br Results During and a total of individuals were
Results During 2014 and 2015 a total of 886,088 individuals were invited to participate in the colorectal cancer screening. 571,918 individuals submitted a sample for analysis. From these 13,814 (2.4%) were excluded due to missing information, previous colorectal cancer diagnosis, deceased prior to the invitation or due to faulty date registration in the register. Of the 10,786 who were excluded due to missing information on sociodemographic variables, 22,27% were non-Western immigrants. 2.58% of excluded non-Western immigrants submitted an ineligible sample. This left 558,104 for analysis, of which 2,164 (0.4%) submitted samples ineligible for analysis (Fig. 2). In 558,104 study subjects, 527,421 were native Danes of which 1,846 (0.4%) submitted ineligible samples, 13,521 were Western immigrants of which 48 (0.4%) submitted ineligible samples and 17,162 were non-Western immigrants of which 270 (1.6%) submitted ineligible samples. Highest origin proportions of non-Western immigrants were Turkey, Bosnia and Herzegovina, Iran, Iraq and Yugoslavia. Highest origin proportions of Western immigrants were Germany, Sweden, Poland, Great Britain and Norway (Table 2). Of the 314,170 individuals who did not submit a sample, 90.43% were native Danes, 3.83% were Western immigrants, 5.72% were non-Western immigrants and 0.02% were of unknown origin. The proportion of individuals in subgroups submitting an ineligible sample varied from 0.2% (highest income tertile and higher educational level) to 1.6% (Non-Western immigrants). Differences within subgroups were seen in immigration status, age groups, marital status, educational level, and income tertiles but not between genders. Differences were most evident in relation to immigration status, educational level, age and income tertiles (Table 1). Of the 2,164 individuals who submitted an ineligible sample, 1,647 (76.11%) later submitted an eligible sample during follow-up, some individuals submitting as many as 5 times before their sample were eligible. 517 (23.89%) individuals either did not submit a new sample or submitted 2–5 samples, all being ineligible. 17.21% of these 517 were non-Western immigrants, 2.13% were Western immigrants and 80.66% were native Danes. The univariate logistic regression models showed an increased odds ratio of submitting an ineligible sample for non-Western immigrants (OR 4.55 CI95% 4.00;5.18), compared to native Danes. Western immigrants’ odds did not differ from native Danes (OR 1.01 CI95% 0.76;1.35). Odds ratio increased with age, decreased with increasing income and educational level and those with a partner were at lower odds compared to singles (Fig. 3). After full model adjustment, non-Western immigrants still had increased odds ratio for submitting an ineligible sample (OR 3.64 CI95% 2.86;4.64) and Western immigrants did not have an increased odds ratio. Interaction was present between immigration status and marital status. The odds ratio was lower for native Danes with a partner compared to others. The odds ratio was not lower for Western or non-Western immigrants with a partner compared to others (Fig. 4). Interaction was also present between income and educational level, but this Conessine did not affect the estimates for immigration status.
Discussion 98.4% of participating non-Western immigrants submitted an eligible sample, but non-Western immigrants were at increased risk of submitting an ineligible sample for analysis in this sample of 558,104 individuals. Educational level, income, age and marital status also affected the risk. Adjustment for these variables and gender decreased the risk estimate for non-Western immigrants from 4.55 to 3.64. As the information distributed in the invitations for screening and the health authorities’ online information were only available in Danish, potential language barrier may have affected the non-Western immigrant's ability to follow directions, and complete and submit the sample correctly. In a qualitative study language barriers and low literacy were found to be barriers for participating in colorectal cancer screening . This is possible in our sample as well as we found lower proportions of ineligible samples among Western immigrants and native Danes as well as with increasing educational level. Several studies have reported lower participation rates by both Western and non-Western immigrants [, , , , , ], but in our study, all individuals included in the analysis tried to participate, although some failed to do second messenger correctly. This adds knowledge to the field, as we now have an indication that immigrants are not only less likely to participate with current intervention strategies, they are also less likely to succeed when they actually try. As immigrants participate less than native Danes do in this sample , and the immigrants who try to participate fail more often than native Danes, the inequality in participation is further enhanced. Individuals who choose to participate should have the possibility to do so within their own abilities. Individuals submitting ineligible samples would receive a new invitation and are therefore not excluded from the screening. They may thus still gain the possible benefits from screening, but a new sample kit and manual are identical to the original and does not include reason for ineligibility of the first sample. Several submitted samples by the same individuals do not only increase costs of screening but may also be inconvenient, frustrating and unpleasant for the citizen. Almost a fourth of individuals with an ineligible sample either never submitted a new sample or had consecutive and exclusively ineligible samples. Higher educational level and income level decrease the risk of delivering an ineligible sample, which may be due to better language or reading skills or may be due to a higher ability to understand health information, which could result in a higher awareness of the colorectal cancer screening and the sampling procedure. Higher awareness can increase the intention to participate  and may also aid participants in the completion of their sample. A manual or an online guide for sample completion in other languages may increase the awareness of screening, the understanding of how to complete the sample and thereby the ability of non-Western immigrants to submit an eligible sample. Translating guiding materials is not expected to eliminate the problem of ineligible stool samples, but could possibly increase the proportion of non-Western immigrants delivering an eligible sample. Other factors, such as self-efficacy, reliance on others for completion , limited health literacy and cultural factors could have affected the odds of submitting an ineligible sample. Visual and verbal information in mother tongue could possibly increase awareness and self-efficacy  and thus may increase eligibility proportions.