Introduction

Despite recent changes in refugee resettlement policy, the United States (U.S.) remains the top refugee resettlement country in the world. [1] The U.S. has admitted over 3.3 million refugees since 1975. [2] During the 2016 fiscal year alone, 84,995 refugees resettled in the U.S. with most arriving from the Democratic Republic of the Congo, Burma, Iraq, and Somalia. Between 2002 and 2017, Texas was the second refugee resettlement destination after California. [3]

Most of the countries of origin of recent refugees are in the midst of an epidemiologic transition characterized by a decrease in infectious diseases and an increase in the incidence of cancers and other non-communicable diseases. [4–6] Cancer prevention services in these low-income countries, however, are either nonexistent or inaccessible for the majority of the population. [7,8] Refugees represent a specific sub-category of immigrants to the U.S. The universal definition of a refugee according to the 1951 Convention Relating to the Status of Refugees contained in Article 1(A)(2) of the 1951 Convention and 1967 Protocol Amendment is someone, “owning to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence, is unable or, owing to such fear, is unwilling to return to it.” [9]

Refugees face multiple challenges following resettlement in the U.S. that may contribute to lower cancer screenings. These challenges include language and cultural barriers, fear of procedures, history of trauma, work schedules, childcare commitments, and a lack of understanding the U.S. medical infrastructure. [10–17] Previous studies found refugees are less likely to get screened for colon, cervical, and breast cancers compared to U.S.-born populations. These studies, however, are generally small and without a comparison of screening uptake with other refugee groups. [18–21]

The purpose of this cross-sectional study was to assess differences in uptake of cervical, breast, liver, and colorectal screens across six immigrant groups participating in the Building Bridges program. Identifying differences in demographic and screening behaviors and acceptance among different refugee groups may help clinicians and cancer prevention program planners recognize the need for culturally sensitive and individually tailored approaches between the ethnic groups that make up the broader refugee community, and thus reduce the cancer burden among these groups.

Methods

Measurements

Data collected prospectively by CHWs throughout the BB Program (2014–2018) was abstracted by investigators in a secondary analysis to assess differences in screening uptake across all cultural groups participating in the BB Program.

Statistical analysis

Descriptive statistics of baseline measures were calculated across cultural groups. Uptake proportions were calculated for each cancer type (cervical, breast, liver, and colon) among age-eligible participants, by gender and by each cultural group. Differences in uptake of cancer screening services across cultural groups were assessed using stratified analysis and logistic regression. Prevalence odds ratios (POR) and 95% confidence intervals (CIs) were calculated for each cultural group to assess the association between screening completion (dependent variable: yes/no) and cultural group (independent variable: using Myanmar/Thailand, the largest among all six groups, as the comparison group). Wald χ² p-values less than 0.05 were considered statistically significant. All descriptive and statistical analyses were carried out using SAS® software.

The University of North Texas Health Science Center North Texas Regional Institutional Review Board approved all study protocols. Written consent forms were translated into 13 languages with on-going consultation with project bilingual staff and community advisors. These consents were field-tested and refined prior to use with members of each respective immigrant community. Key personnel obtained verbal and written consent from all participants. While all participants received and signed a written consent, the consent was also read to some participants who could not read by themselves due to low literacy levels. The community health workers encouraged questions during the consenting process as another strategy to check for comprehension.

Results

Baseline demographic characteristics of 874 refugees enrolled in the BB Program between March 2014 and September 2018 are shown in Table 1, stratified by cultural groups. The majority (81%) of participants were female, and more than half (53%) were younger than 40 years of age. Two-thirds of participants reported living in the U.S. for less than five years. Most (81%) participants reported not speaking English well or not at all. Seventy-six percent never attended school or had less than 12 years of education; 65% reported that they were unemployed; and 51% were without health insurance at enrollment into the BB Program.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Baseline demographic characteristics of Building Bridges program participants, by geographic origin, Fort Worth, Texas, March 2014 –September 2018 (N = 874).

https://doi.org/10.1371/journal.pone.0230675.t001

Overall uptake of screening services for each cancer type (cervical, breast, liver, and colon) among eligible participants by gender is shown in Table 2. The majority of participants screened were women. Among 662 age-eligible women across all cultural groups, 542 (82%) were eligible for a cervical cancer screen. Of those, 200 (37%) completed a cervical exam. Nearly three-quarters (74%) of those screened had never had a cervical cancer screen before. For cervical cancer screening, the Bhutanese had the highest (47%) uptake among all six groups (See Fig 1). Following Bhutan, Myanmar had the second highest (43%), and the Arabic group had the lowest (24%) uptake for cervical cancer screening among all groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Utilization of cancer screenings among Building Bridges program participants, by geographic origin, Fort Worth, Texas, March 2014 –September 2018 (N = 874).

*Colon cancer screening was offered to those enrolled after 1st March 2017 and met the age of eligibility. No participant from ‘Others’ group was eligible for colon cancer screening. ^a Includes participants from both Myanmar (n = 282) and Thailand (n = 3), ^b Includes participants from: Congo (n = 124), Rwanda (n = 50), Burundi (n = 17), Tanzania (n = 1), Uganda (n = 1), and Kenya (n = 1, who reported Swahili as native language), ^c Includes participants from both Bhutan (n = 135) and Nepal (n = 28), ^d Includes participants from both Somalia (n = 138) and Kenya (n = 3, who reported Somali as their native language), ^e Includes participants from Arabic speaking countries: Sudan (n = 43), Iraq (n = 12), Jordan (n = 4), Syria (n = 7), and, Egypt (n = 1), ^f Includes participants from: Afghanistan (n = 3), Angola (n = 1), Chad (n = 4), Eritrea (n = 7), Ethiopia (n = 7), Liberia(n = 1), and, Senegal (n = 1).

https://doi.org/10.1371/journal.pone.0230675.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Utilization of cancer screening services among Building Bridges program participants by gender, March 2014 –September 2018 (N = 874)^*.

https://doi.org/10.1371/journal.pone.0230675.t002

For cervical and breast cancer screenings, we observed statistically significant associations between the cultural groups and screening uptake for each p-value = 0.01 (Table 3). Also, we found a statistically significant difference in utilization of cervical cancer screening for the Arabic group (OR = 0.41; 95% CI = 0.20, 0.83) and Somali group (OR = 0.50; 95% CI = 0.28, 0.87). No statistically significant difference in uptake of cervical cancer screening was observed for Central African region, Bhutanese, and Others when compared with Myanmar groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Utilization of cancer screening services among Building Bridges program participants, by geographic origin, Fort Worth, Texas, March 2014 –September 2018 (N = 874).

https://doi.org/10.1371/journal.pone.0230675.t003

The uptake for breast cancer screening was 50% or more in all cultural groups except one, Central African region (35%). The odds of participants from the Central African region group accepting a mammogram was 63% lower than the odds of accepting a mammogram from members of the Myanmar group (POR = 0.37, 95% CI = 0.18–0.73, p-value = 0.01). More women received a mammogram (54%) and HBV screening (47%) compared to cervical screens, even though most mammogram and HBV completers were first-time screens (79% and 98%, respectively).

Uptake of HBV screens among men (33%) were lower compared to women (47%). Over 90% of men and 90% of women screened for HBV reported never being screened for HBV. The screening uptake for HBV screen was highest (49%) in the Myanmar group and lowest (31%) among participants from Arabic-speaking countries. Overall, no statistically significant association was found between uptake of HBV screens and geographic origin (p = 0.08). However, the odds of HBV screening acceptance from the Somali group was 48% lower than odds of HBV screening acceptance from the Myanmar group (OR = 0.52, 95% CI = 0.31–0.88, p-value = 0.02).

Among all four cancer screenings, the overall uptake was lowest (27%) for colon cancer screening. Colon cancer screens were accepted almost equally between men and women (26% vs. 28%, respectively). All men and women screened for colorectal cancer were first-time screeners. Uptake for colon cancer screening was highest in the Central African region and Arabic speaking countries (41% vs. 40%, respectively) among all six groups. Overall, no statistically significant association was noted between uptake of colon cancer screening and geographic origin (p = 0.19), though numbers were small in these groups.

Discussion

In this cross-sectional study, we investigated differences in uptake of preventive cancer screenings among multi ethnic immigrants participating in the Building Bridges program-a community-based intervention focused on cancer prevention screenings among refugees. Additionally, we assessed the uptake among those reporting to have never been screened. Overall, the results illustrate variation in the uptake of different types of screenings, between the different ethnic groups assessed and evidence of reaching the never before screened. To our knowledge this study provides the most comprehensive evidence of cancer screening uptake among multiple refugee groups enrolled in a community program. There is a limited evidence base on preventive screening uptake among immigrant groups, as well as, how to implement culturally and linguistically appropriate cancer screenings among diverse populations.

Our results show that cancer screening uptake varied among refugee group by type of screening. We found significant differences in cervical, breast, and liver screening completion across the six refugee groups served through the BB program. However, we did not find evidence of differences in colon screening uptake between these groups, although the numbers were too small for meaningful conclusions. Overall, the Myanmar and Bhutanese groups had higher screening completion compared to Somali, Central African, and Arabic speaking groups. These differences may be explained by longer time resettled in the U.S. by the Myanmar, but it fails to explain the Bhutanese group who were more recent to the U.S. Even though the Myanmar and Bhutan groups reported not to speak English well and reported more unemployment than other groups, their uptake was higher than other groups in the study. Mammograms were the most accepted cancer screen among most refugees (54%), with the exception of the Central African group (35%).

Previous research has also shown that refugee groups respond differently to preventive screenings. For example, Samuels et al. found Somalis were significantly less likely to be screened than Vietnamese and Cambodian. [24] Another study comparing Somali immigrants with other African immigrant groups found a significant difference in ever being screened for breast cancer, but not for pap screening. [25]

That our study found a lower cervical screening uptake among Somali women compared to other ethnic groups align with the findings of a randomized control trial pilot study conducted in Minnesota using the same methodological approach of targeting participants outside of a health system or medical home. In this study, only 19.4% Somali women accepted the cervical cancer screening [26] compared to 27% of Somali women in our study. Similar findings were observed in the Boston area where Somali women were found to be less likely to be screened for cervical cancer than Central African women (21.3% vs. 44.1%). [25]

A higher cervical screening proportion of 48.79% among Somali women, however, was found in a clinical chart review in a primary care setting in Minnesota.[21] Although direct comparison is complicated by methodological and setting differences, there is evidence that the uptake of cervical cancer among Somali women is lower than other refugee groups.

For the Arabic-speaking community, the screening uptake for cervical cancer (24%) found in our study was lower compared to other findings from studies conducted in the US [15,18,27] For example, a survey of religion-related factors among American Muslim women conducted by Padela et al. in partnership with the Council of Islamic Organizations of Greater Chicago found that 84% had obtained a pap test in their lifetime. Screening differences may be due to variations in study design, recruitment methods, education type (e.g. Individual education vs. group, clinical chart review, DVD), and health educator used in each study. The lower screening uptake observed in our study may also be due to a difference in method when calculating screening uptake. We excluded women who were up-to-date with their pap exam at the time of recruitment since our program focused on increasing screening to recommended guidelines. However, other studies among Arabic populations in the U.S. have mostly measured women who have ever received a Pap in their lifetime. [15, 18, 27, 28]

Breast cancer screening was higher in our study than what was observed in research conducted in Buffalo, New York (54% vs 35%).[29] This study bears similarities to ours because it included the same culture groups and focused on never screened and not up-to-date participants. The difference in completion proportions between the studies may be explained by the type of education and assistive services offered. Study participants in Buffalo, NY, received health education and one-on-one navigation from an experienced community health educator with the use of interpreters while participants in the Building Bridges program received transportation and language interpretation assistance for screening appointments in addition to culturally tailored health education and one-on-one navigation provided by an individual community health worker from the same cultural group. Having a health educator of the same ethnic group has been shown to reduce significant cultural, linguistic, and logistical barriers to health care. [30]

Like our study, several studies conducted in Canada, found associations between regions of origin and mammogram screening. [31,32] In one of these studies, Vahabi, et al, found that Asian immigrants were more likely to complete screenings (62%) than African immigrant communities (40%).[31] The screening completion proportion among the sub-Saharan African refugees to Canada was 46.6% in another study.[31] In our study, breast cancer screenings exceeded the Vahabi study overall (54%), but the Central African group had lower screening completion (35%) for breast cancer compared to Myanmar group (59%). In our study, the proportion of participants from the Central African region accepting breast cancer screening were lowest among all groups compared to Myanmar. Similarly, in a survey conducted in the Greater Boston and New Hampshire area, only 25% of Congolese women were up-to-date with their mammogram exam. [25]

Colorectal (Colonoscopy and FIT) education and screening started in March 2017. Among the 68 participants eligible for screening, 59 had never received a prior screening (87%). Due to this small sample size, we focused on the overall screening proportions of those in the program. Other studies reported higher proportions of patients screened at baseline. For example, 54% of immigrants and refugees reported a prior screening in Toronto, Canada and in a study of Somalians in Minnesota reported prior screening proportions of 38.5%. [21,32] Data for these studies came from their patient populations collected through focus groups and chart reviews. This study design may reveal higher screening proportions because the participants were current patients of the health system, thereby increasing their exposure to screening opportunities. As discussed, 51% of our population lacked health insurance and were not enrolled in a discounted health program at baseline. This high percentage of uninsured may explain how we were able to reach refugees that had never been screened prior to participation in the program.

Our results from the Building Bridges program suggest that this program is reaching those who report never being screened (66% pap screening; 73% mammogram; 97% hepatitis, 87% colon screen) Additionally, among the never screened, we observed a high percentage of screening completion (74% pap screening; 79% mammogram; 97% hepatitis; 100% colon screen). Other studies among never screened resulted in refugee women in Buffalo (NY) following a community-based educational program, 20 of 58 (34%) women received a mammogram who had never completed a mammogram screening before intervention. [29]

United States domestic screening requirements for arriving refugees mandate refugees to have Hepatitis B Virus (HBV) screening.[33] However, few of our participants recalled a prior testing, nor could they verify their HBV status at the time of enrollment (69% did not know their status). This is similar to other studies that have reported a lack of awareness of HBV and/or HBV status in the general population.[34] When offered an HBV screening, acceptance proportions were <50% in each cultural group. Overall, there were no significant differences among culture groups for HBV screenings. To our knowledge, our study is the first to explore prior HBV screening and awareness of status post arrival in a community setting versus a clinic setting.

Some of these findings are not surprising and suggest that although while not simple to do, the act of making cancer screening outreach, education and screenings available in multiple languages and adapted to different cultures, does not automatically lead to adoption by the targeted communities.

Previous research shows the challenges of providing culturally and linguistically appropriate education to immigrant populations.[11–14] Despite the program removing most known barriers, screening uptake was lower in all populations compared to the U.S. national average and fell short of the Healthy People 2020 Targets of 93% for cervical, 81.1% for breast, and 70.5 for colorectal screening.[35] As of 2015, the screening proportions in the overall U.S. population were 81.2% for pap test, 71.6% for mammogram, and 62.4% for colon screening.[35] The low uptake of screening observed in our study are similar to findings from other studies reporting lower screening proportions among refugees and immigrants when compared to U.S. born populations.[19–21, 36] Acceptance of screening was highest among the Myanmar and Bhutanese groups. The Building Bridges program was most effective in these two culture groups. More research is needed to help explain not only the variation in uptake of cancer type, but also reasons for not screening in general. Additionally, given the substantial financial and staff investment in community health worker models, studies comparing the cost-effectiveness of models, i.e., community vs clinic based, is warranted. As efforts to include diverse immigrant populations expand, variation in uptake by type and population groups should be considered. The results of this analysis provide evidence that uptake in screenings vary among ethnic groups, even with a culturally and linguistically appropriate outreach and education strategy.