Subjects and methods

PROCEDURES

Participating and non-participating couples were recruited from a feasibility study of preconceptional CF carrier couples screening in The Netherlands. Screening was offered to couples who were considering a pregnancy in the future, by five general practices, between May 1997 and June 1998. In total, 5414 people, aged 20 to 35 years, received a letter of invitation signed by their general practitioner (GP). GPs were asked to exclude patients with fertility problems or psychosocial problems from the mailing list to avoid any possible emotional disturbance of these people by the invitation. The letter invited couples, interested in screening, to attend a 45 minute educational session at a nearby location on two evenings in one week. The letters were non-directive, did not encourage couples to participate, and mentioned that this was a scientific study on interest in participation. Both partners were asked to be present. Enclosed with the letter was an information leaflet. The leaflet described the clinical and genetic aspects of CF, carrier prevalence in the population, the implications for the couple of a positive carrier screening test, and how testing is conducted, including information about the imperfect sensitivity of the test (the test sensitivity in this study was approximately 87%). A member of the research group (LH) organised the educational sessions. At the session, attendees were given more detailed information, and an outline of possible advantages and disadvantages of screening was presented. At the end of the session, couples were offered the carrier test, which would be performed by mutation analysis on a mouthwash sample. All of this was offered free of charge. Exclusion criteria for the participation of couples were: pregnancy, a positive family history of CF, and age younger than 18 years. Both partners of participating couples provided a mouthwash sample. If only one partner was able to attend the session, a kit was provided in which the mouthwash sample of the other partner could be collected at home. After the educational session, couples were given an informed consent form to take home and were asked to return it by mail within one week. The Medical Ethical Committee of the academic hospital Vrije Universiteit in Amsterdam approved the study protocol.

UPTAKE

To determine the uptake of screening, a non-response survey by telephone was performed to estimate the percentage of eligible persons in the invited population. Eligibility was defined as having a steady relationship and planning to have one or more children with the partner. A random, age stratified sample of those who did not respond to the initial invitation for screening, that is, did not attend the educational session, was contacted by telephone on different evenings during the week. Other random subjects replaced those who did not answer the telephone after three calls until samples of approximately 10% (n=387) in every practice were reached. Respondents were asked whether they had received the letter of invitation from their GP, whether they were interested in screening, and whether their situation conformed to that of the target population. The non-response telephone survey showed that 19.6% (76/387) of these invited subjects were eligible for participation. In the calculation of uptake rates, the responses of subjects to the invitation were considered, and not the responses of couples, since invitations were sent to individual people and some partners did not receive an invitation because their age group was not included or because they had a different GP. Subjects whose mail was returned by the post office as undeliverable (n=99) were excluded from the study. In total, 108 subjects (related to 79 couples) responded to the invitation for screening. The response therefore was 2% of the total population (108/5315) and 10.4% of the eligible (target) population (108/1042). Seventy eight couples consented to participate in the test (78/79, 98.7% participation) after the educational session.

COMPARISON OF PARTICIPANTS AND NON-PARTICIPANTS

To determine factors influencing the decision to accept the test, differences between consenting couples (participants) and a sample (see below) of eligible couples who did not attend the screening session, identified through the non-response telephone survey (non-participants), were investigated. Data were gathered by means of questionnaires that were identical for both partners and both groups. Questionnaires were administered to all attendants at the beginning of the educational session. Each partner in a couple was asked to complete the questionnaire individually without consulting the other partner. Of the 78 participating couples, 76 completed the questionnaire.

Non-participants who confirmed during the non-response telephone survey a desire to have children with their partner, that is, they were eligible for participation (n=76), were asked whether they and their partner would be willing to complete the questionnaire and return it by mail. Of these 76 non-participants, six did not wish to take part in the study. In total, 76% (53/70) of the non-attending eligible couples returned the questionnaire. Those who did not return the questionnaire (n=17) had previously reported by telephone that their main reason for non-participation was lack of time. Eventually, 76 participating and 53 eligible non-participating couples could be compared.

QUESTIONNAIRE DESIGN

The questionnaire was developed and tested for homogeneity specifically for this study. Subjects were asked, in an open ended question, to indicate the single most important reason for participating or not participating in the screening. In addition, they were asked to report which partner (both partners, man, or woman) had influenced the decision most, and their agreement over the final decision (both agree, partner disagrees, or I disagree). Furthermore, the questionnaire assessed certain components of the Health Belief Model (HBM) including the following:

(1) Sociodemographics

Subjects provided information on age, gender, marital status, highest level of education, number of children, and religiousness. Couples were also asked to indicate whether a pregnancy was planned in the short term (within the next two years) or in the long term (after the next two years).

(2) Familiarity with the disease

Familiarity with the disease was derived from the response to the question: had you heard about the disease cystic fibrosis before receiving the invitation for screening?

(3) Knowledge of the disease

This consisted of seven multiple choice questions, assessing the level of understanding of the medical and genetic aspect of CF and carrier testing. A response of “don't know” was scored as an incorrect answer. The number of questions answered correctly was calculated as a sum score. The answers to the separate questions were also considered.

(4) Health locus of control

The validated subscale of the Multidimensional Health Locus of Control scale with the locus dimension “internal control” was used to indicate the extent to which subjects perceived their behaviour as responsible for their own health (IHLC).24 According to this model, a person is more likely to engage in healthy behaviour if he or she has a strong internal locus of control. The subscale consists of six items, and the answering format was a six point Likert-type scale (completely agree (1) to completely disagree (6)). A total sum score was computed for each subject, with high scores indicating a higher likelihood of engaging in healthy behaviour. Cronbach's α for IHLC on the data was 0.76.

(5) Perceived discomfort

The respondents were asked to indicate the extent to which they agreed that screening requires too much of their time and effort (completely disagree (1) to completely agree (5)).

(6) Constructs of HBM

The questionnaire included 14 items specifically addressing carrier screening, to measure perceptions concerning carrier testing (see Appendix). The newly developed constructs were derived from the four basic HBM dimensions22: (1) perceived risk of being a carrier (couple) and having a child with CF (“Perceived susceptibility”, three items), (2) perceived severity of the disease and the burden of treatment of a child with CF (“Perceived seriousness”, three items), (3) benefits of testing (“Perceived benefits”, five items), and (4) perceived barriers related to screening, such as worries about testing, the perceived impact of carrier status, and the perception that other people will look differently at them when they are identified as carrier (“Perceived impact barriers”, three items). For “Perceived susceptibility”, respondents were asked to indicate the estimated likelihood of their risk on a six point scale (very unlikely (1) to very likely (6)). For all other items, the respondents were asked to indicate the extent to which they agreed with each statement on a five point Likert scale (completely disagree (1) to completely agree (5)). Principal axis factor analysis with varimax rotation was performed to verify that the items loaded on the four factors of the HBM. Subsequently, a reliability analysis was performed on each scale to determine whether all items contributed to the internal consistency of the scale. Summing up items results in a single measure for three subscales with good reliability: “Perceived susceptibility” (Cronbach's α=0.83), “Perceived benefits” (Cronbach's α=0.90), and “Perceived impact barriers” (Cronbach's α=0.64). For “Perceived seriousness”, the items (n=3) were considered separately, owing to the low internal consistency of the total scale.

DATA ANALYSES

To answer the first research question, the answers of couples to the open ended question about the main reason for participating or not participating in screening were coded into general categories. Before trying to answer the second research question, it was necessary to determine whether the data of both partners should be included as a pair, because both the man and the woman of each couple completed a questionnaire. To investigate this, the responses to questionnaires of both partners were compared, using McNemar non-parametric tests for categorical data and pairedt tests for continuous data. Since the sociodemographic status was highly correlated between the partners within a couple, data of only one randomly selected partner were included in the bivariate analyses comparing participants and non-participants. For all the other variables (2) to (6), at first the median split of the sum scores was taken, which resulted in subjects with low scores and subjects with high scores (knowledge low (0-3), high (4-7); IHLC low (6-23), high (24-36); susceptibility low (3-9), high (10-18); benefits low (5-19), high (20-25); impact barriers low (3-6), high (7-14), and separate items (discomfort and seriousness) low (1-3), high (4-5)). Paired data analyses showed that in the responses to these variables, there were moderate to low associations between partners in a couple, indicating that both partners provide different information. Therefore, it was not possible randomly to select one partner of the couple for the analyses. In addition, since the attendance of a couple at the educational session and participation in screening requires a joint decision, it was assumed that the association between a given variable and participation in screening would be stronger when partners have similar attitudes. To quantify the association between the scores of both partners of a couple, a concordance score was formed on a three point scale: 2=“High-High” (both partners in a couple scoring high on the variable), 1=“High-Low” (one partner scoring high (male or female) and one partner scoring low (male or female), and 0=“Low-Low”' (both partners in a couple scoring low on the variable). Subsequently, bivariate analyses were performed to examine the association between participation in screening and the concordance variables (2) to (6). Chi-square analysis and Fisher's exact tests were used to compare responses between participating and non-participating couples. Finally, all variables (1) to (6) that showed statistical significance in bivariate analyses were entered into one multiple logistic regression model simultaneously. All analyses were performed using SPSS for Windows.25 A p value of <0.05 was considered to indicate statistical significance.

Results

MAIN REASON GIVEN FOR PARTICIPATION AND NON-PARTICIPATION

Among the 76 couples who decided to attend the educational session and consented to CF carrier testing, the decision was taken jointly in 85%, while 15% reported that the woman had more influence on the decision to participate. Nevertheless, none of the couples reported disagreement on the final decision. The main reason given by couples for taking the test was that they wanted to know whether they were a carrier couple with a high risk of having a child with CF (97%). The other couples (3%) gave no specific reasons for taking the test. The reasons non-participants gave (in their own words) for not responding to the invitation for screening varied, but the most commonly stated reason for not attending the educational session was “lack of time” or “forgot to attend the educational session” (53%, n=28). Other reasons given were that “the test results would not influence our attitudes towards family planning” (21%, n=11). Furthermore, 15% of the couples were not interested in testing because they “were not concerned” (n=6) or they had “never heard of CF” (n=2). Five percent reported “testing would make us too anxious”. The answers of most partners in a couple could be placed in the same category. However, in three couples (6%), the partners showed disagreement; the women reported that they were interested in screening but they declined the test because their partners were reluctant to participate.

FACTORS ASSOCIATED WITH PARTICIPATION

Sociodemographics of participants and non-participants, both men and women, are shown in table 1. Within the couples included in the study, no statistically significant differences were found between men and women with regard to educational level, time period in which children were planned, and religiousness. The men were significantly older than the women in both participating and non-participating couples. In general, within a relationship, men are older than women in The Netherlands.26 In bivariate analyses, no associations could be shown between participation and the variables age, marital status, level of education, number of children, the time period for planning children, or religiousness.

While the sociodemographics for both men and women within the couples were highly comparable, paired data analyses showed that in the responses to the other variables, such as familiarity with CF, knowledge of CF, and perceptions with regard to carrier screening, there were differences between partners. Therefore, data of both partners were included in the subsequent analyses. Table 2 shows the bivariate associations between components of the Health Belief Model and participation in screening for the three different concordance groups (“High-High”, “High-Low”, “Low-Low”). Participating couples scored higher than non-participating couples on the knowledge questionnaire, and higher on the internal health locus of control (IHLC) scale and perceived higher benefits of testing. In addition, participating couples were more likely than non-participating couples to perceive low discomfort of screening and low impact barriers to screening. No associations were found between participation in screening and familiarity with CF, perceived susceptibility, or the items on perceived seriousness. Furthermore, no differences in responses were found between the five general practices that participated in offering the screening.

The relevant components found in the bivariate analyses were entered into one multiple logistic model simultaneously. Coefficients, odds ratios (OR), 95% confidence intervals (95% CI), and p values for this model are presented in table 3. Perceived discomfort and perceived benefits appeared to be the strongest predictors for participation in screening. Couples in which both partners perceived low discomfort of screening were more likely to participate than couples in which at least one partner perceived high discomfort of screening. Compared to couples perceiving low benefits from screening, couples in which one or both partners perceived high benefits were more likely to participate. Couples in which both partners perceived low impact barriers were more likely to have the test than couples in which both partners perceived high barriers. Couples with high knowledge scores were more likely than couples with low knowledge scores to participate in screening. Finally, couples with one partner scoring high on the internal health locus of control scale were more likely to participate than couples with low scores.

ATTITUDES OF COUPLES WITH DIFFERENT REASONS FOR NON-PARTICIPATION

Over half of the couples who did not attend the educational session (53%) reported that the main reason for non-participation was that they had no time or had forgotten to attend the educational sessions. These couples may differ in their attitude towards testing from the other non-participating couples. To analyse this, couples who did not attend were divided into two subgroups: (1) those who reported that they had no time or had forgotten to attend, but were possibly interested in screening (“lack of time”, n=28), and (2) those who did not attend for other reasons (“other reasons”, n=25). Bivariate analyses of the variables (1) to (6) showed that only one variable was associated with reporting “lack of time”. In non-participating couples reporting “lack of time”, it was more likely that both partners perceived high benefits of screening than in non-participating couples reporting “other reasons” (50%v 16%) (OR=8.7 (2.0-37.7)). In addition, two-thirds of these couples perceiving high benefits and reporting “lack of time” had high knowledge scores.

Of all non-participants, 76% supported the view that genetic carrier testing for CF should be offered to all couples planning to have children, 15% did not favour this opinion, and 9% were not sure. Among those who reported “lack of time”, only two respondents were opposed to offering screening in the general population.

PARTICIPANT AND NON-PARTICIPANT KNOWLEDGE

With regard to more detailed knowledge of CF, a comparison was made of correct answers of participants (before the educational session) and non-participants to five multiple choice questions on the knowledge questionnaire. The answers to these questions could all have been found in the information leaflet, which was sent with the letter of invitation for screening. Participants were significantly more likely than non-participants to be aware of the most important symptoms of CF (64% v 32%), to know that carriers do not need to have a family history of CF (83%v 69%), to recall the risk of being a CF carrier (61% v 39%), and to understand that carriers of CF would not develop CF related health problems (66%v 41%). There was no association between knowledge scores and level of education. In both groups, women scored higher on the knowledge questionnaire than men. In addition, more women than men had previously heard of CF.

Discussion

In this study, factors associated with participation in preconceptional carrier couple screening were assessed. The results suggest that couples who participated in the CF carrier screening programme, as opposed to those who did not, perceived lower discomfort from screening, perceived higher benefits, perceived lower impact of the consequences of screening, knew more about CF, and perceived their own behaviour to be responsible for their health. Overall, the results were more pronounced when partners had similar perceptions. The main reason given for not participating in screening was lack of time to attend the educational session. The results showed that couples who lacked the time to attend perceived more benefits from screening than couples reporting “other reasons” for non-participation.

As expected, higher perceived benefits of the test were associated with more participation in screening. Benefits from testing reflect the importance couples attach to knowing their chances of having a child with CF, which was also found to be the most frequently reported reason for participation in screening. It also reflects the expectations of the couples that they will have more reassurance and that the test results will help them to make decisions about having children. Higher awareness of the benefits of screening may increase the acceptance and participation of couples in screening. In 34% of the non-participating couples, the perception of the discomfort of screening was high for at least one partner. These couples will probably only participate if screening takes less of their time and does not require an extra visit. The results also suggest that those who perceive low impact barriers, such as the impact of carriership on their general health status and the perception that other people will look differently at them, are more likely to accept screening. Only 41% of all non-participants stated correctly that carriers of CF would not develop health problems resulting from their carrier status. These results suggest that anxiety about screening and the burden of carrier status are partly based on misconceptions owing to lack of knowledge.

Knowledge of CF was also a predictor for participation in screening, although this is probably more a reflection of the interest couples have in screening and reading the leaflet before attending the educational session than knowing about CF before the invitation for screening. Nevertheless, inadequate understanding could lead to increased anxiety, and therefore result in non-participation. Bernhardtet al 27 stated that better baseline understanding of human genetics in the public might provide a basis for understanding genetic screening tests and increase interest in learning about screening tests.

The percentage of respondents who were familiar with CF is comparable with the percentages in the Dutch population in general, in which 60% know about the disease and 15% know somebody with CF.28Familiarity with the disease was not associated with participation. This could be explained by the fact that people may have heard about CF, but do not know the clinical implications or the hereditary pattern of the disease. Possibly therefore, this variable was not a motivating factor in participation.

Studies have shown that people are more likely to participate in screening if they consider themselves susceptible to being a carrier or to having an affected child.4 12 29 30 This is in contrast with the finding of the present study, in which no association was found between perceived risk and participation. Most respondents thought it very unlikely that they would be a carrier or that they would have a child with CF. The percentage of respondents knowing the carrier risk was 61% in the participants and 39% in the non-participants. Again, this might imply that non-participation is partly based on lack of knowledge and the assumption that CF is not very common. The question remains as to whether a more accurate understanding of actual risk might have led to a different perception of risk, which would influence participation. A previous study has shown that perceived risk, rather than actual risk, influences the participation of women in prenatal screening tests.31Presumably, couples participate because they want to be reassured that their risk is low, and not because there is a chance of their being carriers or having a child with CF. This is supported by the findings of Loader et al 12 that the desire for reassurance of a low risk of having a child with CF was mentioned twice as often (50.6%) as the intention to avoid having such a child (27.8%).

Other studies4 29 have also reported the lack of an association between perceived seriousness of the disease itself and participation in screening. In the present study, the burden of the disease and the burden of a child with CF was perceived as high for most respondents, whereas the impact of treatment was perceived as moderate to low. This suggests that respondents perceive CF as a very serious disease, but think they can cope with a sick child in practice. In the present study, no influence of the level of education on participation was found, which is in contrast with the findings of other studies.4 12 13 21 29 Furthermore, already having children was not found to be a reason to decline screening, suggesting that the couples understood this aspect of inheriting CF. Overall, 45% of the couples were married, as expected from this age group in the Dutch population.25

The effect of factors predicting participation in screening was more pronounced when couples shared their views. To our knowledge, only one other study29 described the agreement of husband and wife in predicting participation in a carrier screening programme. In this study, it was found that the combined beliefs of couples increased precision in the prediction of who will participate in screening for Tay-Sachs disease.

In the present study it was stressed that both partners should attend the educational session to receive the same information, on which subsequent decision making could be based. Of all participating couples, 15% reported that the decision to accept screening was merely the woman's decision, whereas 6% declined because the male partner did not wish to participate. The results also showed that women had heard about CF more often than men and knew more about the disease. Overall, these results are consistent with the findings of other studies,3 5-7 21 32 that is, that interest in testing is greater among women than among men. This may reflect the greater concern and responsibility in reproductive decisions felt by women. Future analyses of additional data will be performed to determine whether gender differences in the response to the test results can be found, as has been reported in other studies.33 34

Lack of time was the most frequently reported reason for non-participation. The invitation to attend the educational session on two evenings in one week was sent 14 days in advance. This might have been too short for people who already had other appointments on these evenings. Moreover, it has been found that when screening is offered a second time, on more evenings, or when people can make their own appointment, for example with their own GP, participation increases (Henneman et al, in preparation). Although not reported, another reason for non-participation might be that couples prefer personal counselling instead of attending a general educational session with other couples. In a study carried out by Clayton et al,6 non-pregnant couples showed lack of interest in carrier screening. Reluctance to participate was ascribed to worries about factors such as insurability, being at risk, what they would need to learn, abortion, and religious beliefs. However, the authors believe that lack of interest in that study might also be explained by the mode of invitation, letters placed in pockets on signs and not personally addressed to those who were offered screening. Mode of invitation has been found to be the most important factor influencing acceptance.2-4 Invitations that were more personal might have increased interest in that specific screening programme. In prenatal carrier screening, unwillingness to terminate an affected pregnancy was mainly found to be the most frequently reported reason to decline screening,9 12 35although lack of time has also been reported.14

The uptake of the pre-educational screening session of couples who planned to have children was 10.4% in the present study. This uptake is not high, compared with other screening programmes offered before pregnancy2 3 5 and during pregnancy.9 10 13 36 However, the authors are of the opinion that uptake rate is not the most important determinant, if not for economic reasons, of the desirability of screening. Yet, knowledge about motives and barriers for participation is important. As Marteau37 also emphasised, one of the key research questions for the next 10 years should be to find the best way of offering tests to achieve informed choice. In the present study, the main reason couples gave for participation in the screening was to find out whether they were at high risk of having a child with CF. None of the couples reported that they participated in the screening because they were told to do so by their GP or that they felt that they could not refuse. These results suggest that because of the time and effort needed for participation, couples were stimulated to make a decision based on the conviction that screening is important, and not just because it is offered or strongly recommended, as was argued for opportunistic screening programmes with high uptake rates.36 38

For this study, new scales based on the Health Belief Model (HBM) were developed to evaluate the response of couples to CF carrier screening. Several other components of the HBM, that were not addressed in the present study for practical reasons, may also be important, such as interpersonal interactions and mass media communications,22 and the influence of psychological defence mechanisms, such as avoidance behaviour.30 There has also been more general criticism of the model, like other rational choice theories, that is, that it provides an idealised view of how decisions should be made and that it gives insufficient attention to emotion in decision making and the role of cultural standards and values.39 However, the HBM provided more insight into the motives and barriers reported by couples who were offered carrier screening. In addition to HBM related variables, other variables are also considered to be important in affecting participation in carrier screening, such as tolerance of test uncertainty.4

This article specifically addressed the participation in preconceptional CF carrier screening in The Netherlands. Differences between other screening programmes and other countries will exist, but many similarities are evident, and the results of this study can be used in the development of other programmes. Preconceptional screening was chosen because it provides a maximum number of reproductive options for identified carrier couples and involves a minimum of (time) constraint. Furthermore, there are three other reasons why preconceptional screening is highly applicable in The Netherlands. Firstly, prenatal screening is difficult for practical reasons, because many pregnant women visit a clinic late in their pregnancy. Secondly, preconceptional screening meets the requirements formulated by the Committee for Genetic Screening of the Dutch Health Council, whereas prenatal screening does not.40 Thirdly, there are a large number of planned pregnancies in The Netherlands (85-90%), creating an ideal situation for contacting couples before conception. These reasons may also be valid for other countries. In addition, preconceptional screening can be considered for a number of other reasons, as has recently been suggested for Tay-Sachs disease41 and haemoglobin disorders.42 43 For example, in the UK, prenatal screening for haemoglobin disorders is recommended as a routine practice,44 although it has been shown that the present practice does not always meet the needs for early information and leads to the late recognition of risk.45 46 Moreover, at the beginning of a thalassaemia screening programme43in Sardinia in the mid 1970s, the largest category of participants consisted of pregnant women, whereas the number of couples without a pregnancy is currently increasing. In other countries, however, cultural differences influence the approach to carrier screening.47 For example, screening is offered premaritally to communities of Orthodox Jews, to prevent the marriage of two carriers of recessive disorders.48

In this study, interest in preconceptional CF carrier screening, requiring time and effort to participate, is shown, both by the actual participation of couples and by the perceived benefits of screening of couples who did not participate. The results of the study could be used as a model for other screening programmes. Lack of time to attend the educational session was found to be the main influence on participation in carrier screening by non-pregnant couples. However, the results suggest that these couples have positive attitudes towards screening and will attend when screening is offered more conveniently. The results also indicate that participation is influenced by “psychological” barriers, possibly caused by an inadequate understanding of the consequences of carrier testing. These barriers could be removed by providing more clear information about the meaning of carrier status and by increasing public awareness. In addition to participation in screening, the psychological and cognitive consequences of screening are being investigated.