Setting

The study was conducted in Iquitos, the capital city of the Amazonian region of Loreto, in North-East Peru. It has a population of 437,376 (2015) and can only be accessed by boat or plane [27]. The region’s population is sparsely distributed with a population density of 3.2 people per km² (national average 24.8 people per km²) [28] and 35.2% of the population lives below the poverty line (national average 22.7%) [29]. Locals largely self-identify as mestizo (mixed-race) [30].

Study design and guiding framework

The qualitative design allowed for the development of in-depth understanding of the participants’ knowledge, attitudes and perceptions [31]. The five-level socio-ecological model (SEM) was used as the theoretical framework that informed the study (Fig 1) [32]. The SEM model recognizes the intertwined relationship between individuals and their environment and that public health problems, such as low vaccine uptake, are complex and affected by multiple levels of influence, and can therefore not be investigated using a single-level analysis [32,33]. The model informed data collection, analysis, and interpretation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Five-level socio-ecological model.

https://doi.org/10.1371/journal.pone.0255218.g001

Sampling

Nurses and teachers involved in the delivery of the HPV vaccine program were initially purposively sampled [34]. Diversity of both teachers and nurses in terms of participant age, length of time in program, location of work (rural versus more urban environments) was sought. Exponential non-discriminative snowball sampling was then used to facilitate further recruitment [35].

Recruitment

Recruitment took place in the last two weeks of January and throughout February 2019. The principal investigator (A1), and the local supervisor (A3) approached the health centers and schools to explain the study to the nurse in charge or relevant school director. Once they were aware of the study and its implications, they identified eligible participants. Participants were eligible to participate if they met the following criteria: a nurse involved in the HPV vaccine delivery process or a primary school teacher who worked with the HPV vaccination program; Peruvian citizen resident of Loreto; over 18 years of age; speaking either Spanish or English fluently; willing to participate and able to give informed consent. A1 then approached potential participants, checked their eligibility, explained the study, provided a participant information sheet, and answered any queries. If they agreed to participate, written informed consent was obtained and a suitable time for the interview was identified. It was also emphasized that participation was voluntary and that their information would remain confidential.

Data collection

This qualitative study involved semi-structured interviews with nurses and teachers. Individual interviews were chosen as the main form of data collection because of the study’s aim to gain insight into individuals’ perceptions, which are more likely to be elicited in one-to-one interviews rather than in a group setting [36]. Interviews were conducted in a private room at the participants’ workplace. The semi-structured, face-to-face interviews followed a topic guide (S1 File) developed using existing literature [37,38] and informed by the five-level SEM. A1 conducted the interviews in Spanish. All interviews were audio recorded and then transcribed verbatim in Spanish and translated into English by A1. The transcripts were anonymized. Field notes and initial observations were noted to maximize trustworthiness and researcher reflexivity. Before starting each interview, participant socio-demographic information was collected using a demographic questionnaire.

Data analysis

Data analysis was informed by the Braun and Clarke’s 6-step thematic approach [39]. Recruitment was conducted in parallel to analysis and continued until the research team deemed that the data were sufficiently rich to address the study aims [36].

Data immersion was achieved through repeated reading of the transcripts and by subsequently translating them into English. NVivo 12 software was used to support the analysis process. The SEM was initially applied to the data deductively, and then inductive coding was developed within each of the five SEM levels. A thematic mind map was developed to assist with the process of reviewing the initial themes [39]. Analytical triangulation was performed to increase reliability [40] by having a second researcher (A4) code and validate a sample of two data-rich, anonymized transcripts. Any discrepancies between the two researchers were discussed and agreed on. Further analysis and interpretation of data was supported by the wider research team (A2, A3, A5). Descriptive analysis of the demographic questionnaire data was undertaken using IBM SPSS Statistics 24 software.

Ethics statement

Ethical approval was obtained from the University of Birmingham Internal Ethics Review Committee and the Institutional Research Ethics Committee at the Department of Health Loreto. Informed written consent was obtained from all participants prior to data collection.

Participants

Twenty-one interviews (average length 30 min, range 10–56 min) were conducted with eleven nurses working in two primary health centers, and ten teachers from four public primary schools in the area. All professionals were or had been involved in the delivery of the HPV vaccine program. Four teachers were purposively sampled with the remaining six snowballed via their colleagues. For nurses, seven were purposively sampled and four were snowballed via their colleagues. All the participants self-identified as “mestizo” (mixed-race) and most were female (85.7%). To preserve anonymity, a summary of the participant demographics is presented in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of participant demographics.

https://doi.org/10.1371/journal.pone.0255218.t001

Findings

Five core themes were interpreted inductively: (1) knowledge, (2) acceptance, (3) accessibility, (4) communication and (5) resources. All core themes were relevant to more than one level of the socio-ecological model and therefore the results are presented by inductive theme rather than SEM level. Table 2 summarizes the findings and links the subthemes to the five SEM levels. Interpreted subthemes included: well informed participants, perceived limited awareness of parents, high vaccine acceptability, and girls’ fear of the vaccine (individual level), issues with parental consent, school attendance, and the influence of familial customs and religion (interpersonal level), staff shortage, occupational risks, and problems surrounding the program’s funding (organizational level), beliefs and social norms, geography and climate, the influence of media and the internet, and communication between education and health (community level), and strategies such as running awareness raising campaigns, increasing the program’s target group, improving the recording system, making the vaccine compulsory, and improving decision-making (policy level). An interpretation of each of the themes is presented below with illustrative quotations.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Summary of interpreted core themes and subthemes modelled against the five levels of the socio-ecological model.

https://doi.org/10.1371/journal.pone.0255218.t002

Main findings

This study provides important insight into nurses’ and teachers’ perspectives on the barriers and facilitators to successful implementation of the HPV vaccination program in Iquitos. To the best of the authors’ knowledge this had not been explored previously. Barriers to vaccination were identified at multiple levels of the SEM. The study found that participants were generally well informed about the vaccine. However, they also highlighted that they wanted more training and in-depth information. It is not clear how to best deliver this additional training. A recently published Japanese paper found that delivering a lecture to teachers increased their intention to recommend the HPV vaccine [42]. More novel approaches are also being explored. For example, the online course developed by the Catalan Institute of Oncology (ICO) and the Colombian Cancer Institute [43]. Although its efficacy has not been proven yet, it has had a good uptake with nearly 5000 professionals [43] registering to the course. Ensuring appropriate knowledge and high vaccine acceptability amongst professionals is key, since provider recommendation has been repeatedly identified as an important factor in HPV vaccine uptake [19,20,22]. In addition, our participants perceived high acceptability of the vaccine amongst parents. This was found in previous studies both within Peru [18] and in numerous LMICs, such as the Brazilian Amazon or Indonesia [44].

However, many participants thought parents were uninformed about HPV and the vaccine. This was considered a key barrier to vaccination uptake. Importantly fear of side-effects was the most mentioned concern. Such findings were also reported in a study with Peruvian parents [18]. This lack of awareness needs to be urgently addressed as parents with knowledge about the vaccine show higher acceptance rates, highlighting a need for better educational campaigns [44,45]. Furthermore, multiple qualitative studies, one of them conducted in Peru, report that parents themselves desire more information on the topic to aid decision making, highlighting the need for improved educational and information campaigns [18,23,46]. How to best provide the information needs to be further explored. Many participants suggested organizing parent information evenings. However, issues regarding parental attendance to such events were also noted. Finally, a couple participants expressed negative opinions regarding the provision of information leaflets, explaining that parents were unlikely to read them. Another potential barrier to the effectiveness of written information is the 6.4% illiteracy rate (in people aged 15 and above) in Loreto, reported by national government statistics [47].

As mentioned previously, our participants also reported obtaining written consent as a barrier to vaccination. This was also found to be problematic in other settings [48]. In addition, a report comparing different HPV vaccine delivery strategies in LMIC found that programs with opt-out consent were more likely to achieve high coverage rates [48]. This was also reported by the WHO [49]. A recent qualitative study found that parents in Virginia, the first American state to introduce opt-out consent, expressed concerns regarding health policy makers’ transparency and highlighted that deciding whether one decides to give the HPV vaccine should be a parental right [50]. Although an opt-out consent strategy was supported by our data, whether this would be accepted in Iquitos needs to be further studied, especially since distrust in health providers was identified as an important barrier to vaccine acceptance amongst Peruvian parents [18]. Similarly, evidence suggests that having to sign a consent form generates parental distrust of the vaccine, as this is not required for other vaccinations [18,51]. If written consent is sought, research favors getting parents to sign the consent form at school enrolment [48], a moment that was viewed favorably by various participants. Such a strategy should therefore be explored within the Peruvian context.

Interestingly, although all participants viewed school-based vaccination as beneficial, a few also highlighted associated barriers, e.g., school absenteeism. This was associated with girls’ fear of the vaccine or familial reasons e.g., moving away from the city. It affects coverage rates and in the second case, makes completion of dose schedule complicated, due to loss of follow-up. Another barrier to school-based vaccination raised by a participant was teachers’ opposition to vaccines. Although this was an individual case, given the important role of teachers in the vaccination campaign, it is crucial to ensure they are supportive of the program, once again highlighting the importance of providing them with the appropriate information and training, A recent study looking at HPV vaccination programs conducted in low-income countries reported that a mixed model system, including school and health facility-based vaccination, achieved the best coverage [52]. Although some participants supported the use of house to house vaccination campaigns, it is associated with cold chain and logistical challenges [53] and the feasibility and economics of integrating this approach in remote areas of Peru is yet to be established. In addition, the region has a net primary school attendance rate of 91.7% [54]. Research suggests that attempting to reach out-of-school girls by running specific campaigns does not significantly improve coverage rates in countries with over 80% school enrolment [55]. Although not giving these girls an opportunity to get vaccinated might worsen health inequality, it suggests that the region should focus on improving the school-health center-based approaches.

Importantly, low vaccine acceptance in indigenous communities also needs to be addressed. Certain beliefs regarding the causes of cervical cancer present in the local population, for example, associating it with witchcraft, might make it difficult to shift social norms towards acceptance of the vaccination. Low coverage has been found in other studies looking at vaccination in ethnic minorities in South America [56]. In the Amazonian region of Peru, 51 indigenous groups exist, most of which speak their own language [57]. Distrust in western medicine and a strong belief for traditional medicine have shown to hinder vaccination uptake in such groups [56]. A systematic review looking at strategies to overcome vaccine hesitancy recommends focusing on multicomponent or dialogue-based interventions [58]. These includes addressing the community’s distrust by involving the community leaders in the program design, improving social mobilization, and using social and mass media to deliver information. They also highlight the need to tailor the strategies to the target population by considering the cultural context and the population’s reasons for hesitancy [58]. In the absence of access to free or low cost satellite internet for health care purposes, many remote areas do not have access to mass media and the internet and rely on information from either community members or health workers [59], it is therefore important to investigate what interventions might be most effective in such groups.

Finally, although participants were supportive of the program, some felt their opinion was not considered by policy makers. Given that they have the most hand-on experience, they can provide a valuable insight into the health needs of the population and can make an important contribution to shaping and strengthening effective policy [60,61] and should therefore be included in the decision making process.

Overall, this study’s findings suggest that future strategies should be directed towards building knowledge amongst professionals and raising awareness within the general population, especially with the parents of girls eligible for the vaccination. Such campaigns should focus on reducing fear of side effects. New campaigns should potentially target less accessible areas and especially indigenous groups where cultural and social norms, affecting health beliefs and attitudes, may make increasing vaccination coverage difficult.

Strengths and limitations

As far as the authors are aware, this qualitative study is the first to explore professionals’ perceived barriers and facilitators to the uptake of the HPV vaccination in Peru. A1, who is a native Spanish speaker, conducted the interviews in the local language. This facilitated richer narratives and avoided the need to use interpreters, which can be associated with problems related to misinterpretation of meaning or significance of participants’ responses [62].

Furthermore, the use of a theoretical framework, the SEM, allowed the study to examine the importance of numerous factors, at different levels, on vaccine uptake. Given that most vaccination uptake studies focus on interpersonal level variables [63], using such a model provides a useful holistic overview of the problem. This is particularly relevant given that research suggests that public health interventions to improve vaccination uptake targeting multiple levels of the model might be more effective [64].

It is important to be cautious when applying the conclusions beyond the sample and to evaluate whether they are transferable to other settings. Transferability can be defined as the degree to which qualitative studies’ findings can be transferred to other contexts or settings [65]. The use of purposive sampling to obtain a diverse range of opinions and experiences increase this study’s findings’ transferability [65] and so the findings are likely to be transferable to similar populations and settings. The adequacy of the final sample size was carefully monitored during the research process to ensure that the overall sample and associated data had sufficient information power to develop new knowledge in relation to the research questions [66]. Further to this, we had multiple analysts to support investigator triangulation and ensure that the data were viewed and interpreted via multiple lenses [67].

The impact of the main researcher also needs to be reflected upon. A1 had a central role in both data collection and analysis. Researcher bias was minimized by having a second coder review a sample of the transcripts and the wider research team supporting analysis and interpretation. To minimize the risk of A1’s (Spanish, female medical student) preconceptions and cultural values affecting data collection and interpretation, constant communication and support from the local supervisor and wider research team was sought.