© The Author 2005. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org.
Late-Stage Breast Cancer Among Women With Recent Negative Screening Mammography: Do Clinical Encounters Offer Opportunity for Earlier Detection?
Affiliations of authors: Kaiser Permanente Clinical Research Unit, Denver, CO (JM); Divison of Cancer Control and Population Sciences, Applied Research Program, National Cancer Institute, Bethesda, MD (ST); Group Health Cooperative, Seattle, WA (ST, LI, WEB); Kaiser Permanente Southern California, Pasedena, CA (AMG); Kaiser Permanente Center for Health Research, Portland, OR (SW); Kaiser Permanente Center for Health Research, Honolulu, HI (JG); Kaiser Permanente Northern California, Oakland, CA (MMM); Henry Ford Health System, Detroit, MI (MUY)
Correspondence to: Judy Mouchawar, MD, MSPH, Kaiser Permanente, Colorado Clinical Research Unit, 580 Mohawk Dr., Boulder, CO 80303 (e-mail: judy.mouchawar{at}kp.org).
 |
ABSTRACT |
|---|
 TopNotes Abstract IntroductionMETHODSRESULTSDISCUSSIONReferences |
|---|
Background: Opportunities to prevent late-stage breast cancer within the course of usual care are needed. We evaluate whether clinical encounters offer such opportunities. Methods: Within seven health care plans, we identified 1298 women aged more than 50 years with early (<3 cm), late-stage (
3 cm), or metastatic invasive breast cancer diagnosed during 1995–1999, whose first screening mammogram 13–36 months prior to the diagnosis (index) was negative. We audited all care occurring in the health plans up to 36 months prior to the cancer diagnoses. Ordinal logistic regression compared the frequency of events by disease category. We hypothesized that during the 13–36 months prior to diagnosis, women with late-stage or metastatic breast cancer would have more symptoms and be more likely to have breast-related clinical visits but have less breast screening (clinical breast examination [CBE] or mammography) than women with early-stage disease, thereby indicating clinical opportunities for earlier detection. Results: We found no differences in demographic characteristics across breast cancer stage among the 1298 women. Both before and after the negative index mammogram but during the 13–36 months prior to diagnosis, few women had breast symptoms (5% before index, 8% after), but many sought breast care (86% before index, 90% after) and screening CBE (62% before index, 43% after). Only the occurrence of screening CBE (odds ratio [OR] = 0.73, 95% confidence interval [CI] = 0.56 to 0.95) or screening mammograms (OR = 0.74, 95% CI = 0.57 to 0.97) after the negative index mammogram reduced odds of more severe disease at diagnosis. Conclusion: Although the mortality benefit of CBE, or one compared to two year mammography has not been established, we found that women with late-stage breast cancers undetected by screening mammography did not experience opportunities for earlier detection except through CBE or additional screening mammography.
|
INTRODUCTION |
|---|
|
TopNotesAbstractIntroductionMETHODSRESULTSDISCUSSIONReferences |
|---|
Although the reduction of breast cancer mortality by mammography screening has been demonstrated in randomized trials, even insured women with access to breast care preventive services continue to die from late-stage and metastatic breast cancer (1–9). Recent work suggests that 52% of late-stage cases have not been screened but that a substantial proportion (40%) received screening with normal or negative results 13–36 months prior to a breast cancer diagnosis (9).
It is tempting to infer that something goes wrong in the provision of breast care for these women with "missed" diagnoses, including the potential that films are misread or that patient signs or symptoms are overlooked because the mammogram results are normal or negative. Published work demonstrates that 30% of cancers found within a year of a negative mammogram can be seen on reexamination of the film (10–11). Other work shows that negative mammography reassures both patient and providers, even when the patient presents with clinical findings (12–15).
Although some late stage cancers may be rapidly growing during the interval between normal screens, and thereby undetectable with current modalities, it is possible that the prognosis of even these tumors might be changed with earlier detection (10). Thus, the question is raised whether any opportunities exist to prevent the development of late-stage cancer among women having negative mammography.
This study evaluates whether clinical encounters before a breast cancer diagnosis among women with negative screening mammography present opportunities to change the course of care to prevent late-stage disease. Changes in care must occur far enough in advance of the diagnosis to offer an opportunity to reduce the associated morbidity and mortality. Since annual mammography is the shortest screening interval shown to reduce mortality, we assume that changes in care would have to occur more than 1 year from the diagnosis (1,3). Therefore, our analysis evaluates care from 13 to 36 months prior to diagnosis, though we understand that changes in care during this period would need to be assessed in other studies to evaluate their effect on mortality.
We hypothesize that during the 13–36 months prior to diagnosis, women with late-stage or metastatic breast cancer have more symptoms and are more likely to have breast-related clinical visits, but have less actual breast screening (clinical breast examination or mammography) than women with early-stage disease. Our evaluation includes examining care before and after the negative index mammogram. We examine care before the negative mammogram because that might influence actions at later encounters. We also make comparison by breast cancer stage at diagnosis (disease severity).
|
METHODS |
|---|
|
TopNotesAbstractIntroductionMETHODSRESULTSDISCUSSIONReferences |
|---|
Study Setting, Population, and Sample
This analysis is based on data collected as part of DETECT (Detecting Early Tumors Enables Cancer Therapy), a project of the Cancer Research Network (CRN) (9,16). The CRN consists of the research programs, enrollee populations, and databases of 11 integrated healthcare organizations that are members of the HMO Research Network. The health care delivery systems participating in the CRN are: Group Health Cooperative, Harvard Pilgrim Health Care, Henry Ford Health System/Health Alliance Plan, HealthPartners Research Foundation, the Meyers Primary Care Institute of the Fallon Healthcare System/University of Massachusetts, and Kaiser Permanente in six regions (Colorado, Georgia, Hawaii, Northwest [Oregon and Washington], Northern California, and Southern California). The 11 health plans have nearly 10 million enrollees. The CRN conducts collaborative research on variations in cancer prevention and treatment policies and practices. The CRN is a National Cancer Institute–funded consortium of research organizations affiliated with nonprofit integrated healthcare delivery systems. The overall goal of the CRN is to increase the effectiveness of preventive, curative, and supportive interventions that span the natural history of major cancers among diverse populations and health systems, through a program of collaborative research. Seven CRN health systems participated in DETECT: Group Health Cooperative, Henry Ford Health System, and the Kaiser Permanente Regions in Colorado, Hawaii, Northern California, Northwest, and Southern California. All of these participating health plans had guidelines in place during the study period recommending screening mammography and clinical breast examination at least every 2 years (17).
The study population for DETECT has been described previously and includes women aged 50 years and older who were enrolled for at least 33 of the 36 months prior to a diagnosis of breast cancer between 1995 and 1999 (9). The date of the breast cancer diagnosis came from the date of the pathology procedure at which cancer was confirmed. Women diagnosed with late-stage breast cancer (>3 cm or metastatic at diagnosis) were matched on age and year of diagnosis to women with early stage cancers (<3 cm and not metastatic) at the same institution (9). Women with a history of invasive breast cancer prior to 1995 were excluded.
In this study we included only the women whose first screening mammogram was negative (index mammogram or index) 13–36 months prior to their breast cancer diagnosis (Prediagnostic Period or PDP; Fig. 1) and thus did not retain the matched pairs described above. Women with negative mammograms represented 40% of the original study sample (39.5% of late-stage and 56.9% of early-stage cancers) (9). We included women who have breast cancer rather than general population controls. Since all women had breast cancer, all had similar opportunity to have the same clinical events. Since we were not evaluating the effectiveness of mammography or the effectiveness of changing clinical care, a general population control group was not appropriate (18).
|
Data Collection
The details of the original study and data collection activities as well as the quality-control procedures for the audit have been previously described (9). We obtained institutional review board approval from all participating sites prior to data collection, with a waiver of written informed consent. For all patients, we audited the automated databases and performed chart review for the 3 years prior to diagnosis to identify "patient" and "clinical" factors. "Patient" factors included patient descriptors such as age at diagnosis, race/ethnicity, length of enrollment in the health plan prior to diagnosis, family history of breast cancer, documented patient-reported symptoms, patient refusal of care, and previous breast conditions. "Clinical" factors or clinical encounters included the occurrence of mammography, clinical breast examination, and breast-related visits. Breast-related visits were clinical visits in the health care system involving a breast-related complaint or breast-related procedure, as well as its associated findings, recommendations, and pathology.
Definitions
As this evaluation included unique constructs and temporal complexities, we offer the following definitions as orientation to the results. Again, our focus was on whether clinical encounters existed to improve earlier breast cancer detection. We defined a clinical encounter as a breast-related procedure (clinical breast examination or mammography) or visit in the health care system. A clinical opportunity then was defined as a clinical encounter within the health care system that could possibly have changed the course of disease severity at diagnosis.
We defined screening mammography as one using bilateral routine views specifically noted to be for screening. All other mammograms were considered nonscreening. Positive mammograms included those with one of the following: positive image assessment ( "additional imaging evaluation," "probably benign finding," "suspicious abnormality," or "highly suggestive of malignancy"), or a recommendation for immediate follow-up, or if additional imaging was performed within 2 weeks. If both the assessment and recommendation were missing, any noted findings were considered positive. All other mammograms were considered negative mammograms. Again, in this analysis we included only the women whose first screening mammogram 13–36 months prior to their breast cancer diagnosis (index mammogram or index) was negative.
We defined a screening clinical breast examination (screening CBE) as one performed in an asymptomatic woman seen for a complete physical or routine health visit. All other CBEs were considered nonscreening CBEs. Our definition of nonscreening CBE therefore included a CBE performed in a woman with a breast symptom (symptomatic CBE) or a CBE performed in a woman without a breast symptom at the time of a visit for chronic care or some other concern (opportunistic CBE). Positive CBEs included those given a conclusion of "indeterminate finding" or "positive or probably cancer" or a recommendation for immediate follow-up. If both the conclusion and recommendation were missing, any finding was considered positive. All other CBEs were classified as negative CBEs.
Breast symptoms included patient report of lump, pain, nipple discharge, visual change, or odor. Classification of patient refusal of care required a chart notation of patient's active refusal of a breast-related procedure at the time it was recommended. A breast-related procedure included CBE, mammography, ultrasound, and biopsy, with breast-related visit defined as clinical encounters involving a breast related complaint or breast-related procedure.
We assessed family history by degree of relatedness as well as by having any family history versus none. Finding no difference by degree of relatedness, we present the dichotomized category (any versus none).
Categorization of Periods
We used two major periods counting backward from the date of diagnosis: 1) a diagnostic period (DXP), defined as the period of time from time zero (date of diagnosis) to 12 months before diagnosis, and 2) a prediagnostic period (PDP), defined as the time 13–36 months prior to the date of diagnosis (Fig. 1).
We stratified the PDP into two distinct periods: 1) before the negative (index) mammogram and 2) after the index mammogram. In this manner we consider if clinical findings before the mammogram influenced activity afterward. For example, a clinical finding before a negative index mammogram should influence how that provider responds to later symptomatic presentations. Thus, to understand clinical opportunities, we needed to understand clinical care findings occurring in the encounters both before and after the index mammogram.
Time of the index screen indicated whether the index mammogram occurred during the first or second year of the 2-year PDP. We calculated time to diagnosis as the number of months elapsed from the index mammogram to the breast cancer diagnosis.
During the DXP we assessed the presence of patient symptoms and physical examination findings going back in time from the date of diagnosis to the positive "event" that led to the breast cancer diagnosis. The event was identified as the positive CBE or mammogram furthest from the date of diagnosis but no more than 90 days before a positive diagnostic work-up or the cancer diagnosis. If both the CBE and mammogram occurred on the same day, the diagnosis was attributed to the mammogram.
Outcome Variable
For the outcome variable we created three subcategories within our study sample to reflect increasing severity of disease at diagnosis: 1) early (size <3 cm, no metastasis); 2) late-stage, no metastasis (size 3 cm, no metastasis); and 3) metastatic. We include women presenting with metastatic disease in a separate category because their clinical presentation may differ substantively from the presentation of nonmetastatic disease and we were interested in identifying this difference.
Analysis
Frequencies of patient demographics and clinical care factors were calculated for each disease category (early, late, metastatic) and the clinical care frequencies were stratified by timing before and after the index mammogram. The procedures and clinical exams were quantified at the woman level.
We fit an ordinal logistic regression (OLR) model to quantify the association between each patient demographic and clinical care factor and the diagnosis of more severe stage of disease, adjusting for age at diagnosis, health system site, and year of diagnosis. Using defined cutpoints between each disease category, the OLR model computed an odds ratio (OR) for being in a more severe disease category compared to a less severe disease category. For this analysis, the cutpoints were between early-stage and late-stage and between late-stage and metastatic cancer.
Although there was an odds ratio for each cut-point of the outcome variable, a single odds ratio was computed in the OLR model. However, this odds ratio would be valid only if the odds ratios from each cutpoint of the outcome variable were proportional to each other. For all model results reported here, the proportional odds assumption was satisfied. The proportional odds assumption held for all models considered except for those assessing the association between positive exam findings and disease severity, due to small numbers of positives and unstable estimates. We therefore did not include these results.
All models were adjusted for age at diagnosis, health system site, and year of diagnosis. For analysis of activities occurring during the PDP we also adjusted for time of index screen. By so doing, we account for the possibility that the amount of time women have before or after the index mammogram could influence the likelihood that clinical encounters occurred, as well as the amount of time in which benefit could be realized. All analyses were performed using SAS software version 8.0.
|
RESULTS |
|---|
|
TopNotesAbstractIntroductionMETHODSRESULTSDISCUSSIONReferences |
|---|
There were 1298 women whose first screening mammogram was negative (index mammogram) 13–36 months prior to their breast cancer diagnosis. Most were aged 50–64 years (54%) and were predominantly white (82%) and non-Hispanic (91%) (Table 1). More than 94% were enrolled in their respective health plans for at least 5 years prior to their cancer diagnosis. Most cases had no family history of breast cancer documented (58%). The mean and median size of tumor at diagnosis was consistent with our outcome categorization. The mean time to diagnosis was just over 2 years and did not vary by disease severity.
|
Most women received breast-related clinical care before the index mammogram, with no significant differences in the frequency of symptoms, screening CBE, or frequency of breast related visits across the disease categories (Table 2). Overall, 4% of the women had nonscreening mammograms prior to the index (Table 2). Nearly 20% of women had a nonscreening CBE prior to the index mammogram (Table 2). Across disease categories most of these CBEs were opportunistic (60%, data not shown). Most of the non-screening CBEs had a negative result. There was no association between the occurrence of either nonscreening mammogram or nonscreening CBE and disease severity, suggesting no opportunity to detect a cancer earlier.
|
After the index screen, most women (90%) received breast-related clinical care during the PDP. A little more than one-third of women had a mammogram after the index; among these mammograms, most were screening mammograms (Table 3). Women with metastatic disease were less likely to have received any mammography (screening or nonscreening) after the index.
|
After the index mammogram, 60% of women did not have screening CBE, which mirrors the screening mammogram data and likely reflects that these procedures often occur together (Table 3). Among women who did receive screening CBE, most CBEs (95%) were normal. Screening CBE occurred less frequently in the metastatic group, and ordinal logistic regression results showed that screening CBE after the index was associated with less severe disease at diagnosis (OR = 0.73, 95% confidence interval [CI] = 0.56 to 0.95). Only 19% (245/1298) of women had a nonscreening CBE after the index, with few women having more than one CBE in this period. Overall, 41% of the exams (108/263) were opportunistic and most were negative. No association was found by disease severity (Table 3).
In contrast to the similarities in care and asymptomatic status during the PDP, differences among women became apparent at the beginning of the DXP. The time from index mammogram to the start of the DXP was nearly identical in all three groups: 14.3 months for the early-stage group, 14.5 months for the late stage group, and 13.3 months for the metastatic group (data not shown). The women with late-stage and metastatic cancer were more likely to have a positive CBE or symptoms at the beginning of the DXP (Table 4).
|
Overall, nonscreening events led to the diagnosis of most of the late-stage and metastatic women (Table 4). As expected, screening mammogram was the event leading to diagnosis for most of the early-stage group (46.7%), and women presenting with symptoms outside of the breast were limited to the metastatic group.
|
DISCUSSION |
|---|
|
TopNotesAbstractIntroductionMETHODSRESULTSDISCUSSIONReferences |
|---|
To our knowledge, this is the first report to assess whether clinical encounters offer opportunity for earlier detection of breast cancer among women having negative mammography. Although we show equal time frames in which to receive care, our results suggest that symptomatic presentations and number of breast-related visits are similar across the breast disease severity categories. Thus, even though the late-stage and metastatic cancers are presumably at later stages in their natural history than the early-stage cancers, it appears that clinical encounters during the prediagnostic period offer little opportunity to find disease earlier.
After the index screening mammogram, this study found that having additional breast cancer screening (screening CBE or screening mammography) in less than two years reduces the odds of more severe disease at diagnosis. Whether this means that more frequent screening reduces mortality cannot be determined from this study. The current U.S. Preventive Services Task Force recommendation for breast cancer screening frequency, as well as the recommendation in place during the audit period, states that there are insufficient data to advocate for a specific screening interval (annual or biennial) (19,20). The role that screening CBE plays in mortality reduction and recommendations for CBE screening intervals are also not clear (19–21). Although our findings may indicate a potential opportunity for the prevention of late-stage disease by more frequent screening, other work recently suggested that for women 50 years of age or older, there was no increase in late-stage disease at diagnosis for those undergoing biennial versus annual mammography screening (22).
Although possibly indicating an opportunity to reduce late-stage cancer through more frequent screening, this study is not designed to evaluate this likelihood. To do such a study, we would need to include women without an eventual breast cancer diagnosis. Therefore, the odds ratios for the late-stage disease we report cannot be applied to average persons without disease.
We also acknowledge that evaluating clinical opportunities for early cancer detection versus later stage disease is challenging because the period of opportunity likely reflects different points in the natural history of the cancers. We established our comparison to provide insight into the opportunities within clinical encounters to prevent late-stage cancer. Despite the potential differences in time points in the natural history of these cancers, we found that clinical presentation and findings were the same. For women and physicians reviewing care, this finding is extremely important because it is contrary to the intuitive conclusion that something was missed.
More than half of all cases were eventually diagnosed through nonscreening events (57%). The lack of earlier signs and/or symptoms among this group of breast cancer cases might be attributable to missing documentation or to false-negative CBE interpretation. We doubt that failure to document signs or symptoms would occur to any great degree, as delay in breast cancer diagnosis is a common reason for litigation in medicine, and the physicians were highly motivated to document their care (23,24). Further, the tumors eventually diagnosed among the late-stage and metastatic groups are quite large and therefore likely to have been documented if they were palpated. Patient delay in reporting of symptoms might have occurred, but we believe this is minimal due to the high use of breast-related care (25–27). Regardless, we found no variance in report of signs and symptoms by disease severity during the pre-diagnostic period, suggesting little opportunity to prevent late-stage disease. As noted above, whether screening more frequently would change mortality is still not clear. Thus, it appears that the only opportunity to change the course of the late stage cancers would be to improve imaging and hence the likelihood of earlier detection (28–30).
A limitation of this work is that these health care systems may not be wholly representative of community practice. However, we felt it important to begin assessment of clinical opportunity among practices with a history of a strong prevention focus and among organizations where clinic visits and breast care prevention are prepaid services. Our work is further strengthened by the ability to audit many charts across the nation, ensuring the identification of symptoms and exam findings based on actual medical records rather than self-reports in surveys.
Although our sample definition broke the matching of the original case–control study, we felt that it was the best design and would lead to appropriate conclusions. To be certain of this, we performed all ordinal logistic regression models both with and without the matching variables and found no significant differences in the results.
It is unfortunate that we could not categorize the study group into reasons for the absence of earlier disease detection, such as false-negative interpretation of the mammogram, tumors occult to mammography, rapidly growing tumors escaping any level of screen, or patient delay, but such categorization was beyond the scope of this project. Also, all such reasons and the uncertainty of current screening modalities are part of daily clinical care, and thus our findings are representative of potential opportunities in the provision of breast care in routine practice.
Although screening mammography detects cancers at earlier stages for most of the general population, there remains a group of women destined to develop late-stage breast cancer that will go undetected. Our results indicate that better screening technology is needed for these women because clinical encounters are not likely to lead to earlier detection.
|
NOTES |
|---|
Supported by grant number NIH 5 U19 CA79689 from the National Institutes for Health (Edward Wagner, MD, MPH, principal investigator), and its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
We acknowledge the DETECT Study team who worked through many years, telephone calls, charts, and complexities of their systems to help us better understand screening implementation: (Group Health Cooperative) Sarah Parkhurst; (Kaiser Permanente Northern California) Noelle Blick, Rowena Allison, Reggie Jackson, Barbara Rowe; (Kaiser Permanente Colorado), Kimberly Bischoff, Jennifer Ellis; (Kaiser Permanente Hawaii), Denise S. Williams, Mark M. Schmidt; (Kaiser Permanente Southern California) Carmen N. West; (Henry Ford Health System) Karen Wells, Susan McGuinness (deceased), Lisa May, Patricia Baker, Cheryl Spoutz; (Kaiser Permanente Northwest) Weiming Hu, Deborah Reck, and Jill Mesa.
|
REFERENCES |
|---|
|
TopNotesAbstractIntroductionMETHODSRESULTSDISCUSSIONReferences |
|---|
(1) Letton AH, Mason EM, Ramshaw BJ. Twenty-year review of a breast cancer screening project: ninety-five percent survival of patients with nonpalpable cancers. Cancer 1996;77:104–6.[CrossRef][Web of Science][Medline]
(2) Smart CR, Byrne C, Smith RA, Garfinkel L, Letton AH, Dodd GD, Beahrs OH. Twenty-year follow-up of the breast cancers diagnosed during the Breast Cancer Detection Demonstration Projects. CA Cancer J Clin 1997;47:134–49.[Abstract]
(3) Tabar L, Fagerberg G, Duffy SW, Day NE, Gad A, Grontoft O. Update of the Swedish Two-County program of mammography screening for breast cancer. Radiol Clin North Am 1992;30:187–210.[Web of Science][Medline]
(4) Roberts MM, Alexander FE, Anderson TJ, Chetty U, Donnan PT, Forrest P, et al. Edinburgh trial of screening for breast cancer: mortality at seven years. Lancet 1990;335:241–6.[CrossRef][Web of Science][Medline]
(5) Andersson I, Aspegren K, Janzon L, Landberg T, Lindholm K, Linell F, et al. Mammographic screening and mortality from breast cancer: the Malmo mammographic screening trial. BMJ 1988;297:943–8.
(6) Frisell J, Eklund G, Hellstrom L, Lidbrink E, Rutqvist LE, Somell A. Randomized study of mammography screening—preliminary report on mortality in the Stockholm trial. Breast Cancer Res Treat 1991;18:49–56.[CrossRef][Web of Science][Medline]
(7) Nystrom L, Rutqvist LE, Wall S, Lindgren A, Lindqvist M, Ryden S, et al. Breast cancer screening with mammography: overview of Swedish randomised trials. Lancet 1993;341:973–8.[CrossRef][Web of Science][Medline]
(8) Miller AB, Baines CJ, To T, Wall C. Canadian National Breast Screening Study: 2. Breast cancer detection and death rates among women aged 50 to 59 years. Can Med Assoc J 1992;147:1477–88.[Abstract]
(9) Taplin S, Ichikawa L, Yood MU, Manos M, Geiger AM, Weinmann S, et al. Reason for late-stage breast cancer: absence of screening or detection, or breakdown in follow-up? J Natl Cancer Inst 2004;96:1518–27.
(10) Porter PL, El-Bastawissi AY, Mandelson MT, Lin MG, Khalid N, Watney EA, et al. Breast tumor characteristics as predictors of mammographic detection: comparison of interval- and screen-detected cancers. J Natl Cancer Inst 1999;91:2020–8.
(11) Yankaskas BC, Schell MJ, Bird RE, Desrochers DA. Reassessment of breast cancers missed during routine screening mammography: a community-based study. AJR Am J Roentgenol 2001;177:535–41.
(12) Goodson WH 3rd, Moore DH 2nd. Causes of physician delay in the diagnosis of breast cancer. Arch Intern Med 2002;162:1343–8.
(13) Caplan LS, May DS, Richardson LC. Time to diagnosis and treatment of breast cancer: results from the National Breast and Cervical Cancer Early Detection Program, 1991-1995. Am J Public Health 2000;90:130–4.
(14) Mann BD, Giuliano AE, Bassett LW, Barber MS, Hallauer W, Morton DL. Delayed diagnosis of breast cancer as a result of normal mammograms. Arch Surg 1983;118:23–4.
(15) Patel MR, Whitman GJ. Negative mammograms in symptomatic patients with breast cancer. Acad Radiol 1998;5:26–33.[CrossRef][Web of Science][Medline]
(16) Zapka JG, Taplin SH, Solberg LI, Manos MM. A framework for improving the quality of cancer care: the case of breast and cervical cancer screening. Cancer Epidemiol Biomarkers Prev 2003;12:4–13.
(17) Goins KV, Zapka JG, Geiger AM, Solberg LI, Taplin S, Yood MU, et al. Implementation of systems strategies for breast and cervical cancer screening services in health maintenance organizations. Am J Manag Care 2003;9:745–55.[Web of Science][Medline]
(18) Taplin SH, Barlow WE, Ulcickas Yood M, Westbrook E, Geiger AM, Bischoff K, et al. Re: Breast cancer screening comes full circle. J Natl Cancer Inst 2005;97:461.
(19) U.S. Preventive Services Task Force. Screening for Breast Cancer Release Date: February 2002 (available at: http://www.ahrq.gov/clinic/uspstf/uspsbrca.htm). Last accessed: December 10, 2004.
(20) U.S. Preventive Services Task Force. Guide to clinical preventive services, 2nd ed. Baltimore (MD): Williams & Wilkins; 1996. p. 73–87.
(21) Oestreicher N, White E, Lehman CD, Mandelson MT, Porter PL, Taplin SH. Predictors of sensitivity of clinical breast examination (CBE). Breast Cancer Res Treat 2002;76:73–81.[CrossRef][Web of Science][Medline]
(22) White E, Miglioretti DL, Yankaskas BC, Geller BM, Rosenberg RD, Kerlikowske K, et al. Biennial versus annual mammography and the risk of late-stage breast cancer. J Natl Cancer Inst 2004;96:1832–9.
(23) Homer MJ. Breast imaging, standard of care, and the expert. Radiol Clin North Am 2004;42:963–74, viii.[CrossRef][Web of Science][Medline]
(24) Langlands AO, Gebski V, Hirsch D, Tattersall MH. Delay in the clinical diagnosis of breast cancer: estimating its effect on prognosis, with particular reference to medical litigation. Breast 2002;11:386–93.[CrossRef][Web of Science][Medline]
(25) Caplan LS, Helzlsouer KJ, Shapiro S, Wesley MN, Edwards BK. Reasons for delay in breast cancer diagnosis. Prev Med 1996;25:218–24.[CrossRef][Web of Science][Medline]
(26) Meechan G, Collins J, Petrie KJ. The relationship of symptoms and psychological factors to delay in seeking medical care for breast symptoms. Prev Med 2003;36:374–8.[CrossRef][Web of Science][Medline]
(27) Nosarti C, Crayford T, Roberts JV, Elias E, McKenzie K, David AS. Delay in presentation of symptomatic referrals to a breast clinic: patient and system factors. Br J Cancer 2000;82:742–8.[CrossRef][Web of Science][Medline]
(28) Feig SA, Sickles EA, Evans WP, Linver MN. Re: Changes in breast cancer detection and mammography recall rates after the introduction of a computer-aided detection system. J Natl Cancer Inst 2004;96:1260–1.
(29) Destounis SV, DiNitto P, Logan-Young W, Bonaccio E, Auley ML, Willison KM. Can computer-aided detection with double reading of screening mammograms help decrease the false-negative rate? Initial experience. Radiology 2004;232:578–84.
(30) Smith AP, Hall PA, Marcello DM, Emerging technologies in breast cancer detection. Radiol Manag 2004;26:16–24.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||