Article Text
PDF
Abstract
Background Uptake of human papillomavirus (HPV) vaccination is generally high in high-income countries with school-based vaccination programmes; however, lower uptake in certain population subgroups could continue pre-immunisation inequalities in cervical cancer.
Methods Six electronic databases were searched for quantitative articles published between 1 September 2006 and 20 February 2023, which were representative of the general population, with individual-level data on routine school-based vaccination (with >50% coverage) and sociodemographic measures. Titles, abstracts and full-text articles were screened for eligibility criteria and assessed for bias. A second independent reviewer randomly screened 20% of articles at each stage. A narrative synthesis summarised findings.
Results 24 studies based in eight countries (Australia, Belgium, Canada, New Zealand, Norway, Sweden, Switzerland, UK) were included. Studies reported vaccination uptake by individual-level and area-level socioeconomic status (SES), parental education, religion, ethnicity and/or country of birth. 19 studies reported that more than 70% were vaccinated (range: 50.7%–93.0%). Minority ethnic groups and migrants were more likely to have lower vaccination uptake than White groups and non-migrants (11/11 studies). Lower SES was also associated with lower uptake of vaccination (11/17 studies). Associations with other sociodemographic characteristics, such as parental education and religion, were less clear.
Conclusions Even in high-income countries with high coverage school-based vaccination programmes, inequalities are seen. The totality of available evidence suggests girls from lower SES and minority ethnic groups tend to be less likely to be vaccinated. Findings could inform targeted approaches to mop-up vaccination and cervical cancer screening amidst changing HPV epidemiology in a vaccine era.
Trial registration number CRD42023399648.
- VACCINATION
- INFECTIONS
- Health inequalities
- SYSTEMATIC REVIEW
Data availability statement
Data sharing not applicable as no datasets generated and/or analysed for this study.
This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Many high-income countries are on track to achieve the WHO Cervical Cancer Elimination goals, but inequalities in cervical cancer prevention remain. Previous reviews have demonstrated that certain sociodemographic factors are associated with lower uptake of human papillomavirus (HPV) vaccination, including ethnicity and healthcare coverage. However, no study has focused on high-income settings with high vaccination coverage and school-based delivery, which should be the most equitable and effective programme structure.
WHAT THIS STUDY ADDS
This study uses a robust systematic review methodology that identified 24 population-representative studies of factors (socioeconomic status, education, ethnicity and religion) associated with HPV vaccination in girls in eight high-income countries with school-based vaccination programmes. Across these studies, a narrative synthesis identified lower uptake of vaccination was consistently associated with lower socioeconomic status and among ethnic minority groups and migrants, though associations were minor or inconsistent for other factors. This work demonstrates that even in high-income countries with high vaccination coverage and school-based vaccination, inequalities may need to be addressed.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
High vaccination coverage and changing HPV epidemiology will require changes to cervical cancer screening programmes, and it is important that unvaccinated groups do not get left behind. These findings can inform studies of the outcomes of vaccination, and targeting of mop-up vaccination and cervical screening to help achieve WHO Cervical Cancer Elimination targets.
Introduction
Human papillomavirus (HPV) vaccination uptake is generally high in high-income countries with school-based vaccination programmes, with many such countries on track to reach the WHO Cervical Cancer Elimination target of 90% coverage; however, even small differences in access might be important.1 Globally, HPV vaccination of girls has been introduced in 75 of 84 high-income countries (as of 2021), 41 of which have fully or partially school-based vaccination programmes (online supplemental figure 1).2 3 Many countries have also now introduced gender-neutral vaccination. Prevaccination, cervical cancer has been associated with lower socioeconomic status (SES) in some high-income countries, with the combined effects of higher risks of HPV infection and disease, as well as lower attendance for cervical screening in some population subgroups.4 5
Supplemental material
Previous research has demonstrated that school-based programmes are more equitable than other forms of vaccination programmes, such as clinic-based programmes,6 but inequalities can still occur. Lower uptake of HPV vaccinations in certain population subgroups could continue the inequalities in cervical cancer prevention that were observed prevaccination.7 Groups with lower HPV vaccination rates may also be less likely to attend cervical cancer screening, as well as be at higher risk for high-risk HPV infection,8 thereby exacerbating risk of cervical cancer within a population largely protected due to the success of vaccination and screening programmes.7
A previous systematic review of studies primarily in the USA demonstrated evidence of differences in HPV vaccination by ethnicity and healthcare coverage but found no association with parental education or family income.9 However, HPV vaccination in the USA is offered in healthcare clinics and paid for using health insurance, meaning this review cannot be extrapolated to countries with free-of-charge school-based vaccination programmes, such as the UK. Additionally, this review was conducted in 2013, and therefore captured only early experience of HPV vaccination programmes.
This review and narrative synthesis, conducted in 2023, therefore aimed to determine whether any characteristics were associated with not being vaccinated for HPV among girls in high-income countries with high coverage and school-based vaccination programmes, which we would expect to be the most effective and equitable. Population-based studies which reported on individual-level HPV vaccination status among girls by at least one sociodemographic measure were included. Findings could inform current and future school-based vaccination programmes globally to aid in achieving the WHO Cervical Cancer Elimination target.
Methods
Conduct and protocol
This review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, and the protocol was published on the International Prospective Register of Systematic Reviews (PROSPERO) in advance (CRD42023399648).
Search strategy
Six electronic databases (MEDLINE, Web of Science, Scopus, Embase, PsycINFO and The Cochrane Library) were searched for articles published between 1 September 2006 (initial licensure of first HPV vaccine) and 20 February 2023. Searches were conducted separately for each database using database-specific searches. An initial search was conducted to identify relevant keywords and indexed terms in titles and abstracts. Search strings were developed based on three key concepts: HPV vaccination, sociodemographic factors (eg, ethnicity, parental education, religion, socioeconomic factors) and high-income countries with school-based vaccination programmes. Countries were identified based on those listed as ‘high-income’ according to the World Bank.10 Based on this list, vaccination programme structure was ascertained using global summary reports, as well as national health agency websites.11 12 Medical Subject Headings (MeSH) were used where relevant, along with truncated free text and subject headings. The search terms used are shown in online supplemental table 1. Searches were supplemented by searching of reference lists.
Screening
All identified references were exported to Zotero and deduplicated. The first reviewer (ED) conducted title and abstract screening for all results, as well as full text screening for articles included at that stage. A second independent reviewer (RO) screened 20% of articles in each stage, selected randomly. Conflicts were resolved and consensus reached through discussions between the two reviewers at each stage. There was very minimal disagreement between reviewers at each stage.
Inclusion and exclusion criteria
Study inclusion and exclusion criteria were defined using a PICOS (population, intervention, comparator, outcome and study design) table, described in table 1.
Systematic review inclusion and exclusion criteria (PICOS table)
Quality appraisal
Included full-text studies were appraised using relevant STROBE (Strengthening the reporting of observational studies in epidemiology) checklists. The first reviewer assessed the quality of all papers using the Joanna Briggs Institute (JBI) critical appraisal checklist13 for prevalence studies (20% by second reviewer). A meeting was held to discuss any disagreements. Each article was assigned yes, no or unclear for a set of nine criteria (including representativeness, recruitment, sample size, descriptions of participants, response rate, measurement bias, consistency and appropriateness), then given an overall appraisal (include, exclude or seek further information) (online supplemental figure 2). To distinguish between studies using data from vaccination records and those which used self or parent-reported vaccination status, studies were classified as ‘yes’ for ‘Identification of vaccination’ only where a study used data from vaccination records, and ‘unclear’ for those with self-report. Previous research suggests high agreement between parental report and vaccination records for HPV vaccination.14
Data extraction
Data for all included articles were extracted by the first reviewer, and 20% by the second reviewer as a consistency check. Following the bias assessment, a narrative synthesis was used to describe included studies according to country and characteristics associated with being unvaccinated. Findings were summarised according to the following sociodemographic characteristics: area-level deprivation, income, education, religion, and ethnicity/country of birth.
Results
Screening
Of 5238 records initially identified through the database search, 2374 titles and abstracts were reviewed following deduplication, and 208 full-text articles assessed for eligibility (figure 1).
PRISMA diagram showing the screening and selection process for systematic review. HPV, human papillomavirus; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.
Full-text studies were excluded for a variety of reasons (shown in figure 1), but most commonly due to not reporting vaccination status by a sociodemographic measure (n=88).
Quality appraisal
The JBI critical appraisal checklist13 for prevalence studies was used to assess 28 articles (online supplemental table 2). Four articles were excluded. Two of the excluded studies were available only as conference abstracts, so sufficient information was not available to determine study quality. Another study was excluded because of missing data in the analysis (70% of respondents were not included), which introduced potential selection bias. The final excluded study was identified as being unrepresentative of the general population. Excluded studies had at least three categories rated as ‘unclear’.
Study characteristics
Of 24 included studies, the majority were conducted in the UK (n=6), Canada (n=6) and Australia (n=4), followed by Norway (n=3), Sweden (n=2), New Zealand (n=1), Belgium (n=1) and Switzerland (n=1) (table 2). Most studies were conducted during the first decade following introduction of HPV vaccination, with the earliest data collection period in 2007.6 15–20 The most recent data collection period was 2020.21
Characteristics of included studies
Uptake of at least one dose of the HPV vaccination was consistently high across studies, with most reporting greater than 70% vaccinated.6 9 17 20–34 Of the studies reporting lower vaccination rates (50%–70%),15 16 18 19 35 most were conducted in the initial years of HPV vaccination programmes in Canada and Australia.
Of the included studies, 17 used vaccination and sociodemographic data from national registries or vaccination and census records,6 15–22 26–32 36 while seven used data from cross-sectional surveys.23–25 33–35 37 Sample sizes were large for all studies, with 14 studies reporting a sample size larger than 10 000.6 15–17 19 21 22 27–32 36 The smallest sample size was 774 girls,25 and the largest was 311 656 girls.30
HPV vaccination programme structures are similar across the eight countries represented in the included papers (figure 2). In most of the eight countries, at the time of the studies, routine vaccination had been offered to girls aged 11–13 years, while catch-up vaccination had been offered to older girls in their later teens to early 20s. In all these countries/studies, HPV vaccination was free-of-charge at time of school-based vaccination.
Timeline of (A) HPV vaccination programmes in eight included countries (UK, Canada, Australia, Norway, New Zealand, Belgium, Switzerland and Sweden) and (B) data collection periods for 24 included studies. aHPV vaccination offered earlier than demonstrated in the diagram. The diagram illustrates when the free vaccinations became available and offered in schools. HPV, human papillomavirus
Factors associated with non-uptake of vaccination
Socio-economic status
All studies, except one,35 examined the association between vaccination and SES (table 3): 17 of these 23 reported an association.6 15–20 22 26–30 32 33 36 37 Thirteen studies used individual-level indicators of family or parental income (online supplemental table 3),6 17 18 23–25 27–30 33 34 while 11 studies used non-individual-level indicators of SES, including area-level deprivation15 16 20 21 31 32 36 37 and school-level SES (online supplemental table 4).22 26 One study examined both income and area-level deprivation.19 Generally, lower SES (as indicated by family income, area-level deprivation or school-level SES) was associated with lower uptake of HPV vaccination. However, six studies (UK,32 36 Switzerland,37 New Zealand26 and Canada18 19) found higher vaccination uptake among those with lower SES, though the strengths of these associations were not consistent. Strengths of association between SES and vaccination uptake were similar between studies with individual-level versus area-level measures.
Summary of risk factors for non-uptake of HPV vaccination in included studies
Parental education
Nine studies examined the association between parental education and vaccination uptake (table 3). Four were in Canada,23 24 35 three in Norway27–29 and two in Sweden.6 30 Of these, six reported an association between parental education and HPV vaccination uptake, though the direction of this association differed (online supplemental table 5). Among the studies in Norway, as well as one in Canada,35 higher parental education was associated with reduced HPV vaccination uptake. Both Swedish studies reported that lower HPV vaccination uptake was associated with lower parental education.
Religion
Six studies analysed whether there was an association between religion and HPV vaccination; four studies used individual-level religious identity,25 33–35 while two studies reported school-level religious affiliation (table 3).22 31 Associations were mixed. Two studies reported no association between religion and vaccination.31 35 Some studies found higher vaccination uptake among more religious groups compared with non-religious groups. An Australian study found higher uptake in Catholic schools compared with secular government schools. Similarly, a Canadian study found increased vaccination uptake among Christian girls compared with non-Christian girls. Two UK studies reported reduced vaccination among religious groups,33 34 however, these associations did not remain following adjusted analysis. Most associations between religion and vaccination were weak (online supplemental table 6).
Ethnicity or country of birth
Eleven of 12 studies which examined ethnicity or country of birth identified an association between these variables and HPV vaccination (table 3).6 20 23 25–27 30 32–36 Of these, two studies from Canada, two from Sweden and one from Norway reported on country of birth, rather than ethnicity. Strength of association between ethnicity or country of birth, and HPV vaccination for each study is summarised in online supplemental table 7. The association with vaccination was not significant in one of the Canadian studies (p>0.05),35 but the other Canadian study23 found girls whose parents were born outside of Canada were more likely to be unvaccinated against HPV. The Swedish studies also found non-vaccination was higher among girls born outside of Sweden, or whose parents were born outside of Sweden, compared with Swedish-born girls.6 30 The Norwegian study found that girls with at least one parent born in Norway were most likely to receive HPV vaccination.27
All seven studies which reported on ethnicity identified an association with HPV vaccination: one study was from Canada,25 one from New Zealand26 and the remainder from the UK.20 32–34 36 In the Canadian study, uptake was higher among White girls compared with non-White girls, and in the New Zealand study, Pacific girls had higher vaccination rates than Asian, Maori and European girls. Among the UK studies, all reported that ethnic minority groups, especially Black/Black British and other ethnic groups, were less likely to be vaccinated than White groups. Two of these studies adjusted for SES.33 36 Generally, there was a strong trend for reduced vaccination among minority ethnic groups and immigrants in all studies where an association was identified.
Discussion
This systematic review and narrative synthesis of 24 papers from high-income countries with high coverage school-based HPV vaccination programmes highlights some sociodemographic factors that have been associated with lower uptake of the vaccination, including area or school-level SES (area-level deprivation), individual-level SES (income) and ethnicity, and some with variable and weaker association, including parental education and religion. Importantly, this review demonstrates that, even in high-income countries with high vaccination coverage and school-based delivery, which should be more equitable than other programme designs, potentially important inequalities in vaccination uptake can be seen: these should be considered carefully in vaccination impact studies.
Lower uptake of HPV vaccination was associated with a lower SES across most (11/17) of the studies that identified an association between these variables. Additionally, where studies examined ethnicity and country of birth, minority ethnic groups and migrants were found (11/12) to have lower vaccination coverage than White groups and non-migrants. Associations with other sociodemographic characteristics, such as parental education and religion, were less clear. Some characteristics, such as education, may reflect ‘U-shaped’ associations, where lower education levels and higher education levels may both be associated with non-uptake of vaccination in different settings.
Some of our findings are similar to a previous systematic review,9 which found ethnicity to be associated with HPV vaccination initiation, though no associations with parental education or family income were observed. However, our systematic review builds on the previous review by focusing analysis on countries with high coverage school-based vaccination programmes, and limits potential bias attributable to different healthcare systems and lower vaccination coverage that feature in US studies. All included studies had large, representative samples with the findings generalisable to eligible girls in the respective countries. Additionally, many of the studies used registry data and vaccination records, minimising reporting or selection bias given our restriction to studies with limited missing data and a high proportion of data linkage. There may still be some misclassification and bias in the studies we included; however, by restricting inclusion as we did we believe we minimised this making our findings as robust as possible.
There are also some potential weaknesses to consider. First, it is possible that relevant articles were not captured in the search strategy, though this is mitigated by the finding that no further studies were identified by reviewing reference lists of included studies and other relevant systematic reviews. Second, the included studies are from eight countries, but there are at least 33 other high-income countries with school-based vaccination programmes (online supplemental figure 1). Therefore, while this study aimed to produce generalisable findings, not all countries had data available, so a certain amount of caution is needed in applying trends to other countries. Additionally, many of these studies were conducted within the first 5 years of each country’s HPV vaccination programme, so updated analyses would be warranted to determine changes in inequalities over time, as well as to include analysis of gender-neutral vaccination. This review does not include data on vaccination in boys, despite many of the included countries now implementing gender-neutral vaccination. However, this review chose to focus on girls to robustly look at inequalities in vaccination over time, as there is not yet enough comparable data on vaccination in boys, though trends will likely be quite similar, so these findings can still inform gender-neutral vaccination programmes. Additionally, the study uses ‘at least one dose’ of HPV vaccination as the outcome, so findings can inform single-dose vaccination programmes, even though some of the included studies were conducted at a time where countries had a 2-dose or 3-dose regimen, and the effect of one-dose schedules on uptake equitability will certainly need to be monitored. This review focuses only on sociodemographic characteristics associated with HPV vaccination, rather than other childhood vaccinations, because HPV vaccination is unique from other childhood vaccinations due to the older age at vaccination, as well as the association with sexual activity. However, evidence suggests uptake of other school-based vaccinations (Td/IPV and MenACWY) offered to adolescents in England, as an example of a high-income country, is similar to, if not slightly lower than, uptake of HPV vaccination.38 Many countries, including the UK, also framed HPV vaccination as an ‘anticancer’ vaccine rather than an ‘anti-STI’ vaccine to avoid the association with attitudes toward sexual behaviour.
Access to raw, individual-level data to conduct a meta-analysis may also strengthen this work. However, a meta-analysis was not conducted due to the wide variation in study design across the included studies and the lack of individual-level data, which might bias the associations. However, in many studies, vaccination coverage was not presented by population group as a proportion, instead showing only associations with non-uptake (eg, ORs). Rather, study-level associations have been presented and compared. Displaying findings by country, as we have, allows readers to make their own judgements regarding which countries’ evidence is most relevant to them.
Implications for policy
Despite high coverage of HPV vaccination achieved in countries with routine, school-based vaccination programmes, this systematic review demonstrates that inequalities can still occur, with lower SES and minority ethnic groups generally less likely to be vaccinated. Evidence of inequalities in access to vaccination must be considered in the context of herd protection and infection risk. If groups not accessing HPV vaccination are not within the same sexual networks as vaccinated individuals and therefore not protected by indirect effects, or at higher risk for HPV infection due to sexual behaviours, then they may face an even larger increased risk for developing cervical cancer. The combined effects of lower uptake, exposure risk and indirect protection are what matter ultimately: studies of the impact of vaccination should therefore consider these population groups for subanalyses. However, meanwhile, the lower uptake should prompt consideration of mitigating measures.
The changing HPV epidemiology due to vaccination will require major changes to cervical cancer screening programmes in the coming years. Intervals between screens will likely be lengthened for vaccinated cohorts; however, unvaccinated women may still require more frequent cervical screening, and/or a more targeted approach to improve participation in cervical screening. Vaccination and screening programmes may wish to consider implementing campaigns specific to migrants and ethnic minority groups to increase awareness of vaccination in schools and screening opportunities. Additionally, health authorities might consider offering additional vaccination and screening opportunities in areas of lower SES. The use of self-sampling for cervical cancer screening may also help to increase uptake among undervaccinated and underscreened groups.39 A recent systematic review found that culturally sensitivity interventions, information campaigns and self-sampling were promising methods to achieve higher screening uptake among migrant women.40
Therefore, this review’s identification of key groups at risk for lower vaccination uptake as minority ethnic groups and those with lower SES, could inform studies of vaccination outcomes, targeting of mop-up vaccination and targeting of screening to improve disease prevention among these unvaccinated groups and achieve the WHO Cervical Cancer Elimination goals.
Data availability statement
Data sharing not applicable as no datasets generated and/or analysed for this study.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
References
Footnotes
Contributors The paper was conceived by ED, PS, KS and NF. ED wrote the first draft, with further contributions from PS, NF, KS and RO. Title, abstract and full-text screening were done by ED, with RO as a secondary reviewer. Data extraction and analysis were done by ED. All authors contributed to interpretation, reviewed successive drafts and approved the final version of the manuscript. ED is the guarantor.
Funding ED is funded through the UCL Birkbeck MRC Doctoral Training Programme (MR/N013867/1).
Map disclaimer The depiction of boundaries on this map does not imply the expression of any opinion whatsoever on the part of BMJ (or any member of its group) concerning the legal status of any country, territory, jurisdiction or area or of its authorities. This map is provided without any warranty of any kind, either express or implied.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.