Expanded carrier screening for autosomal recessive conditions in health care: Arguments for a couple‐based approach and examination of couples' views

Abstract Background Expanded carrier screening (ECS) is aimed at detecting carrier states for autosomal recessive (AR) or X‐linked conditions in couples from the general population planning a pregnancy. ECS is currently usually offered on an individual basis despite the fact that, for AR conditions, only carrier couples are at risk of affected offspring. In this paper, we present a couple‐based ECS test‐offer for AR conditions, where results are offered as couple‐results only, and describe how couples view such an offer. Methods and results An online survey covering attitudes, perceived difficulty, and intention to take up couple‐based ECS was used to examine couples' views. Results show that in 76% of the participating couples there is no objection at all towards receiving couple‐results only. Most couples display similar views. Observed discrepancies usually involved one of the couple members having a positive view, whilst the other was neutral. Although views stayed strikingly stable after discussion, the partner's opinion was regarded as important in deciding whether or not to have testing. Conclusion This study shows that most couples do not object to receiving couple rather than individual ECS results, have similar views towards the offer, and are able to discuss differences in views and intentions.

practice in the carrier screening offer to high-risk groups is no longer helpful or necessary when switching to population-based ECS for AR conditions. In this paper, we present a couple-based ECS test-offer for AR conditions, where results are offered as couple-results only, and describe how couples view such an offer.

Knowledge of individual carrier status for AR conditions only has
reproductive utility if we know the status of the other partner. Given that the risk of being a carrier of a particular condition included in the ECS offer is generally low, individual carrier status does very little to predict offspring risk. Reporting of individual carrier states with the aim of cascade screening of family members is, in the context of ECS, therefore of little value. It is only the positive "couple-result" which conveys increased risk for future offspring. Disclosing individual carrier status therefore has no clear advantage, whilst it may lead to anxiety and perceptions of illness 11,12 and goes together with high cost of follow-up testing. Lynch et al examined the time needed to provide genetic counselling in the context of preconception carrier screening. 13 They found that 78% of study participants were carriers of at least one condition and that the median time for results disclosure was 64 minutes (range 5-229 min). Whilst some have expressed concern that individual carrier states are important in case couples split up 2,14 , we argue that also in this situation individual carrier status has little reproductive value. In case couples split up, and the new couple wishes to have a child, a new couple-based ECS test can be done with the new partner.
Also, the argument that people have the right to receive individual carrier states does not apply. Because no individual carrier states are being generated by our analysis, there are no results being withheld.
The European General Data Protection Regulation (GDPR) act does grant data subjects the right to receive, free of charge, the personal data they have previously provided in a "commonly used and machine readable format" (https://eugdpr.org/). The GDPR act thus grants couples who have taken an ECS test the right to receive the raw data that is generated, but not the interpretation of this data in terms of individual carrier status.

| Sample and survey design
This study is part of a larger study on potential users' views and intentions towards couple-based ECS and framing of information, for which participants from the general population were recruited online. The study design has been published elsewhere. 16 most couples do not object to receiving couple-results only criteria and received access. Of these, 504 (58%) respondents completed the survey. The study's flow diagram is depicted in Figure 1.
Study participants were asked to fill in three questionnaires: T0 (before framing of information), T1 (after framing of information, before discussion), and T2 (after discussion). Potential participants were invited until we had 500 respondents who matched the abovementioned inclusion criteria and completed T0 and T1. The results of individual participants on T0 have been described in Plantinga et al 16 and the effects of the framing of information between T0 and T1 in Voorwinden et al. 19 After completing T0 and T1, respondents were asked to (1) invite their partners to also participate in the study, (2) discuss couple-based ECS with their partners, and (3) after discussion, fill in questionnaire T2. The partners were asked to fill in one questionnaire independent from the other member of the couple after discussing couple-based ECS. This study presents the results of each respondent and her/his partner. Ethical clearance for the study was granted by the Medical Ethical Review Committee of the UMCG (M14.152635).

| Sociodemographic characteristics
The following sociodemographic characteristics were recorded: sex, age, religion, and educational level. Educational level was categorized as: "low" (finished primary school, lower secondary school or vocational training); "intermediate" (higher level secondary school or intermediate vocational training); and "high" (higher vocational training or university).

| Relationship characteristics
We included duration of relationship (in years), relationship satisfaction, and the expressed desire to have children with this partner.

| Views on the offer of a couple-based ECS test
In exploring couples' views on the couple-based ECS test-offer, we included four measures based on the Theory of Planned Behaviour framework. 22 Attitude towards the couple-based ECS test-offer was measured by the initial response towards the couple-based ECS testoffer. Answers were rated on a 7-point Likert scale with anchors very negative (1) and very positive (7). Objection towards receiving couple-  this test were to be offered, I would be willing to participate", and rated on a 7-point Likert scale with anchors unlikely (1) and likely (7).

| Couples' discussions about the couple-based ECS test-offer
To gain understanding of couples' discussions on the couple-based ECS test-offer, we included six measures. Duration of discussion was measured by asking couples how long they spent discussing the couple-based ECS test with their partner, measured in five 10-minute intervals from <10 minutes to >40 minutes. The scores were categorized as little (1-2), neutral (3), and much (4-5).

| Data analysis
Sociodemographic characteristics, relationship characteristics, and couples' views on ECS were described with N (%) for nominal and ordinal variables and mean (SD) or median (IQR) for interval and ratio variables. Differences between respondents with or without followup measurement and with or without included partner (see Supporting Information) were tested using the chi-square test for nominal/ordinal variables, with the unpaired Student's t-test for interval/ratio variables with approximately normal distributions and with the nonparametric Mann-Whitney U test for interval/ratio variables with skewed distributions.
As mentioned before, respondents from both sexes were recruited first and completed questionnaires both before and after having had a discussion about couple-based ECS with their partner. The partners were recruited at a later point in time and only filled in one questionnaire, after discussion. To examine couples' combined views and similarities and discrepancies within couples, we compared respondent's T2 (measured after discussion) and partner's T0 (measured after discussion), except for the variable attitude towards receiving couple-results only. Because this variable has only been measured in respondents' T0 (measured before discussion), we could only compare the variable at this time-point with the partners' T0 measurement (measured after discussion). Changes in individual respondents' views after discussion with their partner were based on the comparison of T0 (before discussion) and T2 (after discussion). We chose to use T0 instead of T1, because T0 measures respondents' initial reaction and the study of Voorwinden et al 19 showed that the manipulation between T0 and T1 did not affect respondents' scores.     Figure 3A shows how nondiscrepant couples (that is, couples in which both partners displayed the same view) evaluated the couple-based ECS test-offer. The data show that most couples do not object towards receiving a couple-result only: 93% of the couples said they (both) did not object to this, whilst 7% (four couples) did. Most couples also do not perceive the decision to have couple-based ECS as difficult: 40% perceive the decision as easy, 43% is neutral, and 17% perceive the decision as difficult. If the couple-based ECS test were to be offered, both partners intended to have the test in 35% of the couples, in 20% both did not have the intention to participate, and in 45% of the couples both partners were neutral. Couples displayed most reservation in their attitude towards couple-based ECS, measured by participants' initial response towards the couple-based ECS test-offer. In 65% of the couples, both partners displayed a neutral attitude, in 27% of the couples a positive attitude, and in 8% of the couples both partners had a negative initial response towards the couple-based ECS test-offer. Figure 3B shows how discrepant couples (that is, couples in which both partners did not display the same view) evaluated the couplebased ECS test-offer.

| Evaluation of couple-based ECS test-offer by discrepant couples
Overall, we found that discrepancies within couples often involved one of the couple members displaying a positive view, whilst the other was neutral. An exception was seen on the objection towards

| How much did respondents' views change after discussion?
Finally, we examined whether respondents' views towards the couplebased ECS test-offer changed after discussion with their partner. Table 2 shows the differences, as well as their size (small/large) and The results show that most respondents (55%-75%) held similar views before and after discussion with their partner. Where changes were found, these could be in either direction, but overall the magnitude of the change was small (changing from neutral to positive or negative or vice versa). However, the attitude towards the couplebased ECS test-offer, measured as participants' initial response towards this test, was an exception: here, significantly more respondents (23%) reported a decrease in attitude than an increase in attitude (10%). This decrease mostly involved a change from a positive position before discussion towards a neutral attitude after discussion.
The perceived difficulty of the decision to have the test changed most often: 45% of the respondents reported a change in perceived difficulty of the decision. The changes went, however, in both directions, equally for the men and women in the couples, implying that for some the decision was easier after discussion, whilst for others it was more difficult. Male and female respondents did not differ significantly in this respect.
The intention to take up couple-based ECS was most stable: 25% of the respondents reported a change in intention after discussion with their partner. The changes went, in both directions, equally for the men and women in the couples, implying that for some intention increased, whilst for others it decreased. Male and female respondents did not differ significantly in this respect. Finally, we examined whether changes in respondents' views were correlated with characteristics of the discussion that respondents had had with their partner. Table 3 displays the bivariate (Spearman) correlation coefficients between the different discussion characteristics and changes in attitude, perceived difficulty, and intention after discussion. None of the discussion factors is significantly correlated with a change in view after discussion.

| DISCUSSION
To our knowledge, this is the first report of couples' views towards the offer of an ECS test for AR conditions which reports couple-results Looking at the overall evaluation of a couple-based ECS test-offer, we found that the neutral and positive views prevailed among couple members. Furthermore, we found that couples displayed large similarities in views: often, both couple members being neutral and rarely both members being negative. Respondents' views also stayed strikingly stable after having discussed the ECS test-offer with their partner. The observed changes were only few and mostly small. The initial response towards couple-based ECS did, however, significantly change after discussion. This usually involved a change from a positive view to a neutral standpoint. It might be that these respondents did not take certain complexities into account in their initial response. In order to be able to address the issues playing a role in couples' views towards couple-based ECS, more (qualitative) research is necessary.
We are aware that there may be a gap between intended and actual behaviour 25,26 and that our study examined a hypothetical rather than an actual offer. A further limitation is our high drop-out rate: we had complete data from only 74 couples (30%) of the 504 recruited respondents. Another limitation is an overrepresentation of initial responses from female participants. More women (70%) than men (30%) were willing to participate and able to include their partner.
Most couples are therefore characterized by a female respondent and a male partner. Our sample is further characterized by an overrepresentation of highly educated participants (40%). The sample is also A larger score represents a larger change, regardless of the direction. Scale range is 0 to 5 for changes in attitude and change in intention, and 0 to 3 for change in perceived difficulty.
*P < 0.05. **P < 0.01. characterized by couples with a long-lasting relationship who reported their relationship to be good, presumably the couples who are more likely to be actively considering having children.
To summarize, our study sheds light on how participating eligible couples view a couple-based ECS test for serious AR diseases and their intentions to have such a test together if it were to be offered.
Given that the offer of couple-based ECS testing is aimed at enhancing reproductive choice for couples 1,27 and that the test result impacts both partners, one would need couples to reach a joint decision whether or not to participate in testing. In view of the positive results reported here, the next obvious step is to study joint decisions when couple-based ECS is actually offered in a health care setting.