Extent and nature of sources
In total, 41 published articles, 9 conference abstracts, 124 unpublished documents and 27 key informant interviews (KIIs) were included in analysis, covering 72 HPV vaccine delivery experiences in 37 LMICs. A delivery experience was defined as HPV vaccine delivery to a specific target population (e.g., age range in years, school grade), using a specific vaccination venue (e.g., health facility-based, school-based, outreach, or a combination) within a specific project/programme, as defined by funding source. Not all vaccine delivery experiences collected information covering all data extraction themes.
Social mobilisation
Data on social mobilisation activities were available for 47/72 (65 %) of HPV delivery experiences in 30 of 37 countries. Overall, almost half (33/72; 46 %) of delivery experiences reported both social mobilisation and first-dose HPV vaccination coverage, of which 19/33 (58 %) reported over 90 % coverage rates (Table
1). Sub-themes related to social mobilisation were target audiences, messages, communication channels, timing and duration, and rumour management.
Table 1
Social mobilisation and consent procedures by first-dose HPV vaccination coverage
Total number (%) reporting both social mobilisation and first-dose coverage (n = 33)b | 19 (58) | 11 (33) | 3 (9) |
Start of mobilisation prior to vaccination (n = 16)c |
Within 1–2 weeks of vaccination | 2 (50) | 2 (50) | 0 |
3 weeks - less than 2 months | 3 (60) | 2 (40) | 0 |
2–3 months | 3 (43) | 3 (43) | 1 (14) |
Message content (n = 24)d |
Logistics only | 4 (67) | 2 (33) | 0 |
Informational (logistics and cervical cancer) | 9 (82) | 1 (9) | 1 (9) |
Comprehensive (detailed) | 4 (57) | 2 (29) | 1 (14) |
Materials and approaches (n = 32)e |
Interactive | 1 (50) | 1 (50) | 0 |
Non-interactive | 1 (17) | 4 (67) | 1 (17) |
Both interactive and non-interactive | 16 (67) | 6 (25) | 2 (8) |
Consent procedures (n = 32)f |
Written consent by parents/guardians (opt-in) | 9 (50) | 6 (33) | 3 (17) |
Implied (opt-out) consent by parents/guardians (opt-out) | 8 (73) | 3 (27) | 0 |
Changed from written to implied consent | 3 (100) | 0 | 0 |
Target audiences
In total, 38/47 (81 %) experiences reporting social mobilisation mentioned target audiences. Primary target audiences were usually parents or girls. Secondary target audiences included communities and credible influencers. Credible community influencers were identified in nine experiences as the most common information sources for parents and included health-workers, teachers/school directors, community or religious leaders and influential family members [
6,
22]. Credible national influencers included members of royalty, wives of elected officials, national and sub-national political leaders, and entertainers (e.g., television stars), who were encouraged to launch campaigns and champion HPV vaccination to increase confidence and interest among both primary and secondary target audiences [
23‐
25]. When the MoH was involved, social mobilisation typically started at national level, gaining support of national influencers [
26].
Messages
Approximately half (24/47; 51 %) of experiences reporting social mobilisation also reported message content (Table
1). The vast majority of messages were framed around the ‘anti-cancer’ benefits of vaccination rather than prevention of a sexually transmitted infection (STI). Messaging was categorised as (i) logistical only, e.g., vaccination venues, dates, eligibility; (ii) informational, e.g., logistical information plus basic descriptions of vaccine safety and cervical cancer risk; or (iii) comprehensive, e.g., providing further detailed information on cervical cancer epidemiology and/or the extent of vaccine action against HPV. Informational messages, which combined vaccination logistics and simple explanations of cervical cancer and vaccine safety, were associated with highest coverage compared with either simply logistical or fully comprehensive messages (Table
1). For example, the information required was described by one KII:
“Parents thought the vaccine was new, has not been widely used and may have some health consequences. They had also other questions that required explaining, e.g., why children of a certain age and only girls are being vaccinated. Could the vaccine affect their fertility?” (KII, Country 26)
Communication approaches
In total, 33/47 (70 %) experiences reporting social mobilisation also reported on communication approaches used (Table
1). Various approaches and materials were used to reach target audiences. Communication approaches were categorised as interactive (e.g., one-to-one or group meetings at schools or health facilities, home visits by health-workers) or non-interactive (e.g., leaflets, posters, loud-speaker, radio, or television announcements). Over half of experiences using interactive approaches (17/26) achieved high first-dose coverage (over 90 %) compared to (1/6) 17 % for those using only non-interactive approaches (Table
1). For example, parents in one experience who reported they had attended a teacher-parent meeting about vaccination were more likely to have a vaccinated daughter compared to parents who did not report attending meetings [
22]. While interactive approaches were reportedly more successful than non-interactive in influencing HPV vaccination uptake, most experiences (24/33; 73 %) used a combination of communication approaches [
26]. For example:
“We held two meetings at school, but parental involvement was a bit difficult; not many came because they were busy with their jobs. And then through TV or radio, they were informed” (KII, Country 5)
“Vaccination days were announced by radio but everyone has a mobile phone and we could have made better use of texting.” (KII, Country 25)
Timing and duration of activities
In total, 16/47 (34 %) experiences reporting social mobilisation also reported on timing of mobilisation activities. Seven experiences started 2 to 3 months prior to commencing vaccination, while nine others started less than 2 months before vaccination. Timing did not appear to be correlated with vaccine coverage achieved (Table
1); however, KIIs reported they had encountered problems during delivery when mobilisation had been conducted less than a month before vaccination:
“Social mobilisation needed more time (than 2 weeks before vaccination day) - it should start earlier and continue until the last vaccination day.” (KII, Country 3)
Only 4/47 (1 %) experiences described duration and/or frequency of mobilisation activities. Activities, including community drama and/or radio broadcasts, were provided for 1 day or continuously over 1 to 2 weeks at vaccination venues. Social mobilisation activities sometimes intensified as vaccination day approached:
“Mobilisation varied depending on the method used to deliver the message or timing prior to actual vaccination… mobilisation was conducted in the same community once a week before vaccination started, to daily on the vaccination week.” (KII, Country 33)
Rumour management
Rumours, reported in 13/37 (35 %) countries, were generally consistent in content and often spanned more than one delivery experience per country. One country reported rumours spreading from a neighbouring country. Rumours generally focused on whether HPV vaccine affected fertility, caused dangerous side effects, or was experimental (Table
2) [
22,
27].
Table 2
Reported rumours, institutional refusals, and management approaches
Reported rumours | Management approaches (preventative and reactionary) |
HPV vaccine is experimental/untested (Countries 3, 12, 24) | • Rumours resulted from opt-in consent, which was changed to opt-out; • Government and experts immediately addressed rumours. |
HPV vaccination causes fertility problems (Countries 8, 17, 21, 24, 31, 16, 28) | • Mobilisation was started very early and messages built into parent-teacher meetings; • High-level advocacy using parliamentarians from the beginning of the programme; • Intense mobilisation targeted anti-vaccination lobbyists; • A reactive crisis response was organised, including meeting with communities. |
Vaccine causes long-term adverse events, e.g., death, cancer (Countries 28, 33, 35, 26) | • Adverse events were investigated and guardians reassured that it was not due to vaccination. |
There is another cure for cervical cancer other than vaccination (Country 35) | • Rumours were tackled immediately with email newsletter and/or parent meetings. |
Institutional refusals related to the vaccine | Management approaches |
Private/faith-based schools (Countries 23, 24, 31, 35, 37) | • Sensitization through the community and targeted mobilisation using influencers; • Media access to correct information so communities could obtain HPV vaccine information from an independent source. |
Churches/religious groups (Countries 3, 28, 37) | • Increased face-to-face, community, and religious leaders’ meetings. |
Community/parent groups (Countries 1, 5, 6, 10, 14, 18, 23) | • Identified groups opposing vaccination were provided with more information; • Frequent repetition of messages; • Involved leaders and managed vaccination through government system; • Provided additional training and information to health-workers and teachers. |
Anti-vaccination lobbyists, human rights groups, academics (Countries 12, 30) | • Provided additional media information and internet-based information campaigns. |
Teacher and health-worker reluctance to vaccinate girls (Countries 6, 23) | • Provided additional training to healthworkers and used peers to trace missing and out-of-school girls. |
Experiences generally attempted to prevent rumours and institutional refusals (e.g., refusal by schools or religious groups) by conducting intensive and repeated sensitization activities (Table
2). Once rumours arose, credible influencers were often mobilised to counteract misinformation with targeted messages in specific communities (Table
2). Government endorsement was also reported as useful in mitigating rumours and increasing vaccine acceptability if uptake was low [
6,
22,
26,
28], e.g., some experiences distributed government or WHO letters of vaccine endorsement to allay parental concerns (Table
2). However, in two experiences, rumours and misinformation caused the withdrawal of government endorsement and HPV vaccination was suspended [
18]. Delays in addressing rumours, especially those around severe side effects or that the demonstration project was actually a clinical trial, resulted in vaccine delivery being delayed in one country and stopping prematurely in two countries, including India [
18]. A prompt response was critical for success, for example:
“[Rumours were] usually due to confusion around girls being sterilised and other misconceptions. [They said] ‘Once again women are being used to pilot new products’. [Rumours were] overcome through intensive health messages and education.” (KII, Country 2)
Consent
In total, 50/72 (69 %) delivery experiences reported consent procedures, 32 (64 %) of which also reported first-dose HPV coverage. Of these, 18/32 (56 %) used opt-in written parental consent, 11/32 (34 %) used opt-out implied consent, and three (9 %) changed from written to implied consent during implementation. Opt-in written consent was defined as parents or guardians actively giving permission for daughters to be vaccinated by returning a signed consent form. In addition to signing a consent form, some experiences required parents/guardians to accompany daughters to vaccination venues. In three countries, girls were not vaccinated if the parent/caregiver was not present [
29]. Two national programmes had added HPV vaccine to a pre-existing school health programme consent sheet that included a range of health interventions delivered to children over 5 years old. Opt-out implied consent meant all girls were vaccinated except those whose parents or guardians formally refused vaccination or kept girls home on vaccination days [
10].
“[Written consent] is not used routinely, we developed it for the new vaccine. The MOH advised us to use consent as it was the first time the vaccine was used in the country” (KII, Country 6)
Parents questioned why opt-in consent was required for HPV vaccine but not others, such as measles or hepatitis vaccines, and 13 experiences reported that using opt-in consent resulted in rumours that the HPV vaccine was experimental. Ten of these 13 experiences had data on uptake rates; four reported 64–70 % uptake and six reported 70–90 % uptake. In addition to concerns about vaccine safety, one informant described parental concerns that signing a written consent form indicated they took full responsibility for any consequences, including adverse events:
“For the girls to be vaccinated the project required a lengthy consent form to be signed and returned to the school. Parents/guardians felt that by signing the form they would shoulder the blame in case of daughter’s death or emergency following vaccination” (KII, Country 14)
“Consent was opt-in for first year, which caused a lot of challenges and suspicion” (KII, Country 2)
Delivery experiences using opt-out consent were more likely to report higher coverage than those using opt-in consent (Table
1). Lengthy consent procedures were reported to reduce vaccination uptake, and eleven countries proposed simplifying consent forms or changing to implied consent [
25,
27]. One reported uptake increased from 77 to 99 % on switching from opt-in to opt-out consent, although other programme factors also changed in the same time period. Another, on comparing uptake using opt-in and opt-out strategies, switched to opt-out as it drastically increased uptake. Four countries changed to opt-out consent because of rumours about the vaccine being experimental and to better align with national consent policy for routine immunisation [
25].
It should be noted that girls were not always passive in the vaccination decision. Four countries reported that girls who wanted to be vaccinated had persuaded their parents/guardians to sign consent forms [
11]. One experience allowed children aged 12 years and above to legally consent to vaccination [
30]. Four countries reported that, despite parental consent, some girls refused to be vaccinated.
Acceptability
In total, 45/72 (62 %) experiences reported on HPV vaccine acceptability in 23 of 37 countries. Of these, 14/45 (31 %) experiences in nine countries formally measured acceptability in surveys, studies or post-introduction evaluations; the remainder reported acceptability without detailed methods. Studies in three countries measured vaccine acceptability as parents’ and/or girls’ willingness to be vaccinated and subsequent uptake of HPV vaccination [
31‐
33]. One study in one country specifically selected vaccinated and unvaccinated girls to survey [
22]. Acceptability in the remaining five countries was measured in community coverage surveys [
6,
26,
27,
30,
34‐
37]. One study also measured acceptability among health workers as nurse’s willingness to recommend HPV vaccination [
38]. Publications on experiences in Brazil, Cameroon, Kenya, Peru, South Africa, Tanzania, Uganda, and Vietnam reported parental acceptability [
6,
22,
27,
30‐
32,
37‐
39]. Acceptability among girls was reported in Cameroon, Tanzania, Uganda, and Vietnam [
22,
26,
33,
35,
36]:
“Interestingly, girls who did not normally attend schools would often come to the school on vaccination day to receive their vaccines.” (KII, Country 25)
Studies of hypothetical acceptance prior to vaccination were excluded from this study; the reason was illustrated by a Kenyan study that reported low first-dose uptake despite recording high parental willingness to vaccinate at baseline [
32]. The most common reasons reported in post-vaccination surveys for parental acceptance or refusal of HPV vaccine are shown in Table
3. Communicating directly with influencers such as family members, teachers and health-workers was reported as strongly associated with acceptability [
22,
26,
27,
31], as was being well-informed about HPV, HPV vaccination, and cervical cancer [
27,
30,
32,
35,
40]. Exposure to rumours and incorrect or limited information about HPV vaccine safety, side effects and association with cervical cancer, were associated with low acceptability [
27,
33,
35,
37‐
39].
Table 3
Reasons for accepting or rejecting HPV vaccination reported in surveys in 8 countriesa
Reasons for acceptance stated by parents/guardians | Scoreb | Surveys (n) |
Protection from cancer | 23 | 8 |
Vaccination is good for health | 22 | 8 |
Perceived cervical cancer risk or susceptibility | 8 | 3 |
Convincing information | 6 | 3 |
Vaccine is safe | 5 | 2 |
Following others’ advice | 5 | 3 |
Protection from infection | 5 | 4 |
Informed about the programme | 4 | 2 |
Vaccine is free | 3 | 2 |
To avoid shame/stigma of an STI infection | 2 | 2 |
Interest in HPV vaccine and education | 2 | 1 |
Heard of cancer/knowledge of someone with cancer | 1 | 1 |
Perceived severity of infection and consequences | 1 | 1 |
Provided at school to every child | 1 | 1 |
Reasons for not starting stated by parents/guardians |
Lack of motivation |
Fear of adverse effects on fertility and vaccine safety | 16 | 8 |
Girls or parents do not want vaccine | 6 | 3 |
May encourage early sex | 4 | 2 |
Cancer considered low severity/low risk | 3 | 1 |
Concern about vaccine effectiveness | 3 | 1 |
Undisclosed reasons | 2 | 1 |
Perceived low risk of infection | 2 | 1 |
Not good for a child | 1 | 1 |
Lack of information |
Not aware of the programme | 13 | 6 |
Insufficient information | 8 | 4 |
Systems barriers |
Absenteeism | 15 | 7 |
Difficult to determine age eligibility | 9 | 7 |
Location and time not convenient | 2 | 1 |
Health provider didn’t recommend | 1 | 1 |
Vaccination refusal by private schools, religious groups and health professionals was reported in 15 countries (Table
2). In five experiences, some school principals or teachers would not allow vaccinators into their schools due to fear of parent reactions, religious beliefs, or rumours [
22,
41]. In two experiences, the school administration thought that they would be held responsible for any severe adverse effects. In Tanzania, parental refusal was higher in private than public schools, as school management feared loss of revenue if parents objected to vaccination [
42]. In one experience, school administrators in 23 schools refused access to health-workers as some parents did not consent to vaccination. The importance of training key influencers in the community was emphasized by many KIIs:
“At first religious leaders were not specifically targeted for mobilisation but we soon realised they had a powerful influence in the communities so they were included in the social mobilisation during the first year. Low level of social mobilisation of out of school girls yielded low uptake at community outreach events and at the health facility. Rumours may have also had an influence but after this was realised social mobilisation was increased.” (KII, Country 22)
“The anxiety among the health workers in the first campaign was high as they thought it was potentially dangerous. [We] need to spend a lot more time with health professionals to educate them fully” (KII, Country 30)
Systems barriers, not necessarily related to acceptability, also affected uptake. For example, school absenteeism on vaccination days was reported as a major barrier to HPV vaccine uptake [
6,
28,
32,
35,
37,
42,
43]. Additionally, inadequate official communication about vaccination timing and/or venue and difficulties determining age eligibility were reported by experiences in India, Kenya, Nepal, Peru, Uganda, and Tanzania [
6,
28,
32,
42‐
44].