The majority of countries (83% of 24 respondents) had an official vaccination policy for pregnant women. No replies were received from nine countries and it is possible that response bias results from smaller countries and countries without vaccination policies being less likely to respond. Twenty one countries out of the 27 in the EU, corresponding to 93% of the member population, were covered by the combined responses to both questionnaires.
Some countries had existing advance purchase agreements with pharmaceutical companies which allowed early procurement. A lack of capacity to identify and prioritise key population groups quickly may also have influenced policy decisions. Public perceptions, which can be influenced by certain key events in a population, may be significant. For example, in the Czech Republic, a previous key event involving vaccines contaminated with deadly live H5N1 avian flu virus, might have influenced the decision not to formulate an official vaccination policy [
17].
A larger number of European countries have a vaccination policy for pregnant women against pandemic influenza than against seasonal influenza. Ten out of 27 (37%) European countries recommended vaccination against seasonal influenza for pregnant women during the 2008-9 influenza season: Austria, Belgium, Cyprus, Denmark, Estonia, Ireland, Italy, Portugal, Slovakia and Spain [
18]. None of these countries could provide information on vaccine coverage of pregnant women. All these countries had a pandemic vaccination policy in force during the pandemic influenza, and it is possible that this was facilitated by the previous seasonal flu policy. Both in Canada and the United states, vaccination of pregnant women is recommended during the influenza season, while in Australia, vaccination of pregnant women in the second or third trimester of pregnancy is recommended during the influenza season [
1].
Vaccination policy only partially determines what happens in a country. Data collected by the questionnaires demonstrated that countries with similar vaccination policies achieve very different vaccination rates. For example Italy and the Netherlands had similar vaccination policies but 23,016 women in the second and third trimester of pregnancy were vaccinated in Italy (population ~61 million) and 68,400 in the Netherlands (population ~17 million). The differences in vaccination rates detected between countries with similar vaccination policies might be related to different public perceptions of factors such as the vaccination campaign, vaccine safety and the risk of the influenza pandemic. In addition, inclusiveness and funding of different health systems and implementation issues associated with the delivery of different vaccination policies and associated education campaigns might have been influential. Furthermore, the decision to use vaccination centres instead of general practitioners for vaccination may have lowered the uptake of vaccination in some countries, such as France [
19].
Varying pandemic influenza vaccine uptake rates have been reported for pregnant women. In our survey, three countries reported uptake above 50%. Uptake by pregnant women was reported as 6.9% in an Australian survey [
20], 25.2% in a British survey [
16], 76% in a Canadian survey [
21], 22.7% and 37.1% in French surveys [
9,
12] and 8.8% in a German survey [
15]. Factors associated with vaccine uptake in the Canadian survey were late stage of pregnancy, belief in the efficacy of the vaccine, and consultation of the (Canadian) pandemic influenza website. Factors negatively associated with vaccine uptake in the Canadian survey were belief that the vaccine has not been adequately tested and consultation of popular websites. The Australian survey concluded the main reasons for the low vaccine uptake were the vaccine frequently not being offered by the general practitioner and safety concerns both of the women and the general practitioners. In the British study, vaccination was offered to 64.8% of the pregnant women. In their review of the pandemic experience, the ECDC state the lack of vaccine acceptance can partially be attributed to the complex risk message communicated. This risk message communicated on the one hand that for people outside risk groups the risk of severe disease was low, but on the other hand it communicated that there was a small but real risk of severe disease and death for healthy adults and children [
22]. In one American study, vaccination rates of pregnant women with seasonal influenza vaccine were shown to increase from 19% to 31% through increasing patient awareness via education of vaccination providers and posters, underlining the impact of patient awareness in vaccination campaigns [
23]. A literature review found low vaccination rates of pregnant women in countries where seasonal influenza vaccination is recommended, ranging from > 0.1% to 12.8% [
24].
While in most countries it was decided to vaccinate pregnant women of the second and third trimester, one of the main differences between vaccination policies was the decision whether to vaccinate pregnant women in the first trimester of pregnancy. This difference between countries on whether to vaccinate first trimester pregnant women has also been witnessed in seasonal influenza vaccination policies [
1]. The first trimester of pregnancy is the period of major organogenesis and thus important in relation to risk of congenital anomalies, increasing the potential risk of vaccination. In addition, women in the first trimester of pregnancy are less susceptible to adverse health outcomes caused by influenza infection, decreasing the benefit of vaccinating this group. Differences in trimesters covered by the pandemic vaccination policy might have been caused by differences in appreciation of available safety and efficacy data, or more operational reasons such as potential for limited length of vaccine efficacy and wanting it to be most effective in the later two trimesters, or deciding that targeting women regardless of trimester would better ensure high coverage by the start of the second trimester. Post-pandemic immunogenicity data on Pandemrix in healthcare workers showed that three months post vaccination, 84% of the participants still had a protective antibody titre (HI titre ≥40) compared to 97% three weeks post vaccination [
25]. Both figures are still well above the Committee for Medicinal Products for Human use (CHMP) criteria for seasonal vaccines of > 70% of the subjects achieving an HI titre of ≥40.
Whatever the policy regarding first trimester vaccination, where the entire population is offered vaccine, this will include WCBA who are in the early stages of pregnancy but do not yet know they are pregnant. Other target groups such as women with asthma, or health care workers and persons in close contact with vulnerable groups would also include women in very early pregnancy. These women need to be taken into account when monitoring vaccine safety in pregnancy. In addition, 1st trimester pregnant women belonging to priority groups or specifically requesting for vaccination, could receive the vaccination in some countries that did not routinely vaccinate 1st trimester pregnant women.
Differences between countries were found in vaccines used in pregnancy, including use of adjuvanted or non-adjuvanted vaccines and Thiomersal content. Some European pandemic vaccines were developed and registered pre-pandemic using a "mock-up" registration procedure [
26]. This procedure assumes that swapping the strain in a vaccine will not substantially affect the safety and efficacy of the vaccine. The mock up vaccines would be authorized once a pandemic is declared, after which the initial strain is swapped for the pandemic influenza strain. Preliminary data on the immunogenicity of the monovalent non-adjuvanated and MF59-adjuvanated vaccines showed favorable results [
27,
28]. These clinical trials also demonstrated a single dose of the pandemic vaccine would be sufficient for an immunogenic response, allowing for antigen saving vaccination strategies. Several countries had established "advance purchase agreements" with manufacturers, leading to some of the differences in vaccine chosen [
26]. In these contracts, the manufacturer agrees to supply its pandemic influenza vaccine as soon as possible after a pandemic has been declared and agree to reserve an agreed number of doses for the country. Country of origin of the manufacturer might have played a role in vaccine of choice, as is illustrated by Hungary making use of Fluval P, manufactured by the Hungarian company Omninvest. Other possible factors in choice of vaccine might originate in limited supplies available and differing appreciation of available safety and efficacy data. Several of the pandemic vaccine formulations included oil-in-water adjuvants (AS03 for Pandemrix and MF59 for Celtura and Focetria) or aluminium phosphate gel adjuvants (Fluval P) as an antigen saving policy [
29] and to enhance the immunogenicity of the pandemic vaccine. Celvapan, CSL Pandemic Influenza Vaccine and Panenza did not contain adjuvants. Safety information of AS03 and MF59 adjuvants is sufficient in the general population, but limited in pregnancy [
29,
30]. Five countries solely used non-adjuvanted vaccine to vaccinate pregnant women. Some other countries used non-adjuvanted vaccines for the entire population (table ). Celvapan was the only pandemic vaccine not to contain Thiomersal. No consensus has been reached on the risk of prenatal exposure to Thiomersal [
31]. Both the AS03 and MF59 adjuvants contain squalene [
29,
32]. We included questions about the timeframe of the vaccination policy. There was variation in the speed of response and in at least one case the vaccines arrived after the vaccination campaign had been initiated. Vaccination campaigns were initiated between 28
th September 2009 and 27
th December 2009, while the first pandemic wave hit Europe in September-October 2009.
Post marketing surveillance of vaccines and medicinal products is essential, in particular with regard to their use in pregnancy since pregnant women are excluded from clinical trials and much of the prior safety testing. There are a number of different ways of carrying out post marketing surveillance, including case-control studies of adverse pregnancy outcomes, and cohort studies of women receiving the vaccine, but it is also essential to collect basic information about exposure of pregnant women in the population to the vaccine in order to interpret any possible changes in population rates of adverse pregnancy outcomes. The limited availability of exposure data and the variation in quality of the available exposure data collected on pregnant women in relation to the pandemic vaccine are striking, but in line with the observed lack of exposure data of pregnant women to seasonal influenza vaccination in Europe [
18,
33]. This problem is further complicated by the need to distinguish between first trimester of pregnancy and second and third trimester of pregnancy, when exploring the relationship between influenza vaccination and congenital anomalies. Pre-pandemic safety data of the pandemic vaccines and adjuvants used in the pandemic vaccines in pregnancy is fairly limited. In order to conduct population based pharmacovigilance, exposure data is needed to interpret congenital anomaly rates. A number of countries reported cohort studies being conducted in pregnant women e.g. a national cohort in Norway. The lack of comprehensive data on exposure of pregnant women to pandemic influenza vaccine also limits the policy evaluation process, which is required in order to learn from the vaccination efforts for future pandemic infectious disease outbreaks and vaccination campaigns.