Pregnant women's informational needs prior to decisions about prenatal diagnosis for chromosomal anomalies—A Q methodological study

To study pregnant women's subjective viewpoints on what is important when receiving information prior to decision‐making regarding prenatal testing for chromosomal anomalies.


| INTRODUCTION
Prenatal diagnosis (PND), including both diagnostic and screening procedures, for chromosomal anomalies (CA) is offered as part of the public antenatal care in Sweden.Around 70% of all pregnant women undergo a combined ultrasound and biochemistry test (CUB test), 1 which is the primary screening method used in Sweden for CA.Non-Invasive Prenatal testing and invasive testing are usually offered when the CUB test has indicated a high probability of a CA.
According to the laws and guidelines in Sweden, all pregnant women should be offered general information on PND 2 during the first visit at the antenatal care unit. 3The Swedish National Council on Medical Ethics has recommended that the information should include both the medical and the psychosocial aspects of Down syndrome (DS) as well as perspectives of having a child with a CA. 4 Despite this recommendation, the information given usually focuses on the probability assessment of expecting a fetus with a CA, the test methods and risks associated with invasive tests.Often, no or sparse, information about different aspects of DS is given. 57][8][9][10] Research indicates that pregnant women prefer verbal, in-person information from a knowledgeable healthcare professional. 11,12men focus on the risks of the different tests. 13,14Information about DS and the probability of expecting a fetus with DS, is also desired early in pregnancy, prior to PND. 11,15 The perceived quality of life for the child, as well as the impact on parents and family life, are important for the decision-making regarding PND. 13,14The probability of expecting a fetus with a CA in a current or previous pregnancy can affect women's informational needs. 11,16Informational needs may also be dependent on whether the decision to undergo or abstain from PND has already been made when attending the pre-test counseling.Having already made a decision can result in less need of information, whereas being undecided can result in a need for more extensive information. 17Crombag et al. 18 argue that women might use information to justify their intuitive choice rather than basing an informed decision on information received in nondirective pre-test counseling.
Undergoing PND can be seen as a threatening medical condition where the pregnant women may need to cope with a stressful situation and difficult decisions.When dealing with a potentially threatening medical situation, two different psychological coping styles have previously been identified: "monitoring," that is, focusing on and seeking threat-relevant information and "blunting," that is, avoiding threat-relevant information and seeking distraction. 19ping styles have been found to be related to informational needs and information satisfaction in other medical areas. 19,20men with higher levels of knowledge about PND seem to experience not only less decisional conflict and regret concerning decisions regarding screening for CAs, but also a higher level of wellbeing and less anxiety. 21,22This highlights the need for both more and higher quality information.To be able to improve the quality of information given to women (thereby better enabling informed choices regarding PND and decreasing decisional conflict), it is important to know what information they need.The present study aimed to explore pregnant women's subjective viewpoints on what is important to them when receiving information prior to decision-making regarding PND, with a focus on DS.The secondary aim was to investigate if informational needs differ with different sociodemographic backgrounds and coping styles.

| Q methodology
Q methodology is a method developed for studying subjective viewpoints related to a specific area of interest.It is often used to explore highly complex and socially contested concepts. 23The central part of any Q methodological study is the collection of data in the form of Q sorts.By sorting a set of items or statements covering different aspects of the studied issue (i.e.Q set) along a standardized ranking distribution participants create a personal Q sort that reflects their subjective viewpoint.The Q set is sampled from the concourse, which is a collection of knowledge and experience that represents the various opinions and views about the subject of interest. 24ter the collection of Q sorts (i.e.participants' individually sorted Q set), the second key feature of a Q methodological study is the Q factor analysis, which is a by-person factor analysis looking at intercorrelations between the participants' Q sorts. 25,26This is in contrast to the standard use of factor analysis, where data are usually analyzed by variable, statement or trait. 24Factors can be identified through the by-person factor analysis. 25In a Q methodological study, factors are the groups of similarly sorted Q set, that is, a factor is a group of individuals who share a similar viewpoint on the subject (and hence have sorted the statements in the Q set similarly).Guidelines for reporting Q methodological studies have been followed. 26RNBY ET AL.

| Materials-sampling the concourse, Q set design and content
With the question "what is important for you when you receive information about PND?" participants were asked to sort a set of 50 cards with statements (see Table 1 in the Results section) related to information on PND and CAs.Participants assigned each of the statements a ranking in a fixed quasi-normal distribution (Figure 1) on an 11-point scale, ranging from þ5 for statements that are considered "most important" through 0 to −5 for statements considered "least important." A draft for the Q set was sampled from the concourse, based on the current peer-reviewed literature, interviews with 24 pregnant women (as part of another study with focus on women's perspectives of information and decision-making) and written communication with four mothers of children with DS.The statements were reduced and refined through discussions within the research group with clinical experience such as obstetricians, a midwife and a genetic counselor, resulting in the final Q set of 50 statements.The aim was to include a wide range of statements covering different aspects of information about PND, 23 regarding the ways of receiving information and what the information contained.The Q set was not piloted, but was reviewed by one mother with a child with DS.
In addition to the Q set, two questionnaires were used: the Threatening Medical Situations Inventory (TMSI)-questionnaire identifying different psychological coping styles, that is, monitors and blunters 27 and a questionnaire with sociodemographic questions developed specifically for the study.

| Participants, procedure and administering the Q sort
From January 2020-October 2021, pregnant women who called to book their first appointment with their midwife at antenatal care clinics in three Swedish, middle-sized cities (Eskilstuna, Örebro and Uppsala) were asked by the midwife if they were interested in participating in the study.Inclusion criteria were women in early pregnancy, aged 18 years or above and with a good command of the Swedish language.If they consented, they were sent written information together with the Q set and the questionnaires by post.
Eligible participants were only given brief verbal information before receiving the more comprehensive written information with the study material.Many midwives, at many antenatal care facilities helped with the recruitment, making it difficult to fully determine the response rate.
A total of 45 pregnant women were included, completed the Q sort and filled in the questionnaires at home and sent them back by post.Women's timing for filling in the questionnaires and doing the Q sort varied, and some waited until they had received information about PND and some underwent PND before returning the material.Due to the Covid-19 pandemic, it was impossible to conduct interviews with the participants, so qualitative data was derived from written comments that all participants were asked to provide when ranking the Q set.Further, it was difficult to offer an interpreter, which hindered the inclusion of non-Swedish speaking women.

| Analysis
An inductive approach to Q-methodological analysis with exploratory factor analysis was used.Q factor analyses were conducted using the Q dedicated software package PQMethod (version 2.35) with Centroid factor extraction and Varimax factor rotation (Figure 2).
Initially, seven factors were extracted according to the "magic number seven" 25,28 and Watts and Stenner's advice to start with extracting a factor for every six to eight participants. 25After examining the eigenvalues and factor loadings of the seven extracted factors, a three-factor solution was chosen based on the Kaiser-Guttman criterion, factors with two or more significant loadings, and Humphrey's rule. 25e centroid factor extraction was repeated with three factors, followed by a factor rotation with Varimax.Initially, manual flagging of participants' Q sorts loading exclusively and significantly on one of the factors was applied.Since the results were similar to the auto flag option in PQMethod, this was instead used to identify unique loaders, that is, defining Q sorts for each factor in the final analysis.The software used these defining Q sorts to produce factor arrays with an idealized Q sort for each factor that was calculated through weighted averages of the defining Q sorts produced by participants who shared the viewpoint of one factor.Factor arrays were used to produce crib sheets 25 to help in the process of factor interpretation.In-depth factor interpretation was performed using factor arrays, crib sheets, 25 qualitative data from participants' comments on their sorting process and sociodemographic data from the questionnaire.The women's comments gave a more in-depth understanding of their reasoning during the sorting process.Special attention was also placed on distinguishing statements (i.e.statements that were significantly differently ranked between factors) and consensus statements (i.e.statements that did not distinguish between any pair of factors) at a p < 0.01 level.
Intermediate factor analyses were performed at 30, 40, 45 and 50 participants looking at background information to make sure the participant group was heterogeneous and comparing factor loadings and factor arrays.When comparing intermediate analyses at 45 and 50 participants, both background characteristics, factor loadings and factor arrays indicated that saturation was reached at 45 participants.Recruitment was therefore discontinued and in-depth factor interpretation was performed with 45 participants.
The TMSI questionnaire was analyzed by summing up the relevant items for monitoring and blunting for each scenario.The total blunter scores were subtracted from the total monitor scores, giving a total score.Higher scores indicate a monitoring coping style and lower (negative) scores indicate a blunting coping style.

482
- T A B L E 1 Factor array with the 50 statements and the ranking of each statement for the three factors: Factor 1 (F1), Factor 2 (F2), Factor (F3).were compared using the Independent-Samples Kruskal-Wallis test.

No
A significance level of p < 0.05 was used for all comparisons.When comparing women's responses on having had personal experience of someone with a congenital anomaly or not, response alternatives were sorted into "yes" and "no" due to small numbers, where "yes" includes the participants with experience in the close family, extended family and/or among friends.

| RESULTS
In total, 45 pregnant women with diverse sociodemographic backgrounds from three middle-sized cities in Sweden participated.Age ranged from 23 to 38 (mean 29.9) years, almost half of the women were parous and a majority had higher education.  -4 (F3).Since not all participants' Q sorts were defining sorts for any of the factors, Q sorts of 11 participants are not included in the three defined factors.No significant differences between the participants who produced defining Q sorts for the three factors were found when comparing background demographics and coping styles (Table 2).
Table 1 displays a factor array with the ranking of each statement for the three factors, as well as distinguishing and consensus statements.When describing the viewpoint of each factor below, the ranking of statements that inform the interpretation are included in brackets in the text where, for example, (20:þ2) indicates that statement 20 was ranked at þ2.
Seventeen of the statements are consensus statements that do not distinguish between any pair of factors (Table 1).To receive information from someone with sufficient knowledge was ranked as most important from all three viewpoints.To have their thoughts and questions taken seriously was also a priority for the women (31:þ4, þ3, þ4) as was having time to reflect between receiving information and decision-making (33:þ1, þ3, þ2).The opportunity to discuss religious values related to PND was not a priority (14:−5, −5, −5) and neither was verbal information in groups (47:−4, −5, −5).

| Factor 2-Decision-making as a continuous process based on couple autonomy
Factor 2 has an eigenvalue of 2.06 and explains 17% of the study variance.Q-sorts of eight pregnant women exemplify this factor.
The women representing viewpoint 2 strive for an informed decision together with their partner (2:þ4, 8:þ4) regarding the complete concept of accepting or declining screening and how to handle the test result (39:þ3, 3:þ3, 49:þ1).These women wish to receive information about test characteristics, although this is not as highly prioritized as for the women with the other viewpoints

| Factor 3-As much information as early as possible-The importance of personal autonomy in decision-making
Factor 3 has an eigenvalue of 1.85 and explains 17% of the study variance.Q-sorts of 11 pregnant women exemplify this factor.

486
- T A B L E 2 Background socio-demographics and coping styles for all 45 participants and for the participants who produced defining Q sorts for each factor (Factor 1: F1, Factor 2: F2, Factor 3: F3), respectively.

| DISCUSSION
This study aimed to explore pregnant women's subjective viewpoints on what is important when receiving information prior to decision-making regarding PND.Three different viewpoints were found that illustrate the wide range of informational needs.Some This highlights the complexities when providing information to pregnant women.
Receiving updated, fact-based information from a knowledgeable healthcare professional was highly prioritized from all three viewpoints.31]36 This highlights the need to better educate healthcare professionals if the needs of pregnant women are to be met.]34 A woman's probability of a fetal anomaly in a previous or current pregnancy may affect what kind of information she values to receive. 11,16Also, parity and age have been shown to affect information needs. 37However, we found no significant differences in sociodemographic background data between the three viewpoints (Table 2).Moreover, a "monitor" or "blunter" coping style did not seem to affect informational needs.This is interesting as one might expect that "monitors" would like to receive more extensive information (like viewpoint 3), while "blunters" would avoid threatrelevant information (like viewpoint 1). 19,20As this was not the case, women's informational needs seem to be affected by something other than their coping style or sociodemographic background, thus making it difficult to predict a woman's informational needs based on demographics.Neither (total score 0) 1 (2.2) 0 (0.0) 0 (0.0) 0 (0.0) Note: P-values for comparisons of background characteristics between factors, with no significant differences at p < 0.05 level.Comparisons between factors were not analyzed for experience of Down syndrome since there were no Yes-answers in either factor.The numbers on received information and having had PND or not concern the ongoing pregnancy and not previous pregnancies for parous women.

488
- Different individuals have different informational needs, 38 making it difficult to design standardized information that suits every pregnant woman.Information in groups got a low ranking in all viewpoints and is thus not optimal at pre-test counseling.Individual in-person meetings give the healthcare professional a chance to discern both how and what information needs to be provided to this particular woman, so that she can make an informed decision regarding PND.It has been suggested that alternative methods for pre-test counseling should be explored since proper pre-test counseling is resource demanding. 39The use of decisional aids has shown improved decision-relevant knowledge and informed decision-making. 40,41Decisional aids could work as a complement to individual counseling, but should not replace individual counseling as a woman's informational needs cannot be identified in advance and need to be explored during the counseling session.
It has previously been reported that women may use a multistep decision process and that some consider each decision separately, anticipating that they will feel differently at the time of follow-up decisions. 17This is consistent with viewpoint 1's desire for a step-by-step information process, probably because women with this viewpoint also apply a multi-step decision process.Information about conditions screened for is not routinely given prior to testing, 5 which suits these women.However, this stepwise approach, does not meet the informational needs of the women with viewpoints 2 and 3 who also desire information about the effects of conditions screened for.It is not surprising that information about conditions screened for is important to some pregnant women since the perceived quality of life for an individual with a CA and the effect on the child and family seem to affect decision-making regarding PND. 13,14Healthcare professionals need to be aware that this information is important and should be provided if desired.
Previous research indicates that women's decision-making process regarding PND is often influenced by their partners' opinions and many strive for a consensus decision with their partner. 11,12,14,17,42,43The women's need to involve their partner in the process also varied depending on their viewpoints.Time between receiving information and decision-making is needed to have the chance to discuss the decision with their partner.This illustrates the varying needs regarding how to receive information and how to make decisions concerning PND.

| CONCLUSIONS
Pregnant women's informational needs vary, seemingly independent of background characteristics, thus making a standardized information process unsuitable.To enable informed decision-making and to ensure that women's varying informational needs are met personalized information and individual counseling with healthcare professionals trained in pre-test counseling is needed.

| GENERALIZABILITY/STRENGTHS AND WEAKNESSES
We included women with diverse backgrounds, giving a good representation of pregnant women in today's Swedish healthcare system.Therefore, it is likely that similar viewpoints are found among other pregnant women under similar circumstances.
Non-Swedish speaking women were not included in the study because of language barriers, which is a weakness.The Covid-19 pandemic prevented in-person meetings, which might otherwise have facilitated the inclusion of non-Swedish speakers.The pandemic also hindered in-person interviews with the participants, which could have given a deeper understanding of the viewpoints.
A potential limitation is self-selection bias, which is a general problem with this kind of research as those interested in PND may have stronger reasons to participate.However, by including some women who indicated that they would not have PND and some that appeared to be undecided, we think the viewpoints of women not interested in PND have been included.
Comparative analyses between the factors must be interpreted with caution.Due to the small numbers, minor differences may not have been detected.However, comparisons between the viewpoints were not the main aim of the study.Since the Q methodology rarely includes a large number of participants, but rather aims at stopping inclusion once saturation has been reached, large groups will not be attained in this kind of study setup.

F I G U R E 1
The fixed quasi-normal distribution sorting grid that was employed for this study.Participants sorted the 50 statements of the Q set, one statement per square in the grid, to indicate their prioritization of the statements relative to each other.For example, only two statements can be ranked as most important at þ5.F I G U R E 2Overview of the steps of the Q factor analysis process.[Colour figure can be viewed at wileyonlinelibrary.com]TERNBY ET AL.

Factor 1
has an eigenvalue of21.48 and explains 22% of the study variance.Q-sorts of 15 pregnant women exemplify this factor.The women representing viewpoint 1 see information and decision-making as a step-by-step process, where step 1 is focused on information and decision-making regarding PND.Information about conditions screened for is desired in step 2, if screening indicates an increased probability for an anomaly.In a written comment, one woman felt that (prior to a diagnosis) information about conditions could cause unnecessary concern.Compared to the other viewpoints, the women with viewpoint 1 place greater value on being informed that receiving information about PND is optional (12:0).However, they value receiving updated, fact-based (37:þ5) and neutral information (5:þ4), together with their partner (2:þ3) in an individual, separate consultation (6:0, 47:−4, 9:−1).They do not want too much information on a single occasion or too early in the pregnancy (45:−3, 18:þ1, 11:0).They are open to receiving advice from healthcare professionals on making a choice about testing (28:0), but want to make decisions without pressure from healthcare professionals (27:þ3).It seems important to these women that their partner also receives information (2:þ3), though not necessarily involved in the decision-making (8:þ1).Statements concerning information about test methods are highly ranked (7:þ3, 20:þ2, 23:þ2, 26:þ4), while diagnosis-specific information is ranked low, indicating a preference for a step-bystep process (15:−1, 16:0, 22:−2, 24:−2, 32:−3, 34:−4, 35:−2, 43: −5, 48:−4, 50:−2).They do not wish to receive information from patient organizations (46:−3) or about care programs and support from society for an affected child and its family (25:−3).They are interested in information about reproductive choices if a test indicates an anomaly (3:þ3), but not about management and support available during an affected pregnancy (49:0).
prefer a step-by-step process starting with information and decision-making concerning the tests.Others want more extensive information from the start, including taking into account the effects of the conditions screened for and what decisions regarding the pregnancy they may have to make, depending on the test result.
To receive information about how adult life can be like for an individual with down syndrome (medical care, living arrangements, occupation and support from society).that it is not possible to foresee how severe the symptoms will be for a specific individual with down syndrome just by looking at the chromosomes.Distinguishing statements (i.e.statements that were significantly differently ranked between factors) are indicated with F1, F2 and F3 respectively depending on which factor they are a distinguishing statement for (e.g. a statement can distinguish one factor from the others, or all factors from each other).Consensus statements (that do not distinguish between any pair of factors) are also shown.One statement is not distinguishing nor consensus and is referred to as "Neither" in the table.
Three factors were found to represent different viewpoints on what is important to pregnant women when receiving information about PND.Together, they explain 56% of the study variance.Each viewpoint consists of defining Q sorts (i.e.Q sorts of participants who share that viewpoint), with 15 Q sorts in Factor 1 (F1), 8 in Factor 2 (F2) and 11 in Factor 3 T A B L E 1 (Continued) 484 -TERNBY ET AL.