This feasibility study demonstrates that recruitment to a future randomised controlled trial investigating the efficacy of a film-based educational intervention in reducing the risk of acquiring CMV infection in pregnancy would be feasible and has generated essential data upon which to design and power a larger RCT. This single-centre randomised controlled trial has shown that digital antenatal education about CMV is acceptable and accessible to pregnant women and does increase knowledge about CMV, change attitudes towards personal susceptibility and severity, and that pregnant women were willing to adopt risk-reducing behaviour change to reduce exposure to saliva and urine of young children. A future large multi-centre randomised controlled trial would be needed to determine whether such changes in knowledge, attitudes and behaviour would have an impact on seroconversion in pregnancy and therefore prevention of congenital CMV.
In this feasibility study, we have been able to identify factors which would be crucial to the design of a multi-centre randomised controlled trial. To determine the efficacy of an educational intervention, it is necessary to identify and enrol seronegative women in order to demonstrate seroconversion – thus acquisition of infection. We have shown that testing for CMV serology is highly acceptable to pregnant women in the first trimester of pregnancy; 2.86% (n = 144) of women declined testing for CMV antibodies, suggesting that the vast majority of women would be willing to be screened for CMV infection in pregnancy, in the NHS setting, and is consistent with that reported in other studies .
Multiparous seronegative women who have young children are at the highest risk of acquiring infection and transmitting this to their fetus, therefore these women would be the target population for future studies. We have demonstrated the challenges in identifying and enrolling this target population. A large number of women were ineligible for the study (n = 2320; 58.4%) because they were primiparous (this was their first pregnancy) and of those who were multiparous, a further 431 women were excluded because they did not have a child < 4 years of age. Together with the women who were ineligible for other reasons (n = 202) or for whom no sample was obtained (n = 13), only 878 (22%) of the 3975 women approached had a blood sample for CMV screening obtained. These factors are critical to take into account when designing and assessing the feasibility of future studies.
Of the women who consented for CMV screening, 43% were seronegative and therefore at risk of primary CMV infection and eligible for the study, and 57% of women were seropositive. The proportion of women who were seropositive varied considerably with ethnicity. The seropositivity in white women of 39% is similar to that seen in previous studies (45.9% Tookey, 1992; 49% Pembrey, 2013) [21, 22]. However, we found lower seropositivity in women from South Asian ethnicity (78%) compared to that seen in the cohort of pregnant women in Bradford (89%—98%)  and higher seropositivity in black women (94%) than has been observed in a population of women attending antenatal care in London in the 1990s (77%) . Both Tookey et al. and Pembrey et al. found place of birth, as well as ethnicity to be important in seroprevalence, with British born women less likely to be seropositive [21, 22]. We did not collect information about place of birth and so were unable to investigate this aspect. Because of the eligibility requirements of the studies being recruited for, we only screened women living with a child aged less than four years, which may mean that this population is not completely representative of the pregnant population as a whole, but does represent women who are likely to be at the highest risk of infection in pregnancy.
A total of ten women (1.16%) had evidence indicating recent primary CMV infection within the first trimester of pregnancy, this is higher than that observed in an unselected population in a single centre in France (0.42% seroconversion) , but consistent with proportions seen in a population of women in Italy who had a young child or worked with young children (1.2%) . Although this is a small proportion of women, this results in a large number of infants born each year with CMV. Vertical transmission in the first trimester of pregnancy is estimated at 36.8% with nearly 20% of fetuses from these women showing evidence of being affected by CMV . Without interventions to reduce the risk of acquisition of CMV or transmission of CMV, these infants will continue to acquire CMV and a significant proportion of them continue to suffer long term adverse sequelae as a result of congenital CMV infection.
As well as generating essential data to inform a future larger study, we have also been able to describe important differences in knowledge about CMV, perceived severity, susceptibility and CMV risk reducing behaviour of pregnant women in the two study groups before the intervention in early pregnancy and at 34 gestational weeks. By collecting post-intervention data at 34 gestational weeks, we are able to show that these differences were evident even at the end pregnancy, suggesting that women were able to sustain these changes throughout pregnancy.
Before the intervention, most women were unfamiliar with CMV. Previous studies have also shown that only a minority of pregnant women have heard of CMV: 16% in an Australian study, 18% in a Japanese study  and 20% in two separate studies in Singapore , and the US , and that the level of knowledge about CMV is less than for other conditions which affect newborn infants [25, 26, 28, 29]. Despite the fact that CMV is the most common congenital infection in the UK, pregnant women in our study were also less knowledgeable about CMV than other conditions affecting newborns. In our study, 34.7% of women reported being ‘somewhat’ or ‘very’ familiar, a higher proportion than in other studies. This may reflect volunteer bias in which those individuals who are better informed about CMV are more likely to take part in research about it, or it may have been a product of the screening process in which it was necessary to provide some information about CMV in the process of obtaining consent for serological screening.
Participants in the intervention group showed a greater awareness of the ways in which CMV can be transmitted and ways in which congenital CMV can affect children following the intervention, at 34 gestational weeks, compared to those women in the treatment as usual group. This is in agreement with the study by Price et al., who also included change in knowledge as an outcome following an antenatal educational intervention .
The ultimate aim of a CMV educational intervention in pregnancy is not acquisition of facts, but rather to modify behaviours that would place a woman at increased risk of exposure to CMV. In agreement with other studies [12, 20, 23, 30,31,32], we found that an educational intervention in pregnancy was associated with a reduction in the frequency of activities which could expose women to saliva and urine of young children, compared to before the intervention and compared to the treatment as usual group, specifically a reduction in participants eating leftovers from their child’s plate and kissing their child on the lips. These behaviours have previously been identified as being most difficult to change . These changes in reported behaviours may relate to the change in the perception of severity and susceptibility which was seen in the intervention group; change in perception of severity of the condition and an individual’s susceptibility to it has been shown to be an important mediator of behaviour change .
As far as possible, we wanted to have a single intervention early in pregnancy in order to create circumstances as similar as possible to clinical practice, and we therefore provided no reminders to participants about risk reduction, we did not ask them about their behaviours between the first appointment and the questionnaire at 34 weeks and we did not use any objective measures of adherence which is in contrast to some other studies [30, 31, 34]. Whilst all of these measures were important to our ultimate goal of investigating an intervention which would have clinical utility in a routine setting, there are also limitations associated with this approach. Self-reported behaviour may not be the same as actual behaviour, especially when asking participants about their activities over a prolonged period. This may particularly be the case for those behaviours for which there is a perceived ‘right’ answer, for example washing hands after changing a nappy. We were unable to completely simulate real life conditions; in order to screen for the serostatus of potential participants it was necessary to provide some information about CMV which caused many of the participants to seek further information. This may have led to our whole study population being better informed about CMV than the general population and may have limited our ability to detect differences between the groups—although this would have led to an underestimation of the effect of the intervention and if such an intervention were used in routine care there might be an even greater impact on behaviours.
In this study we used a film as our educational intervention that had been designed in partnership with pregnant women and families of affected children. The feedback we received from study participants suggests that this was highly accessible and acceptable to them. Importantly, participants in the intervention group had similar scores on a global measure of distress and on a screening tool designed to identify individuals at risk of perinatal depression compared to those in the treatment as usual group – both pre- and post-intervention.
This study confirmed a finding which has been shown in repeated studies which is that pregnant women want to know about CMV and are often shocked that this has not been discussed with them before [26, 27, 34]. This reinforces the importance of a future large trial to determine the efficacy of an educational intervention to reduce the risk of CMV acquisition in pregnancy and the optimal implementation strategy for CMV antenatal education in routine clinical practice.