This paper examines the status of mammography screening and factors related to its utilization among women ages 40 to 74. Responses from 5,548 women aged 40-74 who participated in the 1992 National Health Interview Survey Cancer Control Supplement were studied. Bivariate and multiple logistic regression analyses were performed to estimate the odds ratios of mammography use as a function of demographic and other predictor variables. The data indicated that 40 percent of women aged 40-74 reported ever having had a mammography. High income, having a regular source of health care, individuals' preventive orientation, and age (50-64) were strong predictors of mammography use. Low educational attainment, being widowed, residence in a rural area, and living in the South were found to be strongly associated with less likelihood of having had a mammography. Despite the anecdotal and empirical evidence that the proportion of women ever having had mammography has dramatically increased in the past several years, our analyses suggest that mammography utilization varies greatly across levels of predictive variables considered in this study. In order to achieve and maintain high rates of mammography utilization, public health programs should especially target those women with low incomes and education and women with no regular source of medical care.

Breast cancer is the most commonly diagnosed cancer and the second leading cause of cancer death among women in the United States. (American Cancer Society, 1995) An estimated 184,200 new cases were expected to be diagnosed in 2000. About 40,800 U.S. women were expected to die from breast cancer in the same year, accounting for about 15.2 percent of cancer deaths among women. (Landis, 2000)

Substantial reduction in breast cancer death rates is possible if the tumor is discovered at an early stage. Mammography is the most effective method for detecting these early malignancies. There is controversy regarding the ages at which women should undergo mammography and necessary frequency of such testing. The American Cancer Society and ten other national organizations issued a consensus statement about breast cancer screening guidelines in 1989. This consensus statement supports mammographic screening every one to two years between the ages of 40 to 49 years with annual screening beginning at age 50. In contrast, the American College of Physicians and the United State Preventive Task Force maintain that regular screening should be initiated at age 50, and mammography should only be considered among younger women at increased risk of breast cancer. (White, Urban, & Taylor, 1993) More recently, the American College of Obstetricians and Gynecologists (ACOG) and the National Cancer Institute recommended that a women in her forties with no risk factors have a mammogram at least once every two years and yearly after the age of 50. (ACOG, 1997) American Cancer Society, however, continues to recommend annual mammograms for a woman in her forties. (ACS, 1997)

During the past decade, several studies have documented the effectiveness of screening mammography in the reduction of mortality from breast cancer among women aged 40 to 74. (White, Urban, & Taylor, 1993) However, limited evidence exist to suggest a mortality benefit attributable to screening older women, particularly among those over 75. (Morrison, 1989) The estimated range for the reduction in mortality for screened women aged 40-74 years is given to be 5-76 percent. (White, Urban, & Taylor, 1993) The results of these studies, improvements in technique, lower cost, and more accessible screening facilities have resulted in substantial increase in the number of women screened with mammography.

While a 1977 sample indicated only 15 percent of women ever having had a mammography, (Reeder, Berkanovic, & Marcus, 1980) the proportion of women having had a mammography increased to 41 percent in 1983. (Howard, 1987) Similarly, in a 1986 national sample, 39 percent of women ages 50 and over reported having ever had a mammography. (Hayward, Shapiro, Freeman, & Corey, 1988) The 1987 Cancer Control Supplement to National Health Interview Survey found 38 percent of women ages 40 and older having had a mammography. Finally, the 1990 National Health Interview Survey of Health Promotion and Disease Prevention (NHIS-HPDP) reported 57.7 percent of women ages 40-75 having had a mammography. (Rakowski, Rimer, Bryant, 1993).

Despite the anecdotal and empirical evidence that the proportion of women ever having had mammography has dramatically increased in the past several years, it has been documented by some of the previous work on this subject that the likelihood of women using mammography screening varies greatly across levels of basic demographic factors and other potential predictive variables. (Calle, Flanders, Thun & Martin, 1993) (Zapka, Stoddare, Costanza & Greene, 1989) A recent analysis of data from the 1990 NHIS investigated the likelihood of women 40 to 75 years of age having regular mammography. The findings of this study suggested the presence of strong associations between the use of regular mammography and higher income, having a regular source of health care, living in the west or a metropolitan area, and engaging in healthy practices. (Rakowski, Rimer & Bryant, 1993)

The purpose of this paper is to examine the status of mammography screening experience and factors related to its utilization among women ages 40 to 74 using more recent data from a nationwide survey of US women.

The data for this study were drawn from the public release tape of the 1992 National Health Interview Survey Cancer Control Supplement. The National Health Interview Survey (NHIS) is a cross-sectional, household interview survey of the civilian, noninstitutionalized population of the United States conducted annually by the National Center for Health Statistics (NCHI). The primary purpose of the NHIS is to obtain information about the amount and distribution of illness, its effects in terms of disability and chronic impairments, and the kinds of health services people receive using a multistage probability sampling design. The survey consists of a set of basic health and demographic questions, and some other supplemental questions on specific health topics. Variables unique to the Cancer Control Supplement included items on cancer knowledge and attitudes, cancer screening knowledge and practice, and other preventive practices related to cancer.

For the present study, we restricted analyses of mammography use to 5,548 women ages 40 to 74. After excluding the cases with cancer history, the sample size was reduced to 4,980. Finally, bivariate and multivariate analyses are based on the final sample size of 2,618 excluding all the cases with missing information on the dependent variable.

Logistic regression analyses were performed using "ever having had a mammography" as the dependent variable for women 40 to 74 years of age. The dependent variable was set equal to 1 if the woman reported ever having had a mammography and set equal to 0 otherwise. Four categories of potential independent variables associated with mammography use were identified based on the literature review on this subject and included in this study: demographic characteristics, health status, resources, and individuals' preventive orientation.

Five demographic variables were age (40-49, 50-64, and 65-74), marital status (married, widowed, divorced/separated, and never married), ethnicity (white, black, and other), region of the country (northeast, midwest, south, and west), and type of residence (MSA-central city, MSA-other, and Non-MSA). Health status measure represented the values of self-rated health by the respondents and coded as a multilevel variable (excellent/very good, good, and fair/poor). Resource variables characterized both the family and community resources and included income (less than $15,000, $15,000-24,999, $25,000-49,999, and $50,000 or more), education (0-8 years, some high school, high school degree, some college, college graduate, and post-college), employment status (in the work force; yes/no), and regular source of care (yes/no). Finally, variable "ever having had a breast physical exam" (yes/no) was included to reflect the individuals' preventive orientation.

Of the 4,980 respondents 40 percent reported ever having had a mammography. Distribution of study variables are summarized in Table 1. Simple and multiple logistic regression were used to evaluate the individual and the simultaneous effects of all predictor variables on the likelihood of ever having had a mammography. Crude and adjusted odds ratios (ORs) and 95 percent confidence intervals (CIs) were calculated using the SPSS Advanced Statistics version 10.0.

Table 2 presents the results of the logistic regression analyses. Percentage of women reported having had a mammography, bivariate and multivariate adjusted odds ratio associations with all predictor variables and the dependent variable, and 95 percent CIs were listed in this table. Odds ratios were obtained by exponentiating the estimated logistic regression coefficients. The multivariate model performed relatively well, classifying 77 percent of the cases correctly. The model chi-square was equal to 238.01 and it was significant at p<0.001.

In the bivariate analyses, age (65-74), income, marital status, education (except the high school graduate category), living in the South, being in the labor force, residence in a rural area, health status (fair/poor), having a regular source of care, and having had a breast physical exam all significantly predicted mammogram utilization. Eight variables remained significantly associated with use of mammography in the multiple logistic regression model after simultaneously controlling for all explanatory variables.

Income strongly predicted mammography use in all categories included in the analysis. Women in the highest income category ($50,000 or more) were nearly two times more likely to use mammography than women in the lowest income category (adjusted OR= 1.87, 95 percent CI= 1.23, 2.85). Having a regular source of care and having had a breast physical exam both remained strong predictors; women who have a regular source of care were almost two and a half times more likely to use mammography than those who do not have (adjusted OR= 2.42, 95 percent CI= 1.78, 3.30) and women who had have breast physical exam were five times more likely to use mammography than those women without it (adjusted OR= 5.29, 95 percent CI= 3.31, 8.46). Similarly women aged 50-64 and those with post-college education were more likely to have mammography screening in the past when compared with women in the reference groups (40-49 years old and 0-8 years of education respectively). Finally, living in the South, residence in a rural area, and being widowed were all associated with less likelihood of using mammography in the multivariate model.

In general, the results of the logistic regression analyses were consistent with the earlier work. However, the lack of influence of health status and race on mammography use in the multivariate model was rather unexpected. Those in poor health were as likely to use mammography as were healthier women, suggesting that current perceived health is not related to a decision to use mammography. This is, however, consistent with the finding reported in a study by Mor, et al. (Mor, Pacala, & Rakowski, 1992) which concluded that "a woman's health status is unrelated to having had a screening mammography" after investigating the associations between a number of predictive variables and the use of screening mammography in older women.

Evidence regarding the importance of race on mammography utilization are mixed in the literature. For example, Calle, et al. (Calle, Flanders, Thun, & Martin, 1993) found Hispanic ethnicity and other race to be strong predictors of mammography underuse based on the data from the 1987 National Health Interview Survey Cancer Control Supplement. Another study by Breen and Kessler (Breen & Kessler, 1994), on the other hand, demonstrated that the importance of race as a potential predictor of mammography screening has declined in 1990 using data from the 1987 and 1990 National Health Interview Surveys. The lack of evidence to find the race variable significant in the present study provides some support to the finding of the second study and may suggest that the racial differences among women are less likely to play a key role in the tendency of women to use mammography. Finally, none of the two-way interaction terms between age, race, and income were found to be statistically significant when they were included in the multivariate logistic model.

In this paper, we have examined the role of several potential predictor variables (demographic, economic, health status, and individuals' preventive orientation) in the use of mammography among women aged 40-74 based on the data from a nationwide survey of U.S. women. Overall, in the study population, 40 percent of women reported ever having had a mammography. Despite encouraging recent increases in mammography utilization, the present data suggest that the likelihood of women using mammography screening varies greatly across levels of predictive variables considered in this study.

High income, having a regular source of health care, individuals' preventive orientation, and age (50-64) were strong predictors of mammography use. On the other hand, low educational attainment, being widowed, residence in a rural area, and living in the South were found to be strongly associated with less likelihood of having had a mammography. A woman's health status and race were unrelated to ever having had a mammography.

Although multivariate analysis did not indicate any significant differences in the use of mammography by women aged 65-74 compared to women aged 40-49, the results of the bivariate analysis suggested that women in the first category (65-74) were significantly less likely to have had mammography, a finding consistent with the an earlier analysis of data from a 1987 national sample. This finding is of particular importance given almost half of all breast cancers are diagnosed in women 65 and older and points to the fact that this age group was still underserved in 1992. In the present study, mammography use was more likely to be reported by women 50-64. On the other hand, the finding that the role of racial differences in the use of mammography has declined in recent years is encouraging given the significant associations reported earlier between race and the mammography use. (Calle, Flanders, Thun, & Martin, 1993)

Another important finding of this study is that women who did not have a regular source of care were less likely to utilize mammography than women who reported having a regular source of care. Although disparities in the use of preventive services were found, even in the presence of universal insurance coverage, (Katz & Hofer, 1994) having access to a regular source of care is an important enabling factor in improving both the availability and potential use of mammography screening.

Finally, individuals' preventive orientation, measured by "ever having had a physical breast exam", was significantly associated with mammography utilization. This finding highlights the importance of developing health education programs aimed at informing women of the benefits of adopting preventive practices and encouraging their adherence to routine mammographic screening. Of course, physicians can make a significant contribution in this regard by educating their patients and recommending them to receive screening mammography on a regular basis.

In conclusion, in order to achieve and maintain high rates of mammography utilization, public health programs should especially target those women with low incomes and education and women with no usual source of medical care. Widowed women should also included in the target population. Likewise, health education programs to promote screening mammography must encourage the active support and participation of physicians and other health care providers in reaching women with low education attainment and preventive orientation. Such efforts are critical in achieving the Healthy People 2000's ambitious goal for the 1990s-to have at least 60 percent of women ages 50 and older receive a screening mammography and a clinical breast examination during a 2-year period. (Public Health Services, 1990)

American Cancer Society. Cancer Facts and Figures, 1995. Atlanta: American Cancer Society, 1995.

American Cancer Society. Mammography guidelines for asymptomatic women. Atlanta, GA: American Cancer Society, 1997.

The American College of Obstetricians and Gynecologists [ACOG]. Press statement on mammography screening guidelines, 1997.

www.acog.com/from_home/publications/press_releases/nr-mammos.htm.

Breen, N., Kessler, L. Changes in the use of screening mammography: Evidence from the 1987 and 1990 National Health Interview Surveys. American Journal of Public Health 1994; 84: 62-67.

Calle, E., Flanders, WD., Thun, MJ., Martin, LM. Demographic predictors of mammography and pap smear screening in US women. American Journal of Public Health 1993; 83: 53-60.

Hankey, BF. Age distribution of breast cancer cases. J Natl Cancer Inst 1993, 85: 892-7.

Hayward, AA., Shapiro, MF., Freeman, HE., Corey, CR.. Who gets screened for cervical and breast cancer? Results from a new national survey. Arch Intern Med 1988; 148: 1177-1181.

Howard, J. Using mammography for cancer control: An unrealized potential. CA Cancer J Clin 1987; 37: 33-48.

Katz, SJ., Hofer, TP.: Socioeconomic disparities in preventive care persist despite universal coverage. JAMA 1994; 272: 530-4.

Landis, S.H., Murray, T., Bolden, S. et al. Cancer Statistics, 2000. CA-A Cancer Journal for Clinicians, 2000. 50(1): 2398-2424.

Mor, V., Pacala, JT., Rakowski, W. Mammography for older women: Who uses, who benefits? The Journals of Gerontology 1992; 47(special issue): 43-49.

Morrison, A. Review of evidence on the early detection and treatment of breast cancer. Cancer 1989; 64: 2651-6.

Public Health Service: Healthy People 2000: National health promotion and disease prevention objectives. DHHS Publication no. (PHS) 91-50212, U.S. Government Printing Office, Washington, DC, 1990.

Rakowski, W., Rimer, BK., Bryant, SA. Integrating behavior and intention regarding mammography by respondents in the 1990 National Health Interview Survey of Health Promotion and Disease Prevention. Public Health Reports 1993; 108: 605-624.

Reeder, S., Berkanovic, E., Marcus, AC. Breast cancer detection behavior among urban women. Public Health Reports 1980; 95: 276-81.

White, E., Urban, N., Taylor, V. Mammography utilization, public health impact, and cost-effectiveness in the United States. Annual Review of Public Health 1993; 14: 605-33.

Zapka, JG., Stoddard, AM., Costanza, ME., Greene, HL. Breast cancer screening by mammography: Utilization and associated factors. American Journal of Public Health 1989; 79: 1499-1502.

Demographic Resources

Age: Income:

40-49 years 1,993 40 Less than $15,000 1,000 24.5

50-64 years 1,929 38.7 $15,000-$24,999 778 19

65-74 years 1,058 21.2 $25,000-$49,999 1,311 32.1

$50,000 or more 997 24.4

Marital Status:

Married 2,976 59.9 Education:

Widowed 585 11.8 0-8 years 580 11.7

Divorced, Some high school 631 12.7

Separated 1,028 20.7 High school gradate 1,818 36.6

Never married 383 7.7 Some College 927 18.7 College graduate 513 10.3

Ethnicity:

White 4,147 83.3 In labor force:

Black 688 13.8 Yes 3,061 61.9

Other 145 2.9 No 1,885 38.1

Region of the country: Regular source of care:

Northeast 1,040 20.9 Yes 4,144 83.6

Midwest 1,246 25 No 809 16.4

South 1,612 32.4

West 1,082 21.7

Preventive orientation

Area of residence: Ever had breast physical exam:

MSA-Central Yes 2,593 93

City 1,663 33.4 No 194 7

MSA- Other 2,144 43.1

Non MSA 1,173 23.6

Health Status Ever having had a mammography

Self-rated health: Yes 1,962 39.4

Excellent/ No 656 13.2

Very good 2,800 56.3 Missing Cases 2,362 47.4

Good 1,317 26.5

Fair/poor 855 17.2

Note: Some of the totals (N) do not add to 4,980 because of the missing values.

Demographic

Age:

40-49 years 1,007 75 Reference Reference Reference Reference

50-64 years 1,018 77.5 1.12 .91, 1.38 1.39 1.09, 1.79

65-74 years 593 70 .75 .59, .94 1.28 .91, 1.80

Marital Status:

Married 1,439 78 Reference Reference Reference Reference

Widowed 457 67 .55 .44, .69 .65 .47, .91

Div./separated 555 74 .78 .62, .98 .89 .66, 1.20

Never married 167 71 .66 .46, .95 .64 .40, 1.01

Ethnicity:

White 2,171 76 Reference Reference Reference Reference

Black 374 72 .83 .64, 1.06 1.09 .78, 1.52

Other 73 71 .80 .48, 1.35 .66 .35, 1.24

Region of the country:

Northeast 531 78 Reference Reference Reference Reference

Midwest 665 76 .87 .66, 1.13 .94 .68, 1.29

South 866 69 .62 .49, .80 .67 .49, .90

West 556 79 1.07 .80, 1.43 1.09 .77, 1.52

Area of residence:

MSA-Central

City 913 77 Reference Reference Reference Reference

MSA- Other 1,099 77 1.0 .81, 1.23 .77 .59, 1.00

Non MSA 606 69 .67 .53, .84 .60 .45, .94

Health Status

Self-rated health:

Excellent/

very good 1,438 77 Reference Reference Reference Reference

Good 723 74 .83 .67, 1.02 .99 .77, 1.27

Fair/poor 453 70 .71 .56, .90 .98 .71, 1.36

Resources

Income:

< $15,000 570 62 Reference Reference Reference Reference

$15,000-$24,999 430 75 1.84 1.39, 2.44 1.54 1.13, 2.09

$25,000-$49,99 696 78 2.18 1.70, 2.80 1.43 1.03, 1.99

$50,000 or more 472 85 3.44 2.55, 4.65 1.87 1.23, 2.85

Education:

0-8 years 266 71 Reference Reference Reference Reference

Some high school 333 62 .67 .47, .95 .46 .30, .70

High school

graduate 1,050 73 1.10 .81, 1.48 .70 .48, 1.02

Some college 509 79 1.53 1.09, 2.15 .87 .56, 1.34

College graduate 238 84 2.21 1.43, 3.44 1.23 .71, 2.14

Post-college 217 90 3.80 2.26, 6.40 1.93 1.02, 3.67

In labor force

Yes 1,450 78 1.39 1.16, 1.66 1.04 .80, 1.34

No 1,149 71 Reference Reference Reference Reference

Regular source of care

Yes 2,300 77 2.41 1.88, 3.10 2.42 1.78, 3.30

No 289 58 Reference Reference Reference Reference

Preventive orientation

Ever had breast physical exam

Yes 2,475 77 7.45 5.05, 11.01 5.29 3.31, 8.46

No 125 31 Reference Reference Reference Reference

Note: OR= Odds ratio; CI= Confidence interval; MSA= Metropolitan statistical area.

Odds ratios significant at p< .05 are in boldface.

* Adjusted for all other variables listed.

Model Chi-square= 238.01 at p<0.001.