Linn Abraham

Individual and Combined Effects of Age, Breast Density, and Hormone Replacement Therapy Use on the Accuracy of Screening Mammography

Context High breast density increases breast cancer risk and the difficulty of reading mammograms. Breast density decreases with age and increases with postmenopausal hormone therapy use. The interplay of breast density, age, and hormone therapy use on the accuracy of mammography is uncertain. Contribution For women with fatty breasts, the sensitivity of mammography was 87% and the specificity was 96.9%. For women with extremely dense breasts, the sensitivity of mammography was 62.9% and the specificity was 89.1%. Sensitivity increased with age. Hormone therapy use was not an independent predictor of accuracy. Implications The accuracy of screening mammography is best in older women and in women with fatty breasts. Postmenopausal hormone therapy affects mammography accuracy only through its effects on breast density. The Editors Mammographic breast density may be the most undervalued and underused risk factor in studies investigating breast cancer occurrence (1). The risk for breast cancer is four to six times higher in women with dense breasts (2, 3). Breast density may also decrease the sensitivity and, thus, the accuracy of mammography. Radiographically dense breast tissue may obscure tumors, which increases the difficulty of detecting breast cancer. In addition, dense breast tissue may mimic breast cancer on mammography (4), which increases recall rates (4-12), reduces specificity, and compromises the benefit of screening in women with dense breasts (such as women who use HRT or who are premenopausal) (6, 8, 13). Breast density is affected by age, use of hormone replacement therapy (HRT), menstrual cycle phase, parity, body mass index, and familial or genetic tendency (4, 5, 14-21). Studies show that the sensitivity of mammography increases with age (6-8), especially in postmenopausal women whose breasts are less dense (8). Earlier research has examined the individual effect of each factor we have described, but most studies could not adequately examine the interaction of these factors because of insufficient sample size (4-15). Studies conducted in the 1970s with data from the Breast Cancer Detection Demonstration Project (22) and New York Health Insurance Plan (23) are based on mammographic examinations that are very different from those performed using current technology. The Mammography Quality Standards Act (24) and the standardized reporting efforts of the American College of Radiology (25) have resulted in important improvements in mammography that necessitate reexamination. We used data from the National Cancer Institutes Breast Cancer Surveillance Consortium (BCSC) (26) on 329 495 women in the United States who had 463 372 screening mammograms, which were linked to 2223 cases of breast cancer. Our goal was to examine the individual and combined effects of age, breast density, and HRT use on mammographic accuracy. This large data set provides a unique opportunity to examine these issues in women undergoing screening mammography in the United States, especially women younger than 50 years of age and older than 80 years of age. We chose to study a sample that had been recently screened (within the previous 2 years) so that the risk for breast cancer would be similar to that in women who receive routine mammographic screening. Methods Data Collection Initially, we included data on women 40 to 89 years of age who underwent screening mammography between 1996 and 1998, as submitted by seven registries in the BCSC (North Carolina; New Mexico; New Hampshire; Vermont; Colorado; Seattle, Washington; and San Francisco, California). We included women who reported having previous mammography or who had a previous mammographic examination recorded in a registry within 2 years of the index mammogram. Women with breast implants or a personal history of breast cancer were excluded. In addition, women with missing data for age (<1%), breast density (27%), or HRT use (21%) were excluded (36% of all data). Demographic characteristics, clinical characteristics, and accuracy measures for women missing any of this information were very similar to those for women with complete data. All registries obtained institutional review board approval for data collection and linkage procedures, and careful data management, processing, and security procedures were followed (27). Consortium mammography registries and data collection procedures are described elsewhere (26). Briefly, seven institutions in seven states receive funding from the National Cancer Institute to maintain mammography registries that cover complete or contiguous portions of each state. Data are collected similarly at each registry. Demographic and history information is collected from women at the time of mammography by using a self-administered survey or face-to-face interview methods. Variables include date of birth, history of previous mammography, race or ethnicity, current use of HRT (prescription medication used to treat perimenopausal and postmenopausal symptoms), and menopausal status. We assumed that women 55 years of age and older were perimenopausal or postmenopausal. For women 40 to 54 years of age, premenopausal status was defined as having regular menstrual periods with no HRT use; perimenopausal or postmenopausal status was defined as either removal of both ovaries or uncertainty about whether periods had stopped permanently. This latter category was further classified into HRT users and nonusers. These definitions recognize that HRT users with intact uteri may have menstrual-like bleeding. Additional data, including mammographic breast density, mammographic assessment, and recommended follow-up (based on the American College of Radiology Breast Imaging Reporting and Data System [BI-RADS]), are collected from the technologist and radiologist at the time of mammography (25). Pathology data are collected from one or more sources: regional Surveillance, Epidemiology, and End Results (SEER) programs, state cancer registries, or pathology laboratories. Design We included all screening examinations for women who met the described criteria and who had at least one screening mammogram in 1996, 1997, or 1998. These years were chosen to ensure 1-year follow-up for cancer reporting and to account for routine reporting schedules in obtaining data from SEER and state cancer registries. We classified mammography as screening if a radiologist indicated that the examination was a bilateral, two-view (craniocaudal and mediolateral) examination. To avoid including diagnostic examinations, we excluded any breast imaging study performed within the previous 9 months. Because our goal was to study routine screening, mammographic accuracy was calculated on the basis of the initial assessment of the screening views alone (only 6% required supplemental imaging). Interpretation codes included BI-RADS assessments of 0 (incomplete), 1 (negative), 2 (negative, benign), 3 (probably benign), 4 (suspicious abnormality), or 5 (highly suggestive of malignancy). In cases in which the initial screening visit included both a screening examination and additional imaging to determine an assessment, the initial screening assessment was assigned a 0 (incomplete assessment) for analysis. When a woman had different assessments by breast, we chose the highest-level assessment for the woman as a whole (woman-level assessment) on the basis of the following hierarchy of overall level of radiologic concern: 1 < 2 < 3 < 0 < 4 < 5. We defined a screening examination as positive if it was assigned a BI-RADS assessment code of 0, 4, or 5. An assessment code of 3 associated with a recommendation for immediate additional imaging, biopsy, or surgical evaluation was also classified as positive. Although the BI-RADS recommendation for a code 3 (probably benign) is short-interval follow-up, immediate work-up was recommended in 37% of code 3s in the pooled BCSC data; therefore, this assessment is more consistent with a BI-RADS code of 0 (incomplete assessment) (28). We defined a screening examination as negative if it received a BI-RADS assessment code of 1, 2, or 3 when associated with short-interval follow-up only or routine follow-up. We classified breast pathology outcomes as cancer if pathology or cancer registry data identified a diagnosis of invasive or ductal carcinoma in situ. Lobular carcinoma in situ (<0.01% of cancer cases in our pooled data) was not considered a diagnosis of cancer in our analyses because it cannot be detected by mammography and is not treated. Examinations were classified as false-positive when the assessment was positive and breast cancer was not diagnosed within the follow-up period (365 days after the index screening examination or until the next examination, whichever occurred first). Examinations were classified as true-positive when the assessment was positive and cancer was diagnosed. A false-negative examination was a negative assessment with a diagnosis of cancer within the follow-up period. A true-negative examination was a negative assessment with no subsequent diagnosis of cancer within the follow-up period. Radiographic breast density was defined according to BI-RADS as follows: 1) almost entirely fatty, 2) scattered fibroglandular tissue, 3) heterogeneously dense, and 4) extremely dense (25). We excluded one registry that collects two categories of breast density (dense or not dense) at some facilities. Statistical Analysis For age, breast density, and HRT groups, we calculated rates of incident breast cancer, rates of breast cancer detected by mammography, and rates of missed cancer. To examine the nonlinear effects of age, we categorized age into 10-year groups, except for ages 40 to 59, which were divided into 5-year groups to explore changes around menopause. Accuracy indices included sensitivity and specificity. Sensitivity was calculated as true-positive/(true-positive + false-negative). Specificity was calculated as true-negative/(true-negative + false

Does Utilization of Screening Mammography Explain Racial and Ethnic Differences in Breast Cancer

Context Breast cancer mortality rates have fallen but still differ by race and ethnicity. One explanation might be differences in mammography use. Content These investigators linked data from mammography registries to tumor registries and showed that African-American and Hispanic women have longer intervals between mammography and are more likely to have advanced-stage tumors at diagnosis and to die of breast cancer than white women. However, in women with similar screening histories, these rates were similar regardless of race or ethnicity. Implications Differences in mammography use may explain ethnic disparities in the incidence of advanced-stage breast cancer and in mortality rates. The Editors Breast cancer mortality rates in the United States began to decrease in the 1990s (1) because of increased use of screening mammography and improved breast cancer treatment (2, 3). However, these decreases have primarily benefited non-Hispanic white women, whereas the mortality rate for breast cancer in African-American women changed little (1). Although racial and ethnic differences in breast cancer mortality rates have been consistently documented (1, 4-9), reasons for the persistence of these differences have been difficult to ascertain (10). Possible explanations include differences in biological characteristics of tumors (11-13); patient characteristics, such as obesity, that may affect prognosis; mammography use (14, 15); timeliness and completeness of breast cancer diagnosis and treatment (16, 17); social factors, such as education, literacy, and cultural beliefs; and economic factors, such as income level and health insurance coverage, that might affect a patients access to and choices for breast cancer screening and treatment (18-22). Stage at diagnosis, the strongest predictor of breast cancer survival (23), is proportionally higher in all non-Asian minority groups than in white women (8). Although minority women have historically undergone less mammography than white women (14), several recent surveys have found only small differences in mammography use between white and nonwhite women (24, 25). These observations raised doubt that tumors go undiagnosed until later stages in minority women because of infrequent breast cancer screening (26). However, the 2 most widely cited surveys of mammography use are based on self-report and only inquire about recent use, not adherence over time (24, 25). We explored stage of disease at diagnosis, tumor characteristics (including size and grade), and lymph node involvement among women of different races and ethnicities whose patterns of mammography use were similar. We hypothesized that differences in tumor characteristics may result primarily from differences in mammography use and that women with similar patterns of mammography use may have similar tumor characteristics. We had sufficient sample sizes within each racial and ethnic group and obtained sufficiently detailed data regarding mammography use to permit stratification of the cohort by pattern of mammography use; this technique enabled us to compare tumor characteristics among women with similar screening histories. Methods Data Source We pooled data from facilities that participate in 7 mammography registries that form the National Cancer Institutefunded Breast Cancer Surveillance Consortium: San Francisco Mammography Registry, San Francisco, California; Group Health Cooperative, Seattle, Washington; Colorado Mammography Project, Denver, Colorado; Vermont Breast Cancer Surveillance System, Burlington, Vermont; New Hampshire Mammography Network, Lebanon, New Hampshire; Carolina Mammography Registry, Chapel Hill, North Carolina; and New Mexico Mammography Project, Albuquerque, New Mexico. The data consisted of information sent to the registries regarding all mammographic evaluations performed at these facilities, including radiology reports and breast health surveys. The surveys, which were completed by patients at each mammography examination, included questions regarding race, ethnicity, presence of breast symptoms, and previous mammography use. Breast cancer diagnoses and tumor characteristics were obtained through linkage with state tumor registries; regional Surveillance, Epidemiology, and End Results programs; and hospital-based pathology services. Previous research has shown that at least 94% of cancer cases are identified through these linkages (27). Each surveillance registry captures most mammography case reports within its respective geographic area, and mammograms in these registries include approximately 2% of mammographic examinations performed in the United States. Each registry obtains annual approval from its institutional review board to collect mammography-related information and to link with tumor registries. Participants This study included women without a history of breast cancer who were 40 years of age and older who had undergone mammography at least once for screening or diagnostic purposes between 1996 and 2002 (n= 1010515). We categorized the race and ethnicity of the participating women (the mammography registry cohort) as non-Hispanic white (n= 789997), non-Hispanic African American/black (n= 62408), Hispanic (n= 90642), Asian/Pacific Islander (n= 49867), or Native American/Native Alaskan (n= 17601). We excluded women who did not report their race or ethnicity (n= 133235 [12%]) or reported mixed or other race (n= 6003 [<1%]). Breast cancer was diagnosed in a subset of the women in the mammography registry cohort (Table 1). Table 1. General Categorization of Study Participants Characterization of Mammography Use We included all mammographic evaluations in eligible women that were performed during the study period. We characterized each mammogram that was included in the study by the time interval between that mammogram and the one most recently preceding it. We determined these intervals by using examination dates that were recorded in the database (data were available for 85% of patients) and self-reported dates that the remaining women provided at the time of their examination. The mammography screening intervals were categorized into the following groups: within 1 year (4 to 17 months); 2 years (18 to 29 months); 3 years (30 to 41 months); and 4 years or longer (>41 months). At the time of each mammogram, women completed a breast health survey and provided the date of their last mammogram. We created 2 classifications for first mammograms. Mammography was classified as a first screening if the radiologist coded the examination as screening and the woman reported no breast symptoms. The mammogram was classified as diagnostic if the radiologist coded the examination as diagnostic or if the woman reported a breast mass or nipple discharge. Women whose first mammogram was diagnostic were assigned to the never screened group. Of note, a woman could have had mammography more than once during the study period and therefore could contribute more than 1 observation to the analyses. A woman could have observations that were categorized into different mammography screening intervals. For example, a woman could have had her first mammographic evaluation in 1998 and had subsequent mammography in 1999 and 2001. Her first mammogram would have been categorized as a first screening or as diagnostic, depending on the radiologists indication for that examination and whether the patient reported symptoms. Her second mammogram would have been categorized in the 1 year group, and her third mammography would have been categorized in the 2 year group. Breast Cancer To determine breast cancer status, we tracked each participants mammogram for 365 days following the date it had been obtained or until the patient underwent her next mammographic examination (whichever came first). Consequently, each tumor was associated with a single mammogramthat obtained closest to the date of diagnosis. We characterized breast cancer as either invasive or ductal carcinoma in situ. Large tumors were defined as invasive tumors that were 15 mm or larger in diameter. We used the TNM (tumor, node, metastasis) system (which is based on the criteria of the American Joint Committee on Cancer) to classify stage at diagnosis as 0 (ductal carcinoma in situ), 1, 2, 3, or 4 (28); advanced-stage tumors were defined as invasive lesions of stage 2 or higher. High-grade tumors were defined as invasive lesions of grades 3 and 4. Lymph node status was defined as positive, negative, or unknown. Advanced disease was defined as the presence of a large, advanced-stage, high-grade tumor or lymph nodepositive tumor at the time of diagnosis. Statistical Analysis We calculated the frequency distributions of various risk factors for all women in the mammography registry cohort. Among the subset of women with breast cancer (n= 17558), we calculated the proportion of tumors that were invasive and, among invasive tumors, the proportion that were advanced-stage or high-grade tumors; we then calculated the distribution by race and ethnicity. For all women in the cohort, we evaluated whether overall and advanced cancer rates per 1000 mammograms were similar across racial and ethnic groups after we adjusted for age and registry by using Poisson regression. We then calculated whether adjusted overall and advanced cancer rates per 1000 mammograms were similar across mammography screening interval groups. Because overall and advanced cancer rates varied across racial and ethnic groups (P< 0.001) and by previous mammography use (P< 0.001), and because mammography use potentially varied by race and ethnicity, we modeled cancer rates among similarly screened women in each ethnic group. We used Poisson regression to adjust for age and registry; an interaction term between race and ethnicity and previous mammography use was included in the Poisson model to allow for possible differences in the association between ethnicity and cancer rates by mammography group

Effectiveness of Computer-Aided Detection in Community Mammography Practice

BACKGROUND Computer-aided detection (CAD) is applied during screening mammography for millions of US women annually, although it is uncertain whether CAD improves breast cancer detection when used by community radiologists. METHODS We investigated the association between CAD use during film-screen screening mammography and specificity, sensitivity, positive predictive value, cancer detection rates, and prognostic characteristics of breast cancers (stage, size, and node involvement). Records from 684 956 women who received more than 1.6 million film-screen mammograms at Breast Cancer Surveillance Consortium facilities in seven states in the United States from 1998 to 2006 were analyzed. We used random-effects logistic regression to estimate associations between CAD and specificity (true-negative examinations among women without breast cancer), sensitivity (true-positive examinations among women with breast cancer diagnosed within 1 year of mammography), and positive predictive value (breast cancer diagnosed after positive mammograms) while adjusting for mammography registry, patient age, time since previous mammography, breast density, use of hormone replacement therapy, and year of examination (1998-2002 vs 2003-2006). All statistical tests were two-sided. RESULTS Of 90 total facilities, 25 (27.8%) adopted CAD and used it for an average of 27.5 study months. In adjusted analyses, CAD use was associated with statistically significantly lower specificity (OR = 0.87, 95% confidence interval [CI] = 0.85 to 0.89, P < .001) and positive predictive value (OR = 0.89, 95% CI = 0.80 to 0.99, P = .03). A non-statistically significant increase in overall sensitivity with CAD (OR = 1.06, 95% CI = 0.84 to 1.33, P = .62) was attributed to increased sensitivity for ductal carcinoma in situ (OR = 1.55, 95% CI = 0.83 to 2.91; P = .17), although sensitivity for invasive cancer was similar with or without CAD (OR = 0.96, 95% CI = 0.75 to 1.24; P = .77). CAD was not associated with higher breast cancer detection rates or more favorable stage, size, or lymph node status of invasive breast cancer. CONCLUSION CAD use during film-screen screening mammography in the United States is associated with decreased specificity but not with improvement in the detection rate or prognostic characteristics of invasive breast cancer.

Accuracy and Outcomes of Screening Mammography in Women With a Personal History of Early-Stage Breast Cancer

CONTEXT Women with a personal history of breast cancer (PHBC) are at risk of developing another breast cancer and are recommended for screening mammography. Few high-quality data exist on screening performance in PHBC women. OBJECTIVE To examine the accuracy and outcomes of mammography screening in PHBC women relative to screening of similar women without PHBC. DESIGN AND SETTING Cohort of PHBC women, mammogram matched to non-PHBC women, screened through facilities (1996-2007) affiliated with the Breast Cancer Surveillance Consortium. PARTICIPANTS There were 58,870 screening mammograms in 19,078 women with a history of early-stage (in situ or stage I-II invasive) breast cancer and 58,870 matched (breast density, age group, mammography year, and registry) screening mammograms in 55,315 non-PHBC women. MAIN OUTCOME MEASURES Mammography accuracy based on final assessment, cancer detection rate, interval cancer rate, and stage at diagnosis. RESULTS Within 1 year after screening, 655 cancers were observed in PHBC women (499 invasive, 156 in situ) and 342 cancers (285 invasive, 57 in situ) in non-PHBC women. Screening accuracy and outcomes in PHBC relative to non-PHBC women were cancer rates of 10.5 per 1000 screens (95% CI, 9.7-11.3) vs 5.8 per 1000 screens (95% CI, 5.2-6.4), cancer detection rate of 6.8 per 1000 screens (95% CI, 6.2-7.5) vs 4.4 per 1000 screens (95% CI, 3.9-5.0), interval cancer rate of 3.6 per 1000 screens (95% CI, 3.2-4.1) vs 1.4 per 1000 screens (95% CI, 1.1-1.7), sensitivity 65.4% (95% CI, 61.5%-69.0%) vs 76.5% (95% CI, 71.7%-80.7%), specificity 98.3% (95% CI, 98.2%-98.4%) vs 99.0% (95% CI, 98.9%-99.1%), abnormal mammogram results in 2.3% (95% CI, 2.2%-2.5%) vs 1.4% (95% CI, 1.3%-1.5%) (all comparisons P < .001). Screening sensitivity in PHBC women was higher for detection of in situ cancer (78.7%; 95% CI, 71.4%-84.5%) than invasive cancer (61.1%; 95% CI, 56.6%-65.4%), P < .001; lower in the initial 5 years (60.2%; 95% CI, 54.7%-65.5%) than after 5 years from first cancer (70.8%; 95% CI, 65.4%-75.6%), P = .006; and was similar for detection of ipsilateral cancer (66.3%; 95% CI, 60.3%-71.8%) and contralateral cancer (66.1%; 95% CI, 60.9%-70.9%), P = .96. Screen-detected and interval cancers in women with and without PHBC were predominantly early stage. CONCLUSION Mammography screening in PHBC women detects early-stage second breast cancers but has lower sensitivity and higher interval cancer rate, despite more evaluation and higher underlying cancer rate, relative to that in non-PHBC women.

Mammography Facility Characteristics Associated With Interpretive Accuracy of Screening Mammography

Background Although interpretive performance varies substantially among radiologists, such variation has not been examined among mammography facilities. Understanding sources of facility variation could become a foundation for improving interpretive performance. Methods In this cross-sectional study conducted between 1996 and 2002, we surveyed 53 facilities to evaluate associations between facility structure, interpretive process characteristics, and interpretive performance of screening mammography (ie, sensitivity, specificity, positive predictive value [PPV1], and the likelihood of cancer among women who were referred for biopsy [PPV2]). Measures of interpretive performance were ascertained prospectively from mammography interpretations and cancer data collected by the Breast Cancer Surveillance Consortium. Logistic regression and receiver operating characteristic (ROC) curve analyses estimated the association between facility characteristics and mammography interpretive performance or accuracy (area under the ROC curve [AUC]). All P values were two-sided. Results Of the 53 eligible facilities, data on 44 could be analyzed. These 44 facilities accounted for 484 463 screening mammograms performed on 237 669 women, of whom 2686 were diagnosed with breast cancer during follow-up. Among the 44 facilities, mean sensitivity was 79.6% (95% confidence interval [CI] = 74.3% to 84.9%), mean specificity was 90.2% (95% CI = 88.3% to 92.0%), mean PPV1 was 4.1% (95% CI = 3.5% to 4.7%), and mean PPV2 was 38.8% (95% CI = 32.6% to 45.0%). The facilities varied statistically significantly in specificity (P < .001), PPV1 (P < .001), and PPV2 (P = .002) but not in sensitivity (P = .99). AUC was higher among facilities that offered screening mammograms alone vs those that offered screening and diagnostic mammograms (0.943 vs 0.911, P = .006), had a breast imaging specialist interpreting mammograms vs not (0.932 vs 0.905, P = .004), did not perform double reading vs independent double reading vs consensus double reading (0.925 vs 0.915 vs 0.887, P = .034), or conducted audit reviews two or more times per year vs annually vs at an unknown frequency (0.929 vs 0.904 vs 0.900, P = .018). Conclusion Mammography interpretive performance varies statistically significantly by facility.

Predictors of urinary incontinence in a prospective cohort of postmenopausal women.

OBJECTIVE: To prospectively assess risk factors associated with occurrence of urinary incontinence among postmenopausal women. METHODS: We followed up 1,017 postmenopausal health maintenance organization enrollees, aged 55 to 75 years, for 2 years. The primary outcome measures were any urinary incontinence and severe incontinence reported at 12- or 24-month follow-up visits. RESULTS: Baseline prevalence of any amount or frequency of urinary incontinence in the past year was 66%. Among the 345 women without incontinence at baseline, 65 (19%) at 1 year and 66 (19%) at 2 years reported any incontinence. Ninety-two of 672 (14%) and 96 of 672 (14%) women with incontinence at baseline reported no incontinence at years 1 and 2. In an adjusted multiple logistic regression model, independent predictors of any incontinence included white race (odds ratio [OR] 1.7, 95% confidence interval [CI] 1.1–2.6), vaginal estrogen cream (OR 2.0, CI 1.1–3.7), vaginal dryness (OR 1.6, CI 1.2–2.2), vaginal discharge (OR 1.5, CI 1.0–2.2), 6 or more lifetime urinary tract infections (OR 1.8, CI 1.2–2.6), and diabetic peripheral neuropathy (OR 1.7, CI 1.0–3.1). In adjusted models, predictors of severe incontinence were history of hysterectomy (OR 1.8, CI 1.1–2.7) and any vaginal symptom (OR 1.7, CI 1.0–2.8). CONCLUSION: A substantial proportion of incontinence-free postmenopausal women developed urinary incontinence during 2 years of follow-up. Because vaginal symptoms are associated with urinary incontinence, their relationship with other risk factors, including vaginal Escherichia coli colonization and vaginal estrogen cream use, warrant additional study. Similarly, diabetic peripheral neuropathy and hysterectomy associations suggest areas for future investigation. LEVEL OF EVIDENCE: II-2

Mammography surveillance following breast cancer

AbstractBackground. To describe when women diagnosed with breast cancer return for their first mammography, and to identify factors predictive of women returning for mammographic surveillance. Methods. Women who underwent mammography at facilities participating in the National Cancer Institutes Breast Cancer Surveillance Consortium (BCSC) during 1996 and who were subsequently diagnosed with ductal carcinoma in situ or invasive breast cancer were included in this study. Data from seven mammography registries were linked to population-based cancer and pathology registries. Kaplan–Meier curves were used to depict the number of months from the breast cancer diagnosis to the first mammogram within the defined follow-up period. Demographic, disease and treatment variables were included in univariate and multivariate analyses to identify factors predictive of women returning for mammography. Results. Of the 2503 women diagnosed with breast cancer, 78.1% returned for mammography examination between 7 and 30 months following the diagnosis. Mammography facilities indicated that 66.8% of mammography examinations were classified as screening. Multivariate analyses found that women were most likely to undergo surveillance mammography if they were diagnosed at ages 60–69 with Stage 0, I or II breast cancer and had received radiation therapy in addition to surgery. Conclusions. While the majority of women return for mammographic surveillance following breast cancer, some important subgroups of women at higher risk for recurrence are less likely to return. Research is needed to determine why some women are not undergoing mammography surveillance after a breast cancer diagnosis and whether surveillance increases the chance of detecting tumors with a good prognosis.

Reactions to Uncertainty and the Accuracy of Diagnostic Mammography

BackgroundReactions to uncertainty in clinical medicine can affect decision making.ObjectiveTo assess the extent to which radiologists’ reactions to uncertainty influence diagnostic mammography interpretation.DesignCross-sectional responses to a mailed survey assessed reactions to uncertainty using a well-validated instrument. Responses were linked to radiologists’ diagnostic mammography interpretive performance obtained from three regional mammography registries.ParticipantsOne hundred thirty-two radiologists from New Hampshire, Colorado, and Washington.MeasurementMean scores and either standard errors or confidence intervals were used to assess physicians’ reactions to uncertainty. Multivariable logistic regression models were fit via generalized estimating equations to assess the impact of uncertainty on diagnostic mammography interpretive performance while adjusting for potential confounders.ResultsWhen examining radiologists’ interpretation of additional diagnostic mammograms (those after screening mammograms that detected abnormalities), a 5-point increase in the reactions to uncertainty score was associated with a 17% higher odds of having a positive mammogram given cancer was diagnosed during follow-up (sensitivity), a 6% lower odds of a negative mammogram given no cancer (specificity), a 4% lower odds (not significant) of a cancer diagnosis given a positive mammogram (positive predictive value [PPV]), and a 5% higher odds of having a positive mammogram (abnormal interpretation).ConclusionMammograms interpreted by radiologists who have more discomfort with uncertainty have higher likelihood of being recalled.

Risk Factors for Second Screen-Detected or Interval Breast Cancers in Women with a Personal History of Breast Cancer Participating in Mammography Screening

Background: Women with a personal history of breast cancer (PHBC) have increased risk of an interval cancer. We aimed to identify risk factors for second (ipsilateral or contralateral) screen-detected or interval breast cancer within 1 year of screening in PHBC women. Methods: Screening mammograms from women with history of early-stage breast cancer at Breast Cancer Surveillance Consortium-affiliated facilities (1996–2008) were examined. Associations between woman-level, screen-level, and first cancer variables and the probability of a second breast cancer were modeled using multinomial logistic regression for three outcomes [screen-detected invasive breast cancer, interval invasive breast cancer, or ductal carcinoma in situ (DCIS)] relative to no second breast cancer. Results: There were 697 second breast cancers, of these 240 were interval cancers, among 67,819 screens in 20,941 women. In separate models for women with DCIS or invasive first cancer, first breast cancer surgery predicted all three second breast cancer outcomes (P < 0.001), and high ORs for second breast cancers (between 1.95 and 4.82) were estimated for breast conservation without radiation (relative to mastectomy). In women with invasive first breast cancer, additional variables predicted risk (P < 0.05) for at least one of the three outcomes: first-degree family history, dense breasts, longer time between mammograms, young age at first breast cancer, first breast cancer stage, and adjuvant systemic therapy for first breast cancer; and risk of interval invasive breast cancer was highest in women <40 years at first breast cancer (OR, 3.41; 1.34–8.70), those with extremely dense breasts (OR, 2.55; 1.4–4.67), and those treated with breast conservation without radiation (OR, 2.67; 1.53–4.65). Conclusion: Although the risk of a second breast cancer is modest, our models identify risk factors for interval second breast cancer in PHBC women. Impact: Our findings may guide discussion and evaluations of tailored breast screening in PHBC women, and incorporating this information into clinical decision-making warrants further research. Cancer Epidemiol Biomarkers Prev; 22(5); 946–61. ©2013 AACR.

Variability of interpretive accuracy among diagnostic mammography facilities

BACKGROUND Interpretive performance of screening mammography varies substantially by facility, but performance of diagnostic interpretation has not been studied. METHODS Facilities performing diagnostic mammography within three registries of the Breast Cancer Surveillance Consortium were surveyed about their structure, organization, and interpretive processes. Performance measurements (false-positive rate, sensitivity, and likelihood of cancer among women referred for biopsy [positive predictive value of biopsy recommendation {PPV2}]) from January 1, 1998, through December 31, 2005, were prospectively measured. Logistic regression and receiver operating characteristic (ROC) curve analyses, adjusted for patient and radiologist characteristics, were used to assess the association between facility characteristics and interpretive performance. All statistical tests were two-sided. RESULTS Forty-five of the 53 facilities completed a facility survey (85% response rate), and 32 of the 45 facilities performed diagnostic mammography. The analyses included 28 100 diagnostic mammograms performed as an evaluation of a breast problem, and data were available for 118 radiologists who interpreted diagnostic mammograms at the facilities. Performance measurements demonstrated statistically significant interpretive variability among facilities (sensitivity, P = .006; false-positive rate, P < .001; and PPV2, P < .001) in unadjusted analyses. However, after adjustment for patient and radiologist characteristics, only false-positive rate variation remained statistically significant and facility traits associated with performance measures changed (false-positive rate = 6.5%, 95% confidence interval [CI] = 5.5% to 7.4%; sensitivity = 73.5%, 95% CI = 67.1% to 79.9%; and PPV2 = 33.8%, 95% CI = 29.1% to 38.5%). Facilities reporting that concern about malpractice had moderately or greatly increased diagnostic examination recommendations at the facility had a higher false-positive rate (odds ratio [OR] = 1.48, 95% CI = 1.09 to 2.01) and a non-statistically significantly higher sensitivity (OR = 1.74, 95% CI = 0.94 to 3.23). Facilities offering specialized interventional services had a non-statistically significantly higher false-positive rate (OR = 1.97, 95% CI = 0.94 to 4.1). No characteristics were associated with overall accuracy by ROC curve analyses. CONCLUSIONS Variation in diagnostic mammography interpretation exists across facilities. Failure to adjust for patient characteristics when comparing facility performance could lead to erroneous conclusions. Malpractice concerns are associated with interpretive performance.