We sampled female breast cancer patients whose mother language is Turkish. The patients were recruited at two different stages in their treatment process. Inclusion criteria for the first stage were that the patient should be diagnosed with breast cancer for the first time and the time of diagnosis should not be longer than 3 months prior to enrollment. For the second stage, patients were included while they were in the rehabilitation process after having completed the initial treatment.

Potential Turkish participants were contacted at one CBC and one rehabilitation clinic in Germany. In the CBC, participants were recruited by a case manager. German or Turkish informational material was provided. Interested patients signed consent to be contacted by phone for further information and eventually to make an appointment for an interview at a place of choice. Interviews were done by one female Turkish research team member (YYA, PhD) experienced in qualitative methodology. Only one participant could be recruited in the CBC (n=1) over the time span of 12 months. In the rehabilitation clinic, a medical doctor recruited eligible participants. German or Turkish informational material was given and when the patient was willing to participate, an appointment was arranged during the patient s stay. The interviews took place in the clinic (n=5).

As indicated, recruitment via medical centers turned out to be inefficient. Hence, a different recruitment strategy was applied. Further Turkish participants were recruited using key persons in social networks22. Here, key persons in the Turkish communities were a leader of a self-help group, an organizer of a neighborhood initiative, and a Turkish medical doctor. Study information and informed consent with contact data was handed out by the key persons to female Turkish persons known to have or have had breast cancer. The inclusion criteria to classify participants into first stage and second stage patients (1^st diagnosis, time of diagnosis < 3 months or being in the rehabilitation process) were abandoned. The women could either contact the researcher (YYA) directly or provide a phone number. The interviews took place mostly at the participants home or in the context of neighborhood initiative meetings (n=12).

Before the interviews took place, potential participants received full study information, and written informed consent was obtained. We used semi-structured face-to-face interviews focusing on several topics such as illness perceptions, treatment expectations and psycho-oncological care. The semi-structured guideline was developed in German language by an interdisciplinary team including researchers from various fields, i.e. pedagogy, health sciences, and one breast cancer patient. Questions related to psycho-oncology cover need, knowledge, utilization, content, attitude and appraisal of psycho-oncological care. After finalization, the guideline was translated and pretested in Turkish language with one Turkish breast cancer patient who had terminated cancer treatment. Minor adaptations such as rearrangement of questions and words were made. Socio-demographic data were recorded before or after the interview in a separate sheet. Field notes were marked in a postscript. Information about non-responder was not obtainable as potential participants were approached indirectly via key persons. During one interview a relative was present as requested by the participant. All other interviews were conducted without other persons being present.

This study was approved by the ethics committee of the Medical Council of Westphalia-Lippe and the Medical School of the University of Munster, and by the ethics committee of the University of Bremen.

Interviews were audio-taped and transcribed in Turkish. Transcripts were anonymized during transcription. For the analysis, we applied the method of qualitative content analysis according to Mayring23. The method allows to work deductive as well as inductive and thereby to assess all issues related to our research questions. We entered MS Word 2010 files of transcribed interview into MAXQDA qualitative analysis software (version 11)24 which facilitates the categorization and the organization of codes and categories. Overall, four researchers (LS, YYA, YG, FKE) were involved in the categorization process of whom two Turkish-speaking researchers (YYA, YG) categorized the Turkish interviews. During the categorization process the four researchers discussed and unified the categories regularly in order to simplify the comparison between all participants.

Finally, three researchers (LS, YYA, YG) discussed meanings of unclear text passages, as well as resemblances and variations of Turkish and German participants in regular analysis team meetings. Only selected citations were translated. For this paper, Turkish citations were translated from German into English and rechecked with the Turkish original citations.

Overall, we conducted 18 interviews with Turkish breast cancer patients between February 2013 and February 2014. On average, the duration of the interviews was 45 minutes.

Table 1 summarizes the demographic information. Participants had been diagnosed 3 months to 2.6 years before the interview. The age range was 30 to 73 years (median 53.5 years). Overall, the educational level was rather low: More than half of the Turkish participants (11 of 18) had either no education or an educational level of elementary school. Almost all participants were married (16 of 18) and had zero to five children; the majority had three or more children. Most of the participants have been living in Germany for a long time (range 5 to 43 years, median 36.5 years). As the recruitment of Turkish participants who were recently diagnosed failed, results are reported irrespective of the time of diagnosis.

This qualitative study is among the first to explore the need, utilization and attitudes towards psycho-oncology among Turkish breast cancer patients in Germany. While the results are specific to the care context in Germany, we believe that some general lessons regarding psycho-oncology in cancer care for Turkish migrants irrespective of country may be drawn from our findings.

To summarize, Turkish participants showed a need for support, notably informational and emotional support. Turkish participants face barriers which prevent them from psycho-oncological care. First, due to the language barrier the communication between patient and psycho-oncologist is impossible or stays at a superficial level. Second, cultural differences between the patient and the psycho-oncologist might lead to misunderstandings or advice given by the psycho-oncologist which does not fit into the daily life of the patients. Due to negative experiences with psycho(-onco)logical care, some participants did not believe in its effectiveness. Third, some Turkish patients did not understand the objective of psycho-oncology and/or had prejudices in mind. However, few participants showed a high level of suffering and therefore openness towards psychological help in general.

These results are somewhat paradoxical. On the one hand, the results clearly show a need notably for emotional and informational support among participants. On the other hand, psycho-oncological care was not used although it - in principle - directly aims to cover the needs the participants expressed. Thus, psycho-oncological care as assessed in our study does not adequately correspond to the needs of Turkish participants yet.

The subsequent claim of the participants for Turkish psycho(-onco)logists is a classical migrant or culture specific intervention and fits into a current discussion on how to deal with culturally diverse patient groups. A different approach is to provide culture-sensitive care within the usual health services, i.e. diversity management2526. Even though evaluation studies are scarce, culture-sensitive interventions are mostly preferred considering practicability and financial feasibility.

Our results suggest changes on both, the provider s and the beneficiary s sites, i.e. by offering culture sensitive psycho-oncology, a better and culture-adapted clarification of the purpose and content of psycho-oncological care as well as actions for a change of mind regarding psychological care in general in the Turkish community.

An earlier study in Germany detected no differences in utilization of psychosocial and informational services between immigrant and non-immigrant cancer survivors18. Equally, in our study most of the Turkish participants had at least one contact to a psycho-oncologist. However, Turkish participants did not benefit from psycho-oncological care as it was not used beyond an initial contact partly due to prejudices or a missing belief in the effectiveness of psycho-oncological care. In conclusion, the attitude towards and appreciation of psycho-oncological care is crucial for effective utilization.

Our findings confirm an earlier study2728 which showed a less positive attitude towards psychotherapy among young Turkish adults living in Germany compared to young German adults. The sex and the grade of acculturation moderated the attitude27. In a systematic review28, further determinants of attitude towards seeking psychological help were sociodemographic variables, fluency of the host country language, and psychological attributions of distress. A study on Moroccan and Turkish migrants in the Netherlands stressed the impact of sociodemographic factors rather than ethnicity29.

The low level of information about their disease among Turkish breast cancer patients might be due to the low educational level, poor language proficiency, and the cultural habit that the diseased person is not told the truth about his or her disease15. Furthermore, the caregivers attitude, lacks of effort to provide information and of cultural understanding may also contribute to an overall low level of disease knowledge in these patients. Similar results could be ascertained among Turkish immigrants in Germany with Diabetes mellitus type 2 who showed lower disease knowledge and outcome quality despite more diabetes training compared to a non-immigrant sample of type 2 diabetics30. Informational needs on the disease and the health system were also detected in Australian migrant studies which mostly were related to poor proficiency of the language of the host country2019.

This study has some limitations. First, as the results of this study are generated by qualitative interviews this sample does not necessarily reflect Turkish or other migrant female cancer patients in general. Hypotheses derived from this study therefore should be tested in a larger sample. Second, it was not possible to recruit Turkish cancer patients recently diagnosed so that differences in utilization and appraisal of psycho-oncological care at the beginning and the end of the intense cancer treatment could not be compared. We assume that differences exist according to the phase of cancer treatment. For this reason, it is worthwhile that different phases of the treatment course are reflected in future studies. Third, bilingual research projects face the problem of lost information due to translation. The meaning of expressions and underlying social constructs might differ between cultures and thus might be interpreted differently. For example, in this study, Turkish participants wished to relax in the psycho-oncological setting which initially would not be defined as an objective of psycho-oncology. However, this expression might contain more meaning for Turkish participants. A methodology to address this problem is missing but should be developed for future bi- or multilingual health research projects.