Offering non-invasive prenatal testing as part of routine clinical service. Can high levels of informed choice be maintained?

Abstract Objectives To assess rates of informed choice among women offered non‐invasive prenatal testing (NIPT) for aneuploidy as part of routine clinical care. Methods A cross‐sectional survey was conducted across 6 antenatal clinics in England. Women with a high risk (≥1/150) Down syndrome screening result were offered NIPT, invasive testing, or no further testing. Pretest counselling was delivered as part of routine care by the local maternity team. Women were given a questionnaire containing a measure of informed choice immediately after pretest counselling. Results In total, 220 of 247 women completed the questionnaire. Seventy‐six percent were judged to have made an informed choice, a significant decline from our previous study (89.0% vs 75.6%; χ2(2) = 20.2, P < .001). Of those making an uninformed choice, 46% had insufficient knowledge, 19% had not deliberated, and 13% had made a value‐inconsistent decision. Multivariate analysis showed women who were highly educated (OR, 4.33; 95% CI, 1.08‐17.36) or had had screening in a previous pregnancy (OR, 0.24; 95% CI, 0.90‐0.65) were significantly more likely to make an informed choice. Conclusions The findings highlight the challenges of ensuring informed choice in routine prenatal care where NIPT is not discussed at multiple points, less time is available for counselling, and written consent is not required.


| INTRODUCTION
Non-invasive prenatal testing (NIPT) using cell-free DNA in maternal blood is an advanced screening test, which has been shown to be highly accurate for Down syndrome with detection rates of over 99% and false positive rates less than 0.1% for singleton pregnancies. 1 The test can also be used to screen for trisomies 13 and 18 as well as sex chromosome aneuploidies; however, detection rates are slightly lower. 1,2 The lower predictive value means that NIPT is not considered diagnostic and invasive testing is recommended to confirm positive NIPT results. 3,4 Non-invasive prenatal testing has become widely available in recent years through the private sector, 5 and a number of studies have been conducted to address implementation within a national prenatal care setting. [6][7][8] Clinical advantages include improved accuracy rates in comparison to combined screening, 9 reduction in the need for invasive tests that carry a small risk of miscarriage, 10 early information about risk status, and the opportunity for early reassurance for low risk women. 11 Nevertheless, a number of disadvantages have been identified such as the smaller diagnostic yield in comparison to invasive testing with array comparative genomic hybridisation and the potential for test failures or inconclusive results. 12,13 Empirical research has demonstrated that the test is viewed favourably, with women valuing the opportunity to have a safe, accurate test that can identify cases of Down syndrome that might otherwise be missed. [14][15][16][17][18][19] However, a number of concerns related to routinisation of testing have been raised. These include women accepting NIPT without due thought because of the ease and risk free nature of the procedure, 20,21 rapid implementation raising concerns regarding the capacity of healthcare providers to counsel women appropriately, 19,22,23 NIPT being performed without women realising that they are having a screening test, 24 and NIPT leading to societal pressures to participate in prenatal screening and stigmatisation of those who forego screening. 16,25 A key principle of prenatal testing is the promotion of reproductive autonomy by providing women and families with information to assist in pregnancy management with informed choice and informed consent procedures seen as protecting this principle. 21 Consequently, much discussion has focused on the importance of facilitating informed choice through provision of balanced pretest information and nondirective counselling. [26][27][28] In 2016, our research group developed and validated a measure of informed choice for women offered NIPT, which was included in a questionnaire assessing women's experience of testing. 29 This work formed part of the RAPID evaluation study, which was designed to investigate implementation of NIPT into the maternity care pathway in the UK National Health Service (NHS). 30 Using the measure, we found that 89% of women had made an informed choice. 29 A key limitation of that informed choice study was that NIPT was offered within a highly controlled research setting where participants were given written information at multiple stages and received up to 30 minutes pretest counselling with a specifically trained research midwife who obtained written consent. 31 At that time, we acknowledged that this degree of information and counselling may be challenging to replicate in a routine clinical setting and recommended further evaluation following NIPT implementation. 30 In this paper, we report a follow-up study designed to assess rates of informed choice among women offered NIPT following a high risk Down syndrome screening result (≥1.150) as part of routine NHS care.

| Participants
Between March 1, 2015, and October 31, 2016, women with a singleton pregnancy who were identified as high risk (≥1/150) following Down syndrome screening in antenatal clinics in London and South East England were offered the option of either NIPT (for T21, T18, and T13), invasive testing (QF-PCR on trisomies), or no further testing. Both NIPT and invasive testing were offered at no cost to parents.
The cfDNA sequencing was performed by our NHS service laboratory (North East Thames Regional Genetics Service) as previously described for the research trial. 6 For women identified as having a Down syndrome screening risk ≥1/150, a member of the local maternity care team (either a fetal medicine midwife or consultant) discussed the options available including NIPT, invasive testing, or no further investigations. This follows the model of implementation favoured by the UK National Screening Committee. 32 Prior to the study, maternity care teams were given group training about NIPT by a member of the RAPID research team, 31 and the NIPT information leaflet developed by the RAPID team was available to give to patients. Written consent for NIPT was not required. The care pathways for the delivery of NIPT for the both the RAPID evaluation study and this study are presented in Figure 1.

| Information and counselling
The health professional training given to local maternity units highlighted that the information given to women during pretest counselling should include the following key points: that it is a blood test and there is no risk of miscarriage associated with the test; that it tests for Down syndrome, Edward syndrome, and Patau syndrome; that it detects around 99% of cases where the baby has Down syndrome but that there is a small chance (0.5%-1%) of a false positive and therefore invasive testing is recommended to conform a positive result; that the NIPT result will be ready within 7 to 10 days; that in a small number of cases, the result is inconclusive or the test fails in which case a repeat NIPT will be offered; that it is less accurate than invasive testing that is over 99.9% accurate; that invasive testing carries a small risk of miscarriage (0.5%); and that invasive test results will be ready in 3 working days. The aim of the training sessions was to provide information about NIPT and answer any questions health professionals had.

| Study procedure
The study was conducted at 6 maternity units to allow recruitment from a wide range of social and ethnic groups including a large South Asian population at 2 of the sites. Four of these units had previously recruited participants into the RAPID evaluation study; two had not.
Between October 2015 and September 2016, women who were able to read and understand English were invited to take part in this study on informed choice. A consecutive sample of women were recruited across each of 6 sites. After the discussion in the clinic, potential participants were given a Participant Information Sheet about the study and a paper copy of the questionnaire ( Figure S1) to complete immediately after counselling at the clinic, irrespective of what prenatal testing option they chose. Participants who were unable WHAT'S ALREADY KNOWN ABOUT THIS TOPIC?
• Non-invasive prenatal testing for aneuploidy is a highly accurate screening test, but concerns exist around potential routinisation.
• Previous evidence indicated high levels of informed choice are possible, but this was a tightly controlled research setting.

WHAT DOES THIS STUDY ADD?
• Non-invasive prenatal testing can be offered within routine prenatal care in a way that facilitates high levels of informed choice.
• However, the decline in rates of informed choice compared with those in the research setting highlight the challenges of offering non-invasive prenatal testing in routine prenatal care.
to complete the questionnaire at clinic were offered the option of taking it home and returning it in the freepost envelope within 1 week.

| Questionnaire
The questionnaire included the measure of informed choice for NIPT, which comprised questions to assess knowledge, attitude, deliberation, 33 and NIPT uptake. The development and validation of the measure are described in detail elsewhere. 29 An informed choice was defined as one whereby the participant had good knowledge, had deliberated, and had a positive attitude and had NIPT or had a negative attitude and had declined NIPT (classified as value consistency). 29 The questionnaire also included the Decisional Conflict Scale, 34 the short form of the State Trait Anxiety Inventory (STAI-6), 35 questions to explore motivations for testing, and questions on parity and sociodemographic questions.

| Data analysis
To assess informed choice, knowledge scores, attitude scores, and deliberation scores were combined with test behaviour. Knowledge scores were dichotomised as good or poor using a preset cut-off of ≥9/12. Attitude scores were classified as positive, negative, or neutral with neutral responders being omitted (approach based on van den Berg et al). 29 Deliberation scores were dichotomised as either deliberated or nondeliberated. For the purposes of the informed choice calculation, we excluded women who had declined NIPT but had opted for invasive testing. This was because whilst many of them had a positive attitude towards NIPT, they had declined the test and opted for invasive testing which, according to the definition of the measure, would have categorised them as having made an uninformed choice. However, a limitation of the measure is that it is unable to account for these more complex situations whereby a positive attitude towards NIPT is not behaviourally implemented, not because the decision is uninformed but for practical reasons such as not wanting to wait 7 to 10 days for the test result or the indication for trisomy being so strong that invasive testing was considered most appropriate. A similar approach was taken by van Schendel et al in their assessment of informed choice. 36 The Decisional Conflict Scale score was calculated as directed by O'Connor 34 with participants scoring ≥37.5 categorised as having decisional conflict. The 6-item State Trait Anxiety Index was scored according to the authors' instructions 35 with scores of ≥50 indicating elevated state anxiety. Descriptive analysis was conducted on single items.
Regression analysis was conducted to determine which independent variables were significant predictors of informed choice. Religion and ethnicity were collapsed into binary variables to strengthen the analysis.
A Mann-Whitney test was used to determine differences in decisional conflict and anxiety between women making informed and uninformed decisions. A chi-square test was used to compare rates of informed choice between the RAPID evaluation study and this study. Missing data on the knowledge scale were treated as incorrect answers. Questionnaires with ≥50% missing data were removed from the analysis. Where there was <50% missing data, missing values were replaced by imputing the mean. Analysis was performed using SPSS 22 (IBM, Chicago, Illinois).

| Sample characteristics
A total of 220 of 247 women invited to take part in the study completed Q1 (89% response rate). Two questionnaires were removed due to ≥50% missing data (N = 218). Participant characteristics are presented in Table 1. The mean age was 35.7 years (range, 23-47). One participant had opted for no further testing, and 10 had chosen invasive testing over NIPT. Mean Down risk scores did not differ significantly between women who opted for invasive testing vs NIPT (mean, 58.4 vs 73.8, respectively; F = 1.10, P = .30).

| Informed choice
Of the total sample (N = 218), 84.9% had good knowledge (n = 185); 81.7% had a positive attitude (n = 178), 12.8% (n = 28) had a negative attitude, and 5.5% had a neutral attitude (n = 12); and 93.1% had deliberated (n = 203). Of the 10 women who opted for invasive testing over NIPT, 6 had a positive attitude towards NIPT, 3 had a negative attitude, and 1 had a neutral attitude.  Note: Not all % add up to 100 due to rounding. Not all participants answered all questions, and therefore, there are some discrepancies with total numbers. a It is not possible to work out "missing" here as some may have had a previous pregnancy that did not result in having a child presently.  Table 3). We then explored the relationship between education level and the subscales of informed choice further. Analysis indicated that there was a positive association between higher education level and higher knowledge score (H(2) = 11.04, P = .004), whereas education was not associated with attitude or deliberation scores.

| Decisional conflict and state-trait anxiety
Decisional conflict occurred in 6.6% (n = 13) of cases. Women who made an uninformed decision were significantly more likely to experience decisional conflict than women who made an informed decision U = 2866, P = .217). There were no significant differences between the women in this study and the high risk women in the RAPID evaluation study when we looked at decisional conflict (3.0% vs 6.6%; χ 2 (2) = 2.4, P = .12) and anxiety (54.5% vs 60.4%; χ 2 (2) = 1.2, P = .261).

| Motivations for accepting or declining NIPT
Of those women who opted for NIPT, the most frequently cited reason was "for reassurance that the baby doesn't have Down syndrome" (28.1%) followed by "to help me make a decision about whether or not continue with the pregnancy" (21.9%). Of those women who chose invasive testing over NIPT, the most frequently cited reason was because they would "get the results more quickly" (46.6%). Only one participant declined any further testing, the reason being "I would never terminate an affected pregnancy so there would be no point taking the test" (Table 4).    study, which could be perceived as an inherent bias in the study. Our sample predominantly comprised older, well-educated women; however, this probably reflects the fact that older women are more likely to be at increased risk and be highly educated as they have delayed child bearing for educational or vocational reasons. Finally, in our study, we found that 62.6% of women had elevated anxiety scores at the time of testing. This finding is in line with other studies looking at anxiety in women identified as high risk through screening. 46 Nevertheless, the absence of a baseline anxiety assessment is a limitation as we are unable to determine whether these women were anxious prior to testing or whether anxiety increased as a result of screening.

| CONCLUSION
Non-invasive prenatal testing is set to become part of routine care in the UK NHS imminently. Maintaining high levels of informed choice will be very dependent on effective training of health professionals to ensure they can provide up-to-date unbiased information and also have the confidence and skills to support parents to discuss prenatal testing optional. 47 Current research shows it is possible to achieve high levels of informed decision-making for NIPT, but given the possibility of NIPT as a first line screening test continued research in this area is important.