Factors Associated with Interval Adherence to Mammography Screening in a Population-Based Sample of New Hampshire Women

Transcription

1 219 Factors Associated with Interval Adherence to Mammography Screening in a Population-Based Sample of New Hampshire Women Patricia A. Carney, Ph.D. Beth G. Harwood, M.Ed. Julie E. Weiss, M.S. M. Scottie Eliassen, M.S. Martha E. Goodrich, B.S. Department of Community & Family Medicine, Dartmouth Medical School, Lebanon, New Hampshire. BACKGROUND. Interval adherence to mammography screening continues to be lower than experts advise. The authors evaluated, using a population-based mammography registry, factors associated with adherence to recommended mammography screening intervals. METHODS. The authors identified and recruited 625 women aged 50 years and older who did and did not adhere to interval mammography screening. Demographic and risk characteristics were ascertained from the registry and were supplemented with responses on a mailed survey to assess knowledge, perceived risk, anxiety regarding breast carcinoma and its detection, and women s experiences with mammography. RESULTS. The authors found no differences in risk factors or psychologic profiles between adhering and nonadhering women. Women who did not adhere had a statistically higher body mass index than women who did adhere (27.6 versus 26.1, P 0.003). Exploration of mammographic experiences by group found that care taken by technologists in performing or talking women through the exam was higher in adhering women than nonadhering women (75.6% vs 65.71% for performing the exam, and 71.6% vs 60.8% for talking patients through the exam, respectively, P 0.05). CONCLUSIONS. The authors found that previous negative mammographic experiences, particularly those involving mammography technologists, appear to influence interval adherence to screening and that patient body size may be an important factor in this negative experience. Cancer 2002;95: American Cancer Society. DOI /cncr Supported by the American Cancer Society (CRTG CCE), the U.S. Department of Defense (DASD17-94-J-4109), and the National Cancer Institute (U01 CA ). The authors thank Deirdre O Mahony for database design, development, and extraction of data for these analyses and Bridgett Luce for administrative support. Address for reprints: Patricia A. Carney, Ph.D., Dartmouth Medical School 1, Medical Center Dr., HB 7925, Lebanon, NH 03756; Fax: (603) ; Patricia.A.Carney@Darmouth.edu Received January 14, 2002; revision received February 13, 2002; accepted February 15, KEYWORDS: breast carcinoma screening, early detection, mammography screening, adherence. Breast carcinoma is a significant problem in New Hampshire (NH). It is the leading cancer in NH women, with over 900 new cases diagnosed each year, representing 33% of female cancers statewide. 1 The state s breast carcinoma incidence rates (109 per 100,000) are comparable to those of the U.S. overall (108.2 per 100,000). 1 However, the breast carcinoma mortality rate is 30.4 per 100,000, which is substantially higher than the national rate of 24.1 per 100, Thus, early breast carcinoma detection may not be optimal in this region. Evidence that routine screening has a significant impact on mortality reduction is mounting. 3 However, interval adherence to mammography screening continues to be problematic. Research indicates that only 40 70% of women aged 50 years and older routinely undergo mammography screening, 4 7 leaving significant room for im American Cancer Society

2 220 CANCER July 15, 2002 / Volume 95 / Number 2 provement. The majority of these reports relied on patient self-report, which may not be as accurate as clinically documented mammographic events. Previous studies on mammography screening reveal that patient knowledge about risk, income status, race, a well-established relationship with a primary care provider, anxiety, and embarrassment are all associated with interval adherence The national annual incidence of breast carcinoma increases as women age, with 48% of new breast carcinoma cases and 56% of breast carcinoma deaths occurring in those aged 65 years and older. 15 Additional risk factors include family history in a first degree relative, 16 previous personal history of invasive or in situ breast carcinoma or atypical ductal hyperplasia, 17,18 late age at first birth or nulliparity, exposure to high-dose radiation, 19 and long term use of estrogen replacement therapy Other studies indicate that breast carcinoma mortality increases with advanced stage of disease at diagnosis, usually resulting from delay in breast carcinoma detection. 26,27 Characteristics of patients as they relate to interval screening adherence have been studied, but those studies did not use clinical documentation of mammographic encounters to link adherence to patient characteristics and risk factors. In addition, detailed information about women s experiences with previous mammography should be included in the analysis when comparisons are made between those women who do and do not adhere to interval screening. Using our population-based mammography registry (which contains clinically documented mammographic encounters) to identify women aged 50 years and older who did and did not adhere to interval screening, we found an adherence rate of 32%. This is lower than has been previously reported. Through the registry, we obtained consenting women s demographic characteristics and health risk factors and supplemented this information with data from a survey we developed, tested, and administered to the study cohort to ascertain information on knowledge, perceived risk, anxiety, and experience with previous mammography. Here we report on factors that we found to be associated with interval adherence to screening in NH. METHODS Recruitment, Data Collection, and Study Measures The development and design of the NH Mammography Network (NHMN) is described in detail elsewhere. 28 Briefly, all women who have a mammogram in a participating NH mammography facility are eligible to enroll. Enrollment includes completing a baseline demographic and epidemiologic survey; participating in an in-person interview with a mammography technologist; consenting to release medical records, including pathology information, for linkage to radiology data; and agreeing to future contact for research purposes. The NHMN data collection instruments include three forms: a general information form, a patient intake form, and a radiologist interpretation form. The general information form is self-administered and covers demographic, risk, and health history information. The patient intake form, which is administered verbally by a mammography technologist, obtains information on breast concerns, history of previous breast procedures (e.g., breast reduction, implants, biopsy, surgery, radiation therapy), personal or family history of breast carcinoma, menopausal status, and current hormone replacement therapy (HRT) use. The radiologist interpretation form records the date and indication for the exam, level of breast density, whether comparison films were used to interpret the exam, the assessment of the exam using the American College of Radiology Lexicon for mammogram assessment, 29 and future recommendations. During the study time period, the NHMN obtained written consent from over 90% of the women who entered a participating mammography facility. Using the NHMN registry, we identified two groups of women: those who returned for a screening mammogram within 24 months of their initial exam and those who did not. Eligible participants were consenting NH women aged 50 years and older whose first recorded mammogram recorded in the registry occurred between May 1, 1996 and December 31, We allowed 24 months to pass from the women s initial exam before classifying them as adhering or nonadhering. Women with a personal history of breast carcinoma and women whose initial screening mammograms were reported as abnormal were excluded. Using Institutional Review Board approved procedures for recruiting study subjects, eligible women were mailed an initial invitation letter outlining the study objectives and inviting them to participate. A followup mailing was sent within one month if no response was received from the first mailing. In addition to the standard NHMN information described above, we developed a survey, which was pilot-tested for this investigation, with adaptations of perceived risk questions by Black et al. 30 and Woloshin et al. 31 Our survey asked women about their knowledge of breast carcinoma screening measures, anxiety, perceived benefits of and barriers to mammography, cancer helplessness, perceived risk, perceived dread of a breast carcinoma diagnosis as compared to other diseases, and past experiences with mammography.

3 Mammography Screening Adherence/Carney et al. 221 To evaluate anxiety, we used the Spielberger State-Trait Anxiety Inventory (form-y). 32 This wellestablished 40 item scale assesses both state and trait anxiety. The state scale was designed to assess a palpable reaction or process that takes place at a given time and level of intensity (such as anxiety related to getting a mammogram). The trait scale is intended to assess a predisposition towards anxiety or the extent to which someone is characteristically anxious about a condition (such as would occur in someone with a family history of breast carcinoma). Both scales are scored on a four point Likert scale. We used the benefits of and barriers to mammography utilization to measure health belief model concepts of health benefits and barriers within the context of mammography screening. 33 The benefits subscale measures perceived benefits (i.e., detecting a lump early, reducing worry, decreasing chances of dying from breast carcinoma, resolution of potential breast problems), while the mammography barriers subscale assesses perceived negative features or attributes related to obtaining a mammogram (i.e., increasing worry, embarrassment, pain, time, and cost). Each variable is scored on a five point Likert scale. Cancer specific anxiety was assessed with a cancer-related anxiety and helplessness scale. This scale is a 12 item self-report inventory that assesses general cancer anxiety and sense of hopelessness. This scale has been used in three previous studies of participants who do not have cancer and who have participated in cancer-screening or community-based cancer prevention and control programs. 34,35 The dread questions were developed and tested by Woloshin et al. In a study to determine how dread of breast carcinoma diagnosis, treatment, and living with this disease compares to other diseases (e.g. heart disease, osteoporosis, blood clots) from a patient vulnerability perspective. 31 Data Analysis Demographic and risk characteristics were evaluated between the adhering and nonadhering women using chi-square for categorical variables and an independent sample t test for continuous variables. Comparisons between adhering and nonadhering women s responses to the survey questions were made using chi-square and t tests. Alpha levels were set at 0.05, and all tests were two-sided. RESULTS The NHMN identified a total of 2243 women who met the inclusion criteria for the current study. Of these, 708 were adhering women and 1535 were nonadhering. Our recruitment efforts identified 27 women who were ineligible (5 deceased and 22 mailings undeliverable), 309 women who actively declined (184 nonadhering women and 125 adhering women), and 1281 women who were lost to followup. Reasons reported by adhering women who actively declined included no time (n 35), not interested (n 24), too ill (n 3), difficulty reading forms (n 8), other (n 44), and no answer (n 12). Reasons reported by nonadhering women who actively declined included no time (n 63), not interested (n 32), too ill (n 14), difficulty reading forms (n 9), other (n 50), and no answer (n 16). The final cohort consisted of 323 adhering and 302 nonadhering women. The participation rate associated with our recruitment efforts was 65%. Once the cohort was established, the survey instrument was mailed to each woman. Of the 625 consenting women, 539 women returned the first survey. Of the 86 women who did not return their first survey, four requested to be removed from the study, five had a recent breast carcinoma diagnosis, 24 were undeliverable, and 53 were lost to followup. To minimize the number of women lost to followup, we r ed the survey if a reply was not received within six weeks of the initial survey mailing. Among the 539 study participants, the mean age was 57 (standard deviation: 7) years. Fewer than 5% had less than a high school education and more than 30% were college graduates. The majority of participants were married (75%). None of these factors were statistically different between the two study groups. Table 1 illustrates health risks in women who did and did not adhere to interval screening. We found no differences between study groups in the number of first degree relatives with breast carcinoma, age at menarche, age at first live birth, or number of previous breast biopsies. We did find that more adhering women reported current HRT use than nonadhering women (51% vs 41% respectively, P 0.04). We also found that adhering women were more likely to be insured than nonadhering women (99% versus 95%; P 0.01), and that their body mass index (BMI) was lower than that of nonadhering women (26.1 versus 27.6, P 0.003). Our survey collected information on the participants knowledge about breast carcinoma screening measures as reported in Table 2. We found that those adhering to interval screening were more likely to identify mammography as benefiting year olds than those who did not adhere (89% versus 83% respectively; P 0.03). The two groups answered similarly to how often does a women need a mammogram in different age groups and at what age should a woman have her first mammogram. Women who adhered to interval screening were more likely to re-

4 222 CANCER July 15, 2002 / Volume 95 / Number 2 TABLE 1 Health Risks of New Hampshire Women Who Do and Do Not Adhere to Interval Mammography Screening Characteristics a n % n % Health insurance coverage Any % % 0.01 Private % % 0.87 Medicare 41 15% 49 19% 0.25 Medicaid 4 2% 2 1% 0.44 HMO/PPO 82 30% 61 23% 0.08 Other 9 3% 11 4% First degree relatives with breast carcinoma 0.06 None % % One 35 13% 18 7% Two or more 5 2% 7 3% Age at menarche years 64 24% 61 24% years % % 14 years 58 22% 50 19% Body mass index (mean/se) 26.1 (0.3) 27.6 (0.4) Current HRT use % 82 41% 0.04 Age at first full live birth years 35 13% 43 17% years % % years 74 29% 75 29% 30 years 15 6% 21 8% Number of breast biopsies 0.57 None % % Any 31 12% 34 13% HMO: health maintenance organization; PPO: preferred provider organization; SE: standard error; HRT: hormone replacement therapy; BMI: body mass index. a Five women were missing data for type of insurance (1 adherer and 4 nonadherers), 16 missing data for age at menarche (7 adherers and 9 nonadherers), 18 missing data for BMI (6 adhering and 12 nonadhering), 136 missing data for HRT use (70 adherers and 66 nonadherers) and 26 missing data for age at first full term pregnancy (15 adherers and 11 nonadherers). port that women 50 years or older need a mammogram every year (P 0.01); their mean responses only differed by 3 months when compared to nonadhering women. More than 70% of adhering and nonadhering women reported that their physician was the person who brought up the subject of mammograms when visiting the doctor. Table 3 outlines the psychological profile of adhering and nonadhering women. No differences were found between the two groups for state and trait anxiety measures or cancer attitudes and beliefs. The nonadhering women reported more barriers to mammography utilization than adhering women (P 0.01), but the groups responded similarly to mammography benefits. Responses to the dread scale indicated that for women in both groups breast carcinoma, along with heart disease and embolism, scored higher for levels of dread in diagnosis of, undergoing treatment for, living with, and dying from the disease, than did osteoporosis and uterine carcinoma. Risk perceptions (Table 4) were compared in the two study groups and both perceived their chances of being told that they have breast carcinoma within the next 10 years or will die from breast carcinoma to be small. Likewise, fewer than 25% of the women in each group thought that their risk of getting cancer in the next 10 years or the threat of breast carcinoma to their health was high. Table 5 outlines women s perceptions of the mammography experience by those who did and did not adhere to interval screening. The adhering women reported a negative experience with previous mammography (29.3%) three times more frequently than nonadhering women (8.8%, P 0.01). When we investigated details about the negativity of their mammographic experience, we learned that the primary reason for negative experiences among adhering women was difficulty getting an appointment, while the primary reason for a negative experience among nonadhering women was pain or discomfort during the pro-

5 Mammography Screening Adherence/Carney et al. 223 TABLE 2 Knowledge Differences in New Hampshire Women Who Do and Do Not Adhere to Interval Mammography Screening Knowledge variables n Mean (SD) (years) Range (years) n Mean (SD) (years) Range (years) How often do women need a mammogram when they are: 40 years old (2.0) (0 10) (1.8) (0 10) years old (1.1) (0 10) (1.1) (0 5) years old (0.2) (0 2.5) (0.6) (0.5 5) 0.01 At what age do you think the average woman should have her first mammogram? (6.7) ( ) (8.1) (1 60) 0.06 n % n % Who brought up the subject of mammograms at your last health care visit? 0.27 Doctor % % Self % % Never discussed 5 1.9% % Which age group of women benefits most from mammograms? year old women % % year old women % % year old women % % and older % % 0.23 SD: standard deviation. cedure itself. Exploration of reasons for positive experiences by group found that care taken by technologists in performing the exam or talking women through the exam was higher in adhering women than nonadhering women (75.6% vs 65.71% for performing the exam, and 71.6% vs 60.8% for talking patients through the exam; both s were 0.05). As indicated in Table 5 women in both groups received information about mammography from the technologists, and the majority of women in both groups indicated that the information was easy to understand (97.6% in adhering women versus 94.9% in nonadhering women; P 0.66, data not shown). However, adhering women were more likely to agree that the information provided by the technologist affected how they felt about getting their next mammogram (62% in adhering vs 57.8% in nonadhering women). For those women who responded to the question about planning to get another mammogram within the next two years, 100% of adhering women were planning to do this, while 96% of nonadhering women were also planning to get another mammogram within the next two years (P 0.01). DISCUSSION In the current study, we were surprised to find no statistical differences between adhering and nonadhering women in many of the demographic risk factors or in psychological profiles that have been previously reported in other studies on interval mammography adherence. were more likely to be HRT users than nonadhering women, which may indicate their connection to regular health care and subsequently screening mammography related to the use of this medication. Even though a statistical difference was found in reported health insurance between adhering and nonadhering women, more than 95% of the women in each group reported having health insurance, which raises questions about the clinical significance of this factor. The same is true for our assessment of women s intentions to obtain future mammograms. Though the question achieved statistical significance, more than 95% women in each group indicated they intended to obtain a mammogram within the next 24 months. A potentially important finding in the current study was the difference in BMI between women who do and do not adhere to

6 224 CANCER July 15, 2002 / Volume 95 / Number 2 TABLE 3 Psychological Profile of New Hampshire Women Who Do and Do Not Adhere to Interval Mammography Screening Psychological perceptions n Mean (SD) n Mean (SD) Range a Anxiety State (10.9) (11.1) (20 75) 0.85 Trait (9.6) (9.2) (20 71) 0.62 Mammography utilization Benefits (3.6) (3.8) (6 23) 0.40 Barriers (2.9) (3.8) (6 25) 0.01 Cancer attitude and beliefs Cancer anxiety (2.8) (3.0) (6 20) 0.68 Cancer hopelessness (2.9) (3.7) (11 32) 0.06 Cancer adjustment (2.8) (2.9) (5 20) 0.37 Dread Being told you have: Breast carcinoma (0.28) (0.29) (0 1) 0.27 Heart disease (0.25) (0.25) (0 1) 0.30 Osteoporosis (0.28) (0.28) (0 1) 0.73 Uterine carcinoma (0.35) (0.34) (0 1) 0.02 A blood clot (0.28) (0.29) (0 1) 0.69 Undergoing treatment for: Breast carcinoma (0.26) (0.28) (0 1) 0.16 Heart disease (0.26) (0.28) (0 1) 0.04 Osteoporosis (0.26) (0.27) (0 1) 0.92 Uterine carcinoma (0.34) (0.32) (0 1) 0.07 A blood clot (0.28) (0.30) (0 1) 0.11 Living with: Breast carcinoma (0.26) (0.28) (0.02 1) 0.28 Heart disease (0.24) (0.25) (0.02 1) 0.21 Osteoporosis (0.26) (0.26) (0 1) 0.65 Uterine carcinoma (0.32) (0.33) (0 1) 0.50 A blood clot (0.26) (0.28) (0 1) 0.08 Dying from: Breast carcinoma (0.30) (0.32) (0 1) 0.23 Heart disease (0.27) (0.29) (0.02 1) 0.29 Osteoporosis (0.30) (0.33) (0 1) 0.41 Uterine carcinoma (0.34) (0.35) (0 1) 0.47 A blood clot (0.30) (0.32) (0 1) 0.40 SD: standard deviation. a Range: anxiety, 20 less anxious; benefits, 6 strongly agree; barriers, 6 strongly agree; cancer anxiety, 5 greater anxiety; cancer hopelessness, 5 greater hopelessness; cancer adjustment, 5 good adjustment; dread, 0 don t dread at all. screening, with women who adhere having a lower BMI than those who do not. A key finding in the current study was the difference in previous mammography experiences between the adhering and nonadhering women. Our mailed survey asked women to describe their experience at their last mammogram as positive, negative, or neutral. Surprisingly, a greater number of nonadhering women (66%) responded positively compared to adhering women (50%). Only 9% of nonadhering women responded that their last mammogram was negative, compared to adhering women at 29%. However, when we investigated the details of these negative experiences, the predominant reason given by the adhering group was difficulty in getting an appointment, while pain was the most frequent response (almost 10%) for the nonadhering group. Not one adhering woman listed pain as a reason for a negative experience. In addition, 11% fewer nonadhering women than adhering women felt the technologist provided comfort in talking them through the exam. We speculate that either a perceived or an actual response from the technologist might be related to body composition of the patient. More than 60% of women in both groups felt that comfort provided by the technologist in talking them through the exam was

7 Mammography Screening Adherence/Carney et al. 225 TABLE 4 Risk Perceptions in New Hampshire Women Who Do and Do Not Adhere to Interval Mammography Screening Risks n Mean (SD) n Mean (SD) Range What is the chance that you will be told by a doctor that you have breast cancer in the next 10 years? (0.2) (0.3) (0 1) 0.23 What is the chance you will die from breast cancer in the next 10 years? (0.2) (0.2) (0 1) 0.68 n % n % What is your risk of getting breast cancer in the next 10 years? 0.28 Low % % Medium % % High % % How big of threat is breast cancer to your health? 0.32 Low % % Medium % % High % % SD: standard deviation. important enough to list as a reason for a positive mammography experience. This also underscores the importance of having clinically documented mammograms, a critical aspect of the current study, as a basis of cohort selection and as a reference point for collecting information about mammographic encounters. The infrastructure now in place for the NHMN in collecting and processing information on mammography encounters can potentially allow us to work with facilities to enhance the quality of care they provide. Previous negative mammography experiences, particularly those involving mammography technologists, may represent a meaningful barrier to mammography screening. Few studies address the affect of the technologist s role in mammography adherence; thus, the current study may lay the groundwork for future intervention research to enhance technologists approaches to patients. Identifying factors that will improve compliance with mammography screening is crucial and can lead to the development of interventions targeted at addressing these factors. At the time of the current study, mammography screening was receiving high profile media attention. The National Institutes of Health held a Consensus Development Conference in 1997 to address the issue of appropriate mammography screening schedules for women ages years. The panel concluded that the available data did not warrant a universal recommendation for mammography for all women in that age group. It is interesting to note that the current survey results reflected the fact that those women adhering to interval screening were more likely to identify mammography as benefiting year olds. Research of this nature is not without limitations. Recruitment from a population-based sample of women proved to be a much more significant challenge than we had anticipated. Our participation rate of 65% was lower than we had hoped. We speculate that because the study was being conducted during a time when a high level of media attention was focussing on confusion around screening intervals for mammography, this might have been discouraging to NH women. We can reflect on differences in recruitment groups by reasons for actively declining. Thirty four percent of nonadhering women stated they were too busy to participate, compared to 28% of adhering women. Three percent of adhering women reported being too ill to participate, compared to 8% in nonadhering women. There were essentially no differences in the remaining decline categories. These findings indicate that nonadhering women may have been more reticent than adhering women about participating, perhaps due to health-related reasons. In any case, consent rate does limit, to some extent, the generalizability of the current study findings. Another limitation involves the fact that NHMN data collection activities depend on the good will of the staff of participating mammography facilities

8 226 CANCER July 15, 2002 / Volume 95 / Number 2 TABLE 5 Perceptions of the Mammography Experience in New Hampshire Women Who Do and Do Not Adhere to Interval Mammography Screening Experience n % n % The last time you had a mammogram, how would you describe the experience: 0.01 Positive % % Negative % % Neutral % % Reason for negative experience: Pain or discomfort during the procedure 0 0.0% % 0.01 Difficulty getting an appointment % 0 0.0% 0.01 Reason for positive experience: Ease in getting an appointment % % 0.02 Care taken by the technologist in performing exam % % 0.04 Comfort provided by the technologist in talking you through the exam % % 0.01 Received information about mammography from tech % % 0.96 The information provided to me by the tech affected how I feel about getting my next mammogram 0.12 Disagree 9 7.3% % Neither agree nor disagree % % Agree % % The information provided to me by the tech made me feel better about getting my next mammogram 0.53 Disagree 3 2.4% 7 6.1% Neither agree nor disagree % % Agree % % Are you planning to have a mammogram in the next two years? 0.01 Yes % % Undecided 0 0.0% 5 2.0% No 0 0.0% 4 1.6% around the state. We recognize that it is possible that some women we classified as nonadhering may have obtained screening exams outside of our participating region. In these cases, we were unable to accurately track the women s screening behavior. However, an important advantage of a registry is its ability to track screening behavior longitudinally for a large number of, albeit not all, women. In a related study, we are continuing to track this cohort after testing interventions to improve interval screening. In conclusion, we found that a previous negative mammographic experience, particularly those involving mammography technologists, appears to influence interval adherence to screening. We found that higher BMI may be an important factor associated with this negative experience, perhaps due to embarrassment or greater sensitivity among heavier women about how they are coached or how their bodies are handled during this very personal procedure. Further research should focus on how best to coach women through the mammographic experience. Using registry resources such as the NHMN can provide unique opportunities to study this topic. REFERENCES 1. Annual Report of the New Hampshire Department of Health and Human Services, Concord, NH. New Hampshire State Cancer Registry, Chen VW, Wu XC. Cancer in North America, Volume II: Mortality. Sacramento, CA: North American Association of Central Cancer Registries, Tabar L, Vitak B, Chen HT, Yen MF, Duffy SW, Smith RA. Beyond randomized controlled trials: organized mammographic screening substantially reduces breast carcinoma mortality. Cancer. 2001;91: Breen N, Kessler L. Trends in cancer screening: United States, 1987 and Oncology. 1996;10: Urban N, Taplin SH, Taylor VM, Peacock S, Anderson G, Conrad D. Community organization to promote breast cancer screening among women ages Prev Med. 1995;24: Meissner HI, Breen N, Coyne C, Legler JM, Green DT, Edwards BK. Breast and cervical cancer screening interventions: an assessment of the literature. Cancer Epidemiol Biomarkers Prev. 1998;7: Anonymous. Self-reported use of mammography among women aged 40 years, United States, MMWR. 1997;46: Dolan NC, Reifler DR, McGrae, McDermott, McGaghie WC. Adherence to screening mammography recommendations in a university general medicine clinic. J Gen Intern Med. 1995;10:

9 Mammography Screening Adherence/Carney et al Zapka JG, Stoddard A, Maul L, Costanza ME. Interval adherence to mammography screening guidelines. Med Care. 1991;2: Rimer BK, Trock B, Engstrom PF, Lerman, C, King E. Why do some women get regular mammograms. Am J Prev Med. 1991;7: Mayer JA, Slymen DJ, Jones JA, et al. Mammography knowledge and intentions among insured women. Prev Med. 1992; 21: Lerman C, Rimer, B, Trock, B, Balshem A, Engstrom PF. Factors associated with repeat adherence to breast cancer screening. Prev Med. 1990;19: Chrvala C. Breast cancer risk status as a predictor for breast cancer screening behaviors. Prog Clin Biol Res. 1990;339: Lerman C, Daly M, Sands C, et al. Mammography adherence and psychological distress among women at risk for breast cancer. J Natl Cancer Inst. 1993;85: Ries LAG, Miller BA, Hankey BF, editors. SEER cancer statistics review, Bethesda: National Cancer Institute, Colditz GA, Wiett WC, Hunter DJ. Family history age and risk of breast cancer. JAMA. 1993;270: London SJ, Connolly JL, Schnitt SJ. A prospective study of benign breast disease and the risk of breast cancer. JAMA. 1992;267: Bodian CA. Benign breast diseases, carcinoma in situ, and breast cancer risk. Epidemiol Rev. 1993;15: Harris JR, Lippman ME, Veronesi U. Breast cancer. N Engl J Med. 1992;327: Laya MB, Gallagher JC, Schreiman JA. Effect of postmenopausal hormone replacement therapy on mammographic density and parenchymal pattern. Radiology. 1995;196: Stomper PC, Van Voorhis BJ, Ravkinar VA. Mammographic changes associated with postmenopausal hormone replacement therapy: a longitudinal study. Radiology. 1990;174: Leung W, Goldberg F, Zee B, Sterns E. Mammographic density in women on postmenopausal hormone replacement therapy. Surgery. 1997;122: Kelsey JL, Gammon MD. Epidemiology of breast cancer. Epidemiol Rev. 1990;12: Grady D, Rubin SM, Petitti DB. Hormone therapy to prevent disease and prolong life in post-menopausal women. Ann Intern Med. 1992;117: Boyd NF, Byng JW, Jong RA, Fishell EK, Little LE, Miller AB. Quantitative classification of mammographic densities and breast cancer risk: results from the Canadian National Breast Screening Study. J Natl Cancer Inst. 1995;87: McCarthy EP, Burns RB, Freund KM, et al. Mammography use, breast cancer stage at diagnosis, and survival among older women. J Am Geriatr Soc. 2000;48: Tabar L, Vitak B, Chen HH, et al. The Swedish Two-County Trial twenty years later. Updated mortality results and new insights from long-term follow-up. Radiol Clin North Am. 2000;38: Carney PA, Poplack SP, Wells WA, Littenberg B. Development and design of a population-based mammography registry: the New Hampshire Mammography Network. AJR Am J Roentgenol. 1996;167: D Orsi CJ, Bassett LW, Feig SA, et al. Breast imaging reporting and data system. 3rd ed. Reston, Va: American College of Radiology, Black WC, Nease RF Jr., Tosteson AN. Perceptions of breast cancer risk and screening effectiveness in women younger than 50 years of age. J Natl Cancer Inst. 1995;87: Woloshin S, Schwartz LM, Byram S, Fishhoff B, Welch H. A new scale for assessing perceptions of chance: a validation study. Med Decis Making. 2000;20: Speilberger CD, Gorsuch RL, Lushene RE. Manual for the State Trait Anxiety Inventory (Self Evaluation Questionnaire). Palo Alto, Ca: Consulting Psychologists Press, Champion V. Development of a benefits and barriers scale for mammography utilization. Cancer Nurs. 1995;18: Schottenfield D, Kerner J. Final Report Cancer Control Development Grant (National Cancer Institute grant- CA ). Bethesda: National Cancer Institute, Kash KM, Holland JC, Osborne MP, Miller DG. Psychological counseling strategies for women at risk of breast cancer. J Natl Cancer Inst Monogr. 1995;17:73 79.