test without due consideration for the possible outcomes of testing. Because of this, respondents indicated ambivalence towards the introduction of NIPT. The need for public debate and appropriate regulation of the use of NIPT was cited,[25] with some participants suggesting that NIPT should only be offered in those pregnancies at increased risk of abnormality,[23, 26] including those that were related to a particular sex;[18, 24, 26, 27] however, use of NIPT for foetal sexing for social reasons was condemned.[18, 23, 26] Only Kooij et al.[18] compared use of NIPT in different conditions: more women in that study felt that its use was justified for foetal sexing in sex-linked conditions than for Down syndrome or late-onset disorders, but no statistical analyses were performed.

NIPT in practice

Participants observed that in practice having an NIPT was ‘easy’,[26] but the implications of the results on the individuals involved could be profound:[27] in essence, the decision to terminate was a difficult one, regardless of the method of obtaining the results.[25] These opinions seemed to be consistent whether the women involved were at high risk of having an affected child or when viewed by members of the general public asked to discuss NIPT for Down syndrome.[25] It was therefore suggested that accuracy of results was vital[21, 24-27] and that NIPT should only be offered with the provision of pre-test and post-test counselling.[21, 25] Although the early detection of a problem was considered an advantage, others stated that having an early termination for a pregnancy that may have miscarried in any case could be seen as a negative outcome.[18] Unsurprisingly, Tischler et al.[21] found that women who would not want to have a termination were less likely to accept NIPT and also that the more information women had about NIPT, the less likely they were to use it to make pregnancy decisions and the more likely they were to request confirmatory invasive testing.

In terms of personnel involved, women expressed a wish for the person who conducted the pre-test counselling to give them the results;[24] this may relate to their lack of confidence in the perceived ability of many health professionals to provide accurate information about the test.[19, 24] Having the results conveyed by telephone, followed by information in writing, was preferred by respondents to one study,[24] although 74% of health professionals believed that information should be given at a different appointment from the one in which consent was sought, giving the parents time to think about the decision.[22] However, significantly less health professionals considered written consent for NIPT was needed, when compared with invasive tests (p < 0.001).[22]

Economic considerations

In two studies, economic modelling was undertaken to determine whether offering NIPT was cost-effective. Hill et al.[17] compared the costs of using NIPT for foetal sexing in pregnancies at risk of Duchenne muscular dystrophy and congenital adrenal hyperplasia, compared with the use of invasive testing, and concluded that the cost of NIPT was offset against the reduced number of invasive test required after foetal sexing. However, in a study focussed on use of NIPT in women who were Rhesus negative, Szczepura et al.[20] suggested that although use of NIPT to determine foetal Rhesus status could result in some savings to the health service (dependent on the level of royalty fees paid), these were minimal and unlikely to warrant the additional risk of sensitization of mothers (due to misdiagnosis of foetal Rhesus status in 0.9–1.5% of cases) if postnatal serology tests were no longer conducted. Other economic considerations included a likely reduction in travel costs for parents if they did not need to attend a specialist centre for invasive testing,[24] although the same authors found that some parents would be prepared to pay for NIPT themselves.

DISCUSSION

Respondents to the study by Kelly[25] emphasised that the ethical and societal issues raised by NIPT were not new. In this review, these concerns and ambivalences have also been elicited from a number of sources. Overall, we identified some tension between individual wishes to use the technology and societal concerns about the potential negative impact on diversity of the population and diminished acceptance of those who have some type of disability or abnormality. These concerns are not restricted to the use of NIPT and have been voiced for decades;[28, 29] however, they appear to be emphasised by the comparative ease of use of NIPT and potential lack of consideration of the potential life-changing decisions that may have to be made.

The goal of prenatal testing programmes is generally described as offering reproductive choice, although whether this is truly achievable has been disputed in other studies[30] and the possibility that ease of the testing could actually impede, rather than facilitate, informed choice has been raised.[22] The challenge to those offering NIPT is therefore to maintain the goal of fully informed decision-making in a context where the testing itself may be procedurally easier, giving less time for reflection and consideration. The importance of the opportunity for considered reflection has been emphasised elsewhere in the literature,[31] and allowing time to consider the choice to accept or decline NIPT should therefore be an aim when introducing models of implementation for NIPT.[10] This is likely to be more difficult if NIPT is introduced as an option for all women, for example, as part of a Down syndrome screening programme. Currently, antenatal screening programmes provide parents with an opportunity for considering the options before consenting to the diagnostic procedure,[6] but this would be reduced if they were offered a diagnostic test, using NIPT, as a one-stage procedure. Although this will be possible if the diagnostic accuracy is acceptable, it is also feasible to use NIPT as an alternative to invasive testing, after initial screening for Down syndrome using the current combination of biochemical analysis and ultrasound. This approach was considered the most ethically sound by Deans and Newson[13] in a paper in which they considered three possible models of service provision in the context of NIPT and aneuploidy screening. The Council of Europe recommendation for all prenatal testing and screening includes the need to provide pre-test and post-test counselling and obtain informed consent.[32] In the situation where NIPT is used for high-risk pregnancies (e.g. foetus is at risk of a monogenic disorder), the parents have often had counselling before conception, or there is a natural time lapse between discussion of the test and readiness to perform it, because of the need for laboratory preparation. Issues identified relating to NIPT in practice include concerns about the nature and timing of the pre-test and post-test counselling and the nature of consent that is obtained, concerns addressed recently by the American College of Obstetrics and Gynecology,[33] which stated that NIPT should be an informed patient choice after pre-test counselling. Direct-to-consumer genetic tests are increasingly available via the Internet, and NIPT may be available via this route in the future; however, participants varied in their views as to whether it was feasible because of the need for counselling by health professionals pre-test.[25] This is somewhat consistent with Modra,[34] who raises the possibility that direct-to-consumer prenatal testing could limit parental autonomy by reducing the chances that the user was adequately informed about the implications of the test.

Other ethical concerns highlighted in a number of papers in this review centred on the use of NIPT for social reasons, such as foetal sexing without a medical indication. These have also been raised by authors such as George,[35] who highlighted the issue in communities where male children are more valued than female children.

The two economic studies[17, 20] identified in this review demonstrated the cost-saving possibilities when the use of NIPT reduces the number of further procedures such as invasive testing. However, the review demonstrates the necessity for considering the technology as part of the whole care pathway, including provision for other costs such as royalty or licensing payments for use of the technology and the potential costs of confirmatory invasive tests, if requested by users. In addition, it is clear that the economic context for use of NIPT will vary greatly between use for high-risk pregnancies and offering it on a population level to women at generally low risk. The opportunity to consider carefully the way in which NIPT could be utilised for detection of foetal aneuploidy was discussed in an editorial by Chitty et al.,[36] who indicated that the economic evaluation of the introduction of NIPT is complex, relating to consideration of not only screening cut-off points and costs per test but also the number of foetal losses due to invasive testing and the detection rate of chromosomal abnormalities other than aneuploidy that is possible using a pathway of screening, invasive testing and karyotyping.

This review was conducted in a systematic way with every effort to obtain and analyse all the available relevant evidence on this topic. However, it is possible that some evidence, particularly in the grey literature, was not identified. It should be noted that there is a dearth of literature relating to the use of NIPT in diverse cultural settings, so the evidence pertains mainly to those living in Western societies. In addition, there were only two studies in which the views of those who had actually experienced NIPT were reported, and this needs to be addressed.

Given the societal concerns about possible impact of the use of NIPT and the need for more evidence on the clinical use of this type of prenatal testing, we suggest the following:

  1. Consideration needs to be made with respect to the needs of specific patient groups before the introduction of NIPT into clinical practice. For example, although it may be appropriate to offer NIPT in high-risk pregnancies, introduction for population screening may not be warranted without further research, public consultation and discussion at policy level to address concerns about the ethical and social implications.
  2. Specific provision needs to be made to ensure parents have the time and information needed to make an informed choice about the use of NIPT. This may require building in time for reflection between information giving and seeking informed consent, as well as ensuring health professionals are appropriately trained.
  3. Further empirical research is needed in a range of cultural settings to ensure that clinical practice is appropriate for the population served.

Further consideration at policy level is needed to determine whether NIPT should be available as a direct-to-consumer test.

CONCLUSIONS

Non-invasive prenatal testing is being introduced into pathways of antenatal care, and therefore, adequate provision for the use of NIPT must be made in terms of both policy and practice. Fears about possible eugenic applications of NIPT need to be aired in public debate[25] to ensure they are ethically and culturally acceptable, and policies are needed to ensure that the technology is used where it is most practicable from a practice and economic point of view. The opportunity to use NIPT should maximise autonomous decision-making; however, practice guidelines are required to ensure that the testing is not accepted without understanding of both the technique and the potential implications of the results.

WHAT'S ALREADY KNOWN ABOUT THIS TOPIC?

  • Non-invasive prenatal testing (NIPT) is available for foetal sex determination in pregnancies at high risk of sex-linked disorders and for diagnosis of a limited number of genetic conditions.
  • There is limited use of NIPT for aneuploidy detection, albeit currently only in the private sector.
  • This method eliminates the risks to the foetus inherent in invasive procedures and can be performed earlier in the pregnancy.

WHAT DOES THIS STUDY ADD?

  • Users and potential users of NIPT regard it positively, but there are wider ethical and societal concerns about the impact of easier access to prenatal testing.

Ancillary