Background
Cervical cancer is the third leading cause of cancer deaths among women in Lao People’s Democratic Republic (Lao PDR), with an estimated crude incidence and mortality rates of 9.8 and 5.3 per 100,000 women annually [
1]. The vast majority of cervical cancers are caused by infection with human papillomaviruses (HPV), particularly HPV types 16 and 18. The high fatality rate associated with cervical cancer in Lao PDR is probably due to various factors. This includes the lack of a national HPV vaccination programme, lack of effective chemo/radiotherapy in the country, and delays in diagnosis [
1]. The delay in diagnosis is in great part due to the fact that there is no national cervical cancer-screening programme in Lao PDR, where it has been estimated that only 5 % of females aged 18–69 in urban areas and 1 % in rural areas are screened every 3 years [
1].
A systematic screening programme might reduce the disease burden, but may not be possible in Lao PDR due to various reasons including financial and sociocultural barriers, poor healthcare infrastructure, and poor performance of laboratory tests [
2]. Given these problems, HPV vaccination might be a more suitable approach for the country. It has been shown to be efficacious, with bivalent and quadrivalent vaccines providing extremely high rates of protection against high grade cervical intraepithelial neoplasia (CIN 2/3) related to HPV types 16 and 18 [
3]. Moreover, Goldie and colleagues [
4] showed that HPV vaccination in preadolescent females is very cost-effective in 72 Global Alliance for Vaccines and Immunization (GAVI)-eligible countries, including Lao PDR, and Jit and colleagues [
5] found similar outcomes on a global scale. However, no nationwide vaccination strategy has so far been implemented in Lao PDR.
A HPV vaccination pilot project, which consists of vaccinating 5th grade schoolgirls (10–13 years old) in the Lao capital Vientiane, and in neighbouring Vientiane Province, is currently taking place. It is likely that such a HPV vaccination programme will become routine practice in the future. However, vaccination coverage might be low. Considering this eventuality, the authors thought it is important to evaluate the benefit of complementing such a vaccination programme with additional interventions, such as adding a catch up vaccination campaign and/or a 10-year-old boy vaccination element. In order to examine these questions, we used a mathematical modelling approach to estimate the cost-effectiveness of various HPV vaccination strategies in the Lao context.
Discussion
Our model suggests that vaccinating 10-year-old girls is very cost-effective even if the vaccine is expensive (I$ 100 per dose). Adding a boy vaccination component produces little additional benefit, with a further reduction in cervical cancers of just of 3.4 %. As a result, adding this component is less effective than a girl vaccination along with a catch-up vaccination component for 11–25 year-old women, which results in a further reduction of 8.9 % in the number of cancers and an additional diminution of 5 DALYs per 1000 women. This catch-up vaccination component becomes the most attractive strategy with a cost per DALY below one GDP per capita. This result is similar to that found in a previous review [
17].
Moreover, the model predicts that adding a catch-up component for females aged 11–25 years old is more attractive than adding a catch-up component for an older age group, if GDP per capita is considered. This age group was also found to be cost-effective in the studies of Elbasha and colleagues [
18] and Dasbach colleagues [
19]. However, to provide more comprehensive information regarding the appropriate maximum limit age in the catch-up component, which was not reported in previous studies [
19], we compared further maximum ages, from 18 to 75, using 5-year intervals. Our study found that a catch-up component for women up to 40 years old was the most attractive option, costing less than one GDP per capita per DALY averted. Several reasons can be proposed. First, the ideal age for a catch-up component might depend on sexual behavior. Second, the prevalence of HPV infection in our model simultaneously decreases after 40 years of age. Finally, one should consider that the incidence of cervical cancer increases after 40 years of age. However, our results should be interpreted cautiously because our model was not calibrated to age-specific HPV prevalence, although the trend of HPV prevalence in Lao PDR seems to be similar to that found worldwide [
20]. Also, a clinical trial showed that the vaccine was safe and that it conferred a high-level of immunogenicity in women up to the age of 45 years [
21].
Nevertheless, in the case of a higher burden of the disease, or waning of natural immunity, or a suboptimal protection from the vaccine in terms of duration, effectiveness, or vaccination coverage, implementing a catch-up component for females aged 11–75 years old is the most attractive option. This parameter’s influence was also reported by Jit and colleagues [
6] and by Van de Velde and colleagues [
22]. Indeed, the effectiveness of the vaccination increases when 1) the incidence of cervical cancer is high or 2) the natural immunity wanes; contrarily, this effectiveness decreases in other cases. However, both situations lead to the same conclusion, which is that it is more efficient to vaccinate larger female populations. The lower effectiveness of vaccines might be true in developing countries due to the fact that the HPV vaccine also requires an appropriate maintenance and delivery process [
23]. In Lao PDR, a low optimal efficacy of vaccination was reported for hepatitis B vaccine, at only around 65 %. The rate is even lower still in rural areas [
24]. Moreover, a low vaccination coverage might be found in rural settings where fewer girls have been found to attend school regularly [
8].
Furthermore, it is more cost-effective to include a boy vaccination component in addition to the catch-up component to the girl vaccination program if the time covered by the simulation is shorter, 30 years for instance. This reflects the insufficient level of vaccination protection in the population in the early stages of implementation.
Our study had some limitations. First, our model did not take into account any cross-protection provided by the vaccine to other HPV-related diseases, such as warts and other cancers. This might underestimate the total DALYs averted related to all HPV types. However, this might not significantly bias our conclusion because of the slight benefits provided by this cross-protection [
25]. Second, we have ignored some items related to screening and treatment. These include the cost of specimen delivery and the cost of treatment complications. This might lead to an underestimation of the total cost per person. However, according to Goldhaber-Fiebert and Goldie [
26], these cost components are small relative to the cost of screening and treatment. Third, it is likely that newer vaccines, active against multiple HPV types, will provide even greater levels of protection [
27]. It has also been reported that two doses of HPV vaccine are equally effective in producing immunity as three doses [
28]. This might further reduce the cost, and subsequently increase the cost-effectiveness of the vaccination, as demonstrated in a cost-effectiveness study in the UK [
29], As our study did not take into account these aspects, future studies might be necessary to investigate these factors for Lao PDR. Finally, DALYs as the outcome of interest might be less interpretable compared to Quality-Adjusted Life Years (QALYs), because they do not reflect as precisely sociocultural considerations regarding health states. Moreover, the discounting rate used for DALY is controversial because it leads to age discrimination. However, there is no valid instrument to measure utilities in the Lao population. A standard gamble method was tried out to produce utilities scores for a range of health problems in a sample of Lao students, but could not come out with sensible results (Daniel Reinharz, personal communication). That is why DALYs were used as recommended by WHO [
15].
Finally, one should stress that the study does not reflect the financial affordability of the health care system in Lao PDR. The threshold ratio used to measure the cost-effectiveness is the GDP per capita, which is controversial. Moreover, the limited resources in the country lead to strong competition among interventions in health care programmes. Accurate data on the burden of this disease in Lao PDR would provide important information for decision makers.
Acknowledgements
We would like to thank the Mathematical and Economic Modelling group in Bangkok and those at Laval University for their contributions to training and analysis. This includes Ben Cooper, Yoel Lubell, Wirichada Pan-ngum, Hina Hakim, Diane Fournier, Julie Duplantie and Léon Nshimyumukiza. We would also like to thank the Department of Health Insurance and national immunisation programme, Lao PDR for providing the cost data related to healthcare expenditure in Vientiane hospitals.