Antenatal screening for cystic fibrosis: a trial of the couple modelBMJ 1994; 308 doi: https://doi.org/10.1136/bmj.308.6942.1459 (Published 04 June 1994) Cite this as: BMJ 1994;308:1459
- J Livingstone,
- R A Axton,
- A Gilfillan,
- M Mennie,
- M Compton,
- W A Liston,
- A A Calder,
- A J Gordon,
- D H J Brock
- Human Genetics Unit, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU
- Department of Obstetrics and Gynaecology, University of Edinburgh, Edinburgh EH3 9EW
- Correspondence to: Professor Brock.
Objective : To assess the delivery and acceptability of antenatal couple screening for cystic fibrosis. Carrier status was notified only when both members of a partnership had cystic fibrosis alleles and therefore a one in four risk of having an affected child.
Design: Mouthwash samples were tested when both partners participated. Results were returned only to positive couples.
Setting : Two large maternity hospitals in Edinburgh.
Subjects : Screening was offered to all couples who booked at one of the two hospitals.
Main outcome measures :(a) The take up of screening, carriers and carrier couples identified, take up of prenatal diagnosis, and numbers of affected fetuses detected; (b) questionnaire measures of patient satisfaction and stress.
Results : Screening was offered to 8536 couples. 714 (8.4%) were regarded as ineligible, usually because of late booking or absence of a partner. 1900 (24.3%) of the remainder declined screening. Among the 5922 screened couples, four tested positive - that is, both partners were cystic fibrosis heterozygotes. All four elected to have prenatal diagnosis. There were three terminations of pregnancy because of an affected fetus, one couple having two successive pregnancies with affected fetuses. The participation rate was 76% for eligible couples (5922/7822) and 69% for all couples (5922/8536). Only 89 screened couples (1.5%) requested information on individual carrier status. No anxiety was detected among a cohort of the screened population, and 99% of questioned participants expressed satisfaction with the concept of couple screening.
Conclusions : Antenatal couple screening is a20satisfactory and acceptable way of screening for cystic fibrosis and has been adopted as routine in the two trial hospitals.
The participation rate in an antenatal couple screening trial for cystic fibrosis is about 69%
Some 24% of couples may decline the offer of screening
All positive couples may be expected to accept the offer of prenatal diagnosis
Around 1.5% of couples may request information on individual carrier status
In this series no anxiety was detected among a representative cohort of the screened population
Over the past three years we have carried out a pilot trial of cystic fibrosis heterozygote screening in the antenatal clinics of a large Edinburgh maternity hospital.1 We used a “two step” mode of testing. Firstly, women were screened for six mutant alleles, representing about 85% of cystic fibrosis chromosomes. If they were negative for these alleles no further action was taken. If they were positive, then their male partners were also tested. When both members of the partnership tested positive the couple was referred to a consultant obstetrician for discussion of possible prenatal diagnosis.
The participation rate in the trial was high and there seemed to be no long term adverse psychological consequences among identified carriers of the disorder.*RF 1-3* However, one in 26 women carried a mutant cystic fibrosis allele and needed counselling. Though most were reassured when their partner was found to be negative, anxiety levels were high as they awaited their partner's test result. It was necessary to have a trained genetics nurse in attendance at all times.
An alternative method of screening for recessively inherited disorders was proposed by Wald.4 He suggested that the couple should be considered as the screened unit and be regarded as positive only when each partner carried a cystic fibrosis allele. When one partner tested positive and the other did not they would be treated as “couple negative” and be given the same residual risk of an affected child as couples in whom only the woman had been tested and found to be negative (fig 1). In this scheme only about one in 676 couples would need counselling and further investigation.
We have carried out trials of couple screening for cystic fibrosis in the antenatal clinics of two maternity hospitals.
Patients and methods
When the booking clinic appointment was sent out the letter included a leaflet explaining the main features of cystic fibrosis and the nature of the trial and inviting participation.5 Two universal containers were enclosed. Couples were told that they would be tested only if both members agreed to take part and that the return of a mouthwash sample from each (together with signed consent forms) indicated their willingness to be screened. The leaflet stated that if they had heard nothing from the trial coordinator 10 days after giving samples they were not a high risk couple. It was emphasised that screening would reduce but not abolish the chance of a child with cystic fibrosis.
The average booking time at both hospitals was 12 weeks, giving most couples about three weeks to consider the invitation. At the clinic samples and consent forms were collected by a midwife. If a woman was not accompanied by her partner and had forgotten to bring his mouthwash sample she was asked to take the containers home so that they could be sent to the laboratory by post. No testing was done unless a sample from each partner was available.
Counselling and reporting of results
Additional information on the screening trial was provided on request by midwives or a genetics nurse in the clinic. Couples were not encouraged to ask for individual results, but if they did these were freely available. When both partners tested positive they were notified by telephone and referred to their consultant for counselling and possible prenatal diagnosis.
All couples entering the trial were asked to complete a 12 item general health questionnaire before testing (GHQ0), 10 days after being tested (GHQ1), six weeks after being tested (GHQ2), and six weeks after their baby was born (GHQ3). This study was discontinued when 300 completed sets of replies were received from each sex. In a separate study participants as well as non-participants were asked a set of simple questions on their attitude to couple screening. This was discontinued after 100 replies were received from nonparticipants.
We used our own assay7 and the Cellmark system,8 which together detect the ΔF508, G551D, G542X, R553X, Δ1507, and 621 + 1G-T mutations, making up 85% of mutant alleles in our population.9 Analyses were first carried out on the female sample, and only if it was positive (or in the event of sample failure) was the male sample tested. Results were obtained on all samples within the specified 10 days.
Couple screening was introduced into the Eastern General Hospital, Edinburgh, in March 1992 and replaced two step screening in the Simpson Memorial Maternity Pavilion in January 1993. As no significant differences in the two hospitals were seen table I gives the combined results to January 1994.
Screening was not offered to 370 couples in whom the woman had tested negative in a previous pregnancy. It was offered but the couple was considered ineligible when booking was after 18 weeks' gestation (470) or when only one sample was provided (234). No containment facilities were available for samples from 10 subjects who were HIV positive. In the Eastern, 26% of eligible couples declined testing, and in the Simpson 21% declined. Screening was done on 76% of eligible couples (5922/7822) and 69% of those (5922/8536) to whom the offer was originally made. In 10% of screened couples the samples were sent in by post.
There were 238 heterozygotes among the women screened, made up of 208 with ΔF508, 16 with G542X, 11 with G551D, two with 621 + 1G-T, and one with R553X. Four women with ΔF508 mutations had heterozygous partners, each also being ΔF508. The four couples opted for prenatal diagnosis and two were found to be carrying affected fetuses (homozygous ΔF508). Both pregnancies were terminated. One couple embarked on another pregnancy during the trial; that fetus also was affected and the pregnancy terminated. All three diagnoses were confirmed in abortus material.
Eighty nine of the 5922 (1.5%) screened couples requested information on individual carrier status. In this group were two carrier women.
Results for the first 300 couples who completed the general health questionnaire at all four points are shown in figure 2. There was no indication of an increase in respondents with positive scores (3 or more) after testing and during the pregnancy (GHQ1 and GHQ2), though anxiety levels rose when parents returned home (GHQ3). The data were very similar to those reported for non-carriers in the two step trial.1,2
Answers to the question on the preferred time of screening are shown in table II. This study was stopped and the responses analysed after 100 had been received from non-participants.
Two studies have shown that the extent of voluntary participation in cystic fibrosis heterozygote screening trials depends on the way in which the programme is presented. Watson et al reported that when the offer of screening was made by a personal approach at either general practice surgeries or family planning clinics take up was between 76% and 87% but that when it was made by a standard invitation letter take up was only 10%.10 This was confirmed by Bekker et al, who found that the direct suggestion by a research worker that a person should immediately be screened was the only effective approach.11 The participation rate was 70%. Other forms of recruitment had very low take up.
In the completed two step antenatal screening trial in the Simpson the take up rate was 83% among eligible women and 71% among all women.1 However, as we had found that nearly half the women agreeing to screening did so without reference to their partners,12 we assumed that couple screening, needing a joint decision between partners, might have a low response rate. This was not the case. The take up rates of 76% for eligible women and 69% for all women were closely comparable to those of two step screening. In the Simpson the changeover from two step to couple screening has led to no decline in the participation rate. If the take up rate continues at or near 70% it should be possible to reach 50% of at risk couples (0.7x0.85x 0.85). It is noteworthy that all four heterozygous couples in this trial and all six in the two step trial2 opted for prenatal diagnosis.
Anxiety and timing of screening
As measured by the general health questionnaire, couple screening appeared to cause little anxiety in participants. In the two step trial1,2 and in other studies in pregnancy13 a fairly high proportion of women presenting to booking clinics had positive general health questionnaire scores before any procedures were carried out (GHQ0). This proportion declined at the two other measurement points during pregnancy (GHQ1 and GHQ2) but rose again, presumably because of the stress of taking a new baby home (GHQ3). The positive responses among both women and men (fig 2) were almost identical with the proportions seen among non-carriers and male controls in the two step trial.2 However, the sharp peak in anxiety affecting over half of the carrier women awaiting their partner's results in two step screening was conspicuously absent in this trial.
There were initial worries that not giving any results at all to most participants might have generated some anxiety. However, only 1.5% of couples asked for information on their individual carrier status. Two identified carriers in this group were satisfied when they learnt that their partners' results were negative.
A substantial proportion of a cohort of women who entered the trial, as well as those who chose not to enter, said that screening during pregnancy was their preferred option. Very similar numbers of women suggested that screening before pregnancy was desirable, provided that it was done with their partner. We found little support for the idea of individual carrier testing (table II). This is powerful support for the concept of the couple as the screening unit. Several respondents asked to write comments on the timing of screening remarked that discovering their cystic fibrosis carrier status before forming a relationship might have inhibited partner bonding or prevented marriage. This aspect of screening in primary care has not been fully investigated.10,11
An aspect of couple screening that may be worrying is whether it is ethical to set the programme up in such a way that people are usually unaware of their carrier status. Our prime reason for doing this was to avoid causing anxiety in couples in whom one partner is shown to be a carrier while the other is test negative. Such couples have a residual chance of an affected child of one in 640. There are currently no practical steps which can be taken to reduce this risk.
In the two step trial these positive-negative couples usually needed two counselling sessions with a trained genetics nurse, and many remarked that they would prefer to have been screened simultaneously and to have been informed only if their risk was one in four.14 Theoretically, prepregnancy screening permits some reproductive choice in that couples could decide not to have children, to resort to artificial insemination by a screened donor, or even to change partners. In practice, experience from other screening programmes shows that these options are seldom pursued.15,16 That so few couples in our trial sought information on their individual status supports this observation.
MacIntyre and Sooman raised the question of the impact of illegitimacy on prenatal carrier screening.17 In a review of reports they cited possible non-paternity rates of 1.4% to 30%. However, they included data from several unreferenced sources which preceded the precision of DNA based systems. Two studies using molecular genetic techniques have shown non-paternity rates of 1.1% (Scotland)18 and 2.8% (France).19 Both have been criticised on the ground that they were based on families attending genetic counselling clinics for estimation of recurrence risks, where any doubts about paternity might have led to self exclusion. This objection cannot be levelled at our broadly based antenatal carrier screening trials, in which information leaflets have exhorted couples with doubts about the identity of the father not to take part. The proportions of couples citing this as the reason for non-participation were 1.4% in the two step trial2 and 2.7% in this trial.
A possible difficulty with couple screening is that it reduces the opportunity for heterozygote testing in the immediate families of identified carriers. In the two step trial about 10% of carriers referred first degree relatives to the genetic counselling clinic for testing.1 Super et al have been enthusiastic advocates of this type of “cascade” testing, pointing out that in terms of the ratio of carriers detected to tests performed it is very efficient.20 This is plainly true, but cascade testing is an addition to and not a substitute for population screening.21 In order to reach all those at risk in an equitable way - an accepted criterion of screening22 - it is necessary to find convenient points through which a substantial proportion of the relevant population passes. We suggest that antenatal clinics are suitable turnstiles and that couple screening is an efficient form of delivery.
We thank the medical and nursing staff of the Eastern General Hospital and the Simpson Memorial Maternity Pavilion, Edinburgh, for making this trial possible. The study was funded by grants from the Medical Research Council and the Cystic Fibrosis Trust. We are grateful to Cellmark Diagnostics for reagent kits and to Professor N J Wald for helpful discussion.