Trends in Self-Reported Use of Mammograms (1989-1997) and Papanicolaou Tests (1991-1997) -- Behavioral Risk Factor Surveillance System

Donald K. Blackman, Ph.D.¹
Eddas M. Bennett, M.B.A., M.P.H.²
Daniel S. Miller, M.D., M.P.H.¹

1Division of Cancer Prevention and Control
²Division of Adult and Community Health
National Center for Chronic Disease Prevention and Health Promotion

Problem/Condition: In 1999, an estimated 175,000 women will be diagnosed with breast cancer, and 43,300 will die from the disease. In the same year, an estimated 12,800 women will be diagnosed with invasive cervical cancer, and 4,800 will die from it. Early detection and timely treatment of breast cancer and cervical dysplasia can alter the progress of and reduce mortality from these diseases.

Reporting Period Covered: 1989-1997 for breast cancer screening and 1991-1997 for cervical cancer screening.

Description of System: The Behavioral Risk Factor Surveillance System is a state-based telephone survey of the civilian, noninstitutionalized adult population (i.e., persons aged greater than or equal to 18 years). In this report, responses for women aged greater than or equal to 40 years are included for measures of breast cancer screening, and responses for women aged greater than or equal to 18 years with an intact uterine cervix are included for measures of cervical cancer screening.

Results: The percentage of women aged greater than or equal to 40 years who reported ever participating in breast cancer screening and the proportion who had participated within the previous 2 years increased during 1989-1997. The percentage of women aged greater than or equal to 18 years who reported ever participating in cervical cancer screening and the proportion who had participated within the previous 2 years were stable during 1991-1997. For both types of screening, substantially fewer women had received screening within the previous 2 years than had ever been screened.

Interpretation: These findings may indicate that some women who participate in initial screening do not seek further screening.

Actions Taken: Initiatives to encourage women to receive initial screening should continue, but additional initiatives specifically aimed at promoting rescreening should be developed. Continued surveillance of the percentage of women who receive regular screening will help public health officials evaluate breast and cervical cancer prevention programs.

In 1999, an estimated 175,000 women will be diagnosed with breast cancer, and 43,300 will die from the disease. In the same year, an estimated 12,800 women will be diagnosed with invasive cervical cancer, and 4,800 will die from it (1). Early detection and timely treatment of breast cancer and cervical dysplasia can alter the progress of and reduce mortality from these diseases (2). This report summarizes trends in the self-reported use of mammograms, the principal procedure for breast cancer screening, and Papanicolaou (Pap) tests, the principal procedure for cervical cancer screening, as tracked by the Behavioral Risk Factor Surveillance System (BRFSS).

In the United States, the incidence of breast cancer increased 25.3% (whites: 26.2%; blacks: 36.7%) from 1973 through 1996 (3). Most of the increase occurred during 1973-1991; during 1992-1996, the overall incidence was stable. From 1973 through 1996, the age-adjusted mortality rate from invasive breast cancer for all women declined from 26.9 to 24.3 per 100,000 women (3).

Decreases in breast cancer mortality can be partially attributed to earlier disease detection and treatment due to greater use of screening, which has been proven effective for detecting breast cancer in early stages (4). Findings from clinical trials evaluating the efficacy of screening mammograms have indicated that deaths from cancer could be reduced by 19%-30% if guidelines for regular breast cancer screening were followed (2,5,6).

From 1973 through 1996, the incidence of invasive cervical cancer decreased 44.0% (whites: 44.5%; blacks: 60.2%) in the United States (3). During this period, mortality due to this disease also decreased, by 47.3% (whites: 45.7%; blacks: 57.3%). The decline in mortality appears to be part of a long-term trend. During 1950-1959, the age-adjusted mortality from invasive cervical cancer per 100,000 U.S. women was 9.1 among whites and 21.2 among nonwhites (7). In 1996, mortality was 2.4 among whites and 5.2 among blacks (3).

Much of the reduction in cervical cancer mortality has been associated with increased use of Pap tests. The effectiveness of this test for reducing cervical cancer mortality has not been formally evaluated in experimental or randomized trials. However, accumulated evidence indicates that Pap tests can detect asymptomatic precancerous lesions (i.e., dysplasia) and preinvasive lesions, which may progress to invasive cervical cancer if untreated (8). Detection and treatment of precancerous and preinvasive lesions can reduce the risk for developing invasive cervical cancer (9). Further, early detection and treatment of invasive cervical cancer may improve the prognosis for women diagnosed with this disease. From 1986 through 1992, the 5-year survival for women diagnosed with localized invasive cervical cancer was 91%; for women diagnosed with distant disease, 5-year survival was approximately 9% (7). Conclusions based on improved survival must be evaluated cautiously because of possible lead time or length bias (8). Nevertheless, an estimated 37%-60% reduction in cervical cancer mortality could be achieved with regular screening for all women (2).

The state-based BRFSS monitors self-reported health behaviors of U.S. adults (i.e., persons aged greater than or equal to 18 years). In this report, trends in BRFSS data on self-reported receipt of breast cancer screening (1989-1997) among women aged greater than or equal to 40 years and cervical cancer screening (1991-1997) among women aged greater than or equal to 18 years with an intact uterine cervix are presented. The report includes state-specific estimates and estimates across participating states by age, race, ethnicity, annual household income, education, and (beginning in 1991) insurance status. Only data from the 38 states that participated continuously in the BRFSS from 1989 through 1997 are included in this report (Figure 1).

The use, history, and objectives of the BRFSS have been previously described (10).

Every month, using random-digit-dialing telephone survey techniques, each state health department selected a probability sample of its civilian, noninstitutionalized adult population living in households with telephones (10,11). Most states used multistage cluster sampling based on the Waksberg method (12); other states used simple random, stratified random, or other sampling designs.

From 1989 through 1997, each adult female respondent was asked whether she had ever had a mammogram. A respondent who answered positively was then asked how long it had been since her last mammogram and whether that mammogram had been performed as part of a routine checkup, because of a breast problem other than cancer, or because she already had breast cancer. Beginning in 1990, each woman was also asked whether she had ever received a clinical breast examination (CBE). If the answer was "yes," the respondent was then asked how long it had been since her last CBE and whether it had been performed as part of a routine checkup, because of a breast problem other than cancer, or because she already had breast cancer. In this report, only women aged greater than or equal to 40 years were included in the analyses of mammogram and CBE use.

From 1991 through 1997, each adult female respondent was asked whether she had ever received a Pap test. Each woman who reported receiving the test was then asked how long it had been since her last test and whether it had been performed as part of a routine checkup or for another reason. In addition, each woman was asked whether she had had a hysterectomy. All women aged greater than or equal to 18 years who answered "no" (i.e., had an intact uterine cervix) were included in the analyses of Pap test use.

BRFSS is designed to provide state-specific prevalence estimates. These estimates are weighted to the age, sex, and race (i.e., white and nonwhite) distribution of each state's adult population by using the most current census or intercensal estimates. Estimates are also weighted to reflect each respondent's probability of selection.

Estimates for mammography use and Pap test use were adjusted to the age distribution of women in the 1989 BRFSS sample for participating states. Time trends were evaluated with logistic regression by using models in which the outcome was regressed on the respondent's age and the interview year. A time trend was considered statistically significant if the beta coefficient for year was non-zero at p less than or equal to 0.01. SUDAAN, a statistical package for analyzing complex survey data, was used to calculate the standard errors for the prevalence estimates (13).

The number of women who participated in the BRFSS increased from 37,009 in 1989 to 62,278 in 1997. The median response rate ranged from 77% to 84% across the 9 years. Most respondents were aged greater than or equal to 40 years, were white, were non-Hispanic, had an annual household income of $10,000-$50,000, had graduated from high school, had health-care insurance, and had not undergone a hysterectomy (Table 1).

The total age-adjusted proportion of women aged greater than or equal to 40 years who reported ever having a mammogram increased from 63.9% in 1989 to 84.8% in 1997 (p for trend less than 0.01) (Figure 2). This proportion increased in each of the participating states, and the range across the 38 participating states shifted from 52.7%-73.9% in 1989 to 74.5%-89.2% in 1997 (Table 2). The age-adjusted proportion of all women who reported that their most recent mammogram was for screening (i.e., part of a routine checkup) was 53.1% in 1989 and 76.9% in 1997 (p for trend less than 0.01). The proportion increased in each state, and the range across states shifted from 42.1%-65.3% in 1989 to 67.4%-85.0% in 1997. In 1989, an age-adjusted total of 54.3% of women reported receiving a mammogram within the previous 2 years; in 1997, 71.3% did so (p for trend less than 0.01). This proportion increased in each state, and the range across states shifted from 43.8%-64.5% in 1989 to 60.8%-77.9% in 1997.

For all three measures of mammography use -- ever (Table 3), as part of a routine checkup (Table 4), and within the past 2 years (Table 5) -- there was an increasing trend over the 9 survey years among all sociodemographic subgroups (p for trend less than 0.01) except for American Indians and Alaska Natives. These trends are generally characterized by substantial increases from year to year in the first 6-7 years and modest or no increases for the final 2-3 years.

Differences in mammography use between sociodemographic subgroups were similar for the three measures (Tables 3-5). From 1989 through 1997, women aged 50-69 years were almost always more likely to report having received a mammogram than were the youngest or the oldest women. Women aged greater than or equal to 70 years were consistently least likely to have received a mammogram within the past 2 years during 1989-1995. In 1996 and 1997, however, women aged 40-49 years were the least likely to have received a timely mammogram. From 1989 through 1997, mammography use was almost always lower among Hispanic women than non-Hispanic women. Reported differences between white and black women were minimal for all years, and in 1996 and 1997, the proportions reporting having had a mammography were about equal. For all 9 years, mammography use was lowest at the lowest levels of annual household income and education and increased as income and education increased. Women without health-care insurance were consistently less likely than those with insurance to have received mammograms. The proportion of uninsured women who reported receiving a mammogram within the previous 2 years did not substantially increase until 1996.

The questions addressing CBE were added to the BRFSS in 1990. The age-adjusted proportion of women aged greater than or equal to 40 years who reported ever having both a mammogram and a CBE rose each year, from 65.2% in 1990 to 79.5% in 1997 (p for trend less than 0.01) (Figure 3) (Table 6). The proportion of women who reported that both tests were part of a routine examination also rose each year, from 55.0% in 1990 to 70.3% in 1997 (p for trend less than 0.01). The proportion who reported they received both a mammogram and a CBE within the past 2 years was 65.2% in 1997, an increase from the 55.0% who reported having both procedures in 1990 (p for trend less than 0.01).

In each year from 1991 through 1997, 91%-93% of women aged greater than or equal to 18 years with an intact uterine cervix reported ever having had a Pap test (Figure 4). The age-adjusted proportions among the states ranged from 86.6% to 95.1% in 1991 and from 81.8% to 96.8% in 1997 (Table 7). In 1991, less than 90% of women in five states reported ever having had this screening procedure, and in 1997, less than 90% of women in only one state did so. For most of the sociodemographic subgroups, the age-adjusted proportions of women who ever received a Pap test were high and changed minimally from 1991 through 1997 (Table 8). In 1997, greater than 90% of women in most subgroups reported ever receiving a Pap test; the exceptions were women aged greater than or equal to 70 years, Asian American and Pacific Islander women, women of "other" races, Hispanic women, women with an annual household income of less than $10,000, women with less than 12 years of education, and women without health-care insurance.

Over the 7 survey years, greater than or equal to 77% of the respondents reported having received a Pap test within the past 2 years (Figure 4). The age-adjusted proportions among the states ranged from 73.2% to 85.0% in 1991 and from 72.4% to 87.2% in 1997 (Table 7). Overall, and for most subgroups, the proportion of women who received a Pap test in the past 2 years was stable over the survey period (Table 9). From 1991 through 1997, women aged less than 60 years were more likely than older women to report having received a Pap test in the past 2 years. Black women were slightly more likely than white women to have received a recent Pap test; both blacks and whites were consistently more likely than women of "other" races to report having had a recent Pap test. Hispanic women were less likely than non-Hispanic women to have received timely tests. In each year, the likelihood of having had a timely test generally increased with annual household income and with education, and women without health-care insurance were substantially less likely than were women with insurance to have received a timely Pap test. Analyses of trends revealed a minimal but statistically significant increase during 1991-1997 in the overall proportion of women having a timely Pap test. However, most subgroups did not demonstrate a substantial change during these 7 years. Even for subgroups for which significant trends were found (i.e., women aged 50-69 years, white women, black women, non-Hispanic women, women with an annual household income of less than $10,000 or $25,000-$50,000, and insured women), the differences between the highest and lowest values were less that 5 percentage points and did not represent substantial change.

Screening for and early detection of breast and cervical cancers are most effective if they are performed for each woman at regular intervals. Yet for both types of screening, the proportion of women who were ever screened and the proportion who were screened within the previous 2 years differed substantially. This difference may indicate that some women who participate in initial screening do not continue to be screened at regular intervals. It may be that the full benefits of breast and cervical cancer screening have not been achieved in the United States.

BRFSS data are consistent with other survey findings that breast cancer screening has increased over the past decade. For example, the proportion of women aged greater than or equal to 50 years participating in the National Health Interview Survey who reported ever having a mammogram increased from 37% in 1987 to 67% in 1992, and the proportion who reported receiving a mammogram within the previous 3 years increased from 23% in 1987 to 49% in 1992 (14). In a separate report from the same survey, the proportion of women aged greater than or equal to 50 years who reported having had both a mammogram and a CBE within the preceding 2 years increased from 25% in 1987 to 51% in 1992 (15). From 1990 through 1995, the proportion of women aged greater than or equal to 40 years who reported regular breast cancer screening as recommended by the American Cancer Society increased from 31% to 47% (16). Despite these substantial gains in use of breast cancer screening, its use continues to be low among several subgroups, including women with low income, less education, and no health-care insurance (17).

Several professional organizations have endorsed guidelines for breast cancer screening. All the guidelines recommend periodic mammograms and CBEs but differ on recommended frequency and age to begin breast cancer screening (6). The U.S. Preventive Services Task Force recommends a screening mammogram, with or without an annual CBE, every 1-2 years for women aged 50-69 years (5). Other women (e.g., those aged less than 50 years who are at high risk for breast cancer) might also be recommended for screening after consultation with their physicians. The American Cancer Society recommends an annual screening mammogram with a concurrent CBE for women aged greater than or equal to 40 years (18). The American Medical Association recommends an annual or biennial screening mammogram and an annual CBE for women aged 40-49 years and an annual mammogram with CBE for women aged greater than or equal to 50 years (19).

Healthy People 2000 objective 16.11 is to "increase to at least 80 percent the proportion of women aged 40 and older who have ever received a clinical breast examination and a mammogram, and to at least 60 percent those aged 50 and older who have received them within the preceding 1 to 2 years" (20). The BRFSS data in this report indicate encouraging increases in the proportions of all women who reported ever having had a mammogram, having their last mammogram as part of a routine checkup, and having a mammogram within the previous 2 years.

The BRFSS findings for use of Pap tests are consistent with results from the 1987 and 1990 National Health Interview Surveys. In the latter survey, the proportion of women who reported ever having a Pap test (approximately 90%) and the proportion who reported having it within the previous 3 years (approximately 75%) did not change substantially from 1987 through 1990 (14,15,21). In both the BRFSS and the National Health Interview Surveys, Hispanic women, women with less than a high school education, and women with the lowest household income were generally less likely than their counterparts to report having received a Pap test. Women without health-care insurance have also been found to be less likely than women with insurance to receive Pap tests (17).

The U.S. Preventive Services Task Force recommends Pap tests for all women beginning when they become sexually active (but no later than age 18 years) and then every 3 years for women at normal risk for cervical cancer; the interval may be shorter for women at high risk for the disease (5). Screening for cervical cancer may be discontinued after age 65 years for women who have had consistently normal findings on previous examinations, and screening is not recommended for women who have had their uterine cervix removed unless the hysterectomy was part of treatment for cancer. The National Cancer Institute, the American Cancer Society, the American College of Obstetricians and Gynecologists, and the American Medical Association endorse annual Pap tests for women who are sexually active or have reached age 18 years (5).

Healthy People 2000 objective 16.12 is to "increase to at least 95 percent the proportion of women aged 18 and older with uterine cervix who have ever received a Pap test, and to at least 85 percent those who received a Pap test within the preceding 1 to 3 years" (20). In the current report, the findings that 93% of women reported having ever received a Pap test and that 80% reported having had a Pap test within the previous 2 years suggest that this objective is achievable.

Some part of the gap between the proportions who have ever had a Pap test and those who have had one within the past 2 years may be attributed to the discontinuation of testing among women aged greater than or equal to 65 years who had a history of regular screening and whose tests results were consistently normal (5). Among women aged less than 65 years, however, the difference most likely results from failure to have Pap tests at regular intervals.

Healthy People 2000: National Health Promotion and Disease Prevention Objectives, which was published in 1990 by the U.S. Public Health Service, includes goals for increasing the use and timeliness of breast and cervical cancer screening procedures among all American women and among specific groups of women (e.g., those aged greater than 70 years, blacks, Hispanics, women with low income, and those with less than a high school education) (20). Several Federal service initiatives for breast and cervical cancer screening were developed concurrently. These initiatives underscored the growing national appreciation of the importance of breast and cervical cancer screening.

The Breast and Cervical Cancer Mortality Prevention Act of 1990 mandated a nationwide program to increase access of medically underserved women to comprehensive breast and cervical cancer screening services (22). The National Breast and Cervical Cancer Early Detection Program (NBCCEDP), which was established as a result of the mandate, is administered by CDC. NBCCEDP is a program of cooperative agreements with state health agencies, the District of Columbia, American Indian and Alaska Native (AIAN) programs, and U.S. territories. By 1998, 50 states, the District of Columbia, 13 AIANs, and 4 U.S. territories had implemented NBCCEDP comprehensive screening programs. NBCCEDP gives them resources to provide screening, follow-up, and referral services to medically underserved women; to disseminate information to health-care professionals and the general public about detecting and controlling breast and cervical cancer; and to evaluate program activities and the quality of screening procedures. Among medically underserved persons, NBCCEDP identified several high-priority groups: women aged greater than or equal to 50 years, women of racial or ethnic minority groups, women with low income, and women without health-care insurance.

Federal medical insurance programs have enabled increased use of breast and cervical cancer screening programs. Since 1991, Medicare has provided insurance coverage for screening mammograms and Pap tests (23,24). These were among the first preventive services covered by Medicare (25). A requirement for participation in the NBCCEDP is that the Medicaid program serving the state (including the District of Columbia), AIAN program, or territory provide coverage for screening mammograms, CBEs, Pap tests, and pelvic examinations.

The BRFSS has several limitations. First, estimates of behavioral risk factors are based on self-reports, which may not agree with reports based on other sources (e.g., medical, laboratory, and imaging center records) (26,27). Second, the BRFSS does not include in the sampling frame persons who do not have telephones. Approximately 5% of U.S. households do not have a telephone (28). Because the geographic and demographic distributions of households with and without telephones differ (28), the trends observed in BRFSS may not reflect trends for households without telephones. Third, approximately 20% of eligible respondents refused to participate, which introduces a potential source of bias. Fourth, because the BRFSS sample reflects the population distribution of participating states, the sample may include only minimal numbers of participants in sociodemographic subgroups of particular interest (e.g., Asian Americans or Pacific Islanders). Estimates for these subgroups are accurate, but they are less precise than estimates for subgroups with larger numbers of respondents.

Not all states have participated in the BRFSS since its inception, and multiyear studies can include data only from states that participated in each year of the study. Twelve states and the District of Columbia did not participate each year from 1989 through 1997 and could not be included in this report. These exclusions may limit the generalizability of these findings to the Nation as a whole.

These BRFSS results reflect the progress the United States has made toward increasing the proportion of women who have participated in breast cancer screening and illustrate the success the United States has had in maintaining the consistently high proportion of women who have participated in cervical cancer screening. These results also indicate that older women, women with a low annual household income, those with a low level of education, and those without health-care insurance are less likely to participate in breast and cervical cancer screening. National goals should emphasize maintaining screening levels among subgroups of women most likely to participate in screening as well as increasing screening levels among subgroups of women who are less likely to participate in screening.

Initiatives such as the NBCCEDP, which encourage women to participate in initial screening, should continue. But the full benefits of screening on morbidity and mortality due to breast and cervical cancers can be achieved only if a substantial proportion of U.S. women receive screening examinations at regular intervals (29,30). The BRFSS findings suggest that national efforts should now aim to preserve current levels of initial cancer screening while emphasizing repeat screening. Additional initiatives specifically to promote rescreening should be developed. Continued surveillance of trends in screening timeliness will help public health officials target and evaluate breast and cervical cancer prevention programs.

References

1. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1999. CA Cancer J Clin 1999;49:8-31.
2. Brownson RC, Reif JS, Alavanja MCR, Bal DG. Cancer. In: Brownson RC, Remington PL, Davis JR, eds. Chronic disease epidemiology and control. Washington, DC: American Public Health Association, 1993:137-67.
3. Ries LAG, Kosary CL, Hankey BF, Miller BA, Edwards BK, eds. SEER cancer statisitics review, 1973-1996. Bethesda, MD: National Cancer Institute, 1999.
4. CDC. Breast cancer incidence and mortality -- United States, 1992. MMWR 1996;45:833-7.
5. US Preventive Services Task Force. Guide to clinical preventive services. 2nd ed. Baltimore: Williams & Wilkins, 1996.
6. Fletcher SW, Black W, Harris R, Rimer BK, Shapiro S. Report of the International Workshop on Screening for Breast Cancer. J Natl Cancer Inst 1993;85:1644-56.
7. Riggan WB, Van Bruggen J, Acquavella JF, Beaubier J, Mason TJ. U.S. cancer mortality rates and trends, 1950-1979. Vol. 2. Research Triangle Park, NC: US Environmental Protection Agency and the National Cancer Institute, 1983; EPA-600/1-83-015b.
8. Cramer DW. Uterine cervix. In: Schottenfeld D, Fraumeni JF Jr, eds. Cancer epidemiology and prevention. Philadelphia: W.B. Saunders, 1982:881-900.
9. Kiviat N. Natural history of cervical neoplasia: overview and update. Am J Obstet Gynecol 1996;175:1099-104.
10. Frazier EL, Franks AL, Sanderson LM. Behavioral risk factor data. In: Using chronic disease data: a handbook for public health practitioners. Atlanta: US Department of Health and Human Services, Public Health Service, CDC, 1992:4-1-4-17.
11. Nelson DE. Validity of self reported data on injury prevention behavior: lessons from observational and self reported surveys of safety belt use in the US. Inj Prev 1996;2:67-9.
12. Waksberg J. Sampling methods for random digit dialing. J Am Stat Assoc 1978;73:40-6.
13. Shah BV, Barnwell BG, Bieler GS. SUDAAN: software for the statistical analysis of correlated data. User's manual, release 6.40. Research Triangle Park, NC: Research Triangle Institute, 1995.
14. Anderson LM, May DS. Has the use of cervical, breast, and colorectal cancer screening increased in the United States? Am J Public Health 1995;85:840-2.
15. National Center for Health Statistics. Healthy people 2000 review, 1993. Hyattsville, MD: US Department of Health and Human Services, Public Health Service, CDC, 1994.
16. Horton JA, Cruess DF, Romans MC. Compliance with mammography screening guidelines: 1995 mammography attitudes and usage study report. Womens Health Issues 1996;6:239-45.
17. Bindman AB, Grumbach K, Osmond D, Vranizan K, Stewart AL. Primary care and receipt of preventive services. J Gen Intern Med 1996;11:269-76.
18. Leitch AM, Dodd GD, Costanza M, et al. American Cancer Society guidelines for the early detection of breast cancer: update 1997. CA Cancer J Clin 1997;47:150-3.
19. American Medical Association Council on Scientific Affairs. Mammographic screening in asymptomatic women aged 40 years and older. JAMA 1989;261:2535-42.
20. Public Health Service. Healthy people 2000: national health promotion and disease prevention objectives -- full report, with commentary. Washington, DC: US Department of Health and Human Services, Public Health Service, 1991; DHHS publication no. (PHS)91-50212.
21. Ackermann SP, Brackbill RM, Bewerse BA, Cheal NE, Sanderson LM. Cancer screening behaviors among U.S. women: breast cancer, 1987-1989, and cervical cancer, 1988-1989. In: CDC surveillance summaries (April 24). MMWR 1992;41(No. SS-2):17-34.
22. Henson RM, Wyatt SW, Lee NC. The National Breast and Cervical Cancer Early Detection Program: a comprehensive public health response to two major health issues for women. J Public Health Manage Pract 1996;2:36-47.
23. General Accounting Office. Screening mammography: higher Medicare payments could increase costs without increasing use. Washington, DC: General Accounting Office, 1993; GAO/HRD-93-50.
24. Power EJ. Pap smears, elderly women, and Medicare. Cancer Invest 1993;11:164-8.
25. Schauffler HH. Disease prevention policy under Medicare: a historical and political analysis. Am J Prev Med 1993;9:71-7.
26. Crane LA, Kaplan CP, Bastani R, Scrimshaw SCM. Determinants of adherence among health department patients referred for a mammogram. Women Health 1996;24:43-64.
27. Zapka JG, Bigelow C, Hurley T, et al. Mammography use among sociodemographically diverse women: the accuracy of self-report. Am J Public Health 1996;86:1016-21.
28. Lavrakas PJ. Telephone survey methods: sampling, selection, and supervision. 2nd ed. Newbury Park, CA: Sage Publications, 1993. (Applied Social Research Methods Series. Vol. 7).
29. Taylor VM, Taplin SH, Urban N, White E, Peacock S. Repeat mammography use among women ages 50-75. Cancer Epidemiol Biomarkers Prev 1995;4:409-13.
30. White E, Urban N, Taylor V. Mammography utilization, public health impact, and cost-effectiveness in the United States. Annu Rev Public Health 1993;14:605-33.