With the introduction of human papillomavirus (HPV) vaccine throughout the world, a major question is whether the target populations in each country can readily access immunization locations. As of September 2016, 67 countries had introduced the HPV vaccine fully into their national immunization schedules,1 with more low-income and middle-income countries projected to introduce it by the end of the decade.2 However, 3-dose coverage for the vaccine has ranged from 17% to 92%, with low coverage present even in high-income countries and often dependent on the delivery strategy for vaccination.3,4 Immunization uptake depends on several factors including travel distance to health centers. As research has shown, distance to the nearest immunization location can ultimately prevent someone from getting immunized.5 The World Health Organization defines “health services access” as the proportion of a population living within 5 km of a health center,6 suggesting that people living more than 5 km away from immunization locations may not be able to access the vaccinations they need. In many health systems, the locations of such centers are not strategically placed to optimize access for the population. Decision makers may prioritize political considerations above optimal access.7 Facilitating the maximum benefits of current and future routine immunization schedules requires addressing the issue of vaccine access.
In anticipation of HPV vaccine introduction in Mozambique, where cervical cancer is the most common cancer in women,8 our team developed Strategic Integrated Geo-temporal Mapping Application (SIGMA) to determine the potential economic impact of immunization against HPV 16 and 18 using bivalent vaccine. We quantified how many people in the relevant target population are reachable by existing immunization locations in Mozambique, how many are not likely to access these locations, and the potential costs and disease burden averted by immunization.
MATERIALS AND METHODS
SIGMA Model of Mozambique HPV Immunization
We developed SIGMA for Immunization, a geospatial information system (GIS), to quantify how many people in the HPV target population are reachable by existing immunization locations in Mozambique and how many cannot access these locations. A previous publication describes our SIGMA platform in greater detail.9 The SIGMA model of Mozambique mapped population density data from the Global Rural-Urban Mapping Project,10 which contains the estimated density of the Mozambique population living in each square km for 2000, projected to the 2015 population of 25,727,91111 assuming a uniform growth rate. The total population was scaled to the population of 10-year-old girls, the target population for HPV immunization in Mozambique Ministry of Health pilot activities who comprise approximately 2.1% of the total population in Mozambique.12
Using geospatial coordinate data from the Mozambique Ministry of Health, the SIGMA model plotted all 1377 health centers that serve as routine immunization locations for the World Health Organization Expanded Program on Immunization in Mozambique. Each immunization location has a corresponding catchment area which represents the surrounding geographic area the location can serve. The catchment area radius can be thought of as the maximum distance some people would need to travel to access vaccines. SIGMA overlaid the catchment areas for all immunization locations onto the geospatially explicit population data for 10-year-old girls to determine the number of people in the target population that the existing health centers can reach. Due to uncertainty around the actual reach of each health center, sensitivity analyses systematically tested a range of catchment area radii. To assess how feasible it may be for the entire population to reach health centers, experiments also determined the area the health centers would need to cover to immunize at least 99% of the target population.
Disease Burden and Costs
For each catchment area radius scenario tested, we calculated the number of people in the target population who lived within any health center’s catchment area. We then computed the estimated number of HPV 16,18-related cervical cancer cases that could be prevented by vaccines in a cohort of 10-year-old girls (Formula 1) and quantified the disability-adjusted life years (DALYs), health care costs, and societal costs (due to lost productivity) averted.
Vaccine-prevented cases = target population within catchment area × vaccine efficacy × disease risk.
We identified several studies on the economics of HPV in Sub-Saharan Africa13–15 and chose to base our estimates on work by Goldie and colleagues16 as it provided aggregated cost estimates for Mozambique that were inclusive of all medical care costs incurred by the disease, adjusted for the proportions of people experiencing different levels of severity of the disease (stage I–IV). Because these costs were originally reported for 2005, we used World Bank gross domestic product deflator estimates for Mozambique between 2005 and 2015 to inflate the costs to 2015.17 We report costs in 2015 USD. Estimates of DALYs averted from HPV immunization also came from Goldie and colleagues and use a discount rate (r) of 0.03 with no age weighting.16 HPV vaccine efficacy is assumed to be 100% against HPV types 16 and 18,16 with unprotected individuals developing HPV 16,18-related cervical cancer at a rate of 23 (16–30) per 100,000 at-risk individuals.8,18 Each case of HPV-related cervical cancer incurs US $1,792 (US $896–US $3584) in health care costs in 2015 US dollars,16 US $2296 in productivity losses (based on a US $639 GNI per capita19), and 3.6 DALYs.16 The health care cost per case represents the average costs for screening and treatment of HPV-related cervical cancer (stages I–IV). Appendix A, http://links.lww.com/OLQ/A151 contains further details about these inputs and their sources.
Impact of HPV Vaccine Introduction in Mozambique
If the entire 2015 population of 10-year-old girls goes without HPV immunization, approximately 125 (range, 111–139) new cases of HPV-related cervical cancer are expected over the lifetime of this cohort. These cases translate to 449 (399–499) DALYs and US $224,002 (US $99,533–497,005) in health care costs, and US $286,983 (US $255,034–318,372) in lost productivity. Figure 1 summarizes the impact of HPV immunization under various catchment area radii.
If each health center covers a catchment radius of 3 km, administering HPV vaccine to a cohort of 10-year-old girls could prevent 19 (17–21) new cases, 68 (60–75) DALYs, $33,901 (US $15,064–75,218) in health care costs, and $43,433 ($38,598-$48,183) in productivity losses. This catchment area size would cover only 15% of the target population. A catchment area radius of 5 km would cover 40% of the target population and could prevent 50 (44–55) cases of HPV-related cervical cancer annually, thereby averting 178 (159–198) DALYs, US $88,926 (US $39,513–197,304) in health care costs, and $113,928 (US $101,245–126,389) in productivity losses in the cohort.
If each health center can instead cover a catchment area with a radius of 7 km, then 49% of the population would be reachable. This scenario could prevent 61 (54–68) cases in the cohort and avert 220 (196–244) DALYs, US $109,846 (US $48,809–243,720) in health care costs, and US $140,730 (US $125,063–156,122). A catchment area radius of 15 km would cover 82% of the target population and could prevent 102 (92–113) cases in the cohort and avert 367 (326–407) DALYs, US $182,744 (US $81,200–405,464) in health care costs, and US $234,125 (US $208,060–259,732) in lost productivity.
Potential Impact of Widespread HPV Immunization in Mozambique
At higher catchment radii, each additional increase in catchment area radius raises population coverage with diminishing returns. A 20-km radius would reach 91% of the target population to potentially prevent 114 (102–127) cases, 411 (365–456) DALYs, US $204,847 (US $91,021–454,504) in health care costs, and US $262,442 (US $233,224–291,146) in productivity losses in the cohort. A 30-km radius is required to reach 98% of the population, and a 35 km would cover 99% of the population. In the 2015 cohort, 99% coverage could prevent 124 (110–137) cases, 445 (395–494) DALYs, US $221,756 (US $98,535–492,021) in health care costs, and US $284,105 (US $252,476–315,179) in lost productivity. Figure 2 displays SIGMA visualizations of a sample of catchment area radius scenarios, showing where some less densely populated areas are not reachable even at high catchment area radii.
Our study suggests that depending on how far girls are willing and able to travel to receive HPV vaccination, a substantial proportion of the target population may not have ready geographic access to immunization. Although “willingness to travel” may vary, existing studies imply that 35 km would be well beyond the travel distance limits of most people in Mozambique, where distance to health centers is a major cause of low immunization coverage.20 Given the proposed HPV immunization schedule in Mozambique,12 each member of the target population would be required to make three visits to an immunization location to receive all recommended doses of the vaccine. After piloting HPV immunization in select districts beginning in 2013,21 training for immunization program staff to scale up HPV immunization commenced in 2014.22 For these activities to achieve the greatest potential impact, geospatial access should be an important part of planning for the introduction.
Geospatial applications can be invaluable tools for planning and evaluating immunization activities by allowing health workers to decide the most effective locations for health facilities, generate more accurate maps of catchment areas, and plan more efficient vaccination routes. The GIS data can provide additional insight into trends in disease burden (dot density maps) and vaccine coverage by location (thematic maps).23 Immunization workers in resource-limited settings often use hand-drawn maps, which rely on local knowledge and usually do not reflect the spatial relationship between major landmarks.24 Additionally, this kind of mapping is not conducive to locating households in high-density, urban areas given overcrowding and the lack of a formal grid structure in many areas. Another complication arising from a reliance on hand-drawn maps is that the boundaries on these maps often do not match between districts, potentially omitting households on the border from vaccination activity plans. Despite the ability of geospatial data to address these issues, implementation in low-income countries with poor immunization coverages remains limited. The costs associated with adopting proprietary geospatial applications and training health workers to have the requisite knowledge and skills needed to effectively use them may be contributing to delayed introduction in resource-constrained settings.23 Accessible, user-friendly tools are needed to help decision makers perform geospatial analyses to improve their systems. Such tools should be flexible enough to work with different degrees of data availability. For example, our analysis mapped catchment areas based on straight-line distances, but additional data would allow SIGMA to take into account regional differences in geography and available transportation methods.
As one of the most common sexually transmitted infections in the world, HPV is responsible for many—largely preventable—diseases, including anogenital warts and cancers of the cervix, vagina, vulva, penis, anus, and oropharynx, which often place a substantial burden on health systems. The HPV vaccine has been highly effective at reducing the prevalence of oncogenic HPV serotypes contributing to these cancers in many populations, even in the presence of below target-level immunization coverage.25 The population-level impact of female HPV vaccination, based on available country-level incidence and prevalence, has been estimated to have reduced HPV 16 and 18 infections by 68%.26 However, these significant reductions were only achieved in settings with at least 50% coverage, underscoring the importance of increasing access to HPV immunization. While this study did not take into account the effects of herd immunity, future modeling may incorporate this aspect to further explore the impact of achieving various coverage levels in a population. Trends in HPV vaccination and coverage have supported both the vaccine's effectiveness and its cost-effectiveness, with studies demonstrating high cost-effectiveness in a majority of countries worldwide when compared with gross domestic product per capita and when compared with existing cervical cancer screening programs.27,28
Data have shown that HPV coverage rates have fallen short of targets after introduction in some countries. A number of factors that this study did not take into account, but could be included in future analyses, may be impeding uptake such as the high cost of the vaccine in settings without public vaccination support.3,29 Although less access to health care is certainly a determinant of low coverage, the low level of information on HPV in the public sphere has been linked to increased cervical cancer incidence, a consequence of high-risk HPV infection. (Appendix B, http://links.lww.com/OLQA/152).30 , 31s, 32s Parents often make the decision to vaccinate based on information from their health providers, with a strong provider recommendation being the key determinant of the choice to immunize.29 , 33s, 34s Negative attitudes and beliefs, as well as trust in the efficacy of the vaccine, also influence this decision.29 , 35s, 36s Finally, the delivery infrastructure for HPV as an adolescent vaccine is often dependent on the setting, with health facility-based or school-based mechanisms differing in their cost and effectiveness.4 , 37s, 38s Without further investigating the population to be targeted, vaccination programs face difficulties in identifying and covering those eligible for immunization. Therefore, it is important that geospatial access is not an additional obstacle. Further modeling to evaluate strategies to increase access to immunization, such as new health facilities and outreach activities, can help identify impactful and cost effective ways to improve immunization coverage.
Each year, Mozambique experiences new cases of cervical cancer that could have been prevented by HPV immunization. Introducing HPV vaccine in Mozambique could avert significant disease burden and costs. However, much of the target population is not able to reach any existing immunization location, thereby reducing the potential impact of this vaccine if geospatial access to immunization does not improve, particularly in rural areas. GIS analysis can assist in planning vaccine introduction strategies to maximize geospatial access and help the population reap the maximum benefits from an immunization program.
1. Loharikar A, Dumolard L, Chu S, et al. Status of new vaccine introduction—worldwide, September 2016. MMWR Morb Mortal Wkly Rep 2016; 65:1136–1140.
2. WHO vaccine-preventable diseases: monitoring system. 2015 global summary [database online]. Geneva: World Health Organization; 2015.
3. Markowitz LE, Tsu V, Deeks SL, et al. Human papillomavirus vaccine introduction—the first five years. Vaccine 2012; 30(Suppl 5):F139–F148.
4. Paul P, Fabio A. Literature review of HPV vaccine delivery strategies: considerations for school- and non-school based immunization program. Vaccine 2014; 32:320–326.
5. Lee BY, Mehrotra A, Burns RM, et al. Alternative vaccination locations: who uses them and can they increase flu vaccination rates? Vaccine 2009; 27:4252–4256.
7. Hipgrave DB, Alderman KB, Anderson I, et al. Health sector priority setting at meso-level in lower and middle income countries: lessons learned, available options and suggested steps. Soc Sci Med 2014; 102:190–200.
8. Bruni L, Barrionuevo-Rosas L, Albero G, et al. Human papillomavirus and related diseases in Mozambique: Summary report. ICO Information Centre on HPV and Cancer (HPV Information Centre); 2016; 26.
9. Haidari LA, Brown ST, Constenla D, et al. The economic value of increasing geospatial access to tetanus toxoid immunization in Mozambique. Vaccine 2016; 34:4161–4165.
10. Global Rural–urban Mapping Project, Version 1 (GRUMPv1): Population count grid [database online]. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC); 2000.
11. Instituto Nacional de Estatística Moçambique 2015. Accessed at http://www.ine.gov.mz/
. Accessed April 2015.
12. Government of Mozambique. In: Application form for country proposals: providing approximately two years of support for an HPV demonstration programme. Manhica District, Mozambique: GAVI Alliance, 2012.
13. Kim JJ, Ortendahl J, Goldie SJ. Cost-effectiveness of human papillomavirus vaccination and cervical cancer screening in women older than 30 years in the United States. Ann Intern Med 2009; 151:538–545.
14. Campos NG, Kim JJ, Castle PE, et al. Health and economic impact of HPV 16/18 vaccination and cervical cancer screening in Eastern Africa. Int J Cancer 2012; 130:2672–2684.
15. Tracy JK, Schluterman NH, Greene C, et al. Planning for human papillomavirus (HPV) vaccination in sub-Saharan Africa: a modeling-based approach. Vaccine 2014; 32:3316–3322.
16. Goldie SJ, O'Shea M, Campos NG, et al. Health and economic outcomes of HPV 16,18 vaccination in 72 GAVI-eligible countries. Vaccine 2008; 26:4080–4093.
17. World Bank. World Development Indicators 2015. Accessed at http://databank.worldbank.org/data/reports.aspx?source=2&series=NY.GDP.DEFL.ZS.
20. Sheldon S, Alons C. A study to describe barriers to childhood vaccination in Mozambique. Final report. Maputo: Ministry of Health, Expanded Program on Immunization, Mozambique; The CHANGE Project, and Project HOPE 2003.
23. Kamadjeu R, Tolentino H. Web-based public health geographic information systems for resources-constrained environment using scalable vector graphics technology: a proof of concept applied to the expanded program on immunization data. Int J Health Geogr 2006; 5:24.
24. Gammino VM, Nuhu A, Chenoweth P, et al. Using geographic information systems to track polio vaccination team performance: pilot project report. J Infect Dis 2014; 210(Suppl 1):S98–S101.
25. Brotherton JM, Ogilvie GS. Current status of human papillomavirus vaccination. Curr Opin Oncol 2015; 27:399–404.
26. Drolet M, Benard E, Boily MC, et al. Population-level impact and herd effects following human papillomavirus vaccination programmes: a systematic review and meta-analysis. Lancet Infect Dis 2015; 15:565–580.
27. Jit M, Brisson M, Portnoy A, et al. Cost-effectiveness of female human papillomavirus vaccination in 179 countries: a PRIME modelling study. Lancet Glob Health 2014; 2:e406–e414.
28. Seto K, Marra F, Raymakers A, et al. The cost effectiveness of human papillomavirus vaccines: a systematic review. Drugs 2012; 72:715–743.
29. Valentino K, Poronsky CB. Human Papillomavirus Infection and Vaccination. J Pediatr Nurs 2015.
30. Kahn JA, Lan D, Kahn RS. Sociodemographic factors associated with high-risk human papillomavirus infection. Obstet Gynecol 2007; 110:87–95.