Stronger responders—uptake and decline of HPV-vaccination in Denmark

Abstract Background The purpose of this study was to identify the stronger responders behind the fluctuating coverage with the HPV-vaccine in Denmark in order to facilitate information campaigns targeted at specific subgroups. Methods Newspaper articles published from 2006 to 2009 with information about coverage with the HPV-vaccine in Denmark were identified from the database Infomedia.dk. Vaccination coverage of recent years was retrieved from the publically accessible statistics from the State’s Serum Institute. Data on average disposable income nationally and for each municipality was retrieved from Statistics Denmark. Results According to numbers published in newspapers, girls residing in municipalities with a high disposable income were the first ones to secure the HPV-vaccine in Denmark. Years later, at the start of the debate about possible side effects of the HPV-vaccine, the decline in vaccination coverage was slightly steeper for girls from high income municipalities than for girls from low income municipalities. Conclusions Girls from municipalities with a high disposable income seem to be the stronger responders of the fluctuating coverage with the HPV-vaccine in Denmark. This was the case both during the initial surge in coverage after the vaccine’s introduction on the market, and during the later decline following the debate on possible side effects. Identification of this dispersion pattern enables health authorities to initiate targeted information campaigns.


Introduction
T he quadrivalent vaccine Gardasil Õ against human papillomavirus (HPV) types 6, 11, 16 and 18 was introduced in Denmark in October 2006. 1 Whilst offering protection against the virus types most commonly causing condylomas and cervical dysplasia, 2,3 it came at a price of 470 euros for the three recommended doses. 4 In October 2008, the HPV-vaccine was introduced into the Danish Childhood Vaccination Programme at the expense of the state, making the vaccine free of charge for the citizens. From January 2009, girls born in 1996 and later were offered the vaccine free of charge from the age of 12-18. A catch-up programme was installed running from October 2008 to December 2010, offering girls born from 1993 to 1995 the vaccine free of charge, whilst another catchup programme running from September 2012 to December 2013 covered women born from 1985 to 1992. This inclusion into the Childhood Vaccination Programme was a popular move as the vaccine was expensive and the risk of developing cervical cancer is highest in the lowest socioeconomic groups. [5][6][7] Furthermore, Denmark has a high background risk of cervical cancer compared to the rest of Western Europe, 8 and a suboptimal participation in the Cervical Cancer Screening Programme. 9 The introduction of the HPV-vaccine was thus a promising new step towards combatting cervical cancer in Denmark.
Data from the national vaccination register show that after the introduction of the HPV-vaccination programme, the first dose coverage for girls born in 1997-2000 was constant at around 90%, and full immunization coverage was around 80%. 10,11 However, a decline in coverage started in 2013, first registered for girls born in 2001, and full immunization coverage reached a mere 15% for girls born in 2004. 11 The decline in coverage followed worries in the lay press and on social media about possible side effects of HPVvaccination. 12,13 We sought to identify the stronger responders in the HPVvaccine's fluctuating uptake in Denmark. First, we searched for data on stronger responders in the uptake of the HPV-vaccination between the marketing of the vaccine in September 2006, and the inclusion of the vaccine into the Danish Childhood Vaccination Programme in October 2008. Subsequently, we investigated the stronger responders to discard the vaccine from 2012 onwards, focussing in particular on girls born in 2000-04. Responders were characterized by geographical distribution and income in order to determine whether information on the HPV-vaccine should be tailored to particular population subgroups.

Methods
Newspaper articles were examined to collect data on stronger responders to accept the HPV-vaccine. A search was made on the database Infomedia.dk for all newspaper articles containing the words 'HPV vaccine' between 1 September 2006 and 1 September 2009. The term HPV-vaccine was well known in the Danish public, and it is unlikely that a newspaper article covering this topic would not have included this term. Infomedia is a comprehensive Danish data base including all daily nationwide, regional and local as well as weekly newspapers. The search was made directly on the Infomedia website by one of the authors (C.L.L.). University students can make search on Infomedia free of costs. The search revealed 1182 articles. The title and the first 20-25 words of each article were read. As newspaper articles start with presenting the 'results' before 'material and methods', the relevance of the content can easily be identified from the first words. The articles expected to provide information on number of and/or characteristics of HPV-vaccinated girls and young women were selected. One of the authors (C.L.L.) selected the relevant newspaper articles. Normally, there are more than one copy of a given article, e.g. from a nationwide newspaper and from several local newspapers. The risk of overlooking a given article is therefore small. If two or more newspaper articles contained identical data, only one was selected and used in the results. Information of date, source, municipality/region, age group if provided and reported purchase/use of HPV-vaccine was abstracted by one author (C.L.L.), and checked by the other author (E.L.).
The website of the State's Serum Institute (SSI) was searched for data on first dose and full immunization of the HPV-vaccine for girls born in 2000, 2001, 2002, 2003 and 2004 nationwide and by municipality. 10,11 These data were retrieved as percentages only. Data, updated on the 8 June 2017, were retrieved on 3 July 2017. Data from municipalities with <100 girls in one birth cohort were excluded due to small numbers. This meant exclusion of the municipalities of Dragør, Fanø, Langeland, Laesø, Samsø, Vallensbaek and AErø. Data from the remaining municipalities were used to identify the stronger responders to discard the HPV-vaccine.
The website of Statistics Denmark was searched for data on average disposable income nationwide and by municipality. 14 Data from years 2006 to 2009 and 2015 were retrieved on the 4 July 2017. The numbers retrieved were rounded to the nearest thousand Danish Krone (DKK).

Results
Stronger responders to secure the HPV-vaccine

Discussion
This study revealed that girls living in municipalities with a high average disposable income appeared to be stronger responders of the varying coverage of the HPV-vaccine in Denmark. They were the first to secure the HPV-vaccine after it was marketed, and they had the steepest decline in vaccination coverage during the debate about possible side effects of the HPV-vaccine.
This trend, that the richer parts of society are the stronger responders in terms of vaccination, is not a new phenomenon. Looking back at the measles, mumps and rubella (MMR) vaccine scare, similar tendencies were observed. Studies from the UK indicated that during the scare about the MMR vaccine, the largest decrease in vaccination participation was seen in wealthy or well-educated areas. 15,16 According to one of the studies, the major part of the decline was seen during the early stages of the controversy with the people from higher social classes having the fastest decline. 15 A similar pattern was evident in our results, where the decline in coverage was slightly steeper in the wealthiest than in the poorest municipalities. One may wonder whether this difference is due to obvious factors such as the wealthier part of the population having better access to information, better pre-requisite for interpreting information, and therefore able to react faster on new information, irrespective of the reliability of the source. This was  investigated in several studies, with many results highlighting that people from higher social classes want to be given the choice, be well informed, and be regarded as competent to make health decisions. 17,18 Concurrently, one study called attention to the fact that whilst rigorous research on the effect and safety of the vaccine is a requirement, research is lacking on factors which affect public trust in vaccines. 19 Concerning the recent discussion of side effects to the HPVvaccine in Denmark, a study from 1997 found that people tend to let information with high-risk indication militate more than information on low risk. 20 This may have played a part in the rapidly declining participation in the HPV-vaccination progamme in Denmark, where the media have been responsible for many headlines concerning severe side effects of the vaccine without presenting any clear evidence. The few press releases from the government about the safety of the HPV-vaccine have seemingly simultaneously been neglected and outweighed. This power of the media is not unheard of. Previous studies suggest that the media coverage was the reason for a dramatic increase in influenza vaccinations in the USA, 21 whilst the media portrayal of possible side effects to the pertussis vaccine in 1974 led to a decline in vaccination participation leading to two major epidemics of pertussis in the UK. 22,23 The media's unsubstantiated statements on risks are reinforced when they are disseminated on social media.
The study has a number of limitations. Firstly, data on stronger responders to secure the HPV-vaccine was based on numbers reported in newspaper articles and most of these numbers came from Sanofi Pasteur; the company marketing the Gardasil Õ vaccine in Denmark. One has to be aware that it could have been in their interest to publish numbers which portrayed the social inequity in HPV-vaccination in order to put pressure on the government to include the HPV-vaccine in the Danish Childhood Vaccination Programme. It is important to point out that the company had this information reported in the standard news sections of the newspapers and not in advertizements. Secondly, numbers from the SSI provided a snapshot of the coverage at a specific point in time and did therefore not show the constant dynamics of the vaccine coverage. We realize that the data illustrate a tendency but do not provide a formal test of a hypothesis. Thirdly, as these number were retrieved as percentages only significance testing of differences was not possible. Lastly, the average disposable income in the municipality was used to characterize all inhabitants in this geographic area. However, the SSI also presents data on coverage by municipality, which means that the two data sets are comparable.
This study portrayed the rather unexpected turn in the coverage of the HPV-vaccination in Denmark, and it uncovered a common characteristic of the stronger responders. In other words, it seems that the girls living in a municipality with a high average disposable income were trendsetters on the HPV-vaccine coverage in Denmark. This common factor could be used in future government vaccination campaigns, where a possibility would be to target information more directly at this group of trendsetters with the hope that they would spread evidence-based information about the vaccine. This however inevitably raises ethical questions on social discrimination especially when it comes to a national vaccination programme, which is seen as a basic right regardless of socioeconomic class.
This study also fostered new inquiries on the mechanism of losing public confidence in a nationwide vaccination programme, to such a severe extent as with the HPV-vaccine in Denmark, and the question remains on how to restore public trust in the vaccine. A previous study suggested that once the MMR scare was over, the wealthier people were the first to resume use of the vaccination programme, 15 whilst several other studies of attitudes towards vaccination indicated that the family doctor played a key role when parents decided on whether or not to vaccinate their children. 21,[24][25][26] Thus, focussing vaccination campaigns more locally at the family doctor, ensuring that they are prepared to give the girls and their parents neutral and evidence-based information, and combining it with information specifically targeting girls in municipalities with high average disposable income may be part of the solution to a vaccination crisis as the one seen with the HPV-vaccine in Denmark.
After massive national efforts to ensure equal access to the HPVvaccine in Denmark, the vaccination coverage has recently seen a dramatic decline due to the public debate about possible side effects. To regain public trust and target vaccination campaigns, it is necessary to unveil the dynamics of these variations in vaccination coverage. This study indicated that girls residing in municipalities with a high average disposable income were the stronger responders of the fluctuation in coverage with HPV-vaccine in Denmark. In other words, according to our results they appeared to be both the first to opt into the newly approved vaccine, when it hit the Danish market, and later they seem somewhat faster to opt out of the vaccination programme, when the debate about possible side effects bloomed. If this is a continuous trend then this subgroup of the population may be crucial, when targeting campaigns to ensure high coverage in national vaccination programmes.

Supplementary data
Supplementary data are available at EURPUB online. Table 3 The five richest and five poorest municipalities with average disposable income (Appendix A1) and percentage of girls fully immunized and girls with first dose of HPV-vaccine by year of birth, 2000-04 (Appendixes A18 and A19)

Disclaimer
Elsebeth Lynge: Roche provides test kits free of charge for a randomized controlled trial. Participated in meetings with Roche with fees paid to the University of Copenhagen. Has participated in one meeting organized by the HPV Prevention and Control Board and in one meeting organized by GSK; both without payment.
Conflicts of interest: None declared.

Key points
In Denmark, girls from high income municipalities were the first to use the HPV vaccine after it was marketed in 2006.
After the HPV-vaccine was included in the Danish Childhood Vaccination Programme in 2008, vaccine uptake become high throughout Denmark. A decline in HPV-vaccine uptake started in 2013 following worries in the lay press and social media about possible side effects. This decline was steeper for girls coming from high income municipalities than for girls coming from low income municipalities.

Introduction
C hildren 0-4 years attending childcare (CC) centres are more prone to acquire infectious diseases. [1][2][3][4][5] They experience $6-12 infectious illness episodes per year, on average resulting in absenteeism of 23 CC days. 4,6-8 CC children also have higher healthcare consumption compared with home-cared children. 5,[9][10][11][12][13][14] Furthermore, due to an average work absenteeism of 1.5 days among parents per illness episode, CC children generate twice as much societal costs compared with home-cared children. 2 Organizational exclusion policies regarding ill children and defensive advices from CC staff drive children's CC absenteeism and healthcare consumption. 9,10,[15][16][17][18][19][20] On the one hand, exclusion policies are important in CC staff's decision-making since CC staff lack formal education on ill children. On the other hand, CC staff perceive that they mainly base decisions on their intuitive judgment. 18,21 These judgments emanate from experiences gained by themselves or by colleagues. However, these experiences may also increase their level of concerns since previous negative affect decision-making and lead to defensive advising. For example, CC staff urge parents to pick up the ill child when they do not feel comfortable anymore to take care of a child with a body temperature exceeding 38 C. 18 CC staff's decisions are also influenced by incomplete knowledge, the belief that an ill child is better off at home, perceived contagiousness of illness symptoms and the belief that ill children increase workload, that they do not feel confident in monitoring ill children, or feel pressured by parents who have to work. 10,16,[21][22][23] Most CC staff fear that a high fever resembles illness Optimizing decision-making among CC staff on fever and common infections