2,999
Views
1
CrossRef citations to date
0
Altmetric
Review

A review of data systems for assessing the impact of HPV vaccination in selected high-income countries

ORCID Icon, , , , ORCID Icon, ORCID Icon, ORCID Icon, ORCID Icon, & ORCID Icon show all
Pages 161-179 | Received 22 Jul 2022, Accepted 21 Dec 2022, Published online: 09 Jan 2023

ABSTRACT

Introduction

The introduction of effective human papillomavirus (HPV) vaccination, screening, and treatment programs has led the World Health Organization to call for the global elimination of cervical cancer. Assessing progress toward this goal is supported through monitoring vaccination coverage and its impact.

Areas covered

We performed a targeted review to assess the characteristics of HPV-related data systems from seven high-income countries (HICs) that represented varied approaches, including Australia, Canada, France, Italy, Scotland, Sweden, and the United States (US). Included data systems focused on preventive and early detection measures: HPV vaccination and cervical screening programs, as well as HPV-related disease outcomes. Differences were observed in approach to development of data systems, along with variation in geographical scope and methods of data collection.

Expert opinion

A challenge exists in how to best follow-up the ongoing global-scale elimination efforts in a comprehensive manner. These sources provide a wealth of information regarding the strengths and limitations of, and notable variation among, current data systems used in HICs. This review can inform improvements to existing prevention programs and the implementation of new programs in other countries, and thus support optimization of cervical cancer prevention policy.

1. Introduction

Human papillomavirus (HPV) infection is well established as a necessary cause of cervical cancer, and an important cause in a proportion of other anogenital (anus 91%, vagina 75%, vulva 69%, and penis 63%), and head and neck cancers (oropharynx 70%) [Citation1–3]. Globally, approximately 4.5% of newly diagnosed cancers are attributable to HPV, with cervical cancer being the fourth most common female cancer [Citation4,Citation5]; an estimated 604,127 cases are diagnosed annually [Citation2,Citation5,Citation6].

Given that cervical cancer is highly preventable through prophylactic HPV vaccines and detection and appropriate treatment of HPV-associated pre-cancerous lesions, the World Health Organization (WHO) made a global call to action for its elimination as a public health issue, by reducing its incidence to the level of a rare disease (less than 4 per 100,000 woman-years) [Citation7–9]. As high rates of vaccination, cervical screening, and treatment are fundamental to achieving this aim, the WHO has established interim targets, to be achieved by 2030, to facilitate elimination (‘90–70-90 strategy’) [Citation9]. These include full vaccination among 90% of girls by age 15, high-performance screening among 70% of women at least by age 35 and 45 years, and appropriate treatment initiated for 90% of women diagnosed with cervical malignancy [Citation9]. Currently approved HPV vaccines are available in various formats including: bivalent (Cervarix, Cecolin) [Citation10,Citation11], quadrivalent (Gardasil, Cervavac) [Citation12,Citation13], and nonavalent (Gardasil 9) [Citation14]. Related to this, it has been predicted that achieving the 90–70-90 targets by 2030 could result in a 42% reduction in cervical cancer incidence by 2045, and 97% by 2120, averting more than 62 million deaths by 2120 [Citation9].

Since 2006, at least 115 countries have implemented HPV vaccination programs to help reduce the burden of cervical cancer [Citation15,Citation16]. While the scope of these programs is varied, at the time of writing, approximately 40 countries had begun gender-neutral vaccination initiatives; this includes the United States and Australia (who were early adopters of this approach and began vaccinating males in 2011 and 2013, respectively) [Citation17–20]. Several efforts have been made to evaluate the impact of these programs [Citation21–24]. However, whilst surveillance of the prevalence of HPV and its related outcomes is not recommended by the WHO as a prerequisite to implementing vaccination programs, it can play an important role in documenting both the benefits and progress of vaccination, as well as informing future policy [Citation15,Citation25–28].

Because of the range of possible tools and processes in preventing HPV-related diseases, there are several interventions and biological endpoints which are relevant for monitoring and analyzing the impact of HPV vaccination programs [Citation29]. Disease prevention through vaccination programs targeting HPV aims to reduce rates of cervical cancer as well as other HPV-related cancers: therefore, precancer and cancer incidence are important outcome variables [Citation25]. However, due to the long latency in disease onset, more immediate outcomes, such as the incidence of HPV infection (especially persistent, high-risk HPV infection), or short-incubation diseases such as genital warts (GW), can be used as surrogate impact measures [Citation25].

Vaccination programs are increasingly demonstrating success through impact and effectiveness studies, which show decreasing prevalence rates of both infection and subsequent HPV-related disease (including cervical precancer and cancer) [Citation30]. However, whilst details are available for individual data systems, an overview detailing the characteristics of data systems that enable such analyses is not currently available.

This review focused on selected high income countries (HIC), to help identify, understand, and summarize the characteristics of existing data systems within countries that have established HPV vaccination programs and associated monitoring systems [Citation31]. These findings offer insight into factors that are common across successful programs established in HICs. This, in turn, can be used to inform the design and development of future monitoring and evaluation systems, which is of relevance for low- and lower-middle income countries (LMICs) where program implementation is ongoing.

2. Methods

To identify data systems that have been used in HICs, we performed an initial targeted literature review and gray literature search. Using the World Bank classification of HICs as having a gross national income of >12,235 per capita, several HICs were identified with potentially relevant HPV-related data systems [Citation31]. Of these, seven countries with established vaccination programs were selected based on the expertise of the contributing authors, to provide illustrative findings on different approaches to HPV outcome monitoring and in program approach (even among countries where vaccination programs were initiated in the same year). The selected countries (and years when national HPV vaccination programs were launched) were the United States (2006), Australia (2007), Canada (2007), France (2007), Italy (2008), Scotland (2008), and Sweden (2012) [Citation32]. Of note, Scottish data are based on public domain descriptions only; however, for the other countries, contributing authors have provided their expert insight to augment the publicly available information.

Publications reporting on the above countries of interest were sourced from PubMed, Google, Google Scholar, and other sources (including websites of public health bodies, and the bibliographies of key references). These searches focused on identifying data systems containing relevant information on the prevalence of HPV infection, related diseases, and vaccine coverage (primarily those detailing prevention and early detection). Systems describing treatments for HPV-related diseases, such as cervical cancer, are also key to the elimination of this public health issue but were not the focus of this review.

Included publications report on records and data systems for surveillance of at least one of the following: cervical screening, abnormalities, and cancer; anal, vulvar, vaginal, or penile lesions and cancer; genital warts; recurrent respiratory papillomatosis (RRP); HPV-related disease, infection, or vaccine coverage.

Data presented in this review include the objectives of key data systems in each country of interest, along with information on data collection (including regularity and timings), data linkage, and scope of included outcomes (e.g. monitoring of HPV infection status or genotype).

Characteristics of data systems captured included: mechanisms of vaccine distribution and assessment of coverage, populations targeted, healthcare system structures, types of surveillance mechanisms, and range of conditions. Qualitative assessment of the characteristics of HPV vaccination data surveillance amongst these high-income countries could thereby allow a comparison of approaches and guide development of less established programs going forward.

3. Results

Data systems containing data relevant to HPV vaccination were reviewed in 2021 (see for an overview of the number of data systems for each outcome; see Supplementary material Table S6 for a full breakdown by country).

Table 1. Summary of the number of identified HPV-related data systems for each outcome of interest from seven HICs.

Identified data systems were classified according to the following reporting categories: vaccination coverage, cervical screening, early-onset HPV-related disease outcomes, and HPV-related cancer outcomes, with several data systems being included in more than one category.

The data sources were categorized as claims databases, databases of electronic health records (EHRs), registries, or surveys. Claims databases are defined as those detailing diagnosis and procedure codes associated with medical, pharmacy and dental claims documented by private and/or public payers. EHR databases systematically collect patient health information in a digitalized format. Registries refer to disease- or intervention-specific databases that collect information indirectly (such as cancer or cervical screening registries, or vaccine registries). Lastly, surveys obtain information directly from respondents by asking them to provide information or insight into a topic of interest such as sexual behaviors or vaccination status.

The characteristics of key data systems are detailed in , to provide an illustration of their varied objectives and outcomes.

Table 2. Key data systems reporting on vaccination coverage in seven HICs.

Table 3. Key data systems reporting on cervical screening in seven HICs.

Table 4. Key data systems reporting on HPV infection and early-onset HPV-related diseases in seven HICs.

Table 5. Key data systems reporting on HPV-related precancers and cancers in seven HICs.

3.1. Data systems on vaccination coverage

All of the HICs included within the review had a system to allow vaccination coverage to be determined (). Australia, Scotland, and Sweden have a national immunization registry or equivalent system, whereas the US, Canada, and Italy have relied on survey-based approaches or regional registries to collect these data. In France, such data are collected by monitoring reimbursement claims, with databases such as the Permanent Sample of Beneficiaries (PSB) supporting determination of uptake [Citation33,Citation34].

3.2. Data systems on cervical screening

Key identified data systems from the different HICs reporting on cervical screening programs are summarized in . Screening systems were present within all of the identified HICs; however, some fundamental differences were identified. National-level, organized monitoring or surveillance systems were present in Australia, Scotland, and Sweden, with national survey reports also being available in Italy. Other HICs (Canada, France, and the US) do not maintain national-level organized screening programs but have jurisdictional coverage via organized or opportunistic screening programs.

Australia and Sweden have established organized screening programs, which invited women from a prespecified population to attend at regular time intervals [Citation35–38]. Meanwhile, only some provinces in Canada maintain organized screening; most other jurisdictions rely on opportunistic screening for the detection of cervical precancer and cancer, i.e. screening based on professional guidelines done by a primary healthcare provider when the opportunity arose, rather than as part of a predetermined program [Citation39,Citation40]. In France, an organized screening program has been implemented since 2018 [Citation41], which coexists with preexisting systems that involve opportunistic screening.

Interestingly, all of the identified HICs currently use cytology-based screening, except for Australia, which switched to national HPV testing in December 2017 [Citation28,Citation42,Citation43]. Data relating to screening were collected through a combination of administrative mechanisms, registry reports, and surveys. Amongst all of the identified surveillance systems, few had reported provisions to enable data linkage i.e. the ability to bring together information from different sources to permit enhanced multi-source surveillance.

HIC: high-income country; HPV: human papillomavirus; NR: not reported.

3.3. Data systems on HPV infection and early outcomes

Data systems relating to early HPV infection and related outcomes (for example, genital warts) were more limited (). In addition, methods of data collection varied between countries and not all data were routinely collected.

Routine collection was particularly lacking in relation to HPV infection status and genotyping; however, sentinel site-based data were identified in several countries, which set a baseline for infection. For instance, in France, the Center National de Référence des HPV (CNR-HPV) biospecimen database for the screening of cervical cancer also recorded data on the prevalence of high- and low-risk HPV types [Citation44]. In Australia, a research project (WHINURS; Women HPV Indigenous Nonindigenous, Urban Rural Study) estimated the genoprevalence of HPV in women presenting for Pap cytology screening before vaccination [Citation45]; thereby contributing to subsequent vaccine impact research. Subsequently, a national surveillance system has been funded by the Commonwealth Department of Health, including monitoring of genoprevalence among males as well as females [Citation46].

Only one routine surveillance system was identified for the identification and monitoring of genital warts: the Kirby Institute Genital Warts Surveillance Network in Australia [Citation47]; genital warts are typically not a reportable condition under regulations on notifiable/reportable health care events. However, several administrative claims databases and drug registries were identified, alongside surveillance activities within sexual health clinics and general practice, which could be used to help determine incidence. For instance, in Sweden, as the therapies used to treat genital warts are only dispensed via prescription, information on the quantity of pharmaceuticals within the Swedish Prescribed Drug Register (PDR) can be used to inform incidence [Citation34]. Surveillance systems for RRP specifically were identified in Australia, Canada, France, and the US [Citation48–52].

GW: genital warts; HIC: high-income country; HPV: human papillomavirus; NR: not reported; RRP: recurrent respiratory papillomatosis.

*While disease registries in Sweden do not specifically capture RRP diagnoses, studies have demonstrated that the ICD-10 code D14.1 (benign neoplasm of larynx) may be used to monitor incidence using Swedish registries [Citation53].

3.4. Data systems on HPV-related cancers

With regards to HPV-related cancers, population-based cancer registries were identified for all of the included HICs. Many countries had multiple databases to help determine both HPV-related cancer incidence and mortality rates (). Australia had a particularly comprehensive approach, with information available both nationally through databases such as the Australian Cancer Database (ACD), and regionally through databases covering regions such as New South Wales, Victoria, and Queensland [Citation35,Citation54]. Similar to Australia, Sweden had nationally-based data systems, with registries such as the Swedish Cancer Register (Cancerregistret), which provided information not only on incidence and mortality, but also (sometimes to a varying extent) on factors such as stage and grade [Citation55,Citation56].

4. Discussion

High rates of vaccination are fundamental to achieving the aim of eliminating cervical cancer, as determined by the WHO. At the time of writing, at least 115 countries have begun vaccination programs for HPV, and at least 40 countries are pursuing gender-neutral vaccination strategies. It is to be noted, however, that for young girls of vaccine eligible age only 13% are fully vaccinated [Citation57]. Therefore, the availability of data systems to monitor vaccination rates and related outcomes is key.

This review summarizes data systems relating to HPV vaccination impact monitoring and evaluation, across a set of high-income countries (HICs). We provide pertinent characteristics for individual data systems, examining the degree of coverage across different process and outcome variables (and population segments of interest). Further, we highlight differences in approach between countries, including associated strengths and limitations. Some of the issues are related to specific features of the prevailing healthcare systems, e.g. national in Australia and Sweden versus regional in a federated system, such as Canada and Italy.

Planning and deliberately setting up a holistic data system that captures linked data on interventions and outcomes is ideal, as has been observed in Australia, especially for assessment of vaccine effectiveness [Citation29]. In addition, the availability of national-level databases for key several outcomes, such as has been observed in Sweden, is likely beneficial. Where impractical or unfeasible, other sources such as administrative data provide opportunities to assess the impact of preventive interventions [Citation25]. Further, while national systems ensure data consolidation, appropriate access and analysis could negate any limitations of a more-fragmented approach. As such, the timely collection and compilation of health data at both the national and sub-national levels can enable progress to be measured toward specified targets, which is of increasing importance given the introduction of performance-based disbursement by several organizations [Citation26].

We identified several data system characteristics that exposed probable limitations in their effectiveness. For example, reliance on voluntary participant inclusion and surveys (such as in Italy), which may affect the breadth and accuracy of available data; in comparison to data recorded in registries by healthcare professionals and/or through structured mechanisms of systematically available administrative systems. Conversely, administrative data systems or data systems using medical records (as seen in the US) may also be inherently limited in terms of coverage (e.g. through collecting information for insured individuals only).

There were also limitations in the time duration of reporting and/or accessibility of the data, such as data available from a one-time event with no follow up i.e. a standalone survey, or data collected within a specific timeframe that has since been discontinued. Continuous consolidated reporting into national registers is preferable for effective monitoring and evaluation [Citation25]. However, country restrictions on patient data use may lead to use of targeted or sentinel-site-based studies.

In addition, the emergence of the COVID-19 pandemic in early 2020 has affected subsequent HPV vaccination or monitoring activities in a way that is difficult to predict. The continuity of data collection activities in areas such as HPV has been disrupted due to the pandemic. However, increased awareness of population health outcomes, and new systems put in place to improve such outcomes in the context of COVID-19, may prove beneficial for control of other conditions such as HPV and associated cancers.

In general, procedures for comprehensive data linkage constitute a significant gap in many countries. This hinders effective data consolidation and comparison, and should be an important consideration moving forward, particularly in light of the elimination goals established by the WHO. There is, however, strong potential amongst several national cancer registries for data linkage, due to the use of unique identifiers. An additional area of note is that most countries did not have systems in place for routine recording and monitoring of HPV prevalent infection, genital warts, RRP and precancer incidence. Due to the long latency associated with HPV infection, collecting data on early endpoints (particularly prevalent infection rates and incidence of genital warts) has been essential to understand the impact of interventions in the short-term, and remains a priority [Citation30]. However, owing to both the scope of information already recorded and the prevalence of existing screening programs, there is potential for a number of the identified cancer programs to be expanded to include precancers. In addition, as organized cervical screening efforts become more common, surveillance of HPV prevalence will also become more feasible.

Despite accumulating experience in many countries worldwide, the practice of recording information about national HPV vaccination programs and outcomes is still developing in many places. As such, relevant data systems may not have been captured by the available literature, and limited information was available regarding many characteristics of identified systems. Further, this review does not describe data systems from all HICs, rather with focus on seven countries that were selected as priority and then characterized through targeted searches. However, whilst limitations exist, this review provides an important information resource to support the strengthening and development of HPV data monitoring initiatives.

4.1. Conclusion

Data systems play an important role in assessing the impact of preventative interventions on HPV-related disease outcomes, particularly in light of the ongoing global cervical cancer elimination effort. This work highlights the need for data collection for all HPV-related disease endpoints, ranging from HPV prevalent infection and genital warts to invasive cervical cancer and beyond, due to the long lag time seen for oropharyngeal and other HPV related cancers to occur. Effective systems will therefore require sustainability and longevity. The characteristics of the established systems identified here thereby provide learnings for countries planning to create or expand data systems based on their local needs. This is of particular pertinence for low- and low-middle income countries where vaccination programs are limited, and where more than 85% of all cervical cancer cases and deaths occur [Citation26,Citation58].

5. Expert opinion

Sustained population-based epidemiologic surveillance is essential for evaluating the short-term impact and monitoring the long-term benefits of public health interventions. In light of the cervical cancer elimination goals enacted by the WHO, understanding effective approaches to real-world data collection on coverage of HPV vaccination, adherence to screening, and treatment recommendations is key to monitoring effectiveness of public health interventions against HPV and associated cancers. This overview of high-income countries’ surveillance of HPV-associated diseases serves as a knowledge base for considering best practices on implementing HPV-related data systems, their similarities and differences, and considering contextual factors in their establishment.

Improving completeness and accuracy of data collection is a priority toward WHO’s elimination goals and country-specific policies to control HPV-associated diseases. However, simply maintaining individual data collection mechanisms specific to individual objectives, such as measuring coverage of HPV vaccination or tracking HPV screening, is not enough. It is imperative that mechanisms be put in place to enable integration of data systems for higher level surveillance of all fronts in prevention, thus allowing linkage at the individual level of HPV vaccination registries with monitoring of HPV prevalence via surveys and with screening and treatment outcomes. Integration and linkage of data systems will also enable robust epidemiologic research on possible bottlenecks and mishaps along the chain from primary prevention to incidence of and mortality from HPV-associated cancers, e.g. by allowing to countries to identify disparities in access, monitor for possible shifts in HPV type distribution in cancer causation, and take policy corrections in screening algorithms. Although most countries in this overview maintain effective surveillance of multiple interventions, processes, and clinical outcomes, at present fully integrated systems are rare because of privacy protection legislation that prevail in most jurisdictions. Data collection relating to interventions (e.g. vaccination), processes (e.g. cervical cancer screening), and outcomes (prevalence of infection, early detection of precancer, and cancer incidence) are currently collected as independent non-linked activities. Comprehensive data systems with valid data sources and reliable linkage among segmented platforms will allow researchers to develop actionable conclusions from the collected data.

Strengths and limitations of existing data systems in high income countries should be closely considered by those who are designing or implementing new systems in LMICs. Such limitations may include fragmented collection of data by sub-national entities (without organized national pooling of data), collection of incomplete or non-validated data that is not ‘research-grade,’ or lack of coverage of surrogate endpoints. In all, these limitations hinder a proper understanding of the burden of disease, or the effectiveness of interventions. The groundswell of support in recent years for universal health care is a positive development, by placing political will in support of the aforementioned tenets for effective, centralized data collection and rational integration of resources and information.

Serendipitously, since the onset of the SARS-CoV-2 pandemic in 2020, countries have had to rapidly adopt disease monitoring, deploy COVID-19 vaccination, and track vaccination coverage using real-time digital platforms. It would be a wasted opportunity not to apply the hard lessons learned with pandemic control and surveillance, as well as the key learnings from research into COVID-19 vaccination, to data collection activities in other preventable infectious diseases, such as those caused by HPV. Improvements in this area may contribute to a greater understanding of the effectiveness of newer intervention paradigms, such as gender-neutral HPV vaccination.

Writing and editorial support was provided by Adelphi Values PROVE and was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. The authors wish to acknowledge Anna Giuliano and Maria Giwa, who contributed to the conceptualization of this research.

Article highlights

  • This review presents an overview of existing data systems of HPV-related outcomes across seven high-income countries, which can be used as a measure of progress against the World Health Organization target for elimination of cervical cancer

  • Data systems capturing HPV-related outcomes such as vaccination coverage, cervical screening, HPV infection rates, and HPV-related cancers were included in the review

  • There were substantial differences observed across collated data systems, including in data collection methods and geographical scope

  • This review presents an overview of current data system practices in selected countries, and can provide a basis for establishing new programs of surveillance in other countries

Declarations of interest

None of the authors received funding for the writing of this manuscript. K Sundstrom has received research grants and consultancy fees to her affiliating institution, Karolinska Institute. S Garland is a member of the Merck Global Advisory board and has received (through her institution) funds for an investigator-initiated grant on HPV in young women, funds for lecture fees and is a member of a Merck vaccine advisory board. EL Franco has served as occasional advisor to Merck, GSK, Roche, and BD. His institution has received funds from Merck in support of a publicly supported, investigator-initiated study. He is co-holder of a patent on methylation markers for cervical cancer screening. P Bonanni received grants for epidemiological and HTA research from different vaccine companies (GSK, MSD, Sanofi Pasteur, Pfizer, Seqirus, Astra Zeneca) and fees for taking part to advisory boards or educational events on different vaccines from the same companies and from Janssen and Moderna. V Wang, S Kothari, and YT Chen are employees of Merck & Co. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or material discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Authors contributions

All authors have contributed to the conception and design of the review and interpretation of results. All authors were also involved in the writing and review of the presented manuscript.

Supplemental material

Supplemental Material

Download MS Word (64.1 KB)

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14760584.2023.2162505

Additional information

Funding

This manuscript was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. The funders of the study had a role in study design, data collection, data analysis, data interpretation, and the writing of the report. The corresponding author had full access to all data in the study and final responsibility for the decision to submit this review for publication. Editorial support was provided by Adelphi Values PROVE and was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. The literature review was completed by Certara and was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA

References