Decision making on antenatal screening results: A comparative Q‐method study of women from two Chinese cities

Abstract Background Although an integral part of ethical and quality health care, little is known about the informed decision making of Chinese women with different socioeconomic backgrounds within the context of antenatal testing. Methods To explore women's viewpoints on informed decision making regarding antenatal screening, a Q‐methodology study that combines both quantitative factor analysis and interviews was conducted between June 2016 and February 2017 in Shanghai and Duyun. A total of 169 women (84 Shanghai and 85 Duyun) participated in the study of 41 ranked statements along a Q‐sorting grid. Results Using by‐person factor analysis, five distinct viewpoints are identified: (a) choice is shared with the partner/husband, but the mother has the right to make the final decision; (b) having antenatal tests is not about choice but about a mother's responsibility; (c) choice is a shared decision led primarily by the partner/husband and secondarily by the doctors; (d) choice should be made using the advice of doctors, but the decision should be made with the partner/husband; and (e) choice is a responsibility shared with the partner, family and doctors. Conclusions The study reveals that women with better education and higher incomes demonstrate more autonomy than those with less education. The nuclear family clearly emerges as the main decision makers in health‐care services in China. Patient and Public Contribution The 169 participants shared their views and stories for at least an hour. They were debriefed after the interviews and contributed their thoughts on our study design and interpretation of the data.


| INTRODUC TI ON
Antenatal screening tests refer to the use of a variety of medical and genomic technologies to help monitor the health conditions of the expectant mother and to check the health status of the foetus during pregnancy. Generally speaking, these tests include (a) an obstetric ultrasound; (b) blood tests; (c) non-invasive prenatal screening technologies, such as maternal serum screening, nuchal translucency ultrasound and cell-free DNA testing; and (d) invasive prenatal diagnostic testing technologies, such as amniocentesis and chorionic villus sampling. Through visualizing technologies and statistical models, antenatal testing calculates and presents the probability of having a child with congenital anomalies, which is factored into their reasoning on whether to keep the foetus and is believed to help relieve their 'anxiety, fears and inner tensions'. 1 In China, the importance of having antenatal tests is mentioned in numerous official guidelines on promoting maternal health care. As a critical integral part of ethical health care, informed choice for antenatal screening has been internationally recognized and discussed. [5][6][7][8][9][10][11][12] Informed choice is based on relevant, high-quality information about the advantages and disadvantages of all possible actions, allowing patients to make autonomous decisions. 13,14 However, this definition has been challenged in recent years. Some argue that the current definition is based only on an interpretation through the lens of biomedical ethics, and it neglects an understanding of the reality that multiple factors will contribute to the different understandings women hold concerning their choices. [15][16][17][18] Informed choice is embedded in the principle of autonomy where the chooser needs to act intentionally, with full understanding of the information and without other influences that affect their understanding. 19 Some researchers argue that individual autonomy is incompatible with authority and cultural value. 20 Beauchamp and Childress, 19 however, point out that there is no full or complete autonomy but substantial autonomy in a practical world. Autonomy is in fact a matter of degree; in order for an action to be substantially autonomous, it must involve intentionality, understanding and voluntariness to an adequate extent. Furthermore, individual autonomy is defined differently across cultural contexts, and women in South Asia, for example, have little autonomy in the decision making of health-care services compared with women in Western countries. 21,22 To make sense of the different understandings of informed choice from a woman's perspective, a series of cross-culture studies have been conducted in recent years. 5,23 Women's understandings of informed choice can be divided into three categories: no autonomy, relational autonomy and individual autonomy. For women with no autonomy, the antenatal test is obligatory due to the responsibility for their foetus, the premise of being a good mother or religious reasons. 18 Women with individual autonomy often make independent decisions, and such an interpretation distinctly reflects the emphasis of individual rights in the West. 5 But, given the increasing emphasis on a shared-decision model, informed choice is increasingly associated with relational autonomy. 24 Relational autonomy means that individuals are embedded in and shaped within the context of social relationships and society. 25 People may not recognize them as independent decision makers. 26 For example, women from South Asia may prefer to follow the doctor's or husband's advice 5,27 and not make decisions themselves. 21,22 Previous research has also shown that Chinese women prefer to make an informed choice with their husband and family and want doctors to provide directive advice. 23,28 Within the Chinese society, an individualistic approach to autonomy would then fail to acknowledge how informed choice is never fully autonomous or overwhelmingly relational.

| Participants
A purposive sampling method was used to recruit subjects. The inclusion criteria were Chinese women living in Shanghai and Duyun who had at least one child aged three years or younger. Recruitment flyers were posted and handed out by local research assistants in public places outside of hospitals, community maternal and child health centres, and other community centres in the two cities. In addition, invitations were sent out via the research team's personal networks. A total of 169 women (84 Shanghai and 85 Duyun) who fit the inclusion criteria and represented diverse backgrounds consented to participate and completed the study (see Table 1 for the demographic characteristics of the participants).

| Materials
Q-methodology was adopted to gauge the women's understanding of informed choice. Q-methodology enables the study of subjectivity. 33 Participants were asked to rank a set of 41 statements on informed choice along a Q-sorting grid (see Figure 1 for the grid and the 41 statements), from the statements they agreed with the most to the least. The 41 statements (Q-set) for this study had been previously developed for a UK research study, 5 which included a Chinese version. To adapt the Q-set to be more appropriate for the Chinese context, the research team compared the Chinese and English version used by the UK study, made minor revisions and pilot-tested the revised Q-set with five Chinese mothers. This Q-set pushes participants to engage and consider the process of informed choice regarding antenatal screening. Each participant's distribution of the statements is called a Q-sort.

| Procedures
The study was conducted between June 2016 and February 2017 in Shanghai and Duyun. Ethical approval was obtained prior to the start of the study from a university ethics committee. Written consent was obtained from all participants before beginning the study. Q-sorts were administrated individually in a location of the participant's choice (health centres, coffee shops or their homes).
Participants were asked to read the statements and rank them from +4 (strongly agree) to −4 (strongly disagree). Then, they were asked to place each statement on the Q-sorting grid. Each participant's sorting result was photographed. Then, a post-sorting interview (audio-taped) was conducted immediately to better understand the reasoning behind each participant's Q-sort distribution. All postsorting interviews were transcribed in verbatim.

| Analysis
Analysis of the Q-sorts was conducted using PQmethod version 2.11.
Factor analysis was used to correlate participants' Q-sorts to identify which participants' Q-sorts clustered together. Factors were extracted using principal component analysis, which maximizes similarities within factors and differences between them. Varimax rotation was used, which rotates factors to ensure that no Q-sort loads significantly at the same level on more than one factor. 34 Q-sorts that were exemplars of each factor were identified, that is only Q-sorts with a loading of ±0.04 (P < .01) on one factor. 34 These exemplar's Q-sorts were merged to create factor arrays (an average score for each item by factor). 35 These factor arrays represent idealized Q-sorts and are interpreted as different viewpoints (see Table 2 for the factor arrays).
Eight factors were originally extracted, with an eigenvalue of 1.00 or more 35 and at least one exemplar. 36 A five-factor solution was reached after inspection of factors 6 to 7 showed that they did not provide distinct viewpoints that were not captured in the other factors. These five factors explain 62% of the total variance.
Interpretations of the five factors were carried out by examining and comparing the factor arrays, with a particular focus on statements in the 'strongly agree' and 'strongly disagree' columns. Statements identified as statistically distinguishing each factor were used to interpret factors by comparing the position of the 'between' factors. In addition, consensus statements on which the levels of agreement/disagreement were similar across factors were also analysed. Transcriptions from the post-sorting interviews were read repeatedly and used to inform, confirm or further clarify the participants' sorting results. Since these interviews were quite short, no specific coding schemes were used.
Initial interpretation was conducted by the first author, and then subsequently discussed with the other authors and continually examined against the qualitative data collected.

| FINDING S
The study participants included 84 women from Shanghai and 85 from Duyun. Most of the participants from Shanghai had higher education levels and family income than those from Duyun. The average age for the five factors was 30 years, with a slight difference between the oldest cohort (factor 1, average 31 years) and the youngest (factor 3, average 27 years) (see Table 3 for details on participants' demographic information for exemplars in the five viewpoints).

| Factor 1: Choice is shared with the partner/ husband, but the mother has the right to make the final decision (21 exemplars)
This viewpoint was held by 21 participants: 16 from Shanghai and 5 from Duyun. Most of these participants had higher educational attainment and income levels compared with participants from the other factors.
In this viewpoint, participants emphasize the mother's autonomy; they believe the decision should be led by the mother (#17, +3), and conversely, they disagree with the husband/partner taking the lead in the process (#39, −1). Nevertheless, in line with a relational approach to autonomy, participants strongly agree that decision making about testing should be shared with the partner/husband (#19, +4), stressing the mutual responsibility of parenthood. Informed choice is then shared with the partner, yet led by the mother.
It was necessary to discuss with him, but since I was the pregnant one, the final decision is mine. If lots of other people are having testing, then testing would be fine by me The decision about these tests is no more difficult to make than routine health tests in pregnancy, such as the mother's blood pressure or diabetes There is no decision for me to make because the tests are just part of good care for pregnant women −1 It is difficult for me to say 'no' to testing when doctors/midwives offer it 0 0

| Factor 2: Having antenatal tests is not about choice but about a mother's responsibility (22 exemplars)
This factor had 22 exemplars: 17 from Duyun and 5 from Shanghai.
Most of the participants from Duyun had a lower education level and lower family income than those from Shanghai. Overall, participants in this factor had the lowest level of income and the second lowest level of education after factor 3.
In this viewpoint, participants emphasize testing as a maternal responsibility rather than a choice.
Participants strongly believe that having the antenatal tests is 'part of good care for pregnant women' (#12, +4) and is 'part of being a good mother' (#26, +4). Testing appears to be a maternal duty rather than a decision, since it fulfils the mother's responsibility for ensuring both the mother's health and the child's health.
Accordingly, the decision about testing is not considered hard (#16, Participants strongly disagree with 'I would accept the child that

| Factor 3: Choice is a shared decision led primarily by the partner/husband and secondarily by the doctors (6 exemplars)
This factor had 6 women: 4 from Duyun and 2 from Shanghai. The participants from Duyun had a lower family income than those from F I G U R E 2 FVisual illustrations of the five factors. Factor 1: Choice is shared with the partner/husband, but the mother has the right to make the final decision (21 exemplars). Note: The mother is in black since she is the final decision maker. All others are in grey because they are not regarded as the decision makers by the mother. Factor 2: Having antenatal tests is not about choice but about a mother's responsibility (22 exemplars). Note: The mother was coloured in grey because she does not think she has choices. Factor 3: Choice is a shared decision led primarily by the partner/husband and secondarily by the doctors (6 exemplars). Factor 4: Choice should be made using the advice of doctors, but the decision should be made together with the partner/husband (11 exemplars). Factor 5: Choice is a responsibility shared with the partner, family and doctors (9 exemplars).
Factor 1: Choice is shared with the partner/husband, but the mother has the right to make the final decision (21 exemplars) Note: The mother is in black since she is the final decision-maker. All others are in grey because they are not regarded as the decision-makers by the mother.

Factor 2: Having antenatal tests is not about choice but about a mother's responsibility (22 exemplars).
Note: The Mother was colored in grey because she does not think she has choices. Shanghai, but their educational levels were the lowest compared with other factors.   (DY048) If your child has a condition… he will be mocked and discriminated against. (DY029)

| Factor 4: Choice should be made using the advice of doctors, but the decision should be made together with the partner/husband (11 exemplars)
This factor had 11 exemplars: 6 from Shanghai and 5 from Duyun.
Participants from Shanghai had higher educational levels and income than those from Duyun.
The emphasis in this viewpoint is on decision making as a process advised by doctors and shared with the partner/husband.  Although they agree about the importance of testing (#35, −3; #36,

(DY037)
Every child is equal, each one is your own. There is nothing to worry about.

(DY039)
He is my child. For me, he is the best even if he is discriminated by whole the society. (DY043)

| Factor 5: Choice is a responsibility shared with the partner, family and doctors (9 exemplars)
This factor had 9 exemplars: 7 from Shanghai and 2 from Duyun.
Participants had the highest average income and educational level, yet participants from Shanghai had higher levels than those from

Duyun.
A distinctive feature of this viewpoint is the shared nature of decision making, where the partner, family members and doctors are all perceived to have different roles. Rather than lacking autonomy, this viewpoint seems to be aware of the responsibility involved, and the participants are thus willing to share that responsibility.
Firstly, this is the only account where parents and siblings would be involved in the decision making and could influence it (#22, +2); in-laws instead would not be so actively involved (#23, −1).
I think the doctor has the right to give advice. I will listen to the doctor (SH062).
Thirdly, similar to viewpoint 3 only, participants agree 'the partner/ husband should make the decision about testing' (#39, +2), thus adopting a more passive role for themselves.
Participants agree that decisions about testing required careful thinking about the effects (#33, +1) and would involve several subjects to guide their choice. They are particularly concerned about being blamed as irresponsible if they were not tested (#34, +1). testing. Yet, participants' accounts range from acknowledging the need for information, through seeking advice, to embracing more directive approaches. Since this research focused on Chinese society specifically, different attitudes towards doctors must be understood and contextualized within the shifting role of health professionals in Chinese control population policies specifically.
Policy implications of this research need to be stated within the context of contemporary China. Starting from the 1980s, the improvement of the quality (suzhi 素质) of the population has been crucial in Chinese modernization project. 38 In an effort to promote a quality population, medical professionals have acted as 'quality inspectors'. 39 They have tended to adopt a directive role, openly promoting antenatal tests and discouraging the birth of a 'defective' child, that would be an economic burden for the family. 34 With the introduction of the two-child policy in 2015, there has been a shift in discourse towards an emphasis on quantity of births. The promotion of antenatal genetic testing is also meant to promote the quality of population. However, the results of the antenatal tests are expressed in terms of risk rather than certainty of diagnosis. To avoid any blame in the case of a wrong diagnosis and shift of responsibility onto parents, health professionals have tended to adopt a less directive approach, aiming to provide information rather than advice, and leave informed choice up to the family.
Our study shows that some Chinese women seem to have em-  38 Entering the antenatal health-care system is equated with women embodying the state discourse of a quality population, hence each individual's responsibility for a healthy birth (yousheng). Some of our participants seem to strongly embrace this discourse; for them, having antenatal tests is rather a mother's responsibility than a choice. Getting tested fulfils the maternal duty of ensuring both the mother's health and child's health. Additionally, women in all the factors acknowledge antenatal testing as an essential tool for informed choice; all would take into consideration termination of a child with a condition, unlike Western populations. 5,40,41 Many of them worry about the discrimination a child with a condition would face, and all would be concerned about the burden such a child would cause to the family. This may relate to the widespread stigma towards disability in Chinese society, the lack of appropriate economic and social support for families, and the consequent challenges they would face. 40 Our findings suggest that without placing enough resources towards disability rights, informed choice in antenatal testing is restricted by societal values of the 'right choice'. Hence, it is important for policymakers to consider improving the rights of people with disabilities and enhancing public awareness of such rights, to provide more social support to families with children with disabilities and to greatly promote barrier-free facilities, so that when the Chinese mothers are making decisions, their decisions are less affected by such concerns.

| LI M ITATI O N S
The statements used in the study were not developed anew but translated from English to Chinese from previous research. When the Chinese version was used during the interview, women with lower education levels had difficulty understanding the statements and researchers had to provide further explanation. Also, the study was only conducted in Duyun and Shanghai, and the conclusions may not be representative of or generalizable to the Chinese population as a whole. In addition, due to the limited time and resources for this research, the participants might not have been diverse enough to represent women with all possible social economic and cultural backgrounds from these two cities. Meanwhile, for the Q-method, our sample size might have been too large to differentiate between the viewpoints. Hence, it raises concerns on noise in the factors, which might have led to misinterpretation.

| CON CLUS ION
This study addressed a series of factors that exert influence on

ACK N OWLED G EM ENTS
The authors would like to thank Ms Zhao Rui for her generous help on data collection.

CO N FLI C T O F I NTE R E S T
All the authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
DD, SA, HY, HJ, YR, MA and JZ made substantial contribution to the conceptualization, design and research protocol for the study. DD and JZ made substantial contributions to data collection and interpretation. DD, EN and JZ wrote the first draft of the manuscript. All authors were involved in revising the article and approved the final manuscript.

E TH I C A L A PPROVA L
The study was approved by the Committee on the Use of Human and Animal Subjects in Teaching and Research of Hong Kong Baptist University, which the first author worked for when the study was conducted. The ethics approval code is FRG1/15-16/057.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.