Vaccination in pregnancy against pertussis and seasonal influenza: key learnings and components from high-performing vaccine programmes in three countries: the United Kingdom, the United States and Spain

A range of methodologies were considered to analyse the performance of influenza and Tdap vaccination programmes in pregnancy. From WHO models on vaccine hesitancy to practical taxonomies for the determinants of vaccine uptake such as the 5As (Access, Affordability, Awareness, Acceptance, Activation), several methodologies have been developed to comprehend the outcome of vaccination programmes [32, 33]. Kassianos et al. study was selected as a methodological backbone for this study given its emphasis on identifying actionable programmatic and policy components contributing to high seasonal influenza VCR amongst high-income countries with adult vaccination programmes [34]. Based on an analysis of the vaccination programmes in the US, UK, Canada, and Australia, this methodological framework consists of five pillars structuring 42 components as contributing factors of high VCR among older adults, some of which are also relevant to vaccination in pregnancy. Since the factors contributing to vaccination in pregnancy may differ significantly from those among older adults, and since our study also includes Tdap vaccination, the five-pillar structure and list of components from Kassianos et al. was adapted and expanded based on data gathered in three relevant benchmark countries (Spain, UK, and US). A three-step data collection approach was thus used in the considered countries including a detailed review of published literature using an academic literature database, a manual search of official sources, and a complementary search of grey literature.

For this analysis focusing on vaccination in pregnancy in high-income countries, Spain, the UK and the US were selected due to their pioneering approach to establishing both influenza and pertussis pregnancy vaccination programmes, their success in improving uptake, and the availability of annual VCR measurements by their respective health authorities to analyse the components that contributed to VCR growth. Even though other high-income countries than the three considered have developed pregnancy vaccination programmes they did not fulfil these three selection criteria and were thus not selected for this study. Furthermore, the US, with a mostly privatized and fragmented healthcare system, the UK with a public and centralized system, and Spain with a public and decentralized system were chosen as three illustrative examples to encompass different health system architectures. Since VCR values are published for England and Wales in separate reports, only England values are considered in this study to analyse trends given that the nation holds 84% of the UK population compared to 5% for Wales [35].

The data collection relied both on global and country-specific published literature, and a manual search of official sources as well as grey literature available online (Table 1). An Embase search was performed to identify relevant published articles. Search terms were refined through Emtree subject headings searches and structured in three categories: Disease, Vaccination, and Population. Disease search terms in each category are detailed in Additional file 1 (Table SI 1). The search was performed for each country and publications from 2010 to 2020 were considered. All the titles and abstracts identified through these searches were screened using a patient, intervention, comparison, outcome (PICO) search strategy (Table SI 2), followed by a screening of the full text. As a result, 86 articles were selected for the US, 33 articles for the UK, and 23 articles for Spain (Table 1).

Additional manual searches of health authority (HA) websites and reports, leading scientific societies and research groups position papers, healthcare professional (HCP) associations training documentation, lay public communication material, and key conferences were conducted to limit potential publication bias in this study (Table 1). Such sources were also used to identify official VCR data and to evaluate the measures and actions implemented over time to improve VCR.

Based on the published literature, official documents and communication materials identified in the three countries considered, all components described as contributing to the performance of pregnancy vaccination programmes were listed. Since VCR is a multifactorial variable and as the impact of some components may take time to materialize, the components were not selected purely on the direct quantitative impact on VCR, but rather based on whether they were described as a key contributor of higher vaccine uptake among pregnant women. The framework was consequently adapted to vaccination in pregnancy by matching identified components with those of Kassianos et al. framework to either add, edit or replace them in the relevant pillars.

Seasonal influenza VCR data was collected from nationwide survey data published in several Mortality and Morbidity Weekly Reports (MMWR) articles and from Kennedy et al. [29, 37, 38]. The US pioneered influenza vaccination in pregnancy with a recommendation for at-risk pregnant women dating back to 1966, which was updated by the Advisory Committee on Immunisation Practices (ACIP) to all pregnant women in their third trimester in 1995, and expanded to the second trimester in 1997 [39‐41]. By the early 2000s the VCR reached 10 to 20% (Fig. 1). In 2004, the recommendation was updated to include all trimesters in accordance with WHO guidance [11, 42, 43]. Despite an endorsement by the American College of Obstetricians and Gynecologists (ACOG) in 2004, the VCR remained below 30% until 2007 [44]. The burden associated with the A/H1N1 2009 pandemic led to an increased mobilization among scientific societies, practitioners and lay public alike [45]. As a result, the VCR reached 50% by 2012 and over 60% in 2019. Nonetheless, current values fall short of the 80% target set by Healthy People 2020, an initiative that develops sets of objectives to improve the health of all Americans [46].

Tdap VCR was gathered from nationwide survey data from the CDC and MMWR articles [28, 29, 47]. A recommendation for the vaccination of at-risk pregnant women with Tdap was first issued in 2007 but the overall VCR remained close to 0% (Fig. 1). Following the 2010 pertussis outbreak in California, a recommendation was issued by the state to expand Tdap vaccination to both at-risk and healthy pregnant women [48]. This was followed by the ACIP which issued a recommendation in 2011 to vaccinate pregnant women with no prior history of Tdap vaccination, updated in 2012 to all pregnancies [49, 50]. This move was actively endorsed by scientific societies and key HCP associations such as the ACOG in 2013 [51]. Guidelines were cascaded to their members and the medical care organizations also followed suit by including Tdap vaccination in the medical protocol of pregnant women [52]. The impact on VCR was substantial, increasing from 9% to 50% within 2 years, and was also driven by the high public awareness at the time fed by communication from multiple stakeholders around pertussis prevention (Fig. 1). Since 2014 however the VCR measured by CDC surveys has plateaued at 55%, suggesting, as for influenza vaccination, persisting issues with regard to access to vaccination, HCP engagement and vaccine acceptance [29].

In the UK, VCR data was collected from PHE (Public Health England) which extracts and aggregates data from General Practitioners (GPs) electronic reports [53]. PHE issued a recommendation for influenza vaccination during pregnancy in 2010, following a Joint Committee on Vaccination and Immunisation (JCVI) decision, in response to the severity of the A/H1N1pdm09 strain among pregnant women [54, 55]. This decision was corroborated by the cost-effectiveness of the intervention [56, 57]. PHE and the NHS (National Health Service) updated their guidance, training kits, and awareness materials accordingly to support GPs. As a result, VCR quickly reached 27% in the first season, 2011/2012, and 40% in the 2012/2013 seasons (Fig. 2). However, the limited vaccine effectiveness and persisting myths around the risks of vaccination have contributed to a plateau of the VCR at sub-optimal levels around 45% since the 2014/15 season, markedly below the WHO and EU target of 75% (Fig. 2).

For Tdap pregnancy vaccination, the VCR increase occurred faster due to high disease awareness among healthcare professionals and lay public alike. The national pertussis outbreak in 2012 caused the deaths of 14 non-immunized infants [58]. In a context of intense media noise on the large increase in cases and deaths reported by the Health Protection Agency (HPA), the JCVI adopted the recommendation for Tdap vaccination during all pregnancies [59‐62]. A rapid roll-out of the guidance, training resources and awareness material for GPs led by PHE and the NHS followed, leading to a VCR above 50% in the first year [63] (Fig. 2). As this programme was proven to be safe, highly efficient, and cost-effective, it was extended for a further 5 years in 2014 [64]. The recommendation of the vaccination timing was expanded in 2016 to include weeks 16–32 of pregnancy and obstetricians, gynecologists, and midwives were increasingly mobilized in addition to the GPs. The programme was eventually adopted as routine in 2019 with VCR having reached a plateau around 70% [63, 65]. (Fig. 2).

Health authorities in Spain included pregnant women in their second or third trimester in the risk groups for influenza vaccination as early as 2005 [66]. This was rapidly extended to all trimesters starting in 2006 in regions such as Catalonia [67, 68]. Yet, influenza VCR remained consistently low among pregnant women, at ~ 5%, as highlighted in several regional studies [67, 69] (Fig. 3). The roll-out of the Tdap pregnancy vaccination programme had a positive impact on influenza vaccination of pregnant women and synergies between the programmes led to a significant increase in influenza VCR, ultimately reaching the national VCR target of 40% for pregnant women in the 2018/2019 season [69‐72] (Fig. 3). VCR data aggregated from regional reporting by the Ministry of Health since 2017/18 has shown a sharp increase [70, 73].

The rapid adoption of Tdap vaccination during pregnancy was driven by a public health response to the rising number of pertussis cases and deaths in infants below 3 months of age due to a “permanent outbreak situation” from 2011 to 2015 [74]. In light of the recommendation from the JCVI in 2012 in the UK, scientific societies including the Spanish Association of Paediatrics (AEP) played an important role in highlighting the burden of pertussis to national and regional health authorities [75, 76]. This led to the adoption of new recommendations by three pioneering regions (Catalonia, Asturias, and Navarra). By the end of 2016, all regions had introduced a Tdap pregnancy vaccination program. An analysis rapidly confirmed the effectiveness of the approach with the Red Nacional de Vigilancia Epidemiológica (RENAVE) highlighting a decreased incidence of pertussis cases in infants under 1 year of age and especially those under 3 months of age [74]. Since 2016, Tdap VCR data aggregated from regional reporting available from the Ministry of Health has been consistently at or above 80% since 2016 [70] (Fig. 3).

As part of this analysis, a set of key components driving vaccine uptake was identified. The pre-existing five-pillar framework for influenza VCR was utilised to organise a list of components tailored to the specificities of seasonal influenza and Tdap vaccination in pregnancy [34]. The 5-pillar structure was confirmed as applicable and the pillars were adapted as follows:

A total of 45 components were identified as relevant contributing factors to the successful roll-out of a pregnancy vaccination programme (Table 2). Of these, 15 components were unchanged from the previous framework, 14 were adapted, 16 were added as specific to the context of vaccination in pregnancy, and 13 removed as not applicable [34]. The national health authority recommendation, the absence of financial barriers to getting immunized, and a structured infectious disease surveillance network are applicable to maternal immunisation programme and were amongst the components that remained unchanged. The components that were adjusted to the context of vaccination in pregnancy include but are not limited to multiple HCPs (Ob-Gyns, GPs, midwives, nurses, pharmacists) being allowed to vaccinate, training of HCPs on pregnancy vaccination by multiple stakeholders, and adapted HCP training material for pregnancy vaccination. The new components specifically identified for the vaccination of pregnant women include the mention of immunization in the pregnant women medical protocol, a clear delineation of HCP roles and responsibilities, and the availability of adapted awareness material at maternal care providers.

The adjusted five-pillar VCR framework of 45 components was used as a methodological basis to analyse the factors that contributed to the successful programmes in the US, the UK and Spain. There was significant variability across the three countries, which is likely to be attributable to the structural differences in the healthcare architecture and strategy applied for the commissioning of the vaccination programme in pregnancy. Overall, the UK had the highest number of components fully implemented (38 out of 45), followed by Spain (31 out of 45), and the US (25 out of 45). In the UK, five components were deemed as partially implemented, compared with 11 in Spain and 18 in the US. Finally, two components were missing in both the UK and the US, and three in Spain (Fig. 4). These differences mainly concerned components relating to the role of VCR targets and monitoring, the convenience of the patient journey, and the importance of the different HCPs involved in recommending and delivering vaccination in pregnancy. The three countries also differed in how adequate training was provided to HCPs as well as the incentivization for vaccinating pregnant women. (Additional file 1, Figs. SI 1–3) Finally, while the framework applied to both influenza and pertussis, some nuances between the two must be considered. The VCR of pertussis vaccination in pregnancy was higher than for influenza in all three countries, possibly because this programme emphasizes the indirect protection of the infant from the potentially serious complications of pertussis in infancy. Although there are benefits of influenza vaccination in pregnancy for infants, these may be less well-known or communicated to pregnant women. Finally, the seasonal nature of influenza vaccination has practical implications which must be considered for both the timing of vaccine administration and the identification of pregnant women.