Elsevier

Social Science & Medicine

Volume 66, Issue 2, January 2008, Pages 260-275
Social Science & Medicine

Contextual analysis of breast and cervical cancer screening and factors associated with health care access among United States women, 2002

https://doi.org/10.1016/j.socscimed.2007.09.009Get rights and content

Abstract

This research explored the relationships between race/ethnicity and area factors affecting access to health care in the United States. The study represents an advance on previous research in this field because, in addition to including data on rurality, it incorporates additional contextual covariates describing aspects of health care accessibility. Individual-level data were obtained from the 2002 Behavioral Risk Factor Surveillance System (BRFSS). The county of residence reported by BRFSS respondents was used to link BRFSS data with county-level measures of health care access from the 2004 Area Resource File (ARF). Analyses of mammography were limited to women aged ⩾40 years with known county of residence (n=91,492). Analyses of Pap testing were limited to women aged ⩾18 years with no history of hysterectomy and known county of residence (n=97,820). In addition to individual-level covariates such as race, Hispanic ethnicity, health insurance coverage and routine physical exam in the previous year. We examined county-level covariates (residence in health professional shortage area, urban/rural continuum, racial/ethnic composition, and number of health centers/clinics, mammography screening centers, primary care physicians, and obstetrician-gynecologists per 100,000 female population or per 1000 square miles) as predictors of cancer screening. Both individual-level and contextual covariates are associated with the use of breast and cervical cancer screening. In the current study, covariates associated with health care access, such as health insurance coverage, household income, Black race, and percentage of county female population who were non-Hispanic Black, were important determinants of screening use. In multivariate analysis, we found significant interactions between individual-level covariates and contextual covariates. Among women who reside in areas with lower primary care physician supply, rural women are less likely than urban women to have had a recent Pap test. Black women were more likely than White women to have had a recent Pap test. Women with a non-rural county of residence were more likely to have had a recent mammogram than rural women. A significant interaction was also found between individual-level race and number of health centers or clinics per 100,000 population (p-value=0.0187). In counties with 2 or more health centers or clinics per 100,000 female population, Black women were more likely than White women to have had a recent mammogram. A significant interaction was also observed between the percentage of county female population who were Hispanic and the percentage who were non-Hispanic Black.

Introduction

Studies of breast and cervical cancer screening in the United States have shown that women with greater access to health care, such as those with health insurance or a higher family income, are more likely to have recent screening tests (O’Malley et al., 2001; Selvin & Brett, 2003). Having had a recent physician visit or a usual source of health care is also predictive of screening adherence (Zapka, Puleo, Vickers-Lahti, & Luckmann, 2002). The accessibility of routine health care is also important. For example, persons who live in areas of the United States with more primary care providers might have greater access to cancer screening (Coughlin & Thompson, 2004).

Adherence to breast and cervical cancer screening is also positively associated with provider factors (Haggerty et al., 1999; Hawley et al., 2000; Pham et al., 2005) and with health care system factors (Fox & Stein, 1991; Haggstrom et al., 2004). For example, several studies have shown that a physician recommendation is one of the strongest independent predictors of a person's decision to have a cancer screening test (Burack & Liang, 1987; Coughlin, Breslau, Thompson & Benard, 2005; Lerman et al., 1990; Nguyen & McPhee, 2003; Zapka et al., 1991). External factors in the health care environment or organizational and contextual factors such as residence in a rural area of the United States or in a county with more minority persons have also been shown to influence use of mammography and other cancer screening tests (Benjamins, Kirby, & Bond Huie, 2004; Coughlin et al., 2002; Ruffin, Gorenflo, & Woodman, 2000). Both health policies and health care system factors may account for higher or lower uptake of screening services. Although the current article focuses on the United States where insurance-based health systems are predominant, health systems vary internationally. Other countries (for example, Canada and Great Britain) have different health care systems. Further, mammography screening is often unavailable in low resource settings outside the United States.

Few population-based studies have examined possible interactive effects between rural residence and race/ethnicity in relation to cancer screening (Coughlin et al., 2002; Duelberg, 1992; Zhang, Tao, & Irwin, 2000). In our previous analysis of breast and cervical cancer screening among women in rural and non-rural areas of the United States, which included data from the 1998–1999 Behavioral Risk Factor Surveillance System (BRFSS) (Coughlin et al., 2002), we found that race/ethnicity modified the association with rurality in relation to recent mammography but not in relation to recent Pap testing. On average, women living in rural areas of the United States were older, more likely to be White, less likely to be single, less educated, more likely to report fair or poor general health status (as compared with good or excellent general health status), less likely to have health insurance, and more likely to have a lower household income, than women in other geographic areas of the country. In an analysis of data from the 1985 National Health Interview Survey, Duelberg (1992) found that Black women residing in urban areas of the United States were much more likely to be screened for cervical cancer than Black women in rural areas. An analysis of 1994 National Health Interview Survey data suggested possible interactive effects between race and urban/rural residence in relation to breast cancer screening, but the association with rural residence was not significant after adjustment for education, household income, and health insurance status (Zhang et al., 2000).

Benjamins et al. (2004) conducted a contextual analysis of data from the Medical Expenditure Panel Survey (MEPS) and the Area Resource File (ARF; 1996–1998) to investigate whether county-level racial/ethnic composition was associated with the use of mammography and other preventive services in the United States. Women residing in counties with more Blacks were more likely to have regular mammograms. In particular, Hispanic women who resided in counties with high percentages of Blacks reported higher levels of utilization of mammograms and other preventive services compared to Hispanic women residing in other counties, suggesting a possible interactive effect between county racial/ethnic composition and individual-level ethnicity.

Rural and non-rural areas in the United States differ in health care workforce and provider supply and in distance to mammography screening facilities (Amey, Miller, & Albrecht, 1997; Guagliardo, 2004; Salsberg & Forte, 2002; Yabroff et al., 2005). Some studies have found that residing in a county with a greater number of physicians is associated with increased Pap testing and mammography (Benjamins et al., 2004) and that primary care physician supply may be associated with earlier stage at breast cancer diagnosis (Davidson, Bastani, Nakazono, & Carreon, 2005).

Few studies have examined both individual-level and ecological measures of health care access (i.e., individual-level factors and contextual factors) as predictors of breast and cervical cancer screening, including possible interactions between the two (Benjamins et al., 2004; Engelman et al., 2002; Rosenberg et al., 2005). A study of Medicare beneficiaries in Kansas found that women in counties with higher median incomes and a greater percentage of residents with high school diplomas had higher utilization rates (Engelman et al., 2002). In a multilevel study of data from the Black Women's Health Study, a large cohort of African-American women, researchers assessed individual-level and group-level socioeconomic predictors of regular mammography use (Rosenberg et al., 2005). Higher neighborhood socioeconomic status was significantly associated with regular mammography use before, but not after control for household income (Rosenberg et al., 2005).

In the current study, we examined data from BRFSS and the ARF (ARF, 2004; Centers for Disease Control and Prevention, 1984–2004) to determine if women who reside in counties with ecological markers of decreased access to health care (i.e., rural residence, fewer physicians, fewer health centers/clinics) are less likely to report a recent mammogram or Pap test. The study represents an advance on prior studies of rural residence and race/ethnicity in relation to breast and cervical cancer screening because it includes additional contextual covariates related to health care access in the analysis. We examined whether associations between cancer screening and individual-level covariates such as household income and health insurance coverage persist after adjusting for contextual factors associated with health care availability and access. We hypothesized that the effects of these individual-level covariates (i.e., household income and health insurance coverage) on cancer screening might be modified by contextual factors associated with health care availability and access including numbers of primary care physicians, health centers/clinics, mammography screening centers, rural/non-rural residence, and racial/ethnic composition.

The conceptual framework that guided the analysis was partly based upon the behavioral model of health services utilization (Andersen, 1995; Phillips et al., 1998) and took into account published research (Benjamins et al., 2004; Davidson et al., 2005; Engelman et al., 2002) and potential linkages between breast and cervical cancer screening and access to health care. The behavioral model of health services utilization and prior studies suggest that individual factors such as age, race, ethnicity, marital status, educational attainment, and household income may predispose to cancer screening. Enabling factors related to health care access, affordability, and availability (higher family income, employment status, having health insurance coverage, and routine physical examination in the past year) may also play a role in screening. Our conceptual framework also took into consideration the fact that residential location may affect screening use through geographic variation in the availability and proximity of providers, clinics, and screening services. Measures of health care availability and access (for example, numbers of primary care physicians, health centers/clinics, mammography screening centers) may be considered enabling factors. The racial and ethnic composition of the areas in which people live and other county-level (contextual) covariates may also be considered enabling factors. For example, women who live in areas with a high population of minority individuals may be more likely to be targeted for community health clinics or to be exposed to health education campaigns (Benjamins et al., 2004). County-level covariates such as racial and ethnic composition may also reinforce predisposing factors. Our conceptualization of possible interactions between individual-level and contextual factors in the prediction of breast and cervical cancer screening was based upon prior research findings suggesting that county-level enabling factors such as racial/ethnic composition may partly influence cancer screening rates by modifying individual-level predisposing factors (Benjamins et al., 2004).

Section snippets

Data sources and analytic samples

We accessed information from 3 data sources, the 2002 BRFSS, the 2004 ARF and Census 2002 county female population estimates (Census Bureau). The BRFSS is a state-based system of household telephone health surveys that targets the US adult population (ages 18 and older) (Centers for Disease Control and Prevention, 1984–2000). This system of surveys is designed to obtain data on behavioral health risk factors and inquires about respondent health and access to and use of health services, as well

Results

A description of the samples (for example, percentage of female BRFSS respondents who reside in rural, suburban, and metropolitan areas of the United States) can be found in a previous publication (Coughlin et al., 2002).

Discussion

The current study adds to the relatively sparse but expanding literature about the role of contextual factors in cancer screening by evaluating county-level factors related to the availability and accessibility of providers and screening services. Although individual-level factors such as health insurance coverage and household income appear to be more influential determinants of both breast and cervical cancer screening than contextual measures of availability of care, we found that all women

Acknowledgments

The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention. This research was supported in part by the appointment of Susan Sabatino to the Centers for Disease Control and Prevention Research Participation Program, administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the US Department of Energy and Oak Ridge Associated Universities.

References (42)

  • Area Resource File (ARF) (2004). US Department of Health and Human Services, Health Resources and Services...
  • L.C. Baker et al.

    The effect of area HMO market share on cancer screening

    Health Services Research

    (2004)
  • M.L. Berk et al.

    The health care of poor persons living in wealthy areas

    Social Science & Medicine

    (1997)
  • Census Bureau population estimates, as modified by the National Cancer Institute's Surveillance Epidemiology and End...
  • Behavioral Risk Factor Surveillance System technical documents and survey data

    (1984–2000)
  • S.S. Coughlin et al.

    Physician recommendation for Papanicolaou testing among US women, 2000

    Cancer Epidemiology Biomarkers & Prevention

    (2005)
  • S.S. Coughlin et al.

    Colorectal cancer screening practices among men and women in rural and nonrural areas of the United States, 1999

    Journal of Rural Health

    (2004)
  • S.S. Coughlin et al.

    Breast and cervical carcinoma screening practices among women in rural and nonrural areas of the United States, 1998–1999

    Cancer

    (2002)
  • P.L. Davidson et al.

    Role of community risk factors and resources on breast carcinoma stage at diagnosis

    Cancer

    (2005)
  • K.K. Engelman et al.

    Impact of geographic barriers on the utilization of mammograms by older rural women

    Journal of the American Geriatrics Society

    (2002)
  • S.A. Fox et al.

    The effect of physician–patient communication on mammography utilization by different ethnic groups

    Medical Care

    (1991)
  • Cited by (0)

    View full text