Conducting qualitative research on cervical cancer screening among diverse groups of immigrant women

Forming an effective research team and building community partnerships

The interdisciplinary research team included family physicians, a medical anthropologist, a sociologist, a pathologist, experienced qualitative researchers, and field workers recruited from each ethnolinguistic group.

Collaboration with a local immigrant settlement agency, the Settlement and Integration Services Organization (SISO), enabled us to enrol participants from diverse groups in the community. Authorization for the partnership was given by the executive director of the agency. Two of the researchers had a previously established relationship with the organization. Through SISO, we identified potential field workers and recruited participants. Linking with an established community-based organization that had strong ties to the target participant groups was critical to the success of the project because this became the central “node” for interaction with potential participants. Other researchers have also found that establishing trust with community organizations that serve the population of interest aids in recruiting people from ethnic minority groups.4,5

Bilingual field workers recruited from each of the selected ethnolinguistic groups brought essential cultural knowledge to the study; however, they also required practical training in research principles, participant recruitment strategies, and qualitative interviewing techniques. They were included in the research team early and participated in 2 to 3 dedicated training sessions with the methodologist, using the interview guides developed for their particular groups. Training included how to identify eligible participants and explain focus group interviews to them; obtain informed consent; co-facilitate focus group interviews in English; and later facilitate group interviews in their own first languages (including encouraging participation by all group members, using probes to obtain more in-depth information, and collecting postinterview demographic information from participants).

Accessing communities and recruiting participants

For each ethnolinguistic group, we conducted 2 separate focus groups: one in English and the other in the first language of participants. Two separate focus groups were done to assess the effects of culture alone and then to assess the effects of culture and language together. Some of the immigrant women with good English-language skills still preferred to participate in the native-language groups. We also included a group of English-speaking, Canadian-born women of low socioeconomic status, in part to test the interview guide and focus group processes, as well as to assess the effects of socioeconomic status, independent of ethnicity, on key research questions.

During this study, we learned that inclusion criteria might need to change for reasons of cultural appropriateness. For example, field workers advised us that because of cultural and religious norms, it might not be appropriate to invite unmarried women to discuss sexual health issues, particularly if the group also included married women; thus, eligibility criteria were altered to include married women only. We also loosened the original requirement for the maximum number of years living in Canada of 5 because of difficulty recruiting newer immigrants in some groups. Participants often preferred to come with a friend or relative who had been in Canada longer, and we welcomed all attendees in the spirit of hospitality that our field workers advised us would be beneficial to recruitment. Although these recruitment methods and the snowball technique might have resulted in groups comprising more acculturated and less representative women, they were essential to ensuring participation.

Our initial strategy of placing recruitment posters in strategic locations, such as community centres and places of worship, did not elicit any responses. The most effective strategy was recruitment by word of mouth, using the snowball technique to reach additional women through each contact, as well as face-to-face contact and scripted telephone calls by field workers and SISO employees who had well established roles in their ethnolinguistic communities. This supports the overall consensus in the literature that many minority groups respond favourably to direct, personal appeals from known individuals as a recruitment strategy.6–8 There was also a need to respect different gender roles and expectations in some of the ethnolinguistic groups; some women’s husbands were “gatekeepers,” speaking with the field workers to ascertain the purposes of the groups, and giving their consent with the condition that their wives be picked up from home or accompanied by a friend or relative.

As reported in other studies, we found that ethnoracial and linguistic differences between the researchers and the participants were not a barrier to research because of steps taken to ensure cultural appropriateness in our recruitment strategies, interview guide development, and hospitality at the focus groups.9 Without the direct involvement of members of the selected ethnolinguistic groups to serve as field workers, however, we would not have known about culturally appropriate methods of securing the trust of potential participants.

Obtaining informed consent

We found that obtaining written informed consent could be complicated by differing norms, concerns, and expectations across cultural groups about signing documents. In 1 of the 2 Somali groups, the women did not understand that the purpose of the consent was assurance of voluntariness before participating, and they expressed concern about signing the form before experiencing the research. A lengthy discussion ensued, resulting in a reduced focus group length, and ultimately the group decided to provide written consent only after the focus group discussion was finished.

Social scientific research in settings where informed consent is not appropriate or advisable (eg, in situations where being identified could result in participants’ arrest or worse) has resulted in growing awareness by research ethics boards that, even in North America, verbal consent might be both appropriate and more comprehensible to participants.10 Where participants are unfamiliar with research and focus groups, extra time should be allotted for discussing written consent and its alternatives.

Sensitive and innovative data collection approaches

To ensure that the research setting was informal, relaxing, and welcoming, most of the focus groups were held at the local immigrant resettlement office, a setting that is familiar to many immigrant groups in Hamilton. Participants affirmed that a culturally appropriate meal and provision of on-site child care by a known member of the community increased their comfort level and signaled that we recognized that many participants were busy wives and mothers whose time, as well as insights, were valuable. We had expected participants to be shy owing to the nature of the topic and the conservative cultures and religions of some groups; however, the women in all groups displayed an unexpected amount of interest, enthusiasm, and willingness to share experiences.

We developed a focus group plan (ie, interview guide or list of topics, questions, and probes, plus activities and materials) that incorporated interactive hands-on activities designed to help participants comfortably engage with the topic of discussion. Field-testing the guide with the Canadian-born group helped us to modify our plan. At the suggestion of field workers, instead of directly questioning participants, we used indirect questioning (eg, “Imagine you have a friend, [common name], who comes to you with some questions about her health. What do you think your friend would know, do, or think about [topic]?”). We also incorporated 3 interactive activities: drawing and labeling reproductive organs on a blank female form on a flip chart as a group, viewing and discussing a videotape showing a conversation between a female physician and a patient contemplating a Papanicolaou smear, and handling and discussing a vaginal self-sample kit for human papillomavirus testing.

The anatomy labeling exercise provided context for discussing cervical screening. Although shy at the beginning, the women in each focus group later became quite animated, joking and teasing one another about their drawings and labels: “That uterus is too big,” said one woman. Another replied, “Yes, but this woman has had 6 children already!” This activity helped “break the ice.” It also revealed that some women do not know any specific words for cervix in their native languages, and that anatomic terminology greatly varies across dialects. We learned that it was important to ensure that the women and field workers understood exactly what was being discussed by providing appropriate descriptions rather than assuming common knowledge from the use of a single term. By doing so, the terminology used by the women could be incorporated into culturally relevant non-English audio and print educational material.

Developing a standardized video, which was used in all groups, demonstrated that such a methodology was acceptable across all groups and useful even for a sensitive health topic. Using additional activities (eg, drawing, videotape, and self-sampling kits) stimulated intensive discussions in all groups, demonstrating the usefulness of different tools to engage the groups.

We learned from field workers that immigrant women want information about cervical screening and related health concerns that they are not receiving from their regular health care providers. To respond to their needs within the research process, we added a postinterview “doc talk”—a question-and-answer period with a female physician. The physician also provided some basic information about cervical health and cancer prevention. Women were very engaged in and appreciative of this part of the group meeting. In fact, the venue proved a missed opportunity to provide services, as several participants indicated that they would have liked to have had Pap smears done right away.

Overall, the focus group process and activities were embraced by women across all groups. This is in keeping with emerging findings that focus groups are a uniquely appealing and effective approach for engaging people from culturally and linguistically diverse backgrounds.1

Managing budget and time requirements

Multilanguage research requires sufficient funds and time. For example, the cost of translation varies considerably, depending on how unique a language is within a given setting (eg, Spanish-speaking translators are easier to find and less costly to hire in Hamilton than Dari- or Somali-speaking translators are). We needed to balance methodologic rigour with costs for both translation and back-translation. It was prohibitively expensive to use a professional translation service ($12 000 for 1 professional translation), and we pursued alternative strategies and hired the field workers or bilingual university students to transcribe and translate the research materials and focus group discussions. The main limitation of our approach was the lack of back-translation to confirm the accuracy of the interview guide terminology and the final transcribed interviews. There was an advantage in terms of local cultural familiarity; similar to our study, Mann et al found that back-translating interview guides is best done by a person who is much like chosen participants.11 Although fluent in the language of their groups, the field workers still encountered difficulties with the vernacular of different regions.

A number of studies have reported the need for flexibility in accommodating participants’ availability and unexpected increases to timelines and budget.9,11,12 Our data collection was extended by approximately 12 months because each new ethnolinguistic group represented in some ways the start of a new process with respect to the need for cultural learning and adaptation of our approach and tools. In addition, the scheduling of groups was often delayed for key seasonal events, religious holidays, and at one time because of violent conflict in the home country of one group. As in the study by Mann et al, we also found that several follow-up and reminder calls by the research team were needed to ensure an adequate number of participants.11

Given the complexity and time demands of the study, harnessing the energies of a group of researchers and community partners was important. More researcher and staff hours were needed to organize and conduct the immigrant groups than the Canadian-born group. The interdisciplinary research team was important in bringing varied and practical expertise to a complex project.

Conclusion

A number of innovative strategies were discovered to be essential to conducting a qualitative study with multiple ethnolinguistic groups. We have highlighted 5 areas of the research process where challenges were encountered, and described some successful approaches. Formation of an interdisciplinary research team, partnership with an agency serving the groups of interest, inclusion of members of the target ethnolinguistic groups, allowing a choice of language for discussion, and incorporating culturally acceptable and interactive activities during the focus groups were among the important factors. Challenges that we did not foresee included the slow pace of recruitment and data collection, changes to our desired methodology to incorporate cultural norms, and high cost of professional translation services in some languages. These challenges, which were not insurmountable, served to underline the need for flexibility and innovation in conducting community-based research with immigrant and ethnic minority populations.

Development of culturally appropriate strategies can enable family medicine researchers and health care providers to understand and meet the needs of diverse primary care populations. Our research approach provides guidance for doing research with multiple ethnolinguistic groups, which can be applied in many content areas.