Recruitment methods and site selection

The executing researcher (JS), who is a young female of Somali origin and studied Health Sciences in Amsterdam, approached Somali organizations and Community Health Services (CHSs) through gatekeepers who had access to the Somali community. Participants were recruited from a wide range of settings, including community gatherings, CHSs, the university, and Facebook. The recruitment was mostly done by JS and via members of the Somali organizations and CHS professionals.

The young Somali women, between 18 and 21 years old, were recruited because they have been a part of the target group for the first (catch-up) HPV vaccination program, and they will be invited for Pap smears when they turn 30. The mothers who were over the age of 30 were recruited in this study because they had either been invited to receive Pap smears or because they had to decide whether their 12 year old daughters would receive the HPV vaccination.

The inclusion criteria has been comprised of being female and of Somali origin, living in the Netherlands, aged between 18 and 65, and having a migration date from the first or second wave of migration (see Table 1). Daughters whose mothers have participated in the study were excluded and vice versa because sexuality issues and health issues are sensitive to discuss in the Somali community, and it would have limited building rapport between the interviewees and the interviewer. Furthermore, convenience sampling has been used at community gatherings of the Somali women.

Purposive sampling has been used to collect information from participants with a varied background. Snowball sampling has been applied and possible participants have been checked to see whether they fulfill the inclusion criteria for this study.

Characteristics of the study population

All of the mothers in the study were born in Somalia. Three mothers were part of the first migration wave (after 1990) and three of the second wave (after 2006). Two mothers from the first migration wave obtained education in Somalia and additional education in the Netherlands. Two mothers had received a Pap smear earlier, while three declined the invitation. One mother did not receive an invitation.

Nearly half of the Somali girls were born in the Netherlands (8 out of 14), the other girls were younger than the age of 5 at the time of migration to the Netherlands. Most of the young Somali women are highly educated (10 out of 14). All the girls had received childhood vaccinations, while only one out of the eight invited girls received the HPV vaccination. Five girls did not receive an invitation because they were turning 17 in 2009 and at that time did not belong to the target age range of 13–16. One girl could not recall the invitation.

Twenty-two out of the 26 mothers, aged 23–66 years, who participated in the natural group discussions came to the Netherlands during the second migration wave.

Ethical considerations

Information and informed consent letters were available in Dutch for the young women and in Dutch and Somali for the mothers. Individual participants were informed about the study during face-to-face recruitment, by telephone and e-mail, and (again) at the start of data collection. Participants were also informed about the voluntariness of the participation and that only members of the research team would have access to the interview data. We obtained informed consent and the women filled in a small demographic questionnaire.

All individual participants signed the consent letter and filled out the questionnaire. Several mothers from group discussions refused to sign and fill out the questionnaire because they were not used to giving this type of written information during their usual group gatherings. It is common that minority communities decline to provide this kind of information during research at the community level [14]. Also, participants did not feel compelled to stay for the whole session. Nevertheless, these mothers gave oral consent for participating and were actively engaged in giving information. Ethical approval is not required for this type of study in the Netherlands [23], as only particular types of behavioral research fall under the Medical Research Involving Human Subjects Act [24].

Measures and procedures

Semi-structured interviews and natural group discussions were held. An interview guide was developed that covered topics based upon intersectionality and earlier studies that explored factors significant for the decision to participate in the prevention of cervical cancer [25, 26]. The HBM, particularly, guided the construction of probing questions.

Since there is a cultural taboo within the Somali community on openly discussing sexual matters, the sequence of the topics was carefully chosen: there were questions that participants easily could answer and then it proceeded to more sensitive topics [27]. In general, the interviews were open in nature: topics were not always addressed according to the sequence of topics presented in the interview guide, but rather to where the participants were taking the interview. However, the interviewer always made sure that all intended topics were covered.

“Natural groups” refers to groups consisting of people who know each other already from other situations, such as sports teams, work, or women’s support groups. Researching with these groups maximizes the interaction between participants, and between participants and the facilitator. It provides access to a shared group culture [14]. In this case, the women gathered regularly to discuss issues that were important to them. With the support of the chair of a Somali women’s organization and a Community Health Service (CHS) educator, natural group discussions were facilitated.

The information in the natural group discussions was collected only after JS was introduced by the moderators and she had become acquainted with the mothers who participated in the weekly group gatherings. Information on (the prevention of) HPV and cervical cancer was presented by JS to the participants at different moments. Facilitating a discussion with mothers who only recently moved to the Netherlands would have been difficult without the provision of any information on HPV and cervical cancer.

The interviews and group discussions were carried out by JS between March and June 2013. The interviews with young Somali women were conducted in Dutch, while most of the interviews with the mothers were conducted in Somali. Interviews lasted approximately 30 to 40 min. Interviewees (with the exception of one) preferred to be interviewed at a location other than home. Interviews were held at cafés, libraries, schools, and community centers. Nearly all the individual interviews were recorded and transcribed verbatim. However, five interviews with mothers were not recorded because those mothers were suspicious of recording. Short notes were taken during those interviews and immediately written out afterwards in field reports.

Two natural group discussions were held and attended by 12 and 14 mothers (aged 23–66 years). The natural group discussions were conducted in Somali, lasted approximately 20 to 40 min, and took place in community centers in two different cities. A group interview protocol was developed by JS, who moderated the focused discussions [28].

All mothers who were interviewed individually also participated in group discussions and/or the member check. Member checking is a method of validity checking, which involves taking the findings back to the participants, presenting the findings for further feedback, and ensuring that the participants agree [29, 30]. The member check, which was carried out in July 2013, was attended by 13 mothers and lasted approximately 40 min. The results were discussed to ensure that JS had comprehended the participants’ views [30]. The group discussions and the member check were recorded and transcribed verbatim.

Data analysis

To improve the credibility and transferability of this study, the RATS guidelines on qualitative research were applied² [32]. Furthermore, the transparency of the study was improved by keeping a research diary and written transcripts.

Somali women and preventive healthcare

Some women suspect that the government is aware of the vaccine’s side effects, yet deliberately uses it on participants to research new vaccines. The young Somali women think that negative information might be concealed by the CIb and in their interviews questioned the government’s cultural norms with regard to sex:

The government thinks: ‘you [a 13 or 14-year-old] probably had sex’, so you [a 12-year old] must take a heavy test or vaccine. I think that you [the government] encourage it [sex at a young age]. As if it is normal?! [N4 (young Somali woman)]

It becomes apparent in this interview that the girl fears that the HPV vaccination campaign normalizes sex at a young age. This is related to young Somali women’s doubts about the vaccination age. However, while they consider the age of 12 too early for girls to be confronted with sex talk, they also see the potential benefits of being prevented from acquiring HPV. The HPV vaccination is also considered as positive, as it protects women from cervical cancer and is considered as similar to other childhood vaccinations. The HPV vaccination is not seen as a measure to prevent Sexually Transmitted Infections (STIs). Aside from this, the HPV vaccine is often assumed to be mandated.

The mothers view Somali, male practitioners as medical experts. Hence, some mothers’ barrier to getting Pap smears is most likely due to having a male practitioner with also a different cultural and religious background. They were not familiar with the possibility of having a Pap smear performed by a female practitioner.

Language, knowledge, and negotiating decisions

Another major barrier is related to language. The Somali mothers from the second migration wave are often not fluent in Dutch, while Somali girls have access to Dutch language and culture through school. Hence, some young Somali women have to translate information about the HPV vaccination to their mothers, which daughters then sometimes perceive as a barrier. As a consequence, they have not always informed their mothers.

Furthermore, daughters and mothers discuss and exchange information on the HPV vaccination with each other, whereby the mothers’ opinions on the HPV vaccination are often decisive. Daughters tend to accept the views of their mother and their peers, and mothers tend to accept the views of their Somali peers.

The mothers’ lack of knowledge and lack of Dutch language skills largely influences their roles in the decision-making process, and this seems more influential than the dissatisfaction with or suspicion towards Dutch culture and healthcare.

Sexual standards, culture, and religion

This idea of eternal shame instills fear in young Somali women and mothers. The mothers use social control and education on Islamic norms as ways to promote chastity among daughters and to prevent stigma. With the cultural double sexual standard on the virginity of girls, the HPV vaccination is not considered necessary. Yet, some girls criticize it by expressing their discontent with the Somali cultural norm that stigmatizes girls who are sexually active before marriage.

However, some participants anticipate difficulties with testing and vaccinating Somali men for HPV. Participants have mentioned that Somali men may assume that their partners do not trust them and feel offended.

Although most participants believe they can control the risk of cervical cancer through sexual behavior, the Somali mothers particularly believe that all diseases are determined by God and indicate low self-efficacy. However, there is also individual religious responsibility to improve one’s health and prevent disease.

Homosexuality is seen as a lifestyle choice for Dutch people, related to the sexual freedom within Dutch culture. The women and girls do not mention the possible transmission of HPV between women.

Strengths and limitations to the study

This study has some limitations. First, discussion within natural existing groups may have inhibited truthful disclosure. However, the women in this study generally have felt safe to discuss sensitive topics within their already established group of cultural peers. Besides, according to the organizations who helped recruit interviewees, the women would not have accepted a structured focus group.

Second, in a few cases, a participant’s friend was present during an interview and some interviews were held in public spaces, which may have influenced the responses to some questions. Third, group discussions and interviews with Somali mothers were conducted in Somali. Meanings of some responses by participants may have inadvertently changed during translation, which could affect the validity of this study [38]. Fourth, some mothers have declined audio-recording, and nuances may have been lost because the information gathered from them is based on extended field reports.

Finally, although the level of education of Somali people in the Netherlands is generally low (Table 1), most young Somali women in this study followed higher education. The researcher has had limited access to lower educated young Somali women, which could be attributed to her own educational background and the short time available for the research. It is difficult to say how such selection bias influences the results. In hindsight, the results do not show major differences in perceptions towards prevention of cervical cancer between higher educated and lower educated girls. Most girls in this study have expressed the importance of these traditional values, independent of their educational background. The recruitment of girls with lower education levels could perhaps have led to more diverse results and saturation [29]. More studies including quantitative research could be used to increase the validity of our findings.

This study has several strengths. First, the recruitment methods applied in this study led to low-cost recruitment and the inclusion of a large number of Somali people. JS’s Somali background has created easy access to the community [39]. Second, this also has helped the participants feel more comfortable and disclose information, enhancing the ecological validity of the study. However, having the same background as the participants also could become a barrier: JS has a reputation to maintain within the community. Possibly, this made it difficult for her to pose questions about sensitive issues, which a Dutch researcher would not have experienced.

Third, data triangulation has been reached by collecting data from different sources, including individual interviews and natural group interviews. Fourth, researcher triangulation and a member check have been applied to increase the validity of the study. During the member check, the researcher established credibility by presenting all findings derived from the interviews and group discussions. These findings were recognized, validated, and further clarified by the participants. Finally, some mothers have been recruited from the first, and other mothers were recruited from the second migration wave. Mothers who are only recently in the Netherlands may be less knowledgeable about the Dutch health system and its preventive measures.

We identified several implications of our study. More information that clarifies misperceptions around the prevention of cervical cancer and the Dutch health care system is important. Furthermore, our study shows how a reversal of parent-child roles can take place as a consequence of parents’ difficulties in navigating the new country in which they live. This study also shows that religion is subordinate to culturally defined gender norms that control the behavior of boys and girls by stigmatizing girls having premarital sex. Moreover, this study shows that Somali women in the Netherlands are often not aware of men’s roles in HPV transmission. Despite the finding that men can be infected with HPV and transmit the virus to women, there are no nationally funded HPV vaccination programs or HPV DNA testing programs targeting men in Europe (with the exception of Austria). Thus, in order to protect the sexual health of ethnic minorities in the Netherlands, it is important to not only consider informing ethnic minorities about men’s roles in HPV transmission, but also to implement nationally funded HPV-related preventive care for men.

This study also shows how social factors influence Somali women’s health beliefs, which underlie participation in cervical cancer prevention. One can question whether the current cervical cancer prevention programs, the ways they are promoted, and by whom, fits with Somali culture. The framing of HPV vaccination messages as specifically preventing cervical cancer affecting Somali women in Somalia and in the Netherlands may increase the perceived severity of HPV and thus increase acceptability of vaccination amongst the Dutch-Somali population [40]. Also, the message that male circumcision has a protective effect against cervical cancer must be presented with caution to Somali women.

Oral education in Somali may reach mothers best [14, 41], as Somalia is known for its long tradition of poetry [41]. Hence, culturally sensitive information on cervical cancer prevention and the Dutch healthcare system can be provided through poetry or theater, and developed together with the target group. In addition, group meetings in which mothers can ask questions, reflect on cultural values, and exchange experiences with each other is recommended. Also, the use of cultural brokers in health care settings, who can identify with and have knowledge of a client’s cultural background as well as the Dutch health care system, might increase the demand for preventive health care [39].

Teachers can provide and discuss culturally appropriate information targeting the double sexual standard in schools. It is also recommended to take the cultural stigma on sexual matters into account and extend the HPV vaccination to an older age, as suggested earlier [33].

This study has unexpectedly increased Somali women’s awareness of their own cultural values and norms, and has stimulated them to reflect on perceptions different from their own. For some, it has even empowered them to change their attitudes towards the prevention of cervical cancer [28]. Forms of participatory research therefore seem appropriate for studying taboo topics, most likely for other marginalized groups as well.