Demographics and Perceptions of Barriers Toward Breast Cancer Screening Among Asian-American Women

ABSTRACT

Objective: The purpose of this study was to identify and compare the differences in perceived barriers for mammography screening among four ethnic groups of Asian-American women.

Methods: Self-administered questionnaires were completed by 315 women in these four groups.

Results: The results from the multivariate analyses indicated that even after taking into account insurance coverage and demographic variables, specific barriers still prevent Asian-American women from seeking mammography.

Discussion: Recognizing similarities and differences in the barriers among demographic variables such as ethnicity, age, and length of U.S. residency among Asian subgroups can assist health professionals to address their needs when promoting adherence to mammography guidelines.

KEYWORDS:

• Breast cancer
• Chinese
• Korean
• Filipino
• Asian Indians

INTRODUCTION

In 2006, approximately 212,000 women in the United States were diagnosed with breast cancer and more than 40,000 of them eventually will die from the disease (American Cancer Society, 2006). In Asian Americans, cancer is the number one cause of death, while heart disease is the leading cause of death in all other racial groups in the United States (National Center for Health Statistics, 2005). Breast cancer is the most frequently occurring cancer for Asian-American women; survey data from the National Center for Health Statistics estimates that 3.9% of total annual deaths among Asian-American women result from breast cancer, compared to 3.3% of deaths among all women in the United States (National Center for Health Statistics, 2003).

Despite the fact that Asian-American women have lower incidence rates of breast cancer than either Caucasian or African-American women, when Asian women move from Asian countries to the United States, their rates of breast cancer increase. In particular, a study by Ziegler et al. (1993) showed that the longer Asian women live in the United States, the greater their risk of developing breast cancer. In fact, their risk of breast cancer rises as much as six-fold compared with women in their native countries. Therefore, it is critical to promote early detection of breast cancer in this population. Critics of breast cancer screening have cited fear, the increased diagnosis of in situ disease, and the high number of false positives as harmful by-products of screening (Gelmon & Olivotto, 2002). Notwithstanding the debates, mammography combined with clinical breast examination (CBE) continues to be the primary means for early detection of breast cancer. For women age 40 years and older, the American Cancer Society recommends yearly mammograms and CBE. However, statistics in Health, United States, 2005, show that only 53.5% of Asian American women age 40 years and older had received a mammogram within the past two years, which represents the lowest screening rate among all U.S. racial and ethnic populations (National Center for Health Statistics, 2005). Furthermore, minority women, including Asian Americans, were far less likely than the population as a whole to repeat mammography (Blanchard et al., 2004). In addition, Asian-American women have been more likely to receive a diagnosis at a later stage and to have larger tumors at the site than white women (Hedeen, White, & Taylor, 1999; Yip, 1997).

According to the United States Census Bureau, the Asian-American population represents the fastest-growing and most diverse ethnic group in the United States (Gerieco & Cassidy, 2001), and it is projected that by the year 2020 the Asian-American population will increase to 20 million (National Asian Women's Health Organization, 1997). Despite this growing presence, little is known about the specific health needs of this heterogeneous population. In addition, the specific health needs of Asian Americans often have been overlooked due to commonly held stereotypes about this group. Among the most misleading of these stereotypes is the “model minority” myth, which perpetuates the impression that all Asian Americans are prosperous, educated, and healthy (National Asian Women's Health Organization, 1997).

The fact that Asian-American women are at a higher risk of dying from breast cancer than other women because of late diagnosis and treatment argues for more effective interventions to promote breast cancer screening in this population. Study findings from Sadler et al. (2003) suggest that efforts to promote screening for breast cancer could have greater efficacy in reducing health disparities if they were focused on the specific attributes of the Asian subgroup to be served (Sadler et al., 2003).

Studies that identified factors influencing women's mammography decisions found factors that were mostly classified as variables from the Health Belief Model (HBM) (Becker, 1974; Skinner, Strecher, & Hospers, 1994). Based on the HBM, an individual's perceptions about benefits and barriers associated with specific health behavior, perceived personal risk and severity of the relevant health threat, and various cues to action influence an individual's likelihood of taking a preventive health action. Research findings from an integrative review examining barriers to breast cancer screening have demonstrated that personal barriers related to knowledge and attitudes are significant predictors of mammography screening among white, African-American, and Tamil women from Sri Lanka (George, 2000; Meana, Bunston, George, Wells, & Rosser, 2001). Although it is evident that HBM variables are associated with breast cancer screening (Yarbrough & Braden, 2001), few studies have examined the barriers to screening among Asian-American women (Wu, Guthrie, & Bancroft, 2005; Wu & Bancroft, 2006).

The purpose of this study was to examine whether socio-demographic factors, such as ethnicity, age, length of residency in the United States, and education level contributed to differences in Asian women's perceived barriers to breast cancer screening. The information obtained from this study could provide health professionals with recommendations for addressing ethnic-specific concerns when designing interventions to improve breast cancer screening among Asian-American women.

METHODS

Study Design and Participants

This descriptive study used a cross-sectional design with self-administered questionnaires completed by 315 Asian-American women from four groups (i.e., 119 Filipinos, 109 Asian Indians, 51 Chinese/Taiwanese, and 36 Koreans) who resided in southeastern Michigan. The selection of these groups to be studied was based on their being the most populous Asian ethnic groups in this community. Approval for the use of human participants was obtained from the university Institutional Review Board prior to conducting the survey.

Four community coordinators (CC) were hired to facilitate participant recruitment that occurred at community-based organizations and events, including community cultural centers, faith-based organizations, senior housing/apartments, and three annual Asian health fairs and expositions. Consecutive sampling was used in this study. This sampling strategy involved recruiting every participant who met the selection criteria over a six-month recruitment period. This sampling technique is considered practical and also the best of the non-probability sampling techniques (Hulley & Cummings, 1988). During these recruitment events, the CCs set up a table and distributed study flyers; they also screened interested women for eligibility on site. Eligibility criteria included women: (1) age 40 years and older; (2) self-identified as being from China/Taiwan, India, Korea, or the Philippines; and (3) able to speak and read English or their native language (Mandarin for Chinese or Taiwanese women, Hindi for Indians, Korean for Koreans, Tagalog for Filipinos). During these recruiting events, because the eligibility was clearly specified and posted for potential participants, for every 10 women who were interested in participation, approximately 1–2 were not eligible (yielding the eligibility rate of 80–90%). During the 6-month recruiting period, 346 eligible women were approached, and 315 completed the consent form and self-administered study questionnaire (a participation rate of 91%).

Procedure

After the eligible women were enrolled, each of the participants received a package that included a cover letter explaining the study, a consent form, and the study survey, and were asked to return the signed consent form with the study survey. The consent form was reviewed and signed by each participant. The self-administered survey took 20–30 minutes to complete; participants were able to elect to take the questionnaire either in English or their native language. Interestingly, more than 85% of the participants in each ethnic group selected the English language version. Participants received a $10 gift certificate for their time and efforts.

Instrument

The survey instrument was first developed in English by the principal investigator and bilingual research associates, and then was translated into Mandarin, Hindi, Korean, and Tagalog, and then back-translated to check each translation. The process of instrument development and validation has been reported elsewhere (Wu & West, 2007).

The survey instrument consisted of both closed and open-ended questions, including demographic information (e.g., age, marital status, education, and income level; length of time of residence in the U.S.; ability to read, speak, and listen in English; health insurance coverage); measures of decisional balance that assessed the perceptions of women about the positive (pros) and negative (cons) aspects of breast cancer screening; measures of mammography stage of adoption (i.e., pre-contemplation, contemplation, action, maintenance, and relapse); and self-report of the frequency of breast self-examination, clinical breast examination, and mammography. The findings of pros and cons subscales among Asian women demonstrated excellent supportive psychometric properties with promising Cronbach's alphas above 0.70, and the results from the confirmatory factor analysis supported construct validity with good model fit indices (Wu & Yu, 2003; Wu & West, 2007).

The open-ended questions asked participants to identify the three most frequent barriers that they encountered that prevented them from having a regular mammogram and the factors that would make them more likely to get a regular mammogram. The participants were instructed to leave the space blank if they did not believe they had significant barrier to using mammography. Qualitative analytic techniques were employed for the analysis of the open-ended questions. Two researchers (TW & HH) reviewed and coded participants' responses separately to uncover emerging themes and categories of the responses. To develop the set of codes and procedures for coding, the first 50 responses were initially coded, and the coding of these responses was then discussed and verified by two senior investigators who had expertise in cancer control and qualitative research. Once consensus on coding was reached by the two researchers and senior investigators, the set of themes/categories were finalized. The two researchers then coded the rest of the responses. To complete each content area, quotes from the responses have been highlighted to illustrate the category. Discrepancies between the two researchers were resolved through iterative discussions until consensus was reached. Both the content and frequency of the responses then were analyzed. This study focused on the analyses of the identified barriers.

Data Analysis

Data analysis was performed using the Statistical Package for the Social Sciences (SPSS, version 16 for Windows; Chicago, IL). Frequency distributions were calculated within each ethnic group for participants' responses on barriers preventing them from obtaining a regular mammogram.

Statistical differences in the indication of barriers among the four ethnic groups and by other socio-demographic variables (i.e., age, length of U.S. residency, and education level) were assessed using chi-square tests, and the Bonferroni correction was applied when testing the significance of the chi-square statistics for each individual chi-square analysis.

Based on chi-square test results, the specific barriers for which the prevalence differed significantly (p < .05) based on ethnicity, age, length of U.S. residency, and education were selected as dependent variables for the subsequent multivariate analysis. Due to the highly conservative effects of the Bonferroni correction, selecting barriers with significant associations with the four socio-demographic variables for subsequent multivariate analyses was based on p values less than 0.05.

Multiple logistic regression models were fitted to estimate the relationships of the following demographic variables with a particular barrier: (1) ethnicity, (2) age, (3) length of U.S. residency, and (4) education level. The models also identified the following variables to be associated with women's perception of barriers toward mammography (which were included in the multivariate analyses): insurance coverage for mammogram, mammograms were up-to-date (within one year vs. longer than one year), and knowledge about breast cancer screening (specific intervals recommended by the American Cancer Society for breast self-exam, clinical breast exam, and mammography).

Next, the two-way interactions between ethnicity and each of three demographic variables were tested in multiple logistic regression analyses to determine whether the effects of ethnicity on the barrier outcomes were consistent for each demographic group. However, interaction effects could not be estimated reliably due to the lack of sufficient variability in the dependent variables by the selected independent interaction variables (e.g., ethnicity × age, ethnicity × education, etc.).

RESULTS

Sample Characteristics

In total, 315 women completed the survey (Table 1). Ninety-seven percent of the Asian Indian participants were born in India. For the Chinese/Taiwanese group, 75% were born in Taiwan, 16% were born in China, 4% were born in Hong Kong, and the remaining 4% were born in the United States. All the Korean women were born in Korea. In the Filipino group, all but one of the women was born in the Philippines. The ages of the participants ranged from 40–84 years, with a mean of 55.7 years (SD = 9.8). The majority of the participants (78%; n = 244) were married. Eighteen percent of the participants (n = 55) reported a family history of breast cancer. Self-reported educational levels ranged from 4–27 years, with an average of 15 years (SD = 2.9). The women in this sample were highly educated: 84% had a college/university education or higher. The length of time that the women had resided in the United States ranged from 1–55 years (M = 21.4, SD = 11.7). Ninety-two percent of the women had health insurance, and 83% reported that their insurance covered mammograms. In terms of mammography screening, about 84% of participants reported ever having had a mammogram, and 59% reported that their mammogram was up-to-date.

TABLE 1. Demographic Characteristics of Asian-American Participants Overall and by Ethnicity (n = 315)

Characteristics	Total n	Sample %	Chinese %	Filipino %	Indian %	Korean %
Ethnicity
Filipino	119	37.8
Asian Indian	109	34.6
Chinese	51	16.2
Korean	36	11.4
Age (years)
Education
High school or lower	49	15.9	9.8	16.1	17.6	18.8
College/university or higher	260	84.1	90.2	83.9	82.4	81.2
Employment status
Full-time	198	62.9	78.2	50.5	33.3	62.9
Part-time or none	117	37.1	21.8	49.5	66.7	37.1
Marital status
Married	244	77.7	82.4	67.8	84.4	83.3
Length of residence in the U.S.
≤10 years	62	21.1	16.3	14.0	6.5	45.7
>10 years	232	78.9	83.7	86.0	93.5	54.3
English speaking ability
Yes	259	82.7	76.5	97.5	86.2	71.4
English reading ability
Yes	271	87.1	76.5	99.2	90.7	48.5
English writing ability
Yes	262	84.2	72.5	97.5	90.7	33.3
Health insurance
Insured	288	91.7	98.0	91.6	93.5	77.8

Analyses between respondents and non-respondents showed that about 26% of the sample respondents did not provide a codable response or left the item blank when they were asked about their perceptions of barriers toward mammograms. Of the 93 responses that were not codable, 78.5% left the question blank, while the rest of the responses were unrelated to the question (for example, self-exam, doctor check-up, exercise, etc.). As a result, additional analyses were performed to compare respondents with non-respondents on demographic factors (i.e., age, ethnicity, marital and insurance status, education, income, and length of U.S. residency). These analyses revealed that, compared with non-respondents on average, respondents were younger (mean age for respondents: 54.9; mean age for non-respondents: 57.8; t(299) = 2.16, p < .05) and less likely to be Asian Indian (response rates were for Filipinas 85%, for Koreans 75%, for Chinese/Taiwanese 74%, for Asian Indian 59%; χ²(3) = 19.24, p < .001).

Overall Barriers to Screening

The survey asked the respondents to name up to three of the most important reasons that prevented women from having a yearly mammogram. In this sample, 222 women (70.6%) gave 579 responses related to their perceptions on barriers toward mammography screening. Eleven themes emerged from the participants' responses to the question of barriers to breast cancer screening (Table 2). If a woman had multiple responses that corresponded to the same theme, it was only counted once.

More than one-fifth (23.3%) of the valid responses indicated barriers that could be classified as “logistical.” These included “lack of time,” “scheduling,” “location,” “poor facility,” and related responses. The second most frequently named barrier was discomfort (19.0%), which included experiencing physical pain during the mammogram procedure, feeling embarrassed or ashamed, and feeling uncomfortable if the mammogram was conducted by a male or strange health professional. About 18% of responses were related to the cost of having a mammogram, or lack of insurance was a major concern. Interestingly, of these 315 respondents, 30 participants reported that their mammogram was not covered by their health insurance, and about two-thirds reported that such issues (financial or lack of insurance) were barriers for them.

TABLE 2. Percentages of Responses to the Open-Ended Question About Reasons Preventing Asian-American Women from Having a Regular Mammogram, with Sample Responses, by Theme

Theme and Sample Responses	Percentage (Valid Responses, n = 579)
Logistical barrier	23.3
“Time conflict”—Taiwanese, age 45
“No transportation”—Asian Indian, age 55
“Difficult to make an appointment”—Asian Indian, age 53
Discomfort	19.0
“Pain during the procedure”—Filipino, age 41
“Embarrassment of exposing the breast”—Korean, age 72
“Taking off clothes in front of male doctor”—Asian Indian, age 46
Cost	18.0
“No health insurance”—Filipino, age 59
“Financial problem”—Korean, age 46
“Insurance won't cover (mammogram)”—Filipino, age 57
Low priority	11.2
“Ignorance”—Filipino, age 63
“Health negligence”—Asian Indian, age 43
“Forgetting the appointment”—Korean, age 60
Fear of finding cancer	7.6
“Afraid of being diagnosed with cancer”—Korean, age 52
“Afraid a mammogram will find cancer”—Taiwanese, age 44
“Fear of the result”—Filipino, age 61
Lack of knowledge	6.9
“No knowledge about mammogram”—Filipino, age 62
“Lack of public awareness”—Asian Indian, age 49
“Unaware of importance of regular mammogram”—Hong Kong, age 59
Feeling OK	6.4
“No symptom”—Korean, age 40
“Not recommended by doctor”—Filipino, age 73
“Good health condition”—Taiwanese, age 50
Culture barrier	2.6
“Communication barrier”—Korean, age 57
“Uncomfortable speaking English to doctor”—Taiwanese, age 44
“Religious reasons”—Chinese, age 70
Radiation exposure	1.4
“Fear of unnecessary radiation exposure”—Asian Indian, age 49
“Radiation is bad for health”—Asian Indian, age 78
Health provider issue	1.0
“Poor management from MD”—Filipino, age 56
“Lack of a health provider”—Filipino, age 66
Other issue	2.6
“Lack of encouragement from family”—Asian Indian, age 50
“My husband passed away”—Taiwanese, age unknown
“Health condition”—Asian Indian, age 81
Total	100

Comparison by Ethnicity

Significant differences were observed among the four ethnic groups in cultural barriers (Table 3). Cultural barriers were related to communication, speaking English, and religious reasons. No significant differences were observed by ethnicity for the other barriers.

Comparison by Age

Based on Bonferroni corrected p values, no significant differences were observed by age in perceptions of barriers. Nevertheless, compared to women younger than 60 years (9.8%), more than twice as many women age 60 or older (20.0%) named feeling OK as a reason preventing them from having an annual mammogram (χ²(1) = 4.24, p = 0.04) (Table 4).

Comparison by Length of U.S. Residency

A significant difference (p = .001) by length of residence in the U.S. was observed in the probability of reporting the barrier, with fear of finding cancer with a 10-fold greater proportion in those who had been in the U.S. for 10 or more years compared to those who had resided in the U.S. for less than 10 years. Cultural barrier and feeling OK were reported significantly (p < .05) more (two- to three-fold) among women who had resided in the U.S. less than 10 years (Table 5).

TABLE 3. Women's Perceived Barriers to Mammogram by Asian Ethnicity (n = 315)

Barriers	Filipino	Asian Indian	Korean	Chinese/Taiwanese	χ^2 Value (p value)
Cost					4.92
Yes	49.0%	33.3%	48.1%	35.1%	(p = .18)
No	51.0%	66.7%	51.9%	64.9%
Fear of finding cancer					9.17
Yes	27.6%	8.3%	14.8%	21.6%	(p = .03)
No	72.4%	91.7%	85.2%	78.4%
Discomfort					11.83
Yes	46.9%	25.0%	25.9%	48.6%	(p = .01)
No	53.1%	75.0%	74.1%	54.1%
Low priority					4.14
Yes	28.6%	25.0%	33.3%	13.5%	(p = .25)
No	71.4%	75.0%	66.7%	86.5%
Culture barrier					18.13*
Yes	2.0%	6.7%	25.9%	8.1%	(p < .001)
No	98.0%	93.9%	74.1%	91.9%
Logistic barrier					0.03
Yes	49.0%	50.0%	48.1%	48.6%	(p = 1.00)
No	51.0%	50.0%	51.9%	51.4%
Feeling OK					3.05
Yes	9.2%	16.7%	14.8%	18.9%	(p = .39)
No	90.8%	83.3%	85.2%	81.1%

Comparison by Education Level

One significant difference was observed between women with at least a college degree and those women without a college degree, with 44% of the respondents with a college degree or higher naming discomfort as a barrier, while only 17% of those respondents without a degree did so (Table 6).

TABLE 4. Barriers to Having an Annual Mammogram by Age Group (n = 315)

Barriers	Younger than 60 years	60 years or older	χ^2 Value (p value)
Cost			0.59
Yes	40.5%	46.2%	(p = .44)
No	59.5%	53.8%
Fear of finding cancer			2.02
Yes	22.2%	13.8%	(p = .16)
No	77.8%	86.2%
Discomfort			2.23
Yes	43.1%	32.3%	(p = .14)
No	56.9%	67.7%
Low priority			0.33
Yes	26.8%	23.1%	(p = .57)
No	73.2%	76.9%
Culture barrier			0.25
Yes	7.2%	7.7%	(p = .62)
No	92.8%	92.3%
Logistic barrier			3.06
Yes	52.9%	7.7%	(p = .08)
No	47.1%	92.3%
Feeling OK			4.24
Yes	9.8%	20.0%	(p = .04)
No	90.2%	80.0%

Multiple Logistic Regression Results

Based on the statistically significant results (p < .05) from chi-square tests, four self-reported barriers—fear of finding cancer, feeling OK, discomfort, and cultural barriers—were included as dependent variables for multivariate analyses using multiple logistic regression. Results from fitting the logistic regression models showed that after taking into account the relationships of the four demographic variables (i.e., ethnicity, age, U.S. residency, and education), access to mammography, knowledge of breast cancer screening, and recent mammogram, the odds of reporting the barrier of “fear of finding cancer” were about 14 times higher for those women who had resided in the U.S. for 10 years or more (OR = 13.86, 95% CI 1.57–122.49), compared with those women who had been in the U.S. for a shorter period of time (Table 7). The odds of reporting feeling OK as a barrier were more than three times higher (OR = 3.40, 95% CI 1.40–8.28) for women who were age 60 years and older relative to the younger women, when controlling for the same covariates. Interestingly, for those women who had a college degree or higher, the odds of identifying discomfort as a barrier were nearly three times higher (OR = 2.92, 95% CI 1.07–7.93) compared to those women who did not have a college degree. In addition, relative to Chinese women, the odds of indicating discomfort as a barrier were about 63% lower for Asian Indian women (OR = 0.37, 95% CI 0.14–0.93). After adjusting for other socio-demographic variables, insurance coverage, and knowledge scores, the effect of ethnicity no longer was significantly associated with cultural barriers.

TABLE 5. Barriers to Having an Annual Mammogram by Length of Residence in the U.S. (n = 315)

Barriers	<10 years	≤10 years	χ^2 Value (p value)
Cost			1.55
Yes	32.6%	43.0%	(p = .21)
No	67.4%	57.0%
Fear of finding cancer			10.74*
Yes	2.3%	24.8%	(p < .001)
No	97.7%	75.2%
Discomfort			1.22
Yes	32.6%	41.8%	(p = .27)
No	67.4%	58.2%
Low priority			0.14
Yes	23.3%	26.1%	(p = .71)
No	76.7%	73.9%
Culture barrier			4.50
Yes	14.0%	4.8%	(p = .03)
No	86.0%	95.2%
Logistic barrier			1.41
Yes	39.5%	49.7%	(p = .24)
No	60.5%	50.3%
Feeling OK			5.47*
Yes	25.6%	11.5%	(p = .01)
No	74.4%	88.5%

TABLE 6. Barriers to Having an Annual Mammogram by Education Level (n = 315)

Barriers	Lower than College	College or Higher	χ^2 Value (p value)
Cost			0.00
Yes	42.9%	42.3%	(p = .95)
No	57.1%	57.7%
Fear of finding cancer			0.93
Yes	14.3%	21.4%	(p = .34)
No	85.7%	78.6%
Discomfort			9.15*
Yes	17.1%	44.5%	(p < .001)
No	82.9%	55.5%
Low priority			0.17
Yes	28.6%	25.3%	(p = .68)
No	71.4%	74.7%
Culture barrier			2.78
Yes	11.4%	4.4%	(p = .10)
No	88.6%	95.6%
Logistic barrier			2.44
Yes	60.0%	44.6%	(p = .12)
No	40.0%	54.4%
Feeling OK			0.01
Yes	14.3%	13.7%	(p = .93)
No	85.7%	86.3%

The results from the multiple logistic regression analyses showed excellent agreement between the observed and predicted outcomes, as evidenced by the goodness of fit of Hosmer-Lemeshow tests (p values ranged from .07 to .85; an insignificant p value suggests reasonable fit of the predicted probabilities to the observed outcomes.

TABLE 7. Multivariate Analysis for Demographic Variables and Dependent Variables: Barriers to Annual Mammogram

n = 315	Fear of Finding Cancer		Feeling OK		Discomfort		Culture Barrier
	Odds Ratio	95% CI	Odds Ratio	95% CI	Odds Ratio	95% CI	Odds Ratio	95% CI
Demographic variables
Ethnicity
Filipino vs. Chinese	1.35	(0.59, 3.63)	0.46	(0.14, 1.52)	1.34	(0.57, 3.11)	0.11	(0.01, 1.18)
Asian Indian vs. Chinese	0.38	(0.11, 1.35)	0.73	(0.22, 2.46)	0.37*	(0.14, 0.93)	0.55	(0.10, 3.02)
Korean vs. Chinese	2.00	(0.41, 9.69)	0.64	(0.14, 2.95)	0.52	(0.16, 1.74)	1.37	(0.25, 7.41)
Age	0.50	(1.40, 8.28)	3.40**	(1.40, 8.28)	0.61	(0.30, 1.24)	1.49	(0.37, 6.08)
U.S. residency	13.86*	(1.57, 122.49)	0.44	(0.13, 1.49)	0.86	(0.34, 2.22)	0.42	(0.09, 2.01)
Education	1.68	(0.49, 5.76)	2.57	(0.64, 10.36)	2.92*	(1.07, 7.93)	0.29	(0.07, 1.27)
Access to mammography	1.29	(0.29, 5.68)	0.89	(0.25, 3.22)	1.33	(0.49, 3.65)	1.00	(0.18, 5.46)
Knowledge of breast cancer screening	1.19	(0.72, 1.97)	0.91	(0.56, 1.48)	1.24	(0.84, 1.82)	1.24	(0.59, 2.61)
Recent mammogram practice	1.27	(0.58, 2.80)	0.82	(0.32, 2.11)	0.95	(0.49, 1.83)	0.53	(0.13, 2.21)
Hosmer* Lemeshow test result	χ^2(8) = 4.04	p = .85	χ^2(8) = 9.34	p = .31	χ^2(8) = 7.51	p = .48	χ^2(8) = 14.68	p = .07
Note: Age and U.S. residency were entered in the regression analyses as numerical variables (instead of categorical).

DISCUSSION

The goal of this study was to understand the nature of barriers to breast cancer screening among Asian-American women. While the majority of earlier studies have focused on mammogram utilization rates and determinants of mammography adherence using a pre-developed instrument, the current study used open-ended questions to explore the views of Asian women concerning barriers to obtaining regular mammography.

The study participants reported in their own words the following top three barriers to receipt of breast cancer screening: logistical barriers, discomfort, and costs. Previous studies of women of other ethnicities have indicated that logistical barriers were the largest barrier for low-income, African-American women (Rawl, Champion, Menon, & Foster, 2000). The findings of the current study suggest that Asian-American participants also reported logistical barriers as the most frequent barrier. Similar to the older African-American women in Rawl's study, these Asian study participants also reported factors such as scheduling conflicts, lack of time, and transportation as contributing to the logistical barriers. The present study's findings support the continuous efforts to assist women to overcome logistical barriers by making changes in healthcare systems, such as those designed to increase convenience: decreased waiting times and improved scheduling.

Based on the multivariate analyses, the study results verified that after taking into account socio-demographics, insurance coverage, and other behavioral and knowledge covariates, specific barriers still prevented Asian women participants from seeking mammography. For example, women older than 60 years were more likely not to be screened for mammography because they “feel OK.” In addition, Asian women in this study who had lived in the United States for 10 years or longer tended to identify the barrier of fear of finding cancer as a factor that prevented them from having a mammogram. A possible explanation of this finding could be that Asian women who have acculturated to American lifestyles in the U.S. may increase their susceptibility of developing breast cancer by changes in dietary patterns, alcohol intake, sedentary lifestyles, and greater exposure to breast cancer awareness campaigns. Interestingly, Asian women participants who had higher education levels tended to identify discomfort as a barrier to seeking mammography, which may be due to the fact that these highly educated women are being more assertive. These findings may guide clinicians and policy makers in implementing interventions to promote breast cancer screening for women with different demographic characteristics. Tailored messages can be offered by healthcare providers that will directly address the specific barrier encountered by the women.

In addition, barriers specific to women's attitudes and beliefs identified by the study participants included low priority, belief that mammography is not needed if they feel OK, and fear of being diagnosed with cancer. The participants also reported lack of specific knowledge about screening as another barrier to receiving mammography, which probably was due to lack of awareness of the importance of early detection of cancer in their native countries. The study findings clearly demonstrate a great need to provide these women with comprehensive cancer screening education that includes the benefits of early breast cancer detection and information about recommended screening guidelines in the U.S., including those for mammography. Given the high esteem that Asian women have for physicians, and that attitudinal barriers prevail in the study population, it is essential for primary care health professionals to be vigilant about encouraging women to take advantage of regular breast cancer screening and to dispel the misconceptions shared by women of these ethnicities.

Several limitations of this study should be noted. Due to the non-representative nature of the convenience sample used, the study findings should not be generalized beyond the sample to a population with similar demographic characteristics. The study participants were recruited from Asian organizations in southeastern Michigan and had higher than average educational background with greater overall socioeconomic capacity. The study sample also represented a more acculturated group, with more than 80% of them having the ability of speaking, reading, and writing in English. The study should be replicated to examine barriers to breast cancer screening in a larger and more representative sample before the study findings can be generalized. The cross-sectional design also made impossible assessment of the temporal sequence of independent and dependent variables. In addition, the use of indirect phrasing of questions in the current study was intended to avoid conveying criticism or social desirability as reported in previous studies (Webber, Schoenbaum, & Bonuck, 1997; Garbers, Jessop, Foti, Uribelarrea, & Chiasson, 2003). The relatively small sample size limited the ability to examine the barriers in each ethnic group or other strata and conduct interaction analyses of sufficient statistical power in multivariate analyses. Further, the potential for social acceptability bias in self-reported information (e.g., mammography practices) might have resulted in erroneous findings. The statistical tests conducted may have resulted in finding significant results due to chance.

Despite the study's limitations, the findings offer important practice implications for comprehensive health education programs and individual patient counseling targeted to promoting breast cancer screening for Asian-American women. Identifying that the perceptions of barriers vary among different Asian-American groups is the first step for developing mammography promotion programs that will effectively improve breast cancer screening rates in specific populations of Asian Americans. The study findings support the need for overcoming the identified barriers specific to the targeted Asian groups through tailored health communication messages that consider socio-demographic factors (e.g., age, education level, and length of U.S. residency). Simple translation of the existing “one-size-fits-all” educational materials into Chinese/Taiwanese, Hindi, Korean, or Tagalog, without incorporating culturally appropriate content tailored for these women, may not be successful for transferring the additional health knowledge into health behaviors. This study points out the need for further research efforts focused on identifying specific factors that prevent Asian-American women of different ethnicities, length of U.S. residency, and educational levels from obtaining breast cancer screening. In particular, acculturation may be a significant factor that deserves further assessment in studies of breast cancer screening behaviors. Furthermore, the study findings about logistical barriers also illustrate the importance of enhancing health systems to be more accessible and accommodating to women from other cultures.

Finally, this study is the first to use open-ended questions to help understand and compare the differences in perceived barriers to breast cancer screening among four ethnic groups of Asian Americans. This study provides in-depth perspectives from the participants' own words on barriers to screening, which can inform future interventions that address individual, cultural, and environmental barriers to receipt of breast cancer screening. The information obtained through this study thus can serve to guide development of culturally sensitive and competent counseling interventions designed to improve screening among this underserved minority population.

This study examined the attitudes and beliefs of women regarding barriers toward breast cancer screening among diverse Asian populations in a Midwestern community in the United States. The findings emphasized the importance of identifying similarities and differences in the perceived barriers as assessed among socio-demographic variables such as ethnicity, age, and length of U.S. residency. Consequently, healthcare professionals should be more sensitive to logistical, attitudinal, and cultural barriers faced by Asian women. Accommodating their specific needs can be the key toward reducing health disparities in breast cancer screening in this population group.

Notes

*With Bonferroni correction for seven bivariate analyses, the corrected significant p value is set at 0.05/7 = 0.007.

*With Bonferroni correction for seven bivariate analyses, the corrected significant p value is set at 0.05/7 = 0.007.

*With Bonferroni correction for seven bivariate analyses, the corrected significant p value is set at 0.05/7 = 0.007.

*p < .05

**p < .01.