Cervical cancer can be detected early by effective screening, such as Papanicolaou testing, which can detect any precancerous lesions, increasing the opportunity for early treatment and preventing the cancer from progressing to a more advanced stage.1 Papanicolaou tests can help to reduce the lifetime risk of cervical cancer by nearly 40%.2 Public health policy on screening guidelines for cervical cancer in different countries is based on the best evidence available, to determine who should have screening and when it should start. In the United States, the Department of Health and Human Services suggests that it should start from the age of 21 years and be repeated every 3 years until the age of 65 years.3 In Canada, the guidelines are a little different from those in the United States. The Canadian Task Force on Preventive Health Care recommends that women aged 25 to 69 years should be tested every 3 years.4 In Britain, the National Health Service program recommends that women aged 25 to 49 years should be screened every 3 years and those aged 50 to 64 years should be screened every 5 years.5 In Australia, the starting age for screening is lower than those in other countries. The Department of Health suggests that women aged 18 to 70 years should have a Papanicolaou test every 2 years.6 With screening guidelines issued by well-known organizations and government and the mounting popularity of Papanicolaou tests in the United States and their increased uptake rate among the general population (now 80%), the incidence and mortality rate of cervical cancer have both been significantly reduced.7,8
This increased uptake rate among the general population does not, however, apply to ethnic minorities (EMs). In the United States, the uptake rate of Papanicolaou tests among Asian women (69.7%) is lower than that of their white counterparts (77.4%).8 A similar phenomenon is observed in Canada and Britain. McDonald and Kennedy9 found that the uptake rate ranged from 39% to 64% among foreign-born Asian women in Canada (South Asian, Southeast Asian, Filipino, Chinese, Korean, and Japanese) and 40% to 65% among native-born Asian women (excluding Japanese this time), with rates significantly lower than those of white women—77.3% foreign born and 80.7% native born. In Britain, only 11.8% of white women do not take Papanicolaou tests, whereas 44% to 71% of the EM population (Caribbean, African, South Asian) do not.10 These results make it clear that ethnic disparities still exist despite the increased popularity of Papanicolaou testing.
These disparities could be the result of various barriers minorities encounter in the country where they are living, preventing them from using the available screening services. “Ethnic minorities” refer to those people who are ethnically different from the majority population and form a numerically inferior group. They have a culture, language, and religion distinct from those of the majority. Ethnic minorities are not simply minorities in a statistical sense—relatively small in number—but also in the way that they are discriminated against because of their ethnic identity.11 Although they are of diverse ethnicities, they often encounter similar barriers to screening, such as language difficulties or lack of a physician’s recommendations.12 Previous reviews in the United States have identified a number of factors that affect screening uptake, such as socioeconomic disparities, low levels of acculturation, lack of knowledge, no sources of standard care, and a lack of insurance. Belief in personal invulnerability to cervical cancer, neglecting the perceived benefits of Papanicolaou tests, fatalistic attitudes, and fear of testing are further reasons for not undergoing screening. In addition, unique beliefs exist in specific ethnic groups, often concerned with misconceptions about personal susceptibility to cancer.13,14 However, both these reviews include studies focusing on EM groups (eg, African Americans, Hispanics, and Asian Americans), which were conducted solely in the United States. Their results may not, therefore, adequately reflect the actual barriers to taking up screening services that minorities may encounter in other countries. Ethnic minority women in different parts of the world may experience different barriers because of variations in healthcare systems and living environments across countries. Furthermore, one of the reviews14 examines the sociocultural factors influencing Papanicolaou test uptake according to the Health Belief Model (HBM), which is limited in only representing the individual level. A person’s behavior is influenced not only by individual characteristics but also by factors from multiple dimensions, such as health insurance, language, and acculturation, which are not explicitly illustrated in the HBM. The inadequacy of the model to deal with factors other than those at the individual level means it cannot provide a comprehensive view of how cervical cancer screening behavior is affected.
Such behavior can indeed be affected by multiple factors, as viewed from different perspectives. McLeroy et al15 identified 5 levels of factors that affect health behavior in an “ecological model”: intrapersonal, interpersonal, organizational, community based, and policy related. The intrapersonal level describes an individual’s characteristics, such as knowledge and attitude, which influence behavior. The interpersonal level describes interaction with the primary group, family, and friends, which provides social support to health behavior. The organizational level includes rules or regulations (and physician’s recommendations) that may constrain or modify health behavior. The community level concerns the influence of social networks or standards that exist either formally or informally among groups or organizations (eg, acculturation). The policy-related level describes the existence of policies that support disease prevention and detection, control, and management (eg, health insurance).15,16 The ecological model, extending beyond the individual level, provides more information on how various factors influence a person’s behavior. In addition, the model emphasizes that there will be some interaction between factors within and across all levels of health behavior.15 Understanding the relationships between these different factors could aid the involvement of appropriate stakeholders at all levels in decision making and resource allocation. Such information also leads to the design of multilevel interventions that combine behavioral and social components to promote screening uptake.
The aims of the present systematic review are therefore to examine the factors that influence cervical cancer screening uptake worldwide among EM women without a history of cancer, to examine similarities and differences in associated factors across different EM groups, and to explore the interrelationships of multilevel factors in influencing screening uptake. The 5 levels in the ecological model discussed previously will be used in the present review to classify the factors identified and offer an explanation of the screening behavior of EM women in general.
The Preferred Reporting Items for Systematic Reviews and Meta-Analysis was used to guide the review.17 Its relevant parts are reported.
The inclusion criteria are quantitative or qualitative studies exploring the factors influencing women’s cervical cancer screening behavior (Papanicolaou test uptake—ever had, had in the past year, or had in the past 3 years; adherence to screening) and studies covering EM women in the country where they were conducted. To ensure that the target participants were really from a genuine EM, the population census of the country in question was consulted, and the relative size of the EM group was checked. To be included in the present review, the EM population should be numerically smaller than the majority population and less than 20% of the total population in the country. Criteria for eligible subjects should also include those without a history of cervical cancer or hysterectomy. (Alternatively, these restrictions are clearly stated in the results sections.) Articles are included that cover both EM and general populations, with a subgroup analysis of the former, as well as studies with outcomes where there are associated factors, barriers to, or facilitators of screening uptake among EM women. The exclusion criteria cover publications not peer reviewed, unpublished studies, or those inaccessible online and articles other than those in English, with resources to handle other languages not being available.
A comprehensive literature search was conducted in October 2015 using 5 electronic databases: Ovid MEDLINE (since 1946), Cumulative Index to Nursing and Allied Health Literature Plus (since 1937), Scopus (since 1823), PsycINFO (since 1806), and PubMed. To locate relevant studies, combinations of keywords or phrases were used: “factor or predictor or barrier or facilitator or determinant or motivation,” “ethnic minority or minority group or ethnic group or vulnerable or underserved,” “cervical cancer or cervical neoplasm or cervical carcinoma or cervix cancer or gynaecological cancer or gynaecological neoplasm or gynaecological carcinoma,” and “prevention or screening or cervical smear or Pap test or Pap smear or Papanicolaou test.” A secondary search was conducted of the studies’ reference lists to identify any additional work. The studies finally identified were then evaluated against the inclusion and exclusion criteria outlined previously.
To select relevant studies, the titles and abstracts were first screened for eligibility and relevance. If they met the conditions, the full text was retrieved. If the title and abstract did not provide adequate information, the full text would be retrieved.
Data Collection Process
Two independent reviewers extracted the data, which included citation details, study design, sample size, theoretical/conceptual model, study population, outcome measures, and results. Any disagreements were resolved through discussion and consensus.
Risk of Bias
The methodological quality of studies was assessed by 2 reviewers using a 16-item assessment tool consisting of 2 items specifically intended for quantitative or qualitative work. Each study would score 1 point for each item that was properly fulfilled, with higher scores indicating higher quality.18,19 Any disagreement was again resolved through discussion and consensus.
Synthesis of Results
The aim of the review was to present integrated findings from both quantitative and qualitative studies. Both types were therefore analyzed and incorporated within themes at the respective levels.
A total of 1390 citations were identified through the 5 electronic databases and a secondary search. Four hundred eighty-nine duplicates were removed, leaving 901 citations for abstracts and titles to be reviewed. Another 776 articles were then removed, leaving 125 full-text articles to be screened for relevance. After a detailed examination of the articles against the inclusion criteria, a further 101 were excluded, leaving a total of 24 articles10,20–42 (23 studies) in the final review (Figure 1).
All 23 studies were published between 1996 and 2015 and conducted in Western countries—20 in the United States and one each in Britain, Canada, and Australia. Among these, 21 adopted quantitative and 2 adopted qualitative approaches. The target participants were EM women in the countries where the studies took place, and they were grouped into 4 major categories: Hispanic or Latino (n = 7), black (African, Caribbean, or Haitian) (n = 5), Asian (South Asian, Vietnamese, Filipino, Korean, Japanese, or Chinese) (n = 15), and Inuit (n = 1). The sample size of the studies ranged from 18 to 3787, and participants’ ages ranged from 18 to 83 years. Ten studies described the use of a model as the basis for examining the factors influencing behavior.23–25,30–32,34–36,40 The models varied from individual to policy levels and included the HBM (n = 4), Social Cognitive Theory (n = 2), Andersen Behavioral Model (n = 2), Health Behaviour Framework (n = 1), Behavioural Model for Vulnerable Populations (n = 1), Ecological Model (n = 1), and Ecological Model for Health Behavior (n = 1). Two interventional studies used a combination of the HBM and Social Cognitive Theory models24,32 (Table 1).
Risk of Bias
Quality assessment was undertaken, with results ranging from 7 to 13 and a mean of 11.29 (Table 2). Most studies described the theoretical framework/literature review and the sample in detail. The items that were poorly discussed were the psychometric properties of the measurement instrument and the validity and reliability of results. With most studies adopting a self-report survey specifically developed for their own purposes, the use of tools not previously validated might limit the validity and reliability of results.
Analysis of the studies identified multilevel factors among EM women associated with ever having undergone cervical cancer screening. These included intrapersonal factors (knowledge, attitude and perceptions, sociodemographics), interpersonal factors (recommendation from friends), organizational factors (quality of care and services, usual source of care, physician’s recommendation, religion), community factors (language, acculturation, culture related), and policy factors (health insurance) (Table 1). Similarities and differences in the factors across EM groups are shown in Table 3.
It was noticed that some factors were common across ethnic groups in different countries (Table 3). They included knowledge, sociodemographics (age, marital status, childbirth, place of birth, employment, educational level), perceived susceptibility to disease, and perceived barriers to screening. Although there are similarities, some factors were only found in certain EM groups. Sociodemographics such as immigration status, a family history of cancer, and adherence to breast cancer screening were only observed in the black groups. Household income, immigrating at a younger age, health literacy, perceived severity of the disease, perceived benefits from and self-efficacy in obtaining screening, and fatalism were only observed when related to Asian groups’ screening behavior (Table 3).
Although it was reported in 5 studies that knowledge of cancer and cervical screening was a common influencing factor associated with Papanicolaou test uptake among black and Asian women,20,29,30,32,40 it was not found to be an associated factor among 3 Asian groups (Vietnamese, Korean, and Filipino) in the United States.23,36,42
The belief among black women in the United States that screening could help to maintain personal health was a major facilitator to screening.20 Among the Asian groups, Chinese women in the United States who possessed more knowledge on cervical cancer were positively associated with Papanicolaou test uptake in the past year or in the past 3 years.30 Chinese women in Australia believed that knowing about the importance of early detection facilitated Papanicolaou test uptake, whereas lack of knowledge and considering Papanicolaou test as a diagnostic test for signs and symptoms were barriers to screening.29 One study reported that Vietnamese Americans who did not know that cancer could occur without symptoms or that a Papanicolaou test could detect cancer early were less likely to have ever undergone screening.32 However, Do,23 Nguyen-Truong et al,36 and Yoo et al42 found that Vietnamese women in the United States who knew that Papanicolaou tests were necessary for women who were asymptomatic, sexually inactive, and postmenopausal and those who possessed correct views on the risk factors of cervical cancer and human papillomavirus knowledge were not associated with Papanicolaou test use. Such inconsistent results among Vietnamese women reflected that understanding the risk factors of cervical cancer was not in itself sufficient to encourage women to undergo screening. Furthermore, the simple knowledge of who should use the Papanicolaou test but ignoring the benefits of such testing formed another reason for the absence of association between knowledge and Papanicolaou test uptake.23,32,36,42
Age, marital status, and level of education were factors commonly explored by the studies and found to have some influence on the screening behavior of certain EM groups.
Age was a common influencing factor identified by the studies. Both young Asian and black women in Britain (aged 30–40 years) and Hispanic women (younger than 40 years) in the United States were more likely to have ever had a Papanicolaou test or adhere to annual Papanicolaou testing, respectively.10,28 However, older Asians in the United States (older than 35 years) had taken more Papanicolaou tests than their younger counterparts in the past year and in the past 3 years.23,31 Despite these, both Jennings-Dozier and Lawrence28 and Taylor et al40 found that age was not associated with any Papanicolaou test uptake in Hispanic, black, and Asian women in the United States.
Marital status was another factor that exerted a certain influence on women’s attitudes toward cervical screening. Unmarried Asian women in the United States were significantly less likely to have taken a Papanicolaou test in the past 3 years,31,40 but married Asian and black women in the United States seemed more likely to have done so in the past year or past 3 years.23,37,41 Such findings could be related to the mistaken belief that only married women should take a Papanicolaou test.23 However, Jennings and Lawrence28 found that it was not a significant associated factor in test uptake among Hispanic and black women in the United States, a result that could be related to the marital status of the participants: 44% single, 27.5% married, and 28.5% divorced/widowed. Divorced/widowed women might or might not consider using the service in the absence of sexual activity.
Educational levels formed another important factor that contributed to Papanicolaou test uptake.28,31,32,41 Lin et al31 found that Asian women in the United States without a bachelor’s degree were less likely to have adhered to Papanicolaou test guidelines in the past 3 years. Black women in the United States whose education went beyond the secondary level exhibited a higher rate of test uptake in the past 1 and 3 years.28,41 However, Marlow et al,10 Dettenborn et al,22 Taylor et al,40 and Jennings and Lawrence28 did not find educational level to be a significant associated factor in black, Asian, and Hispanic groups in the United States or in Britain. Such results might be related to different criteria being used to assess educational levels in different studies.
Household income and a family history of cancer were factors only found in specific ethnic groups. Household income was only associated with test uptake among Asians in the United States in the past years31,35 and was not observed in other ethnic groups.22,28,31,37 One possible reason was that nearly 60% of Asian women’s income was less than twice the Federal Poverty Level, a percentage significantly higher than those in other EM groups.31 A family history of cancer was reported to be a facilitator to screening among black women in the United States,20,41 and it was noticed that women with a family history of cancer were more likely to have had a Papanicolaou test in the past 3 years.41
ATTITUDES AND PERCEPTIONS
Individual perceptions of screening were investigated and found to have a significant influence on women’s screening behavior. It was noticed across EM groups that perceived barriers constituted a factor significantly associated with decreased odds of having had a test.10,20,24,29,32,33,36,39 Black and Asian women in the United States and in Britain felt that fear of the unknown and being scared to find out about cancer, busy work schedules, and difficulties encountered in matching the appointment schedule with their daily commitments at home, resulting in their unavailability for screening, were major perceived barriers to screening.10,20,32 Moreover, Asian women in the United States revealed that difficulty in remembering to schedule a Papanicolaou test appointment was one of their perceived barriers to screening.36
Fifteen percent of Hispanic women in the United States claimed that a feeling of embarrassment during the test itself was their main reason for not undergoing cervical screening.39 Asian women in the United States also revealed that embarrassment and feeling uncomfortable with a stranger performing the test would be negatively associated with screening.24,32 Consistently, Do23 and Ma et al32 noticed that Asian women in the United States reported perceived prohibitive test costs and time constraints as reasons for not ever having the test or for not having a test in the past year. It seemed that perceived psychological and organizational barriers, in addition to women’s roles at home, were common reasons for not being screened. However, Taylor et al40 did not find that “fear of cancer being discovered” was associated with having a Papanicolaou test in the past 3 years among Asians in the United States. Similarly, Marlow et al10 revealed that fear of cancer and concerns about the test (embarrassment and pain) were not associated with ever having the test among Asian and black women in Britain.
Perceptions of the risks of the disease also affected screening uptake and were common across EM groups. Skaer et al39 and Kwok et al29 noticed that Hispanic and Asian women in the United States and in Australia who thought it unnecessary to take a Papanicolaou test in the absence of problems or pain and with a low perceived risk of cancer had the main barriers to screening. Similarly, Marlow et al10 identified 2 perceived variables that were significantly associated with both Asian and black women in Britain being overdue for screening: “I’m not sexually active so I don’t need to go for a smear test” and “I do not need a smear test if I do not have any symptoms.” Asian women in the United States who perceived a need to have a Papanicolaou test because of poor/only fair health and those who perceived themselves at risk of developing cancer and who thought they were more susceptible to the disease than average women were also more likely to have ever had screening or to have screening in the past 3 years.31,33 However, Taylor et al40 found that US Asian women’s own perceived susceptibility to the disease in comparison with white women was not associated with Papanicolaou test adherence. Such results suggest that these women had misconceptions about cervical cancer with regard to its risks and to the benefits of testing for early detection of the disease.
Some attitude and perceptions were largely confined to Asian groups (Table 3). Two studies discussed perceived self-efficacy in obtaining Papanicolaou tests. Asian women in the United States with a low level of confidence in taking tests were significantly less likely to have ever used them,33 whereas after attending the intervention, Asian women with higher perceived self-efficacy were positively associated with Papanicolaou test uptake.24 This result underlines the fact that health-promoting interventions have a positive reinforcement effect on participants’ self-efficacy in obtaining the test. In addition to perceived self-efficacy, the perceived benefits of testing were explored. Asian women who thought that Papanicolaou tests could help to decrease their risk of cervical cancer and that cancer could be cured if detected early were associated with previous screening.32,40 Moreover, women who agreed that Papanicolaou testing was important to maintain good health were independently and significantly associated with ever having had a Papanicolaou test.33 The perception of cancer fatalism may affect screening behavior. A belief in cancer fatalism was certainly a barrier to screening among Asians.29 However, Dettenborn et al22 found that fatalism was not associated with Papanicolaou test adherence among Hispanic and black women in the United States, although there might well be other certain factors moderating the effect.
RECOMMENDATION FROM FRIENDS
Two studies revealed that both black and Asian women in the United States agreed that recommendations from friends facilitated screening and were positively associated with ever having had a test.20,36
Some factors were common across ethnic groups at the organizational level (Table 3). They included quality of care and services provided by healthcare organizations, the physician’s gender, physicians’ recommendations, and religion. A few factors were largely confined to some EM groups (Table 3). A positive relationship with doctors by maintaining trust and confidentiality was a factor only observed in black groups. Among the Asian groups, asking the doctor for a test, the physician’s ethnicity, usual sources of care, and practical factors (such as childcare and transport) were observed to affect screening behavior.
QUALITY OF CARE AND SERVICE
The quality of care and service provided by healthcare organizations affected women’s decisions on screening. Asian women in the United States thought that the pain and vulnerable feelings associated with vaginal insertion of a medical instrument affected their decisions on future screening uptake,29 and feeling comfortable with a previous test was a major reason for ever having had a test.42 Black women in the United States also mentioned that feelings of discomfort with the medical environment and instruments were a major barrier to screening.20 Furthermore, if there was flexibility in scheduling appointments, black women were encouraged to go for the test.20 Asian women in Australia appreciated receiving a reminder letter from a healthcare organization, and such letters encouraged them to go for screening.29 In addition, the provision of and access to low-cost or free screening would persuade black and Asian women in the United States and in Australia to go for testing.20,29 However, Nguyen-Truong et al36 found that the quality of care received from the healthcare system was not associated with Asian women’s test uptake. Concerns about the test due to a bad experience in the past or due to worries about seeing a male doctor were not associated with ever having had the test among both Asian and black women in Britain,10 whereas similar groups in the United States and in Australia would only consider screening with a female physician or technician in the healthcare organization to avoid embarrassment.20,29 Concerned about the service provided by the healthcare organization, black women in the United States revealed that having a positive relationship that maintained trust and confidentiality had encouraged them to attend screening in the past years.20
Only Asian groups also considered the ethnicity of healthcare providers. Lee-Lin et al30 found that having a non-Asian healthcare provider was positively associated with ever having had a Papanicolaou test among Chinese women in the United States. However, Chinese women in Australia preferred to have a female Chinese doctor as they thought that she would understand their needs and be more communicative in their own language.29 Apart from this, Chinese women in Australia also mentioned that practical matters, such as a lack of facilities for taking care of children or the absence of public transport, were other barriers to screening.29
PHYSICIAN’S RECOMMENDATION AND USUAL SOURCE OF CARE
Recommendation by a physician was the most common influencing factor reported among Hispanic, black, and Asian groups in the United States and in Australia.20,23,29,30,32,36,39,40 By following such a recommendation, women were 3 to 8 times more likely to have ever taken a test or had one in the past 3 years.30,32,36,40 However, Dettenborn et al22 did not find it to be a variable predicting cervical screening among black and Hispanic women in the United States.
It was found that only among certain Asian women (Chinese, Korean, and South Asian) in the United States was the lack of a regular source of care from a primary healthcare provider related to lower test uptake in the past 3 years.30,31 However, this was not observed among other Asians (Vietnamese, Japanese, and Filipino Americans) and Hispanic women25,31,36 perhaps because only 6% and 11% of the Japanese and Filipino populations, respectively, reported a lack of availability in the usual source of care.31
Religion was mentioned as an important factor influencing the screening behavior of women in 2 quantitative studies focusing on Asian women in the United States.32,35 Ma et al revealed that women who did not think that God or a higher power could control cancer were 2 times more likely to have ever had a Papanicolaou test. Nguyen and Clark35 used the validated Religious Orientation Scale to measure religiosity,43 a 20-item measure that assessed 3 factors affecting religiosity: intrinsic, social, and personal extrinsic. The study revealed that there was no direct relationship between religiosity and Papanicolaou test uptake.35 However, when the acculturation level of women was also considered, the study found that it was a significant moderator between the 3 types of religiosity and Papanicolaou test uptake. That is, less acculturated women with increased levels of intrinsic religiosity were significantly associated with a decrease in the likelihood of ever having had a Papanicolaou test. Highly acculturated women, however, with increased levels of personal extrinsic religiosity were associated with increased testing.35 Such results suggest that the effect of religion on test uptake is less significant when women are more adapted to the local culture of the country they are living in. Furthermore, a qualitative study by Brown et al20 noticed that black women in the United States felt more comfortable with a Papanicolaou test administered by a female because of religious teachings that forbade them to display their bodies to a male.
Language and acculturation were common features among Asian and Hispanic groups in the United States. However, culture-related factors were only observed to be related to screening behavior among the Asian groups in the United States and in Australia (Table 3).
LANGUAGE AND ACCULTURATION
Language was a common factor explored in the studies and found to be a significant factor influencing women’s decisions on taking Papanicolaou tests across EM groups. Asian women in the United States with better English-speaking abilities showed increased odds of ever having had cervical screening or been screened in the past 3 years.30,32 Those Asian and Hispanic women in the United States who could only read or speak their native languages were less likely to have had a test in the past year because of the perceived language barrier.27 Asian women in Australia also experienced similar problems and would not undergo screening if the physician in the healthcare organization could not communicate in their native language.29 However, the ability of Asian and black women in Britain to read English was not associated with ever having had a test.10 Women who spoke the language that the country commonly used were able to express their needs to healthcare professionals clearly and were thus likely to receive more relevant health information.
The level of acculturation was also examined in a number of studies and found to be a common significant factor influencing screening uptake, acculturation being interpreted in different ways by different studies.23,26,31,32,34–36,38 English proficiency and duration of residence in a country were commonly used indicators to determine acculturation levels.31,32,36 Asian women in the United States who could speak some English were more likely to have ever had a Papanicolaou test, whereas those women with poor English were less likely to have done so in the past 3 years.31,32 Women who used English in daily communication showed that they had adapted to the local culture, with such adaptation facilitating their use of screening services. Asian women who had lived for less than 10 years in the United States, and who were less acculturated, were associated with lower testing rates in the past 3 years.31 Shah et al38 adopted a validated Short Acculturation Scale,44 with questions on language use in their daily lives, to measure acculturation among Hispanic women and found that more highly acculturated women were more likely to have ever had a Papanicolaou test than their less acculturated counterparts.38 Nguyen et al34,35 used the 18-item Suinn-Lew Asian Self-Identity Acculturation Scale45 to measure acculturation among Asian women in the United States. The scale covers language, ethnic identity, friendship choices, generational behavior, geographical history, and attitudes. Nguyen et al found that highly acculturated Asian women in the United States were positively associated with ever having had a Papanicolaou test. Staying in a country for a longer period provided more opportunities and time for individuals to expose themselves to the local culture and to learn the language commonly used in that country. This might change their way of behaving and cause them to adopt the practices of the majority population.
Four quantitative and qualitative studies explored the influence of culture on Asian women’s screening behavior.29,30,36,40 Two30,36 used a validated scale, the Cultural Barriers to Screening Inventory,46 to assess the influence of culture. This is a 17-item scale that measures modesty, crisis orientation, use of Eastern medicine, and lack of family support. Among the Asian groups, maintaining modesty was found to be important to Chinese women in both the United States and Australia,29,30 with modesty being concerned with the levels of comfort associated with talking about one’s own body and the exposure of body parts during a health examination. Lee-Lin et al30 revealed that lower modesty scores were significantly associated with ever having had a Papanicolaou test and having had a test in the past 3 years. However, this phenomenon was not observed in another ethnic group, the Vietnamese,36,40 who did not place much emphasis on modesty. If a woman needed a Papanicolaou test, the gender of the physician or technician administering the test would be a concerning factor. To maintain modesty, Chinese Australian women preferred female physicians or technicians to avoid embarrassment and thought that female doctors were more sensitive to and had a better understanding of their health needs.29
Although Vietnamese Americans did not pay much attention to modesty, the lack of family support and adherence to Eastern medical practices were important factors negatively associated with them ever using screening services.36 They tended to use Eastern medicine to treat their health problems before turning to Western methods, resulting in reluctance to use cervical screening services. Furthermore, the lack of family support for their decisions on screening was another barrier to uptake. Some women tended to rely on their family or friends’ advice before they took action on screening. Although the use of Eastern medicine and lack of family support were important factors in Vietnamese women’s screening behavior, this was not found to be the case among Chinese Americans30 possibly because of where the Chinese participants in the study immigrated from: 54% from China, 30% from Taiwan, 7% from Hong Kong, 6% from Vietnam, and 1% from Macau, Indonesia, or Thailand. They were likely to have been exposed to more than 1 culture before they settled in the United States.
On the crisis orientation subscale, Chinese women recorded high scores, meaning that they were prevention oriented in view of the benefits of the Papanicolaou test and personal risk factors, and were associated with Papanicolaou test uptake during bivariate analysis. The effect disappeared after entering multivariate logistic regression. A similar situation was noticed among the Vietnamese, possibly related to the existence of some contextual variables such as personal finance and emotional states associated with prevention orientation.30,36
Policy Level Factor
The possession of health insurance was suggested as a major facilitator to cancer screening (n = 8)20,23,24,28,30,32,36 common across EM groups in the United States. Lin et al31 examined the influence of insurance and found that some Asian (Chinese, Korean, and South Asian) women without insurance were less likely to adhere to the Papanicolaou test guidelines, but such an association was not observed among the other Asian groups (Filipino, Japanese, or Vietnamese), and no association was found among Hispanic groups either.25,28 Inconsistent results were obtained from Asian (Chinese and Vietnamese) and black women, and no conclusions could therefore be drawn.23,28,30–32,36,37,40,41 However, it might well be related to the types of health insurance that the women had and the terms involved in its coverage of Papanicolaou test costs.
Interrelationships Within and Across Levels
Apart from the direct influence of each factor on screening behavior outcomes, an analysis of the results obtained from the selected studies revealed that certain factors, such as acculturation and other culture-related matters, had significant interrelationships within and across levels and could act on individual-level factors to cause changes in screening behavior.
RELATIONSHIP BETWEEN ACCULTURATION AND RELIGION
Apart from the direct influence of acculturation on screening behavior, it also acted as a moderator of the relationship between religion and screening behavior among Asian women.35 People who held strong beliefs about their own ethnic identity and cultural background (low acculturation) would resort to their deep faith in God’s power to control the occurrence of disease, thereby lowering screening uptake.32
RELATIONSHIP BETWEEN ACCULTURATION, LANGGUAGE, AND PERCEIVED BARRIERS
The level of acculturation was closely related to women’s daily language use, and poor proficiency in the official language of the country they lived in was a perceived barrier to their screening rate.31,32,36 Less acculturated women with less proficiency in the official language experienced difficulties in communicating their needs and had a lower level of exposure to publicly available health information. They would be likely to choose services whose providers could speak their native language. However, the rarity of such providers in healthcare organizations was a perceived barrier to screening.29
RELATIONSHIPS BETWEEN CULTURE-RELATED FACTORS, PHYSICIAN’S GENDER, AND PERCEIVED BARRIERS
Culture exerted an influence on women’s perceived barriers. Among the Asian groups, an emphasis on modesty was suggested as the reason for Chinese women putting aside the thought of having a Papanicolaou test.29,30 They would need to overcome the psychological barrier concerned with showing their bodies to a physician during any examination.29,30 To minimize embarrassment and maintain modesty, they preferred a female physician for the examination.20,29 The unavailability of female physicians in the healthcare organization was perceived as a barrier, and such unavailability gave them no choice but to avoid screening altogether.39
RELATIONSHIP BETWEEN CULTURE-RELATED FACTORS AND RECOMMENDATIONS FROM FRIENDS OR FAMILY
A lack of family support was negatively associated with Papanicolaou test uptake,36 and recommendations from friends or family could therefore be considered as a support for any decision to take screening. Some women relied on their family or friends for advice on health matters, such as the Asian women mentioned by Nguyen-Truong et al.36 In the absence of such support and recommendation, they would not act to check whether they had developed cancer.20,36
To the best of our knowledge, this is the first review to use an integrative model to examine multilevel factors associated with cervical cancer screening behavior among EM women. The findings of the review identify a number of factors other than those on the individual level—the interpersonal, organizational, and community levels are also associated with screening behavior in different EM groups in different countries.
In addition, the present review reveals the interrelationships between factors within and across levels resulting in the modification of screening behavior.
Although there is an increasing amount of data supporting the effect of culture on EM screening behavior, the role of culture-related factors has not been made clear in previous health behavior analyses and reviews.9 Although 1 previous study has reported the influence of cultural beliefs and attitudes on behavior at the individual level,10 the present review suggests that culture-related factors are better considered to be at the community level. Culture is of an inherited nature, providing the ethnic group with a psychological sense of community due to the common features of the individuals concerned, such as shared origin, norms, or practices that shape one’s way of thinking and behavior.15,47 The findings of this review also suggest that culture-related factors have an independent influence on perceived barriers at the individual level; and recommendations from friends and family, at the interpersonal level. Such information is valuable as it increases the health professional’s awareness of an EM’s cultural practices that actually affect perceptions of the disease and screening test. Culturally relevant interventions may therefore be designed to fit such practices and, with the influence of friends and family, to change perceptions of screening and of the disease, thereby improving screening uptake. Despite this, it is of note that research exploring how culture explicitly influences health beliefs and behavior is limited as the influence of culture is only revealed among Asians in the present review. Further examination of culture-related factors’ direct and indirect effects on EM women’s attitudes, perceptions, and screening behavior is certainly warranted.
Consistent with previous reviews, acculturation is another community-level factor commonly influencing women’s screening behavior. In the present review, it is noted that some studies used the duration of residence in a country and English proficiency as the indicators. They are easy to obtain and give us a preliminary idea of study participants’ acculturation. Residing longer in a country implies a longer period of exposure to and increased contact with the local population, and there is a higher potential for the EM to adopt its values and beliefs and follow its general patterns of behavior. Earning a living in the country and using the language that the general population uses to communicate are good ways for them to obtain relevant health information and updated daily news. However, simply understanding the level of English proficiency and duration of residence is inadequate. Some people may choose to communicate with others in the community using English and yet to use their native languages at home as they would like to retain their cultural affiliation.30 Using comprehensive validated scales is therefore a better way to understand acculturation because they cover not only linguistic areas but also ethnic identity, friendship choices, generational behavior, geographical history, and attitudes.45 They provide a comprehensive view of how the physical and social environment influences EMs’ choices in adapting to the local culture and their screening behavior.
Although a range of EM groups is included in the present review, it is found that some factors, ranging from intrapersonal to policy levels, are similar and currently experienced across ethnic groups in different countries. Some factors are only identified in specific ethnic groups. A family history of cancer and having a positive relationship with doctors are only identified in black populations. This could be due to the high incidence and mortality rate of cervical cancer among black women in the United States, where having a family history of cancer may expose them to a higher risk of contracting the disease.7 Unhappy experiences of previous medical research and clinical trials, where they were treated as guinea pigs, created mistrust between healthcare providers and black people when using health services. If a positive relationship could be built up and maintained between the two, with proper levels of trust and confidentiality, participation would be enhanced, and patients would feel more confident that they would not experience the same misfortune.48 Lack of usual sources of care, practical issues, and culture-related factors were only observed in the Asian groups. Asian women’s role in taking care of children at home gave them no time to attend screening, but, if they could find a children’s care center, they would be more likely to attend screening. It is commonly found that Asian women in both the United States and Australia have language barriers affecting their access to health information or regular care from healthcare providers. Being unable to overcome this barrier gives them no choice but to avoid seeking appropriate care. Otherwise, they may try other remedies such as Eastern medicine before turning to Western medicine for their health problems.29–32,35
Some factors are only found in specific ethnic groups, which is likely to be due to the uniqueness of particular factors in particular ethnic groups. However, it should also be noted that not all the same factors are investigated in all studies. There are differences in those chosen and examined in each study, and different criteria are set by the authors, such as the age range. In addition, most of the factors investigated in the studies mainly focus on the intrapersonal level, and there is a lack in the case of factors related to policy change. It is therefore difficult to compare and provide a full picture of the pattern of factor change in relation to updates in health policy. In addition, there are methodological limitations to different methods of data analysis used in the studies. Some of the earlier work uses descriptive rather than regression analysis—a method that may be inadequate to provide information about the strength and significance of the associations between factors and screening behavior.
In the past, most studies have been based on individual-level health behavior models to develop their interventions. It is suggested in the review by Golden and Earp49 that all the interventions on disease screening focus on the individual level—60% on the interpersonal level and only 13.3%, 26.7%, and 6.7% on the organizational, community, and policy levels, respectively. It is well recognized in individual-level models that the influence of attitudes and perceptions on screening behavior is significant because, without these, behavioral change may not take place. However, this raises another concern that environmental and social influences on health and behavior are being neglected. The use of individual-level models may not be adequate to provide a comprehensive understanding of women’s screening behavior from multiple perspectives. The factors identified in the present review—cultural and religious and those concerned with acculturation—that are clearly related to women’s screening behavior are not explicitly reflected in individual-level models.
McLeroy et al15 suggest that ecological models for health promotion could be a way to encompass both individual-level and socioenvironmental factors that play significant roles in health behavior. In addition, the World Health Organization has stated that personal health is a result of the interplay between individual-level and environmental factors.50 Models that are able to encompass multilevel factors would therefore lead to a better understanding of personal behavior. In the present review, some studies use models to explore factors extending to organizational and community levels, but the interaction between factors at different levels is not clearly indicated—for example, religion and acculturation, which are factors identified in the present review as actually moderating behavior. Further exploration of the interaction between factors across levels and the strength of interrelationships between these factors could lead to a better understanding of how factors at different levels interact or affect individual attitudes and perceptions and generate a certain form of behavior. Such information also facilitates the allocation of resources to develop multilevel interventions that may be the most effective way to change behavior and produce sustainable effects. A modified model is therefore needed and should be tested in future studies (Figure 2).
There are limitations to this review that need to be acknowledged. One is that all the selected studies were conducted in Western countries, mostly in the United States, probably because nearly half of the US population is composed of varied EM groups. The diversity of ethnic groups increases the burden of cervical cancer on the healthcare system.3 The results presented in this review may not be applicable to other parts of the world, such as Asian countries, because of differences in the ethnic groups, the living environment, and healthcare systems. Further studies among EM women in other countries are clearly warranted. Another limitation is the inadequate exploration of some other factors, such as health policy, prevalence of cervical cancer, treatment, and vaccination, that may be associated with women’s screening behavior. Changes in health policy, such as screening recommendation or the implementation of a national breast and cervical cancer early detection program in the United States that supports the screening of low-income, uninsured, or underinsured women, may influence women’s Papanicolaou test uptake by means of lowering or eliminating testing costs.8 In addition, the prevalence of cancer in the country would influence women’s perceptions of their own susceptibility to cancer. The advanced treatment and vaccines available may also affect women’s perceptions of their susceptibility and the severity of the disease. Further studies exploring the relationship between these factors and women’s screening behavior are clearly warranted. In the present review, we used 4 major categories (Hispanic or Latino, black, Asian, and Inuit) to classify the EM groups and compare them in different countries. However, ethnicity was poorly defined in some studies, hindering comparisons of women’s health behavior in a single EM.
Nevertheless, the strength of the review is the inclusion of both quantitative and qualitative studies, with the latter’s findings promoting understanding of the former’s. The factors identified in the review provide insights into how a modified health behavior model might be developed to meet the health needs of EM women.
The review has summarized factors, either common to or unique in EMs, that are associated with screening uptake. These factors not only affect screening decisions but also have an impact on the early diagnosis of cancer. Various factors exist at each level of the health behavior model and interact across levels to influence behavior. Because only limited work has been done outside the United States, further studies conducted on ethnic groups in other countries are clearly needed.
Implications for Research
The individual-level model is inadequate to encompass multilevel factors that act on women’s behavior and is unable to provide a comprehensive overview to increase health professionals’ understanding of EM’s current barriers and needs. A comprehensive model (Figure 2) addressing the multilevel factors identified should be developed and tested in future research studies. Each level represents the influence from the physical and social environment that guides screening behavior. Understanding the strength of the interrelationship between these factors could inform the design of multilevel interventions and the level to which attention should be directed to produce the greatest effect. Although some factors are similar or common across EM groups, it should be noted that others are unique to a specific EM group. The proposed model should therefore be tested in different EM groups with modifications to the factors at each level that best fit the particular minority. In addition, health policy, prevalence of cancer, treatment, and vaccination could be factors associated with women’s screening behavior. Therefore, further exploration and understanding of the impact of these factors on and their association with screening uptake will certainly benefit the design of any relevant intervention.
Implications for Practice/Policy
The multilevel model provides important information on how the various factors are interrelated. Such a model provides clues for policy makers in directing interventions at an appropriate level that will have the greatest effect on the public and result in a better allocation of existing resources. The similarities and differences in factors across EM groups also provide the important information that, apart from the intrapersonal level, those at organizational, community, and policy levels also play a significant role in women’s screening behavior. Government and policy makers might revise their strategies to promote screening uptake in other ways, rather than simply supporting testing costs and insurance coverage. Interventions directed at these levels may help to improve the screening uptake by increasing the facilities available and accessible, supporting childcare, having female physicians speaking the minority’s language, and overcoming culture-related influences that deter screening, which are especially important among Asian groups in both the United States and Australia. For example, information about healthcare organizations where female physicians are available would encourage Asian women to use the cervical screening services provided by those organizations so that they would be able to overcome the embarrassment the screening process seems to cause them and maintain their modesty. Hispanic women are also concerned about the language healthcare providers’ use. Providing information about where the service providers could speak their language would probably enhance their use of the screening services. Policy makers or service providers may also consider their users’ concerns, such as the issues of trust and the confidentiality of personal information that US black women emphasize. Service providers should state clearly the strategies they use to maintain the confidentiality of personal information and implement the strategies as needed to help in building a trusting and positive relationship between the healthcare providers and black women. It may also be important to revise the public health education currently directed at black people. For example, a family history of cancer is one of the risk factors of cervical cancer and enhances their screening behavior. Providing more information on the risk factors associated with cervical cancer would probably enhance their knowledge of the disease and arouse their awareness of the risk factors that made them prone to the disease.
1. Center for Health Protection, Department of Health. Health advice on prevention and screening of cancer for the healthcare professionals, 2004. http://www.chp.gov.hk/files/pdf/Advice_on_Cancer_Prevention_professionals.pdf
. Accessed February 15, 2016.
2. Kim JJ, Leung GM, Woo PPS, Goldie SJ. Cost-effectiveness of organized versus opportunistic cervical cytology screening in Hong Kong. J Public Health
3. Agency for Healthcare Research and Quality, U.S. Department of Health and Human Services. Women: stay healthy at any age, 2016. http://www.ahrq.gov/patients-consumers/prevention/lifestyle/healthy-women.html
. Accessed April 6, 2016.
4. Canadian Task Force of Preventive Health Care, the Public Health Agency of Canada. Screening for cervical cancer
(2013), 2016. http://canadiantaskforce.ca/ctfphc-guidelines/2013-cervical-cancer/
. Accessed April 6, 2016.
6. Public Health England. Cervical screening: programme overview, 2015. http://www.nhs.uk/Conditions/Cervical-screening-test/Pages/Introduction.aspx
. Accessed April 6, 2016.
7. National Cancer Institute. SEER stat fact sheet: cervix uteri cancer, 2014. http://seer.cancer.gov/statfacts/html/cervix.html
. Accessed February 15, 2016.
8. Centre for Disease Control and Prevention. Use of Pap smears among women aged 18 and over, by selected characteristics: United States, selected years 1987–2010, 2015. http://www.cdc.gov/nchs/data/hus/2013/084.pdf
. Accessed February 15, 2016.
9. McDonald JT, Kennedy S. Cervical cancer
screening by immigrant and minority women in Canada. J Immigrant Minor Health
10. Marlow LAV, Wardle J, Waller J. Understanding cervical screening non-attendance among ethnic minority women in England. Br J Cancer
11. Gill PS, Kai J, Bhopal RS, Wild S. Black and minority ethnic groups, 2007. http://www.birmingham.ac.uk/Documents/college-mds/haps/projects/HCNA/04HCNA3D4.pdf
. Accessed March 22, 2016.
12. Division of Cancer Prevention and Control of Centre for Disease Control and Prevention. Moving forward in an era of reform new directions for cancer screening 2016. http://www.cdc.gov/cancer/nbccedp/pdf/newdirections_screening.pdf
. Accessed March 23, 2016.
13. Ackerson K, Gretebeck K. Factors influencing cancer screening practices of underserved women. J Am Acad Nurse Pract
14. Johnson CE, Mues KE, Mayne SL, Kiblawi AN. Cervical cancer
screening among immigrants and ethnic minorities: a systematic review
using the Health Belief Model. J Low Genit Tract Dis
15. McLeroy KR, Bibeau D, Steckler A, Glanz K. An ecological perspective on health promotion programs. Health Educ Q
16. National Cancer Institute. Theory at a Glance: A Guide for Health Promotion Practice
. Washington, DC: US Department of Health and Human Services/National Institute of Health; 2005.
17. Moher D, Liberati A, Tetzlaff J, Altman DG; Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med
18. Harden A, Thomas J. Methodological issues in combining diverse study types in systematic reviews. Int J Soc Res Methodol
19. Long AF, Godfrey M. An evaluation tool to assess the quality of qualitative research studies. Int J Soc Res Methodol
20. Brown DR, Wilson RM, Boothe MAS, Harris CES. Cervical cancer
screening among ethnically diverse black women: knowledge, attitudes, beliefs, and practices. J Natl Med Assoc
21. Cerigo H, Coutlée F, Franco EL, Brassard P. Factors associated with cervical cancer
screening uptake among Inuit women in Nunavik, Quebec, Canada. BMC Public Health
22. Dettenborn L, Duhamel K, Butts G, Thompson H, Jandorf L. Cancer fatalism and its demographic correlates among African American and Hispanic women. J Psycho Oncol
23. Do M. Predictors of cervical cancer
screening among Vietnamese American women. J Immigrant Minor Health
24. Fang CY, Ma GX, Tan Y, Chi N. A multifaceted intervention to increase cervical cancer
screening among underserved Korean women. Cancer Epidemiol Biomarkers Prev
25. Gonzalez P, Castaneda SF, Mills PJ, Talavera GA, Elder JP, Gallo LC. Determinants of breast, cervical and colorectal cancer screening adherence in Mexican-American women. J Community Health
26. Harmon MP, Castro FG, Coe K. Acculturation and cervical cancer
: knowledge, beliefs, and behaviors of Hispanic women. Women Health
27. Jacobs EA, Karavolos K, Rathouz PJ, Ferris TG, Powell LH. Limited English proficiency and breast and cervical cancer
screening in a multiethnic population. Am J Public Health
28. Jennings-Dozier K, Lawrence D. Sociodemographic predictors of adherence to annual cervical cancer
screening in minority women. Cancer Nurs
29. Kwok C, White K, Roydhouse JK. Chinese-Australian women’s knowledge, facilitators and barriers related to cervical cancer
screening: a qualitative study. J Immigrant Minor Health
30. Lee-Lin F, Pett M, Menon U, et al. Cervical cancer
beliefs and pap test screening practices among Chinese American immigrants. Oncol Nurs Forum
31. Lin MK, Moskowitz JM, Kazinets G, Ivey SL, Kim Y-B, McDonnell DD. Adherence to Pap test guidelines: variation among Asians in California. Ethn Dis
32. Ma GX, Fang CY, Feng Z, et al. Correlates of cervical cancer
screening among Vietnamese American women. Infect Dis Obstet Gynecol
33. Ma GX, Gao W, Fang CY, et al. Health beliefs associated with cervical cancer
screening among Vietnamese Americans. J Womens Health
34. Nguyen AB, Clark TT. The role of acculturation and collectivism in cancer screening for Vietnamese American women. Health Care Women Int
35. Nguyen AB, Hood KB, Belgrave FZ. The relationship between religiosity and cancer screening among Vietnamese women in the United States: the moderating role of acculturation. Women Health
36. Nguyen-Truong CKY, Lee-Lin F, Leo MC, et al. A community-based participatory research approach to understanding pap testing adherence among Vietnamese American immigrants. J Obstet Gynecol Neonatal Nurs
37. Sentell TL, Tsoh JY, Davis T, Davis J, Braun KL. Low health literacy and cancer screening among Chinese Americans in California: a cross-sectional analysis. BMJ Open
38. Shah M, Zhu K, Wu H, Potter J. Hispanic acculturation and utilization of cervical cancer
screening in the US. Prev Med
39. Skaer TL, Robison LM, Sclar DA, Harding GH. Knowledge, attitudes, and patterns of cancer screening: a self-report among foreign born Hispanic women utilizing rural migrant health clinics. J Rural Health
40. Taylor VM, Yasui Y, Burke N, et al. Pap testing adherence among Vietnamese American women. Cancer Epidemiol Biomarkers Prev
41. Williams KP, Reiter P, Mabiso A, Maurer J, Paskett E. Family history of cancer predicts Papanicolaou screening behavior for African American and white women. Cancer
42. Yoo GJ, Le MN, Vong S, Lagman R, Lam AG. Cervical cancer
screening: attitudes and behaviors of young Asian American women. J Cancer Educ
43. Allport GW, Ross JM. Personal religious orientation and prejudice. J Pers Soc Psychol
44. Marin G, Sabogal F, Marin BV, Otero-Sabogal R, Perez-Stable EJ. Development of a Short Acculturation Scale for Hispanics. Hispanic J Behav Sci
45. Suinn RM, Rickard-Figueroa K, Lew S, Vigil P. The Suinn-Lew Asian Self-Identity Acculturation Scale: an initial report. Educ Psychol Meas
46. Tang TS, Solomon LJ, Yeh CJ, Worden JK. The role of cultural variables in breast self-examination and cervical cancer
screening behavior in young Asian women living in the United States. J Behav Med
47. Leininger M. Culture care theory, research, and practice. Nurs Sci Q
48. Seto B. History of medical ethics and perspectives on disparities in minority recruitment and involvement in health research. Am J Med Sci
49. Golden SD, Earp JAL. Social ecological approaches to individuals and their contexts: twenty years of health education & behavior health promotion interventions. Health Educ Behav
50. Blas E, Kurup AS (Eds). Equity, Social Determinants and Public Health Programmes
. Geneva, Switzerland: World Health Organization; 2010.