The epidemiological and economic impact of a quadrivalent human papillomavirus (hpv) vaccine in Estonia

Evaluations of the cost-effectiveness of high-risk human papillomavirus (HPV) vaccination as a cervical cancer prevention measure are mostly limited to countries in Western Europe and North America [1]. Few studies have assessed the cost effectiveness of HPV vaccination in Central and Eastern European countries [2–5]. Countries in this region tend to have less-developed cervical cancer screening programs, and subsequently have considerably higher age-standardized cervical cancer incidence rates (14.7 per 100,000, 2008) than those in Western Europe (6.9 per 100,000, 2008) or North America (5.7 per 100,000, 2008) [6].

In Estonia the cervical cancer incidence (age-standardized incidence rate 15.8 per 100,000, 2008) and mortality (age-standardized mortality rate 6.2 per 100,000, 2008) are significantly higher than in many developed countries including neighboring Scandinavian countries [6]. Current coverage with any type (opportunistic or systematic) of cervical cytology (PAP smear) testing-based cervical cancer screening is relatively high (3-year coverage of ~72%) in the country [7]. Systematic screening implemented in parallel with ongoing opportunistic screening was introduced in 2006 [8] but has achieved only limited coverage (24% of the target population, women aged 30 to 59 years in 2009). Both bivalent and quadrivalent HPV (HPV2 and HPV4) vaccines are approved for use in Estonia [9].

This study investigated the clinical benefits and economic consequences of routine quadrivalent HPV vaccination of females by the age of 12 years in Estonia. Specifically, this study was designed to evaluate the potential impact in Estonia of prophylactic quadrivalent HPV vaccination on the incidence of cervical intraepithelial neoplasia (CIN), cervical cancer, genital warts, and cervical cancer mortality when added to the current cervical cancer screening and standard of care.

We adapted a previously developed HPV dynamic mathematical model to Estonia (Elbasha & Dasbach, 2010: [10]). Details of the model structure and equations have been published previously [10]. Individuals enter the model as they are born, move between successive age groups at an age- and gender-specific rate per year, and exit the model as they die. The model estimates health benefits and costs in a dynamic population. The model also estimates the impact of vaccination on vaccinees and their contacts (via herd immunity impact).

This is one of the first studies to examine the potential impact of HPV2 and 4 vaccination in Estonia [15]. The incremental cost-effectiveness ratio of the vaccination compared with no vaccination was €4,889 per QALY gained over a time horizon of 100 years. Given that base case as well as all scenarios tested in sensitivity analysis remained below one GDP per capita in Estonia (2011: €11,900 [20]) HPV vaccination can be deemed cost effective (based on WHO recommendations on cost effectiveness [21]) in Estonia.

The analysis of HPV vaccination by Liiv et al. [15] derived a cost effectiveness estimate of €10,019 /QALY (at the 3% discount rate). Variations in the ICER estimates between the current work and that of Liiv et al. are probably explained by the different models used (dynamic transmission model incorporating the effect of herd protection vs Markov model) and several differences in the assumptions made (non-life-long immunity and booster dose inclusion; higher (95%) vaccine coverage; lower proportion (70%) of HPV 16/18 related cervical cancer contribution in Liiv et al.). Such factors are informative for contextualizing the results.

However, both studies are in agreement with the recent systematic review [1] in concluding that while different model structures, input parameters and baseline assumptions have been used to assess the cost effectiveness of HPV vaccination, studies that focused on female-only vaccination programmes usually find this intervention to be cost effective compared with cervical cancer screening alone [1].

Our results document significant health effects related to the combined use (HPV4 vaccination + cervical cancer screening) of cervical cancer prevention methods. The model projected that at year 100, HPV4 vaccination would lead to a relative reduction of HPV 16/18 related deaths and cervical cancer incidence by over 97% and the incidence of HPV 6/11 related genital warts among Estonian women and men by over 94% and 81%, respectively. In addition, there are relevant immediate public health impacts of HPV4 vaccination on genital warts among both women and men. Every year, approximately 751 women and 276 men receive treatment for genital warts in Estonia (based on HIF treatment statistics during 7 years from 2004–2010, ICD 10, diagnosis code A63.0, data not shown). Our model projected a substantial reduction in the incidence of genital warts in the first 25 years.

Based on the sensitivity analysis, the incremental costs per QALY are about 20% higher if vaccination is implemented against a background of improved cervical cancer screening. We found that the cost-effectiveness ratio was heavily impacted by the discount rate and health utility values used. Assuming a limited duration of vaccine protection and reducing the costs of the vaccine led to modest changes in ICER in opposite directions (i.e. limiting duration increased ICER and reducing costs decreased ICER).

Our findings should be interpreted with due consideration for a number of limitations. First, we did not conduct an in-depth literature review to obtain data on the natural history of site- and type-specific HPV infection and disease or on health utility requirements of HPV disease states. Instead, we abstracted relevant data from published studies [10] and we also assessed the cost per life year saved with no quality of life adjustments in the sensitivity analysis. However, for most of the key parameters Estonian specific data were used, and the model was calibrated using cervical cancer incidence/mortality in Estonia. Second, our model is based on a number of efficacy assumptions regarding transient and persistent infection which are not completely available in the published literature. Third, we did not incorporate additional potential benefits of protecting against HPV diseases such as vulvar, vaginal, and anal dysplasia/cancers in this analysis. Data from clinical trials supports the efficacy of the vaccine against these diseases [22–24]. This potentially underestimates the benefits of vaccination. On the other hand, the duration of vaccine protection is uncertain (although we assumed life-long vaccine protection in the base case). Further, the model simulates only the heterosexual transmission of HPV and does not incorporate transmission between homosexual and heterosexual individuals. Only direct medical costs were included in the model, ignoring potential savings in or outside the health care sector. Excluding indirect costs from the analysis most likely leads to undermining of the vaccination effect, given that preventing the disease (cervical cancer) adds to the productive activity. Importantly, the model could not distinguish systematic from opportunistic cervical cancer screening. Systematic screening is generally more effective and cost-effective than opportunistic screening: countries that have organised screening programmes have much lower lifetime testing and subsequent treatment rates than those with only opportunistic screening [25–27]. However, we used sensitivity analysis to evaluate the potential effect of vaccination in combination with the hypothetical/improved (systematic) screening. Our finding of a higher ICER value in the case of hypothetical/improved screening is in agreement with the current understanding of HPV disease and prevention (improved screening results in a lower incidence of cervical cancer, therefore a higher ICER for vaccination strategy). Last but not least, our sensitivity analyses were limited to the inputs related to vaccine properties and coverage, US based health utilities, health care (diagnosis, treatment) costs, and discount rate and did not include all demographic, behavioural, clinical, and other natural history parameters. This was done mainly to manage complexity in the analysis and obtain results in a timely manner for policy evaluations. As a result, the uncertainty results do not reflect the full range of possible factors.

Finally, there are several unresolved issues for successful implementation of cervical cancer prevention, including delineating the optimum set of components (primary prevention, screening, vaccination) to reduce cervical cancer incidence and mortality. Mathematical modelling studies and statistical evidence synthesis are methods that are best suited to investigating the optimal combination of different cervical cancer prevention strategies and should be developed to guide policy decisions about cancer control.

In general, routine vaccination of 12-year-old females appears to be cost-effective in Estonia and may provide both short-term and long-term health gains.