Breast cancer incidence and mammography screening among resettlers in Germany

The present work used data obtained from two different studies conducted in the federal state of North Rhine-Westphalia (NRW), Germany. The InEMa study (Informierte Entscheidung zur Teilnahme am Mammographie-Screening-Programm, N = 4828 [20]) is a cross-sectional study and data were used to observe MSP participation among resettlers. Within this study, questionnaires were sent to a random sample of women residing in the area of Westfalen-Lippe (sub-area of NRW) who had their 50th birthday between October 2013 and July 2014. The sample was randomly chosen by a computer algorithm from a database of the target populations’ addresses retrieved from local registration offices within the study area. Based on the results of previous sample size calculations, it was sufficient to randomly select and contact 56% (17,349 women) of women in the target population rather than contacting all women in the database. Women were sent two questionnaires: one very soon after their 50th birthday (t1), when women in Germany are receiving their first invitation to participate in the MSP, and another one three months later (t2). The t1 questionnaire was used to collect sociodemographic data, as well as data regarding the core elements of an informed decision (participation, knowledge and attitude towards MSP) [21]. These core elements of informed decision were measured also after the women made their first decision whether or not to participate in the MSP (t2). With the information regarding the country of origin and the nationality (German, German due to resettler status, not German), resettlers from the FSU could be identified [20]. The analyses excluded all women with a breast cancer diagnosis earlier in life, because they receive mammograms as part of their follow-up visits and are not the primary target group of the MSP. In addition, women were excluded when participation status at the MSP was unclear, for example due to missing information.

The AMIN study (Aussiedler in Münster - Incidence Cohort Study, N = 32,972 [19]) is a retrospective, register-based cohort study that provides information on breast cancer incidence among resettlers from the FSU and the general German population. The study was conducted in the administrative district (AD) of Münster in cooperation with the federal cancer registry of NRW. The cohort includes a sample of resettlers who immigrated to the AD Münster between 1990 and 2001 and were identified with the help of local registration offices. Person-years were estimated using a validated procedure for cohort studies with an incomplete follow-up [22]. Incidence data as well as population figures of the general population in the AD Münster were available from the federal cancer registry of NRW.

Even though this study compared results from two different study populations, the same target population was investigated. Table 1 is summarising the study characteristics from the InEMa and the AMIN study.

In the InEMa study, participation in mammography screening was recorded in the t2 questionnaire and participants were asked whether the mammogram was performed within the MSP or within an opportunistic screening by a gynaecologist or a radiologist. In some cases, women reported already in the t1 questionnaire that they had participated in the MSP. These women were included in the analyses of this manuscript, even though they did not send back the t2 questionnaire. Additionally, further variables were recorded, which might be possible determinants of MSP participation: educational level (low / medium: no educational qualification, less than 12 years at school or an equivalent degree (in German: Haupt- oder Realschulabschluss); high: at least 12 years at school or an equivalent degree (in German: (Fach-) Hochschulreife / Abitur)), invitation status (invitation to MSP already received yes vs. no), length of stay in Germany (since birth, since ... years) and the language that is mainly spoken at home.

The AMIN study provides information on all cancer diagnoses from 1994 to 2013 of the population of the AD Münster. The analyses were limited to invasive breast cancer diagnoses among women (ICD-10: C50). The cancer diagnoses of resettlers from the AMIN study were identified in the federal cancer registry of NRW using a pseudonymised record-linkage procedure [23, 24]. In addition to the resettler status, the dataset contained information about the age at diagnosis, date of diagnosis and TNM status (T: tumour size, N: lymph node involvement, M: metastases). The tumour stage was defined using the Union for International Cancer Control (UICC) classification system [25]. Therefore, tumours that were smaller than 20 mm (T1) with either no or low lymph node involvement (N0 or N1mic), and in which there were no distant metastases (M0) were classified as UICC stage I (local). All larger tumours (T2-T4), tumours with lymph node involvement (N + (with the exception of N1mic)), and tumours with distant metastases (M +) were classified as UICC stage II + (advanced). Tumours with neoadjuvant therapy were classified as UICC stage 0 [25].

For the period 1994 to 2013, age-specific and age-standardised breast cancer incidence rates (ASRs) based on the old European Standard Population were calculated for female resettlers in the AMIN cohort and the female population of Münster [26]. The rates of resettlers were calculated for 3 years combined. The age-specific rates were combined for three different age groups: (a) < 50 years, (b) between 50 and 69 years, and (c) ≥ 70 years. Differences in breast cancer stages between resettlers and the general German population were tested for significance using Chi-Square tests. Two different periods (pre- and post-MSP implementation) were considered: 1994–2006 and 2007–2013. The cut point 2006 was used even though the MSP implementation began in October 2005, because fewer than 50% of eligible women were invited until 2006 [13]. All analyses were performed using SAS version 9.4.

Using logistic regression, ORs for participation in the MSP between resettlers and the other study participants (women with and without migration background) were calculated, adjusting for the educational level and the invitation status. It was found that resettlers participated more frequently in MSP than the other study participants (OR 2.42, 95% CI 1.08–5.42), regardless of their educational level and whether or not they had received the invitation to MSP. Results showed that women with a lower or medium educational level and who already received the invitation to participate in the MSP had higher odds to participate in the MSP compared to women with a higher educational level and to women who did not receive the invitation for MSP participation yet. Only minor differences were observed between crude and adjusted ORs (see Table 3). In a sensitivity analysis, we excluded women with a migration background from the reference group (N = 345) and thus, compared resettler women with German women, however, the results differed only slightly from the previously reported results (OR 2.48, 95% CI 1.11–5.57).

A total of 199 cases of invasive breast cancer were diagnosed among resettlers of the AMIN cohort, of which 154 (77.4%) were between 50 and 69 years of age. Figure 1 shows ASRs for all age groups.

Prior to the year 2005 (when the MSP was introduced), ASRs were stable for both groups, but lower for resettlers than in the general German population. During the implementation phase of the MSP, a strong increase in incidence rates was observed in both groups, and incidence rates among resettlers rose to a level comparable to that in the general German population. After the MSP was implemented nationwide, the incidence rates in both groups remained roughly stable.

Age-specific incidence rates of breast cancer, separated by three age groups can be found in Fig. 2. It shows that the observations in Fig. 1 may be explained mainly by the breast cancer incidence of women eligible to participate in the MSP. The incidence rates in under-50-year-old women remained almost unchanged. The incidence rates in over-70-year-old women showed no substantial increase during the MSP implementation phase but after 2009, the breast cancer incidence rates of resettlers and the general German population approached values similar to one another.

Table 4 presents the distribution of UICC stages for resettlers and the general German population, separated for two time periods (1994–2006 and 2007–2013).

During the time period of 1994–2006, the proportion of advanced breast cancer stages (UICC II +) was higher in resettlers than in the general German population (68.5% vs. 53.5%). Additionally, an unknown stage was reported less frequently in resettlers. During the time period of 2007–2013, the proportion of an advanced breast cancer stage (UICC II +) of the two groups have converged to similar values and were comparable. The difference in the unknown stage was reduced, but resettlers still showed a lower proportion of unknown stages than women in the general German population. Looking only at women of the eligible age range for MSP (50 to 69 years), a similar distribution of proportions of breast cancer stages between resettlers and the general German population was observed over the two periods (data not shown).

When comparing the immigration periods and the countries of origin of the InEMa and AMIN study, the two study populations mostly immigrated after 1990 and came primarily from the Russian Federation and Kazakhstan. Therefore, both studies reflect the expected immigration pattern of resettlers coming to Germany since the early 1990s. As shown in Table 1, both study populations are residing in the north-western part of the federal state NRW. While the results of the AMIN study cover the time period 1994–2013, the results of the InEMa study reflect the participation behaviour of 50-year-old women in 2013 and 2014. The AMIN cohort is a representative sample of resettlers from the FSU, as resettlers were quasi-randomly assigned to their first place of residence (using the Königsteiner Schlüssel), where they had to live for at least 2 or 3 years [16]. Possible name changes of the cohort were considered by using a name thesaurus.

In this study, we used data from two independent studies, which means that we investigated two different study populations and the comparison of the results should be treated with caution. It is to be noted that the InEMa questionnaires were distributed in the German and Turkish language. Hence, it is possible that a small proportion of women were not able to read the material as some resettlers speak only Russian at home (see Table 2). As it seems reasonable that people with poor language skills are less likely to participate in the MSP [8], it is possible that we have overestimated the association between MSP participation and resettler status. Additionally, it is possible that women with a positive attitude towards the MSP were more likely to have participated in the InEMa study. This is suggested by the high proportion of study participants who have participated in the MSP (about 80% InEMa study vs. 55% Germany-wide). Therefore, the results from the InEMa study could also be explained by selection bias, which we cannot rule out.

It should also be noted that the number of resettlers participating in the study was low, resulting in inaccurate estimates with wide confidence intervals. Previous surveys among resettlers in Germany showed a relatively low response as for example 36% [30]. An analysis of response among individuals with foreign background in Germany by Winkler and colleagues showed that study participation among resettlers is only slightly lower compared to Germans [31]. Unfortunately, it is not possible to calculate the response rate of resettlers in the InEMa study. But given the small proportion of resettlers among the overall study population, the interpretation of the results is limited.

When comparing the cancer stages between the two time periods (“1994 to 2006” vs. “2007 to 2013”), it needs to be considered that improved diagnostics may also have led to a decrease of advanced cancer stages. Furthermore, the number of 50 to 69-year old resettler women with a breast cancer diagnosis was very low in the AMIN cohort, thus, the results from the analysis of the cancer stages should be treated with caution. In the AMIN cohort, it was not possible to perform a mortality follow-up, therefore, person-years of the cohort were estimated [22]. However, sensitivity analyses showed negligible differences in the results. A detailed discussion of the strengths and weaknesses of the AMIN study has been published elsewhere [19].

It needs to be emphasised that resettlers are a specific migrant group in Germany. They are ethnic German migrants who were invited by the German government, they received German citizenship at immigration and therefore, full access to the German social and health systems [16]. In contrast to resettlers in Germany, other non-western migrants in Denmark were found to have a considerably lower willingness to participate in the MSP [6]. Therefore, further research is needed to identify the specific factors explaining the good MSP participation among resettlers which may help to examine whether we could derive our findings to other migrant groups. Furthermore, it should be investigated how resettlers are participating in other screening programmes (such as colorectal cancer screening, skin cancer screening, etc.). Results from the NAKO study, a large prospective cohort study in Germany investigating about 200,000 representative study participants, could be useful for this investigation [32].