Cervical cancer is the fourth most common cancer among women worldwide resulting in an estimated 569,847 new cases and 311,365 deaths , per 2018 estimates [1]. Over 85% of deaths occur in low- and middle-income countries, because access to cervical cancer screening and treatment services is often lacking or non-existent [2,3]. Almost all cervical cancers are caused by persistent infection with sexually-transmitted human papillomavirus (HPV), with HPV types 16 and 18 being responsible for approximately 70% of cancerous lesions [4,5].

Three vaccines against HPV are prequalified by the World Health Organization (WHO), and available for purchase by United Nations agencies – a bivalent vaccine against HPV types 16 and 18, a quadrivalent vaccine against HPV types 6, 11, 16 and 18, and a 9-valent vaccine against additional types 31, 33, 45, 52, and 58. All vaccines are recommended by WHO as a two-dose series administered to girls age 9–14 years, separated by a minimum 6-month interval, for primary prevention of HPV infection [6]. All vaccines have been shown to be safe, efficacious and effective in clinical trials and field evaluations [7]. As of May 15, 2018, 80 (41%) of 194 WHO member states had introduced HPV vaccine into their national schedules; however, the majority of low- and middle-income countries, where the burden of cervical cancer is the highest, have yet to introduce HPV vaccine [8]. Gavi, the Vaccine Alliance (Gavi), began providing support for HPV vaccine introduction to Gavi eligible countries in 2012. To learn how to best reach this target age group that does not routinely access immunizations or other health services in many countries, Gavi initially supported subnational demonstration programs to provide countries an opportunity to gain the experience needed for successful future national introduction. Gavi provided financial support for 2-year demonstration programs in countries targeting approximately 15,000 girls in two geographic areas.

In Cambodia, cervical cancer is the most common cause of female cancer with an estimated 993 new cases and 708 deaths annually [2]; the age-standardized estimated incidence rate is 13.5 cases per 100,000 women per year [2]. In 2016, Cambodia received Gavi support to conduct an HPV vaccine demonstration project program [9], to administer two doses of bivalent HPV vaccine to all 9-year-old girls (in-school and out-of-school) residing in all four operational districts (ODs) in Svay Rieng province (hereafter called Svay Rieng) and two of four ODs in Siem Reap province (hereafter called Siem Reap), during January and July of 2017. Cambodia’s Ministry of Health - National Immunization program (MOH-NIP) purposefully selected these two provinces for having extensive rural areas, representation of high-risk communities (e.g. migrant populations), and being accessible for and collaborative with national level staff. The objective of the program was to demonstrate ability to achieve high vaccination coverage among this new target population and to understand the successes and challenges of HPV vaccine delivery; lessons learned could further inform the country for national scale-up. Schools served as fixed vaccination sites, and out-of-school girls were identified by community health volunteers and invited to come to schools for vaccination during the specified vaccination dates. Prior to the vaccination campaign, a list of eligible girls for each school was created by teachers and school staff, with the assistance of community health workers who identified out-of-school girls living in the area. HPV vaccine was also available at health centers for one month after the campaign ended for any eligible girls (in-school or out-of-school) who missed school vaccination days.

Administrative coverage for the first dose of HPV vaccine was 98% in Svay Rieng and 91% in Siem Reap, as reported by the Cambodia MOH-NIP. Coverage based on administrative data is subject to denominator bias (i.e., underestimation or overestimation of the target population) and is often inaccurate[10]. Therefore, a community-based household-level coverage survey was conducted to more accurately estimate HPV vaccination coverage. Additionally, the survey examined factors associated with vaccine acceptability, vaccine hesitancy, and community understanding of HPV infection and associated conditions. Findings from this evaluation will be used to assist key stakeholders in understanding the uptake of vaccination and acceptability of the demonstration program, related communications and delivery strategies, and community knowledge and attitudes towards HPV and HPV vaccination, to assist in planning for national HPV vaccine introduction in the future.

All 64 sampled clusters were visited (32 in each province). Among the 7,594 households visited (3,784 in Siem Reap; 3,810 in Svay Rieng), 6391 (84%) had persons at home and willing to participate, 1176 (15%) had no persons at home and 27 (0.4%) had persons at home, but not willing to participate (16%). Of those visited, at home and willing to participate, 313 (5%) households had at least one eligible girl and completed the survey (187 HH in Siem Reap, 126 HH in Svay Rieng); two households had two eligible girls each. Three girls in each province were interviewed independently, with permission from an adult residing in the household. In total, 315 eligible girls (188 in Siem Reap; 127 in Svay Rieng) were enrolled in the survey, with interviews completed by either a caregiver, the girl, or both. The median numbers of girls enrolled per cluster were 6 (range 2, 12) and 4 (range 1, 8) in Siem Reap and Svay Rieng, respectively.

This community-based household-level coverage survey indicated that the HPV demonstration program in Cambodia achieved high (>80%) two-dose coverage among 9-year-old girls in the six ODs targeted for vaccination in 2017, through a primarily school-based delivery strategy. Coverage survey estimates were slightly (4%−10%) lower than reported administrative coverage for the first dose of HPV vaccine. Our survey found very high self-reported school enrollment in these areas (100% in Svay Rieng and 99% in Siem Reap), and the majority of girls in the survey who received at least one dose of HPV vaccine reported receiving it in school. This was a likely a contributor to the high vaccination coverage observed and supports the use of school-based vaccine delivery in this setting. The proportion of girls enrolled in school is somewhat higher than what was found in the 2014 Cambodia Demographic and Health Survey which estimated a net attendance ratio for primary school (age 6–12 years) to be 83.4% for females (82.7% in Siem Reap and 88.5% in Svay Rieng) [11]. School-based vaccine delivery has been shown to attain higher coverage than health facility only-based strategies, but can be resource intensive, (though there are very few data on health facility-only based strategies) [12–14]. A list of girls developed for each school, supplemented with information from community health volunteers, facilitated comprehensive enumeration of the target population both in and out of school and helped ensure that a high number of eligible girls received HPV vaccine.

Overall, vaccine acceptability was high. Most caregivers discussed HPV vaccination with other adult family members prior to the campaign. Although knowledge of HPV infection and associated diseases was poor among caregivers and girls, protection from cervical cancer was cited as the most common reason for accepting the vaccine. Studies in other low-income countries in the Western Pacific region have reported similar findings, including high willingness among caregivers to accept a vaccine to prevent cervical cancer, despite low knowledge of cervical cancer and its causes [15,16]. Similarly, other HPV demonstration programs have shown strong trust in HPV vaccination among caregivers and found that parents are often motivated by a general understanding of the health benefits of vaccines [17–19]. Because there is no specific translation for the term “HPV” in the Cambodian language, written materials included English terminology (i.e. HPV) to describe the virus. Prior to the campaign, verbal communication activities described HPV vaccine as the “vaccine to prevent cervical cancer” [20]. Anecdotal feedback on written communication materials from parents after the campaign recommended defining the term “HPV” in the local language to aid in understanding, as well as inclusion of images to aid with community understanding[21]. It is possible that the way these questions were worded on the questionnaire or how the data collectors conducted the interview may have impacted these findings. Consistency in terminology during both campaign communication activities and coverage survey interviews would facilitate a more accurate measure of full understanding around HPV infection and HPV vaccine. As with other self- and caregiver-reported surveys, the potential for social desirability bias remains a possibility.

Girls were identified as a primary source of information about the HPV vaccination campaign for caregivers. Teacher training with strong messaging to girls in schools was utilized to transmit messages to parents and caregivers prior to the HPV vaccination campaign. In other HPV vaccine demonstration programs, collaboration with teachers and community health volunteer networks contributed to effective and efficient service delivery [12,22]. In future HPV vaccination activities, girls should be considered as a key target audience for communication activities [16]. Among the few surveyed girls who received zero doses of HPV vaccine, not being aware of the campaign was the primary reason cited for not receiving the vaccine. Fear of injections and concerns about safety were infrequently identified as reasons for not receiving the vaccine; no widespread instances of vaccine refusal, hesitancy or rumors were found. As the country considers scale-up, equipping healthcare workers, teachers and community volunteers with standard information about HPV vaccine including eligibility, benefits and safety could ensure preparedness to address rumors that could potentially arise and reinforce key knowledge among the community.

Findings from the coverage survey may not be generalizable to other countries, and it is possible that the areas selected for the demonstration program differed from other provinces in Cambodia. Household non-response was 16%. If the same proportion of non-responding HHs had eligible girls as did responding HH (4.9%), then then we estimate that 59 eligible girls were in the non-responding HH. If eligible non-responders were more likely not to receive vaccination, this could bias coverage estimates toward the high side. In addition, very few out-of-school girls were identified during the survey; it is unclear whether this was a result of a true absence of out-of-school girls in the community or whether the survey methodology was ineffective for identifying this sub-population. It is also possible that the sample frame used to select clusters, though based upon the most recent census, may not have included some marginalized populations such as migrants or other groups potentially containing out-of-school girls.

More research is needed to evaluate the effectiveness of strategies to identify and enumerate out-of-school girls [23]. If the target population for the HPV demonstration program is truly higher than initially estimated due to poor enumeration of out-of-school girls, estimated coverage may be falsely elevated. Other countries have reported similar challenges in quantifying out-of-school girls [12]. Aside from conducting a complete census, school headcounts have proved to be a reliable means for calculating target population, though unverified estimates of out-of-school girls remain an ongoing challenge [24]. As more low- and middle-income countries begin to introduce HPV vaccine nationally, it will be important to identify ways to enumerate and reach out-of-school girls as well as populations that may be vulnerable due to geographic or socioeconomic isolation, cultural or language barriers or migrant status. Some of these populations may be at higher risk for sexually-transmitted infections including HPV and most may have limited access to cervical cancer screening and treatment [23].

In 2017, WHO recommended that all countries proceed with national introduction of HPV vaccine, irrespective of availability of cancer screening or treatment, in a way that is affordable, cost-effective, sustainable, compatible with health infrastructure, and capable of achieving the highest possible coverage [6]. Overall, findings from the Cambodia coverage survey reflect that the country was successful in delivering HPV vaccination to the target population of girls during the HPV demonstration program. This evidence greatly contributed to the country’s decision to scale-up vaccination and introduce HPV vaccine into the national immunization program. However, national introduction should involve thoughtful consideration of key implementation factors including target age group, vaccine delivery strategy, dosing schedule and respective training and social mobilization approaches. WHO now recommends vaccinating multiple age cohorts (e.g., 9–14 year olds) in the first year of vaccine introduction, if feasible [6]. Though a school-based vaccination strategy was successful in Cambodia’s demonstration program targeting 9-year-old girls, national introduction to older age groups may pose different challenges. School enrollment is high in primary school but drops significantly in secondary school. While 9-year-old girls may be easily reached through school-based vaccination in primary schools, more older girls (age 13 and 14) may be missed through a similar strategy in secondary schools. As Cambodia plans for national introduction, recommendations from this survey, along with deliberations around new challenges that may arise with scale-up, should be thoroughly considered.