SEARCH

SEARCH BY CITATION

Keywords:

  • breast carcinoma;
  • cervical carcinoma;
  • cancer prevention and control;
  • rural health;
  • rural populations;
  • screening mammography;
  • Pap tests

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

Prior studies have suggested that women living in rural areas may be less likely than women living in urban areas to have had a recent mammogram and Papanicolau (Pap) test and that rural women may face substantial barriers to receiving preventive health care services.

METHODS

The authors examined both breast and cervical carcinoma screening practices of women living in rural and nonrural areas of the United States from 1998 through 1999 using data from the Behavioral Risk Factor Surveillance System. The authors limited their analyses of screening mammography and clinical breast examination to women aged 40 years or older (n = 108,326). In addition, they limited their analyses of Pap testing to women aged 18 years or older who did not have a history of hysterectomy (n = 131,813). They divided the geographic areas of residence into rural areas and small towns, suburban areas and smaller metropolitan areas, and larger metropolitan areas.

RESULTS

Approximately 66.7% (95% confidence interval [CI] = 65.8% to 67.6%) of women aged 40 years or older who resided in rural areas had received a mammogram in the past 2 years, compared with 75.4% of women living in larger metropolitan areas (95% CI = 74.9% to 75.9%). About 73.0% (95% CI = 72.2% to 73.9%) of women aged 40 years or older who resided in rural areas had received a clinical breast examination in the past 2 years, compared with 78.2% of women living in larger metropolitan areas (95% CI = 77.8% to 78.7%). About 81.3% (95% CI = 80.6% to 82.0%) of 131,813 rural women aged 18 years or older who had not undergone a hysterectomy had received a Pap test in the past 3 years, compared with 84.5% of women living in larger metropolitan areas (95% CI = 84.1% to 84.9%). The differences in screening across rural and nonrural areas persisted in multivariate analysis (P < 0.001).

CONCLUSIONS

These results underscore the need for continued efforts to provide breast and cervical carcinoma screening to women living in rural areas of the United States. Cancer 2002; 94:2801–12. © 2002 American Cancer Society.

DOI 10.1002/cncr.10577

Previous studies have suggested that women living in rural areas of the United States may use preventive health care services less frequently than women living in urban areas of the country. 1 Studies also have found that women living in rural areas are less likely than those living in urban areas to have had a recent mammogram or Papanicolau (Pap) test. 2, 3 Furthermore, women in rural areas of the United States have been found to have higher rates of breast carcinoma and late-stage disease than women in nonrural areas. 3, 4

Possible explanations to account for the less frequent use of preventive services by rural women, compared with nonrural women, include greater distances to medical facilities and less accessibility of services; both of these factors are associated with lower education and income levels in rural areas. 1, 3 Inadequate health insurance coverage may also be an important barrier to the use of preventive health care services for people living in rural areas. 1

Previous studies of mammogram and Pap test utilization among women living in the United States have suggested the possible importance of race in assessing associations with rural and nonrural residence. In an analysis of data from the 1985 National Health Interview Survey, Duelberg found that no difference existed between white women living in urban areas and white women living in rural areas in regard to their likelihood of obtaining a Pap test. Black women living in urban areas, however, were much more likely than black women in rural areas to be screened for cervical carcinoma. 5 Thus, urban-rural differences in the use of cancer screening tests may be more pronounced among minorities. 2

The current article describes both breast and cervical carcinoma screening practices of women living in rural, suburban, and metropolitan counties in the United States. The study data were obtained by population-based probability samples during 1998 and 1999. Rates of screening mammography, clinical breast examinations, and Pap tests were examined among women living in rural and in nonrural areas as well as correlates of these types of cancer screening practices. We also examined whether the effects of rural and nonrural residences on screening were modified by specific factors, such as the respondent's race and ethnicity, age, education, and health insurance coverage, as well as whether the respondent had seen a physician within the past year.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The data used in the current study were obtained from 175,054 women who were interviewed during 1998-1999 as part of the Behavioral Risk Factor Surveillance System (BRFSS), a state-based telephone survey of adults aged 18 years or older. 6, 7 The BRFSS used a random-digit dialing technique and multistage cluster sampling in each participating state in order to sample noninstitutionalized adults living in a residence that had a telephone. 8 Trained interviewers administered the computer-assisted interviews.

The current study sample was drawn from women aged 18 years or older who responded to BRFSS surveys in 50 states and the District of Columbia. All eligible women were included regardless of their self-identified race (i.e., white, black, or other) and ethnicity (i.e., Hispanic versus non-Hispanic). Because the number of women of racial backgrounds other than white or black were too small for separate analyses, those women were grouped together in the category other. The latter included 5,287 Asian and Pacific Islander women and 3,065 American Indian and Alaska Native women aged 18 or older. In addition, 950 women whose ages were unknown were excluded from the sample. Data from the two year study period (i.e., 1998 and 1999) were pooled to increase the sample size available for this analysis.

Analyses of the use of screening mammogram and clinical breast examination were limited to women aged 40 years of age or older regardless of their hysterectomy status (n = 108,326). Analyses of Pap test use were limited to women aged 18 years of age or older who had not had a hysterectomy (n = 131,813).

Telephone coverage for the survey ranged from 87% to 98% across states. 9 Survey coverage also varied by subgroup, although specific data for each of the three geographic areas—rural, suburban, and metropolitan—were not available. 9 Weights were used to adjust for differences in probability of selection, nonresponse, and noncoverage. The estimated median response rates according to the Council of American Survey Research Organizations for 1998 and 1999 were 59.1% and 55.2%, respectively. 10 The numerator denotes the number of completed responses and the denominator is an estimate of the number of households in the sample. 11

The study interviews included questions about general health status, both demographic and socioeconomic characteristics, screening mammography, clinical breast examinations, and Pap tests. Each female respondent was asked whether she had ever had a mammogram; participants who responded positively were then asked when they had received their last mammogram. Similar questions were asked concerning clinical breast examination and Pap tests. The women also were asked whether they had undergone a hysterectomy.

The geographic area of residence was defined using the U.S. Department of Agriculture's Beale codes. 12 Beale codes 0-3 correspond to metropolitan areas (including areas with populations of about 250,000 to greater than 1 million); Beale codes 4,5 correspond to urban populations of 20,000 or greater (but less than 250,000); and Beale codes 6-9 correspond to rural populations and to nonmetropolitan urban populations of up to 19,999. Based on their county of residence, the survey respondents were categorized as residents of either more populated metropolitan areas, suburban areas and smaller metropolitan areas, or rural areas and small towns.

Age-adjusted rates of screening test use were calculated for the two year study period of 1998-1999. The direct method was used to adjust estimates of the proportion of women screened for cancer for both participant's age and calendar year using the distribution for women in the overall analytical sample as the standard. In examining the bivariate associations between screening and both the demographic and health factors, the levels of statistical significance were obtained using Pearson chi- square tests and SUDAAN; P values were obtained with adjustment for age and year. 13 All conducted analyses used both SAS (SAS Institute, Inc., Carey, NC) and SUDAAN to calculate the 95% confidence intervals (CIs) and P values and to allow for weighting of the estimates. 13 To better represent the overall population and enable the different samples to be combined and compared, the samples were weighted to compensate for the unequal sampling probability resulting from the unique number of telephones per household; the number of unique telephone numbers per primary sampling unit; and poststratification by age, gender, and race.

A multivariate analysis of correlates of screening test use was conducted using logistic regression techniques and SUDAAN. 13 Indicator variables for the survey year and age categories were included in all models. Two or more indicator variables were included for categorical variables, such as age, and the Wald chi- square test was used to examine the overall statistical significance of those variables. Covariates for categories of educational attainment, rather than for household income, were included in the models to avoid problems with colinearity and missing data. In addition, covariates for the number of persons in the household, rather than the number of children, were included.

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Characteristics of women aged 18 years or older who did not have a history of hysterectomy were categorized according to whether they lived in a rural, suburban, or metropolitan area (Table 1). On average, women living in rural areas were older, more likely to be white, less likely to be single, less educated, more likely to report fair or poor general health status, less likely to have health insurance, more likely to smoke cigarettes, and less likely to consume alcohol, and had a lower household income, than participants living in other geographic study areas. Similar results were obtained for women aged 40 years or older regardless of hysterectomy status (results not shown).

Table 1. Characteristics of Rural and Nonrural Women in the United States, Aged 18 or Older
 Rural % (95% CI)Suburban % (95% CI)Metropolitan % (95% CI)
  • CI: confidence interval.

  • a

    P < .001.

  • b

    P < .01. Weighted population estimates adjusted for age and year of survey; women who responded don't know or not sure or who refused are excluded along with those with a history of a hysterectomy.

Agea
 18 to 29 years24.3 (23.4–25.2)27.9 (26.3–29.5)26.0 (25.5–26.5)
 30 to 39 years22.5 (21.7–23.3)22.6 (21.4–23.8)25.1 (24.7–25.5)
 40 to 49 years19.4 (18.7–20.2)19.4 (18.2–20.5)19.7 (19.3–20.1)
 50 to 69 years21.9 (21.1–22.7)19.9 (18.7–21.1)19.6 (19.2–20.0)
 ≥ 70 years11.8 (11.3–12.4)10.2 (9.4–11.1)9.6 (9.3–9.9)
Race/ethnicitya
 White86.2 (85.5–86.8)84.5 (83.4–85.6)72.8 (72.3–73.3)
 Black7.4 (6.9–7.9)5.7 (5.1–6.3)11.0 (10.7–11.4)
 Hispanic4.7 (4.3–5.2)6.1 (5.3–6.9)12.1 (11.7–12.5)
 Other1.7 (1.5–2.0)3.7 (3.1–4.2)4.1 (3.8–4.3)
Marital statusa
 Currently married62.5 (61.7–63.4)59.5 (58.2–60.9)55.7 (55.2–56.2)
 Divorced or separated11.0 (10.5–11.6)12.3 (11.4–13.1)13.4 (13.0–13.7)
 Widowed12.6 (12.0–13.1)11.5 (10.7–12.2)11.2 (10.9–11.5)
 Never married12.0 (11.4–12.5)14.8 (13.9–15.7)17.2 (16.9–17.5)
 Living as unmarried couple1.9 (1.6–2.2)1.9 (1.5–2.2)2.6 (2.4–2.7)
Educational attainmenta
 < High school graduate16.9 (16.1–17.6)13.4 (12.2–14.5)12.6 (12.2–13.0)
 High school graduate/GED40.2 (39.2–41.1)37.3 (35.8–38.8)31.7 (31.2–32.2)
 Some college/technical school25.7 (24.9–26.5)27.2 (25.9–28.4)27.9 (27.5–28.4)
 College graduate17.2 (16.5–17.9)22.2 (21.0–23.3)27.8 (27.3–28.2)
Household incomea
 < $15,00018.2 (17.4–18.9)17.9 (16.5–19.2)14.1 (13.6–14.5)
 $15,000 to $34,99945.9 (44.8–46.9)40.8 (39.2–42.4)37.1 (36.5–37.7)
 $35,000 to $49,99918.2 (17.4–18.9)19.6 (18.3–20.9)18.2 (17.8–18.7)
 ≥ $50,00017.8 (17.0–18.6)21.7 (20.5–23.0)30.6 (30.1–31.1)
Employment statusa
 Currently employed58.4 (57.5–59.2)59.9 (58.6–61.1)60.2 (59.8–60.7)
 Homemaker or retired33.5 (32.7–34.2)33.5 (32.3–34.6)32.7 (32.3–33.1)
 Unemployed3.6 (3.2–4.0)3.6 (3.0–4.1)4.1 (3.8–4.3)
 Unable to work4.6 (4.1–5.0)3.1 (2.5–3.7)3.0 (2.8–3.2)
Number of persons in householda
 115.8 (15.2–16.3)16.9 (16.0–17.9)16.8 (16.5–17.2)
 233.2 (32.3–34.0)34.0 (32.6–35.3)31.7 (31.2–32.2)
 ≥ 351.1 (50.3–51.8)49.1 (47.9–50.3)51.5 (51.0–51.9)
General health statusa
 Good to excellent83.2 (82.5–83.9)86.5 (85.5–87.5)86.3 (85.9–86.7)
 Fair or poor16.8 (16.1–17.5)13.5 (12.5–14.5)13.7 (13.3–14.1)
Saw a physician in the past yearb
 Yes74.7 (73.9–75.5)74.7 (73.4–76.0)76.5 (76.0–76.9)
 No25.3 (24.5–26.1)25.3 (24.0–26.6)23.5 (23.1–24.0)
Any health insurance coveragea
 Yes83.2 (82.5–83.9)85.9 (84.9–86.9)87.5 (87.2–87.9)
 No16.8 (16.1–17.5)14.1 (13.1–15.1)12.5 (12.1–12.8)
Current cigarette smokinga
 Yes22.3 (21.5–23.0)21.0 (19.8–22.1)19.6 (19.2–20.0)
 No77.7 (77.0–78.5)79.0 (77.9–80.2)80.4 (80.0–80.8)
Current alcohol consumptiona
 Yes37.5 (36.2–38.8)42.4 (40.2–44.9)49.1 (48.1–50.1)
 No62.5 (61.2–63.8)57.6 (55.5–59.8)50.9 (49.9–51.9)

Pap Testing

About 94.5% (95% confidence interval [CI] = 94.1% to 95.0%) of rural women aged 18 years or older without a history of hysterectomy reported that they had received a Pap test at least once, and 81.3% (95% CI = 80.6% to 82.0%) of rural women aged 18 years or older had received a Pap test in the past three years, after adjusting for age and year of survey (results not shown). A similar proportion (93.9%; 95% CI = 93.6% to 94.1%) of women aged 18 years or older who lived in metropolitan areas reported that they had received a Pap test at least once, and as many as 84.5% (95% CI = 84.1% to 84.9%) had received a Pap test in the past three years, after adjusting for age and year of survey (results not shown).

For strata-specific Pap test rates (adjusted for age of participant and calendar year) presented in Table 2, the strata correspond to rural and nonrural residence and selected demographic characteristics, medical history, and cancer screening practices. Having had a Pap test in the past three years was associated with age, race/ethnicity, marital status, higher education level, higher household income, number of persons in the household, employment status, better general health status, having seen a physician in the past year, health insurance coverage, and use of other cancer screening tests (Table 2).

Table 2. Percentage of Rural and Nonrural Women in the United States Aged 18 or Older Who Had Received a Pap Test in the Past Three Years, According to Selected Demographic Characteristics, Medical History, and Cancer Screening Practices
 Rural % (95% CI)Suburban % (95% CI)Metropolitan % (95% CI)
  • CI: confidence interval.

  • a

    P < 0.001.

  • b

    Limited to women aged 40 or older. Weighted population estimates adjusted for age and year of survey; women who responded don't know or not sure or who refused are excluded along with those with a history of a hysterectomy.

Agea
 18 to 29 years87.9 (86.5–89.3)85.8 (83.1–88.4)84.8 (83.9–85.6)
 30 to 39 years87.3 (85.9–88.8)88.3 (86.4–90.2)90.7 (90.2–91.3)
 40 to 49 years84.1 (82.6–85.6)86.3 (83.9–88.6)88.3 (87.5–89.0)
 50 to 69 years80.0 (78.4–81.5)80.7 (77.9–83.5)85.1 (84.3–85.9)
 ≥ 70 years61.2 (58.7–63.6)67.8 (63.7–71.8)69.6 (68.1–71.1)
Race/ethnicitya
 White81.4 (80.7–82.2)83.2 (82.0–84.5)85.3 (84.8–85.7)
 Black85.0 (82.6–87.4)84.6 (80.8–88.3)86.8 (85.6–88.0)
 Hispanic74.5 (69.8–79.2)75.6 (71.2–80.1)81.4 (79.5–83.2)
 Other76.8 (71.3–82.2)77.0 (72.4–81.5)76.1 (73.2–79.0)
Marital statusa
 Currently married85.1 (84.2–86.0)87.0 (85.5–88.5)88.6 (88.0–89.1)
 Divorced or separated80.8 (78.5–83.1)82.8 (79.8–85.8)84.5 (83.3–85.7)
 Widowed78.0 (74.3–81.8)69.8 (65.3–74.3)84.3 (82.0–86.6)
 Never married71.1 (67.6–74.7)73.8 (68.8–78.8)76.6 (75.0–78.3)
 Living as unmarried couple82.7 (79.3–86.0)89.4 (84.9–93.9)86.9 (82.4–91.3)
Educational attainmenta
 < High school graduate71.0 (68.6–73.4)72.2 (68.3–76.0)74.9 (73.4–76.4)
 High school graduate/GED81.4 (80.3–82.6)82.4 (80.5–84.2)83.2 (82.5–83.9)
 Some college/technical school83.9 (82.5–85.3)85.3 (83.3–87.3)86.3 (85.5–87.0)
 College graduate88.1 (86.6–89.6)87.2 (84.9–89.5)89.3 (88.5–90.0)
Household incomea
 < $15,00073.3 (71.2–75.4)75.3 (72.1–78.6)75.7 (74.2–77.2)
 $15,000 to $34,99980.3 (79.1–81.6)83.0 (81.0–84.9)82.5 (81.8–83.3)
 $35,000 to $49,99987.5 (85.7–89.4)87.7 (84.9–90.4)87.6 (86.5–88.7)
 ≥ $50,00087.3 (84.0–90.5)85.9 (82.6–89.2)89.4 (88.3–90.5)
Employment statusa
 Currently employed80.5 (78.5–82.5)82.0 (79.2–84.7)85.7 (84.8–86.6)
 Homemaker or retired81.5 (80.1–82.9)82.7 (80.3–85.0)84.6 (83.6–85.5)
 Unemployed78.8 (73.4–84.2)81.8 (77.5–86.1)81.8 (78.8–84.9)
 Unable to work77.7 (74.3–81.0)80.3 (75.6–85.0)80.3 (77.6–83.0)
Number of persons in householda
 178.2 (76.1–80.3)80.3 (77.4–83.2)83.3 (82.5–84.0)
 284.2 (83.1–85.4)84.7 (82.8–86.7)86.4 (85.8–87.0)
 ≥ 377.5 (75.4–79.6)84.5 (82.0–87.0)83.1 (82.0–84.3)
General health statusa
 Good to excellent82.0 (81.2–82.8)82.9 (81.6–84.3)85.2 (84.7–85.6)
 Fair or poor77.2 (75.1–79.3)80.4 (77.1–83.7)80.7 (79.4–82.0)
Saw a physician in the past yeara
 Yes89.0 (88.3–89.7)90.8 (89.9–91.7)90.1 (89.8–90.5)
 No58.8 (57.1–60.6)57.9 (55.0–60.7)65.8 (64.6–67.0)
Any health insurance coveragea
 Yes83.8 (83.1–84.5)84.6 (83.4–85.8)86.4 (86.0–86.8)
 No68.0 (64.8–71.2)77.0 (73.9–80.1)68.9 (66.7–71.1)
Current cigarette smokinga
 Yes75.9 (74.1–77.7)79.8 (76.7–82.9)80.9 (79.8–81.9)
 No82.5 (81.7–83.3)83.4 (82.1–84.7)85.3 (84.8–85.7)
Current alcohol consumptiona
 Yes85.9 (84.3–87.5)85.6 (83.2–88.1)87.6 (86.5–88.6)
 No79.1 (77.8–80.5)81.7 (79.7–83.7)82.4 (81.3–83.4)
Clinical breast exama
 Ever87.2 (86.6–87.9)88.0 (86.9–89.1)89.6 (89.3–90.0)
 Never46.8 (44.0–49.7)48.0 (44.1–52.0)54.2 (52.6–55.9)
Clinical breast exam in past two yearsa
 Yes95.3 (94.9–95.8)96.3 (95.6–96.9)95.3 (95.0–95.6)
 No44.3 (42.5–46.1)42.3 (39.3–45.2)49.5 (48.3–50.6)
Mammogramab
 Ever86.9 (86.0–87.8)88.3 (87.0–89.7)88.4 (87.9–89.0)
 Never39.4 (36.9–41.8)38.8 (34.5–43.1)44.4 (42.4–46.3)
Mammogram in past 2 yearsa
 Yes94.0 (93.3–94.8)95.3 (94.4–96.2)94.2 (93.7–94.6)
 No45.7 (43.7–47.7)44.8 (41.3–48.3)48.2 (46.8–49.6)

Mammography and Clinical Breast Examination

About 81.5% (95% CI = 80.7% to 82.2%) of rural women aged 40 years or older reported that they had received a mammogram at least once, and 66.7% (95% CI = 65.8% to 67.6%) had received a mammogram in the past two years, after adjusting for age and year of survey (results not shown). In contrast, after adjusting for age and year of survey, almost 87.3% (95% CI = 86.9% to 87.7%) of women aged 40 years or older residing in metropolitan areas reported that they had received at least one mammogram, and 75.4% (95% CI = 74.9% to 75.9%) had received a mammogram in the past two years (results not shown).

For strata-specific mammogram rates (adjusted for age and calendar year) presented in Table 3, the strata correspond to rural/nonrural residence and selected demographic characteristics, medical history, and cancer screening practices. Having had a mammogram in the past two years was associated with age, race/ethnicity, marital status, higher education level, higher household income, number of persons in household, employment status, better general health status, having seen a physician in the past year, health insurance coverage, lack of cigarette smoking, and use of other cancer screening tests (Table 3). A similar pattern was observed for strata specific clinical breast examination rates (Table 4).

Table 3. Percentage of Rural and Nonrural Women in the United States Aged 40 or Older Who Had Received a Mammogram in the Past 2 Years, According to Selected Demographic Characteristics, Medical History, and Cancer Screening Practices
 Rural % (95% CI)Suburban % (95% CI)Metropolitan % (95% CI)
  • CI: confidence interval.

  • a

    P < 0.001.

  • b

    Excludes women who had had a hysterectomy. Weighted population estimates adjusted for age and year of survey; women who responded don't know or not sure or who refused are excluded.

Agea
 40 to 49 years61.1 (59.4–62.9)63.4 (60.6–66.2)68.6 (67.7–69.5)
 50 to 69 years73.2 (72.0–74.5)75.9 (73.7–78.0)81.0 (80.3–81.7)
 ≥ 70 years62.3 (60.5–64.1)71.4 (68.6–74.1)74.2 (73.2–75.3)
Race/ethnicitya
 White67.4 (66.4–68.3)72.1 (70.5–73.6)75.9 (75.4–76.4)
 Black62.7 (59.3–66.2)68.1 (63.0–73.2)75.6 (74.0–77.2)
 Hispanic59.4 (53.7–65.0)58.0 (50.3–65.7)72.1 (69.7–74.5)
 Other70.7 (64.5–76.8)62.9 (56.4–69.4)73.8 (70.4–77.2)
Marital statusa
 Currently married69.8 (68.6–71.0)74.8 (72.9–76.7)78.9 (78.3–79.6)
 Divorced or separated62.7 (59.9–65.5)66.1 (62.3–69.9)71.3 (69.9–72.7)
 Widowed59.0 (54.7–63.3)65.6 (60.9–70.3)71.6 (69.5–73.6)
 Never married55.0 (50.1–59.8)59.6 (52.0–67.1)70.1 (67.8–72.4)
 Living as unmarried couple62.7 (53.3–72.0)68.0 (60.3–75.7)76.1 (69.5–82.8)
Educational attainmenta
 < High school graduate55.8 (53.4–58.1)57.2 (53.0–61.4)63.7 (61.8–65.5)
 High school graduate/GED66.9 (65.5–68.3)70.2 (67.9–72.6)74.0 (73.1–74.9)
 Some college/technical school70.8 (69.0–72.5)74.5 (71.9–77.2)77.2 (76.2–78.1)
 College graduate76.4 (74.4–78.5)78.0 (75.3–80.8)82.3 (81.5–83.2)
Household incomea
 < $15,00055.9 (53.5–58.3)53.8 (49.7–58.0)62.7 (60.9–64.5)
 $15,000 to $34,99963.5 (62.0–65.1)69.6 (67.1–72.2)71.6 (70.7–72.6)
 $35,000 to $49,99975.0 (72.6–77.4)79.0 (75.7–82.3)79.9 (78.6–81.2)
 ≥ $50,00078.5 (74.9–82.2)79.8 (75.9–83.6)83.1 (81.9–84.3)
Employment statusa
 Currently employed67.5 (66.1–68.8)70.0 (67.8–72.3)76.3 (75.2–77.5)
 Homemaker or retired65.2 (63.3–67.1)69.4 (66.4–72.3)74.3 (73.2–75.4)
 Unemployed63.5 (57.0–70.0)52.8 (44.2–61.5)69.3 (64.9–73.7)
 Unable to work63.2 (59.4–67.0)62.7 (57.1–68.2)70.3 (67.4–73.3)
Number of persons in householda
 163.3 (61.2–65.4)69.0 (66.2–71.8)73.6 (72.8–74.5)
 269.2 (67.7–70.6)74.7 (72.5–77.0)78.2 (77.4–78.9)
 ≥ 363.4 (60.7–66.0)68.5 (64.3–72.7)72.6 (71.2–74.0)
General health statusa
 Good to excellent67.9 (66.9–68.9)71.9 (70.2–73.5)76.4 (75.9–77.0)
 Fair or poor62.6 (60.6–64.6)66.7 (63.5–70.0)70.8 (69.5–72.1)
Saw a physician in the past yeara
 Yes75.3 (74.3–76.2)79.6 (78.1–81.0)82.3 (81.8–82.8)
 No36.5 (34.5–38.5)38.0 (34.7–41.3)47.6 (46.1–49.0)
Any health insurance coveragea
 Yes69.6 (68.7–70.5)73.7 (72.2–75.3)77.7 (77.2–78.2)
 No47.3 (43.3–51.4)49.5 (42.1–57.0)54.4 (51.3–57.6)
Current cigarette smokinga
 Yes55.7 (53.5–58.0)62.2 (58.4–65.9)66.0 (64.6–67.3)
 No69.2 (68.2–70.2)72.6 (71.0–74.2)77.5 (76.9–78.0)
Current alcohol consumptiona
 Yes71.2 (69.0–73.5)75.3 (72.0–78.5)78.7 (77.3–80.0)
 No63.5 (61.8–65.1)70.1 (67.6–72.7)71.0 (69.8–72.2)
Pap testab
 Ever66.2 (65.0–67.3)70.4 (68.5–72.4)75.4 (74.7–76.0)
 Never14.0 (10.0–17.9)23.5 (17.2–29.7)27.6 (23.8–31.3)
Pap test in past three yearsab
 Yes78.3 (77.2–79.5)81.5 (79.8–83.2)84.1 (83.5–84.6)
 No14.1 (12.3–15.9)13.6 (11.0–16.3)21.1 (19.7–22.6)
Clinical breast exama
 Ever71.6 (70.7–72.5)75.1 (73.7–76.6)79.2 (78.7–79.7)
 Never32.8 (29.9–35.7)33.7 (29.0–38.4)45.3 (43.3–47.3)
Clinical breast exam in past two yearsa
 Yes83.0 (82.1–83.8)85.0 (83.7–86.4)87.4 (87.0–87.8)
 No22.6 (21.0–24.2)23.4 (20.7–26.1)32.2 (31.0–33.5)
Table 4. Percentage of Rural and Nonrural Women in the United States Aged 40 or Older Who Had Received a Clinical Breast Examination in the Past Two Years, According to Selected Demographic Characteristics, Medical History, and Cancer Screening Practices
 Rural % (95% CI)Suburban % (95% CI)Metropolitan % (95% CI)
  • CI: confidence interval.

  • a

    P < 0.001.

  • b

    Excludes women who had had a hysterectomy. Weighted population estimates adjusted for age and year of survey; women who responded don't know or not sure or who refused are excluded.

Agea
 40 to 49 years77.0 (75.5–78.6)79.2 (76.8–81.5)80.6 (79.7–81.4)
 50 to 69 years75.8 (74.6–77.0)78.8 (76.8–80.9)80.6 (79.9–81.4)
 ≥ 70 years62.1 (60.3–63.9)70.3 (67.4–73.1)70.7 (69.6–71.8)
Race/ethnicitya
 White74.1 (73.2–74.9)78.2 (76.8–79.6)79.9 (79.4–80.4)
 Black69.7 (66.4–72.9)71.3 (65.8–76.7)77.2 (75.5–78.8)
 Hispanic59.3 (53.4–65.2)68.0 (60.3–75.7)70.2 (67.7–72.7)
 Other69.6 (64.6–74.7)66.0 (59.4–72.6)68.0 (64.0–72.0)
Marital statusa
 Currently married75.8 (74.7–76.9)80.4 (78.6–82.2)80.6 (79.9–81.3)
 Divorced or separated69.8 (67.1–72.5)73.2 (69.4–76.9)75.9 (74.5–77.2)
 Widowed68.0 (64.0–71.9)67.5 (62.9–72.1)74.5 (72.6–76.4)
 Never married60.7 (56.0–65.5)69.9 (62.7–77.2)71.6 (69.2–74.0)
 Living as unmarried couple79.5 (71.7–87.3)83.0 (77.5–88.4)73.6 (66.3–80.9)
Educational attainmenta
 < High school graduate61.2 (58.8–63.6)66.0 (62.0–70.1)63.7 (61.9–65.6)
 High school graduate/GED73.1 (71.8–74.4)75.2 (73.1–77.4)76.6 (75.7–77.4)
 Some college/technical school77.2 (75.6–78.9)81.3 (79.1–83.6)81.3 (80.4–82.2)
 College graduate82.8 (80.8–84.7)84.3 (81.7–86.8)85.1 (84.2–86.1)
Household incomea
 < $15,00061.3 (58.9–63.7)61.3 (57.2–65.5)64.6 (62.8–66.5)
 $15,000 to $34,99971.8 (70.3–73.2)74.9 (72.4–77.3)74.2 (73.2–75.2)
 $35,000 to $49,99980.3 (77.9–82.8)84.3 (81.0–87.7)83.5 (82.2–84.8)
 ≥ $50,00083.4 (79.8–87.0)85.7 (82.3–89.2)86.7 (85.5–87.8)
Employment statusa
 Currently employed71.7 (69.4–74.0)78.1 (75.0–81.3)80.0 (78.8–81.2)
 Homemaker or retired71.4 (69.7–73.1)74.9 (72.0–77.8)77.7 (76.6–78.7)
 Unemployed72.8 (67.1–78.5)66.3 (60.0–72.7)71.9 (67.4–76.4)
 Unable to work64.9 (61.1–68.7)72.5 (67.0–78.1)72.0 (69.0–74.9)
Number of persons in householda
 169.4 (67.3–71.6)72.7 (69.8–75.5)77.2 (76.4–78.0)
 276.5 (75.3–77.7)79.7 (77.6–81.8)80.5 (79.8–81.2)
 ≥ 367.9 (65.2–70.5)76.9 (73.1–80.7)75.9 (74.4–77.3)
General health statusa
 Good to excellent74.6 (73.6–75.6)77.8 (76.2–79.3)79.3 (78.7–79.8)
 Fair or poor67.5 (65.6–69.5)74.2 (71.3–77.1)73.5 (72.2–74.8)
Saw a physician in the past yeara
 Yes81.5 (80.7–82.3)85.5 (84.3–86.7)84.9 (84.4–85.3)
 No43.0 (40.9–45.0)45.2 (41.7–48.7)51.3 (49.8–52.7)
Any health insurance coveragea
 Yes75.5 (74.7–76.4)79.1 (77.7–80.5)80.6 (80.1–81.0)
 No54.8 (50.5–59.0)60.7 (54.2–67.3)56.0 (52.7–59.3)
Current cigarette smokinga
 Yes64.4 (62.2–66.6)68.8 (65.0–72.5)72.0 (70.6–73.3)
 No75.0 (74.1–75.9)78.4 (77.0–79.9)79.7 (79.1–80.2)
Current alcohol consumptiona
 Yes78.3 (76.2–80.4)82.6 (79.8–85.3)82.8 (81.5–84.2)
 No70.8 (69.2–72.4)74.9 (72.4–77.3)75.6 (74.4–76.8)
Pap testab
 Ever73.7 (72.5–74.8)77.1 (75.3–78.9)79.1 (78.4–79.7)
 Never17.3 (12.8–21.8)22.9 (16.0–29.8)21.3 (18.3–24.2)
Pap test in past three yearsab
 Yes86.5 (85.4–87.5)89.0 (87.5–90.4)87.7 (87.2–88.3)
 No16.7 (14.9–18.5)16.2 (12.9–19.5)21.0 (19.6–22.5)
Mammograma
 Ever81.0 (80.1–81.8)84.4 (83.2–85.6)83.6 (83.2–84.1)
 Never36.3 (34.2–38.4)35.8 (32.1–39.6)38.1 (36.4–39.8)
Mammogram in past two yearsa
 Yes90.6 (89.9–91.3)92.3 (91.3–93.3)90.5 (90.1–91.0)
 No36.7 (35.2–38.3)38.5 (35.6–41.5)38.3 (37.1–39.4)

Multivariate Analysis of Correlates of Screening

A multivariate analysis was conducted to determine whether associations with rural residence persisted after adjusting for a variety of variables. Using multivariate analysis, several factors were associated with recent mammography, although confidence intervals sometimes overlap (Table 5). For example, an increased level of urbanization was associated with having had a mammogram in the past two years (P < 0.001 from test for trend). Rural women were less likely to have had a recent mammogram than were women living in either suburban or metropolitan areas. Metropolitan women also were likelier to have had a recent mammogram than were suburban women. Having seen a physician in the past year, a higher education level, health insurance coverage, good or excellent health status, fewer than three persons living in the household, and being currently married were all associated with having had a recent mammogram (P < 0.001 in each instance). Age (P < 0.001) and race/ethnicity (P = 0.001) were also significant correlates of mammography use; women of other races were least likely to have had a recent mammogram. The effect of age was bimodal; women aged 50 to 69 years were more likely to have had a recent mammogram. In a separate model that included variables for other cancer screening tests (results not shown), having had a clinical breast examination in the past two years (adjusted odds ratio [OR] = 6.45, 95% CI 5.89–7.07) and having had a Pap test in the past three years (adjusted OR = 6.70, 95% CI 6.02–7.45) were both strongly associated with having had a recent mammogram. To assess whether rurality modified associations with other variables, interaction terms for race/ethnicity by rural/ nonrural residence, education by rurality, having seen a physician by rural/nonrural residence, insurance status by rurality, and age by rural/nonrural residence were added to the initial model (Table 6). Both the race/ethnicity by rural/nonrural residence interaction and having seen a physician by rural/nonrural residence interaction were statistically significant (P < 0.05 in each instance). The results indicated that with the exception of women of other races, women in metropolitan areas were more likely than women in rural areas to have had a recent mammogram, and the association with rural/nonrural residence was stronger among black and Hispanic women compared with white women (Table 6). The association between mammogram use and living in a suburban area was only observed among women who had seen a physician in the past year.

Table 5. Multivariate Results for Having Had a Mammogram in the Past Two Years Among Women Aged 40 or Older in Rural and Nonrural Areas of the United States
 Adjusted odds ratio95% CI
  • CI: confidence interval.

  • a

    P < 0.001 from Wald chi square test.

  • b

    P < 0.001 from test for trend. Women who responded don't know or not sure or who refused to answer are excluded.

Survey yeara
 19981.00
 19991.121.06, 1.17
Agea
 40 to 49 years1.00
 50 to 69 years1.761.65, 1.89
 ≥ 70 years1.080.98, 1.19
Race/ethnicitya
 White1.00
 Black1.111.01, 1.22
 Hispanic1.191.05, 1.36
 Other0.830.70, 0.99
Area of residenceab
 Rural1.00
 Suburban1.191.08, 1.30
 Metropolitan1.461.38, 1.54
Marital statusa
 Currently married1.00
 Divorced or separated0.810.75, 0.88
 Widowed0.690.63, 0.75
 Never married0.600.54, 0.68
 Living as unmarried couple0.940.71, 1.23
Educational attainmentab
 < High school graduate1.00
 High school graduate/GED1.481.37, 1.60
 Some college/technical school1.741.60, 1.89
 College graduate2.192.00, 2.40
Number of persons in householdab
 3+1.00
 21.311.23, 1.40
 11.221.12, 1.33
Employment status
 Currently employed1.00
 Homemaker or retired0.970.91, 1.04
 Unemployed0.950.81, 1.11
 Unable to work1.020.89, 1.16
General health statusa
 Good to excellent1.181.10, 1.26
 Fair or Poor1.00
Current cigarette smokera
 Yes0.670.63, 0.71
 No1.00
Saw a physician within the past yeara
 Yes5.134.84, 5.44
 No1.00
Any health insurance coveragea
 Yes2.111.93, 2.31
 No1.00
Table 6. Multivariate Results for Having Had a Mammogram in the Past Two Years, Showing Significant Effect Modifications by Rurality
  Adjusted odds ratio95% CI
  • Women who responded don't know or not sure or who refused to answer are excluded. Other factors controlled for in the model include survey year, age, marital status, education, number of persons in household, employment status, general health status, current cigarette smoking, having seen a physician in the past year, and health insurance status.

  • a

    P < 0.05.

Race/ethnicityaArea of residence
 WhiteRural1.00
Suburban1.211.10, 1.33
Metropolitan1.421.34, 1.51
 BlackRural1.00
Suburban1.200.87, 1.66
Metropolitan1.721.42, 2.09
HispanicRural1.00
Suburban1.160.75, 1.79
Metropolitan2.041.48, 2.80
 OtherRural1.00
Suburban0.670.38, 1.19
Metropolitan0.990.65, 1.50
Saw a physician within the past yeara
 YesRural1.00
Suburban1.251.13, 1.40
Metropolitan1.441.35, 1.53
 NoRural1.00
Suburban1.020.86, 1.22
Metropolitan1.511.35, 1.69

Using multivariate analysis, numerous factors were found to be associated with having had a recent Pap test (Table 7). An increased level of urbanization was associated with having had a Pap test in the past three years (P < 0.001 from test for trend), but the association was not as strong as that observed with mammography (Table 5). Rural women were less likely to have had a recent Pap test than were metropolitan women. Having seen a physician in the past year, a higher education level, health insurance coverage, good or excellent health status, fewer than three persons in household, and being currently married were all associated with having had a recent Pap test (P < 0.001 in each instance). Age (P < 0.001) and race/ethnicity (P < 0.001) were also significant correlates of Pap test use. Women of other races were least likely to have had a recent Pap test. Women aged 30 to 39 years were more likely to have had a recent Pap test. In a separate model that included variables for other cancer screening tests (results not shown), having had a clinical breast examination in the past two years (adjusted OR = 13.45, 95% CI 12.48–14.50) and having had a mammogram in the past two years (adjusted OR = 4.57, 95% CI 4.09–5.12) were both strongly associated with receipt of a recent Pap test among women who were at least 40 years old. Interaction terms for race/ethnicity by rural/nonrural residence, education by rural/nonrural residence, having seen a physician by rural/nonrural residence, insurance status by rural/nonrural residence, and age by rural/nonrural residence were added to the model (Table 8). Both the age by rural/nonrural residence interaction and having seen a physician by rural/nonrural residence interaction were statistically significant (P < 0.05 in each instance) as shown in Table 8. No other significant effect modifications by rurality were observed. The results suggest that the association between Pap test use and increasing urbanization occurred only among women who were at least 30 years old. Among women aged 18 to 29 years, an inverse association with increasing urbanization was observed. The association between Pap test use and living in a metropolitan area was clearly indicated only among women who had not seen a physician in the past year (Table 8).

Table 7. Multivariate Results for Having Had a Pap Test in the Last Three Years Among Women Aged 18 or Older in Rural and Nonrural Regions of the United States
 Adjusted odds ratio95% CI
  • CI: confidence interval.

  • a

    P < 0.05 from Wald chi square test.

  • b

    P < 0.001 from Wald chi square test.

  • c

    P < 0.001 from test for trend. Women who responded don't know or not sure or who refused to answer are excluded.

Survey yeara
 19981.00
 19991.091.03, 1.16
Ageb
 18 to 29 years1.00
 30 to 39 years1.050.96, 1.15
 40 to 49 years0.700.64, 0.77
 50 to 69 years0.440.40, 0.49
 ≥ 70 years0.180.15, 0.20
Race/ethnicityb
 White1.00
 Black1.571.41, 1.76
 Hispanic1.080.97, 1.21
 Other0.400.34, 0.47
Area of residencebc
 Rural1.00
 Suburban1.080.97, 1.21
 Metropolitan1.181.10, 1.26
Marital statusb
 Currently married1.00
 Divorced or separated0.760.69, 0.83
 Widowed0.540.49, 0.61
 Never married0.280.26, 0.31
 Living as unmarried couple0.840.69, 1.02
Educational attainmentbc
 < High school graduate1.00
 High school graduate/GED1.541.41, 1.68
 Some college/technical school1.931.75, 2.13
 College Graduate2.732.45, 3.05
Number of persons in householdbc
 3+1.00
 21.171.08, 1.26
 11.281.16, 1.40
Employment statusa
 Currently employed1.00
 Homemaker or retired0.920.85, 1.00
 Unemployed1.191.02, 1.38
 Unable to work0.910.76, 1.09
General health statusb
 Good to excellent1.201.10, 1.31
 Fair or Poor1.00
Current cigarette smoker
 Yes1.040.97, 1.11
 No1.00
Saw a physician within the past yearb
 Yes5.355.02, 5.69
 No1.00
Any health insurance coverageb
 Yes1.661.53, 1.80
 No1.00
Table 8. Multivariate Results for Having Had a Pap Test in the Past Three Years, Showing Significant Effect Modifications by Rurality
  Adjusted odds ratio95% CI
  • CI: confidence interval.

  • a

    P < 0.05. Women who responded don't know or not sure or who refused to answer are excluded. Other factors controlled for in the model include survey year, race/ethnicity, marital status, education, number of persons in household, employment status, general health status, current cigarette smoking, having seen a physician in the past year, and health insurance status.

AgeaArea of residence
 18 to 29Rural1.00
Suburban0.850.63, 1.15
Metropolitan0.750.63, 0.88
 30 to 44Rural1.00
Suburban1.090.90, 1.32
Metropolitan1.401.24, 1.59
 45 to 59Rural1.00
Suburban1.080.85, 1.36
Metropolitan1.351.18, 1.56
 60+Rural1.00
Suburban1.191.00, 1.42
Metropolitan1.231.10, 1.37
Saw a physician within the past yeara
 YesRural1.00
Suburban1.120.96, 1.31
Metropolitan1.050.96, 1.14
 NoRural1.00
Suburban1.030.86, 1.23
Metropolitan1.351.22, 1.49

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The current study contributed new information and built on the work of previous authors 5, 14–18 by examining various associations with cancer screening across three categories of residence (rural and small town, suburban and smaller metropolitan, and larger metropolitan areas) rather than only two (metropolitan versus nonmetropolitan) and by reviewing associations with rural/nonrural residence across more than two racial/ ethnic groups, with more recent data. According to analysis from the 1985 National Health Interview Survey, Duelberg 5 found that women living in urban areas may be more likely than women living in other areas to have had a recent Pap test because, on average, women who obtain Pap tests on a regular basis have higher levels of education and income; however, income level does not explain the effect of urban residence on obtaining a mammogram. 5 In addition, race (i.e., black versus white) modified the effect of urban/rural residence on receiving a Pap test. 5 Zhang et al. reported similar findings based on results from the 1994 National Health Interview Survey. 14

The current analysis provided important information about breast and cervical carcinoma screening rates among minority women in rural and nonrural areas of the United States. Age-adjusted breast and cervical carcinoma screening rates were found to be relatively low among Hispanic women and among nonrural women who were neither white nor black. According to multivariate analysis, Hispanic women were found to have been screened as frequently as non-Hispanic white women, after adjusting for area of residence, education, whether a woman had seen a physician in the past year, and other factors. In addition, an apparent interactive effect with associations between mammography and race/ethnicity was only observed among women residing in metropolitan areas. 7 Future studies should examine cancer screening rates among American Indian and Alaska Native women and among Asian and Pacific Islander women by rural/ nonrural residence.

The greater use of preventive services by urban residents may be explained by the greater availability of medical services in urban areas. 5, 19 Women living in rural areas may have limited access to health care practitioners and, consequently, to fewer preventive health care services. 19 Physician recommendation, which is known to positively influence mammogram utilization, 20–24 has been reported to vary significantly between urban and rural women. 1 In the current study, the observed association between patient use of mammography and having seen a physician in the past year was more pronounced in suburban areas. In addition, the association between having a Pap test and having seen a physician in the past year was less pronounced in metropolitan areas than in rural or suburban areas. Previous researchers have noted that geographic accessibility of mammography facilities may also be a barrier in some geographic areas. 25

An important methodologic issue in the current study was the optimal approach to defining geographic area of residence. Classifications focusing exclusively on rural and metropolitan areas pose some limitations to health services research. For example, the rise of suburban areas in the United States, an important demographic shift that has occurred during the past 50 years, is not taken into account in studies that use such classifications. 26 We defined geographic area of residence using U.S. Department of Agriculture codes, which allowed us to categorize the study respondents into three broad areas of residence (i.e., rural areas and small towns; suburban areas and smaller metropolitan areas; and larger metropolitan areas). Although the use of Beale codes is an accepted classification methodology, rural/nonrural residence based on population sizes of geopolitical units may be defined inconsistently. 27 The geographic size of counties varies widely. In some large counties, many residents may live in urban areas but large numbers of women may still live in rural areas of those same counties. Rural populations also may exist within the boundaries of metropolitan areas and urban areas may overlap geopolitical boundaries and extend into areas classified as rural or nonmetropolitan. 27 Consequently, misclassifications may have partly obscured the observed differences in cancer screening test utilization in the current study.

Results from this nationwide, state-based survey may not be generalizable to selected rural populations such as farm communities, to rural areas of Appalachia or of the deep South, to rural areas that have sizeable American Indian and Alaska Native populations, or to rural areas that have sizeable Hispanic populations, such as those in the San Joaquin Valley of California. All rural populations are not equivalent and rural status does not always signify lower levels of health care access or use. 19 Rural areas vary in terms of both population density and socioeconomic structure. 26

With respect to other limitations of the current study, response bias is a possibility because the telephone survey excluded women living in households without telephones, and, among women with household telephones, numerous potential respondents did not participate (40.9% to 44.8%). 10 According to study findings, women who lacked household telephones were more likely to have lower incomes or to live in rural areas. Finally, self-reported information about cancer screening practices may differ from information obtained from records of health-care providers. Validation studies have suggested that patients tend to over-report their use of screening and under-report the time lapse since their last screening. 28–30 Some of the statistically significant differences in screening observed in the present study were small, which may reflect the large sample size. Because BRFSS data are anonymous and have not been individually linked to cancer registry data across the United States, the current study did not explore how screening practices may affect the incidence of breast and cervical carcinoma or result in a stage shift.

These results underscore the need for continued efforts to provide both breast and cervical carcinoma screening to medically underserved women throughout the United States, including rural communities. Successful approaches to increase breast carcinoma screening among women in rural communities include community education interventions and low cost mobile mammography van services. 31–33 Community- based, culturally sensitive educational programs have been shown to be particularly effective in increasing breast and cervical carcinoma screening among African American women and Hispanic women in rural communities. 34–40

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES