Elsevier

Vaccine

Volume 40, Issue 32, 30 July 2022, Pages 4644-4653
Vaccine

Disparities in Human Papillomavirus vaccination coverage among adolescents in Australia: A geospatial analysis

https://doi.org/10.1016/j.vaccine.2022.06.030Get rights and content

Abstract

Aim

This ecological study aimed to examine the geographic patterns in Human Papillomavirus (HPV) vaccination rates among boys and girls aged 15 years across locations in Australia, in addition to assessing contextual area-level factors that may explain the variations in HPV vaccination coverage.

Methods

Aggregate HPV vaccination data for Australian girls and boys aged 15 years from 2015 to 16 was obtained from the Australian Institute of Health and Welfare for each Statistical Area level 4 (SA4). A Gradient Boosting Machine learning model was applied to assess the predictors’ importance for the study outcomes. Geographically weighted regression (GWR) models were run to assess whether substantially different relationships between predictors and outcomes occur at different locations in space.

Results

Completed HPV vaccination across the 88 SA4 regions ranged from 57.6% to 90.6% among girls, and from 53.6% to 85.5% among boys. The 2016 SEIFA Index of Economic Resources was the variable with the highest contribution to the predictions of both girls’ and boys’ HPV vaccination rates. Selected predictors explained 45% and 72% of the geographic variance in vaccination rates among boys and girls, respectively. Normalised coefficients for both GWR models showed a high variation in the associations between predictors and HPV vaccination rates across regions.

Conclusion

Socioeconomic and education factors were important predictors for HPV vaccination rates among Australian boys and girls aged 15 years, although no variable presented a uniform effect on HPV vaccination across SA4 regions. Important spatial heterogeneity in the effect of predictors was identified across the study area.

Introduction

Australia has been a pioneering country in establishing innovative public health strategies to reduce the burden associated with human papillomavirus (HPV) infections [1]. Launched in 2007, the Australian HPV Vaccination Programme initially provided immunisation for girls through a school-based approach and for young women aged up to 26 years. More recently, in 2013, Australia became the first country to expand the eligibility for HPV vaccination by implementing a strategy of gender-neutral HPV immunisation for adolescents [1]. This step towards providing universal HPV vaccination for boys and girls represented a major advance in terms of health equity and program acceptability in Australia [2]. While an increasing number of countries have introduced HPV vaccination (approximately 55% of the World Health Organisation Member States), coverage rates remain extremely low, reaching only 15% of girls in the target age for HPV vaccination worldwide [3]. Only one-third of the HPV vaccination programs worldwide are gender-neutral, mostly in high-income countries [4].

Positive health and economic effects of HPV vaccination have been widely documented [5], [6], [7]. The introduction of HPV vaccines and the implementation of high-coverage HPV immunisation programs have created a promising opportunity to eliminate the burden of HPV-related diseases within a few decades [8]. It is estimated that 630,000 new cases of cancer per year are attributable to HPV worldwide, accounting for 4.5% of all cancers [9]. Although the vast majority of HPV-related malignant lesions are cervical cancer (84%), there has been an increase in the incidence rates of HPV-attributable cancers involving other sites such as oropharynx, anus, and penis [10], [11], [12]. While HPV-attributable oropharyngeal cancer is particularly predominant in high-income countries, approximately two-thirds of cervical cancer cases are concentrated in low- and middle-income countries [9].

The recent development of cutting-edge geospatial technologies has allowed social epidemiologists to examine the effects of area-level factors on health outcomes, contributing to better understanding of mechanisms that lead to health disparities within and across populations [13]. Geographic heterogeneity in factors related to population characteristics (such as ethnicity, age, and educational status) and the context in which people live (such as neighbourhood-level deprivation and proximity to health services) can help explain disparities in vaccination coverage, in addition to identifying at-risk populations in need of specific interventions [14]. Trogdon and Ahn used geospatial methods to investigate the geographic distribution of HPV vaccination uptake among Children in North Carolina, showing a clustering effect of neighboring areas for boys but not for girls [15]. A strong spatial dependence of HPV vaccination rates across ZIP codes in Minnesota was reported by Nelson et al, demonstrating the need to employ spatial models to accurately estimate associations with HPV vaccination [16]. Staples et al. revealed contrasting effects of area-level socioeconomic factors for HPV vaccine initiation and completion among adolescents [17]. Other studies have shown that area-level characteristics such as ethnic composition, population density, gender distribution, area-level socioeconomic status, geographic location, and local health services are associated with the geospatial patterns of HPV vaccination coverage across different populations [18], [19], [20].

In Australia, the school-based HPV vaccination program has achieved meaningful outcomes, reaching 80.5% and 77.6% of all boys and girls aged 15 years in 2020. However, inequalities in vaccine course completion remain substantial across states and between Indigenous and non-Indigenous groups (for instance, the HPV vaccine course completion rate for Indigenous boys in South Australia is as low as 47.5%) [21]. Regional disparities in HPV vaccination coverage are likely to further expand the gaps related to the incidence of HPV infections and HPV-attributable cancers between groups. In this ecological study, we adopted a machine learning approach and a sophisticated geospatial technique to model the geographic distribution of HPV vaccination rates among Australian adolescents. This study aimed to 1) examine the geographic patterns in HPV vaccination rates among boys and girls aged 15 years across locations in Australia; and 2) assess contextual area-level factors that may explain the variations in HPV vaccination coverage.

Section snippets

Methods

This study adopted the 2011 Australian Statistical Areas 4 (SA4) as the geographic scale for the analyses. SA4 are the largest regions within states in the Main Structure of the Australian Statistical Geography Standard (ASGS) and include a minimum population of 100,000 people [22]. Non-spatial SA4s (Migratory - Offshore - Shipping and No Usual Address) were not considered.

Main outcomes

Area-level data for this secondary analysis were obtained from official Australian Government agencies. Publicly available data on 2015–2016 HPV vaccination rates for boys and girls aged 15 years were obtained from the Australian Institute of Health and Welfare (AIHW). Records are managed by the National HPV Vaccination Program Register (HPV Register) and include data on boys and girls who had completed the three-dose vaccination scheme for HPV by 30 June 2016. Vaccination rates were computed

Data processing

A correlation matrix was estimated to assess the multicollinearity among the explanatory variables. Correlation coefficients greater than 0.75 were used as the criterion to drop redundant predictors [26].

Gradient Boosting Machine model

The Gradient Boosting Machine (GBM) model is an ensemble machine learning algorithm that trains decision trees in a gradual, progressive, and additive, approach. By iteratively training new trees based on the revised residuals of previous steps, GBM can successfully identify areas where data is

Results

Overall, 216,575 (76.9%) Australian adolescents aged 15 years had completed the recommended HPV immunisation scheme in 2015/2016. Girls (79.9%) had a slightly higher vaccination rate compared to boys (73.9%). A different pattern was observed in the remote areas of South Australia, where Boys presented higher coverage (65.7% against 62.2% for girls). HPV vaccination rates across the 88 SA4 regions varied from 57.6% and 90.6% among girls, and from 53.6% and 85.5% among boys. Areas with higher

Discussion

This study examined the geospatial patterns of HPV vaccination rates among Australian girls and boys aged 15 years. We selected a range of area-level factors to help elucidate regional disparities in HPV immunisation and applied machine learning algorithms to identify the most relevant predictors. Geographically weighted regression models were employed to explore variations in the importance of each predictor across different SA4 regions. Findings demonstrate that there is an unequal geospatial

Conclusion

Area-level socioeconomic and educational factors were the most important predictors of regional HPV vaccination rates among Australian boys and girls. No factor presented a uniform effect on HPV vaccination across SA4 regions. Instead, important spatial heterogeneity in the effect of predictors was identified across the study area. A higher proportion of adults who never attended school was strongly associated with lower vaccination rates of boys and girls living in Western Australia. The SEIFA

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (53)

  • M.A. Smith et al.

    Effective HPV vaccination coverage in Australia by number of doses and two-dose spacing: What if one or two doses are sufficient?

    Tumour Virus Res

    (2021)
  • G. D'Souza et al.

    Understanding personal risk of oropharyngeal cancer: risk-groups for oncogenic oral HPV infection and oropharyngeal cancer

    Ann Oncol

    (2017)
  • G.M. Bryden et al.

    The privilege paradox: Geographic areas with highest socio-economic advantage have the lowest rates of vaccination

    Vaccine

    (2019)
  • L.-A. McNutt et al.

    Affluence as a predictor of vaccine refusal and underimmunization in California private kindergartens

    Vaccine

    (2016)
  • S. Kurani et al.

    Socioeconomic disadvantage and human papillomavirus (HPV) vaccination uptake

    Vaccine

    (2022)
  • J. Sisnowski et al.

    Differences in school factors associated with adolescent HPV vaccination initiation and completion coverage in three Australian states

    Vaccine

    (2021)
  • C. Patel et al.

    The impact of 10 years of human papillomavirus (HPV) vaccination in Australia: what additional disease burden will a nonavalent vaccine prevent?

    Euro Surveill

    (2018)
  • K.L. Rosettie et al.

    Cost-effectiveness of HPV vaccination in 195 countries: A meta-regression analysis

    PLoS ONE

    (2021)
  • C. de Martel et al.

    Worldwide burden of cancer attributable to HPV by site, country and HPV type

    Int J Cancer

    (2017)
  • T.A. Berman et al.

    Human papillomavirus in cervical cancer and oropharyngeal cancer: One cause, two diseases

    Cancer

    (2017)
  • World Health Organization

    Human papillomavirus vaccines: WHO position paper, May 2017-Recommendations

    Vaccine

    (2017)
  • X. Shi et al.

    Introduction: advances in geospatial analysis for health research

    Ann GIS

    (2020)
  • S.F. Awad et al.

    Implementation of a Vaccination Program Based on Epidemic Geospatial Attributes: COVID-19 Pandemic in Ohio as a Case Study and Proof of Concept

    Vaccines

    (2021)
  • Trogdon JG, Ahn T. Geospatial patterns in human papillomavirus vaccination uptake: evidence from uninsured and publicly...
  • E.J. Nelson et al.

    Geospatial patterns of human papillomavirus vaccine uptake in Minnesota

    BMJ Open

    (2015)
  • Do EK, Rossi B, Miller CA, Ksinan AJ, Wheeler DC, Chukmaitov A, et al. Area-Level Variation and Human Papillomavirus...
  • Cited by (3)

    View full text