BREAST CARE / BREAST CANCER

Overview The Kaiser Permanente Breast Care Management Algorithm provided on this site was developed by the Inter-Regional Breast Cancer leaders group (IRBC). This multidisciplinary group includes physicians from Primary Care, Surgery, Oncology, Obstetrics and Gynecology, Radiology, Mammography, Genetics and Women’s Services and representatives from various regional Breast Cancer Task force groups, Clinical Nursing, Quality Resource & Risk Management, Public Relations & Issues Management, Prevention Services, and the Permanente Federation. The algorithm was developed to:

• Improve the quality of care for our members with breast complaints, • Improve the timeliness of the identification of breast abnormalities and diagnosis of breast cancer,

• Improve the satisfaction of members with breast complaints, and • Respond to the increase in malpractice allegations of failure to diagnose breast cancer. In 2002, the IRBC group held periodic conference calls to develop information to assist primary care clinicians in improving the quality of care for patients with breast complaints. A multidisciplinary consensus-based method was used to develop the content of the algorithm. The group also identified additional information and resources available internally and externally which would support implementation.The Breast Care Leaders in each Region have been encouraged to review and modify the algorithm to reflect local operations. Therefore, prior to use, PCPs are advised to contact a Regional member of the Inter-Regional Breast Care leaders group about revisions for your Region This site is for use within Kaiser Permanente only. What is Available on this Site? The IRBC group and the project management staff from the Permanente Federation worked together to define the project scope and develop the following products and information: I. Rationale II. Breast Care Management Algorithm: The Algorithm provides suggestions to help primary care providers along a care path for evaluating a patient’s breast complaint (e.g., Clinical Breast Exam, abnormal screening mammogram follow-up suggestions, inflammation, breast mass/lumps, spontaneous nipple discharge, and breast pain) to the point where cancer is ruled in or out. III. Information Related to Clinical Practice Guidelines for Breast Cancer Screening: Contact information regarding regional Clinical Practice Guidelines for Breast Cancer Screening. IV. New Technology Report: This section provides information regarding new technologies

used to diagnose breast cancer. V. Breast Cancer Tracking System : Information regarding various computer tracking systems for breast cancer is provided in this section.

Version 2.0 October, 2003 -1-

VI. Sample Presentations, Additional Resources and Regional Continuing Medical Education (CME) Contact information : Sample presentations describing Issues in Breast Cancer and Risk Management information and other available resources to support Breast Care are listed in this section. Contacts for regional CME are provided. VII. Bibliography : This section contains the list of references utilized for development of the algorithm VII. Inter-Regional Breast Cancer leaders group (IRBC): Contact information for the IRBC.

Version 2.0 October, 2003 -2-

I. Rationale • Breast cancer is the most frequently diagnosed cancer and the second leading cause of cancer death in women. • 205,000 new cases of breast cancer will be diagnosed in the U.S. in 2002 • An average women has a 2.5% chance of developing breast cancer between the ages of 35 and 55. • Approximately 70-75% of breast cancers are diagnosed in women over the age of 50. • There are 127 cases per 100,000 women age 40-44 years compared to 450 per 100,000 cases in women age 70-74 years. • Although a woman, between 25 and 34 years of age, has a relatively low risk of developing breast cancer compared to older women, nationally, KP can expect 100 women in this age group to be diagnosed with breast cancer in 2002.1,2, Therefore, complete follow-up and documentation of care to the point of normal findings, is essential for all breast complaints, regardless of a woman’s age. • Family history of breast cancer predicts approximately 15% of all cases. A positive family history of breast cancer is defined as having a first-degree relative who developed breast cancer before the age of 50. Those with relatives whose onset of breast cancer was after age 50 are not considered at higher risk. The risk for women with a first-degree relative with early breast cancer is increased fourfold. Other risk factors include history of previous breast cancer, women with atypical hyperplasia on breast biopsy, late age of first pregnancy, nulliparity, and high socioeconomic status. 1. Actual membership from third quarter 2002 per Performance Analysis, Program Offices. 2. Surveillance, Epidemiology and End Results Program, National Cancer Institute, SEER Incidence Crude Rates, 11 Registries, Years of Diagnosis 1992 -1999, All Races, Females.

Version 2.0 October, 2003 -3-

II. Breast Care Management Algorithm • The Kaiser Permanente Breast Care Management Algorithm provides suggestions to help primary care providers along a care path for evaluating a patient’s breast complaint (e.g., Clinical Breast Exam, Abnormal Screening Mammogram Follow-up Suggestions, Breast Mass/Lumps, Inflammation, Spontaneous Nipple Discharge, and Breast Pain) to the point where cancer is ruled in or out. This algorithm is not an evidence based clinical practice guideline. A multidisciplinary consensusbased method was used to develop the algorithm. The suggestions provided in the algorithm do not replace the reasonable exercise of independent clinical judgment in any particular set of circumstances for each patient encounter. The Breast Care Leaders in each Region have been encouraged to review and modify the algorithm to reflect local operations. Therefore, prior to use, PCPs are advised to contact a Regional member of the Inter-Regional Breast Care leaders group about revisions for your Region. The algorithm may be viewed and navigated on this sight or viewed and printed using Adobe

Acrobat:

Algorithm: • Introduction/ Clinical Visit for Breast Complaint • Abnormal Screening Mammogram Follow-up Suggestions, • Breast Mass/Lumps • Inflammation • Spontaneous Nipple Discharge • Breast Pain

Please direct any questions about the algorithm to Robin Cisneros, Director of Medical Technology Assessment, Quality and Performance Improvement, The Permanente Federation at 510-271-5863.

Version 2.0 October, 2003 -4-

Breast Care Management Algorithm

CLINIC VISIT Evaluation of the Breast Complaint Obtain history including previous mammogram(s) (plus other imaging) findings, and follow-up; medication history including hormone and birth control; and onset and duration of symptoms of current complaint. If prior mammogram was abnormal determine if there is documented completion of follow-up. If not, proceed to page 3 for abnormal mammogram follow-up algorithm. Identify risk factors personal or family (first-degree relative) history of breast cancer including age of diagnosis, atypia on previous breast biopsies, age of menarche, age at first completed pregnancy, and breastfeeding history. Conduct a Clinical Breast Exam (CBE) Inspect both breasts (unaffected breast first) for ulceration or contour change, nipple skin changes, or nipple discharge. Palpate breast in both the upright and supine positions to determine the presence of a palpable mass/lump. Evaluate nodes (axillary, supraclavicular). Assess risk Consider history and physical findings, risk factors and patient concerns Link to NCI Risk Calculator: http://bcra.nci.nih.gov/brc/q1.htm Document all characteristics of the history and physical findings Document symptom characteristics such as size, location, texture, mobility and character for reference in follow-up examinations. A breast clinic record template is available for use. Order breast imaging if indicated or per regional screening/diagnostic guidelines.

Signs of Inflammation

No Signs of Inflammation

(Redness, Warmth, Swelling, Purulent Discharge, Draining Wound)

Inflammation

Characterize Primary Complaint

Breast

Mass/Lump

If not palpable,

• Educate on risk and screening intervals • Re-examine in one month.

See page 4

©2003 Kaiser Permanente Medical Care Program

If palpable, See page 5

Version: January 28, 2003 Next review: January 2004

Spontaneous

Nipple

Discharge

Breast Pain

See page 6

See page 7

(without mass)

(without mass)

—For use within Kaiser Permanente only—

Breast Care Management Algorithm

Abnormal Screening Mammogram Follow-up Abnormal Screening Mammogram

Full Diagnostic Mammogram and/or Ultrasound

Benign

Probably Benign

Suspicious of Malignancy

Highly Suggestive of Malignancy

BI RADS™ Assessment Categories 1,2

BI RADS™ Assessment Categories 3

BI RADS™ Assessment Category 4

BI RADS™ Assessment Category 5

Short-term follow-up per appropriate specialist’s recommendations

Clinical breast exam (f not previously completed) Biopsy

Benign

Indeterminate

Insufficient sample Lack of imaging/ pathology correlation

Second biopsy OR Surgical Follow-up

Malignant OR Any type of atypia

Additional Additional Considerations Considerations If the patient is If the patient is pregnant or lactating, pregnant or lactating, consult the appropriate consult the appropriate specialist prior to specialist prior to following this care following this care path. path.

SCPMG Breast Cancer Member Satisfaction Survey findings suggests that the time frame from suspicion to diagnosis should not exceed 14 days.

Surgical Follow-up

Document mammogram and pathology findings

Document surgical outcomes and other treatment, if applicable

Inform all providers and patient of findings and treatment results

Inform all of appropriate follow-up including next screening date (mammogram and CBE)

©2003 Kaiser Permanente Medical Care Program

Version: January 28, 2003

Next review: January 2004

—For use within Kaiser Permanente only—

Breast Care Management Algorithm Inflammation (Inflammatory carcinoma should always be considered)

Treat with Warm Packs, NSAIDS, Antibiotics1,2

Breastfeeding

Not-Breastfeeding

Advise to continue breastfeeding Consider Ob/Gyn consultation

Continue treatment 5-7 days or less

Continue treatment 7-10 days or less

SCPMG Breast Cancer Member Satisfaction Survey findings suggest that the time frame from suspicion to diagnosis should not exceed 14 days.

Evaluate signs and symptoms

Signs and Symptoms Alleviated

Signs and Symptoms Unchanged or Increased

Annotations to Algorithm Annotations to Algorithm

Consider biopsy and additional imaging If breastfeeding, consider discontinuing

Follow-up to resolution of symptoms

Documentation of care to the point of normal physical findings

(1) Antibiotics (1) Antibiotics

Cephalexin – 500 mg po QID x 10 days Cephalexin – 500 mg po QID x 10 days Dicloxacillin – 500 mg po QID x 10 days Dicloxacillin – 500 mg po QID x 10 days E-mycin– 333 mg. TIC x 10 days if allergic E-mycin– 333 mg. TIC x 10 days if allergic to penicillin (if not breastfeeding) to penicillin (if not breastfeeding) (2) If implants, consider longer course of (2) If implants, consider longer course of antibiotics or referral to Plastic Surgery. antibiotics or referral to Plastic Surgery.

Inform patient of next screening date (mammogram and CBE)

Inform all involved providers of resolution

©2003 Kaiser Permanente Medical Care Program

Version: January 28, 2003 Next review: January 2004

—For use within Kaiser Permanente only—

Breast Care Management Algorithm EVALUATION OF BREAST MASS/LUMP

Pre-menopausal

Cyclical Re-examine after next cycle

(or 4 weeks)

Resolved

Post-menopausal

Non-Cyclical

Annotations to Algorithm Annotations to Algorithm (1) If patient is under 35, (1) If patient is under 35, consider referring the patient consider referring the patient to the appropriate specialist to the appropriate specialist before ordering any breast before ordering any breast imaging. imaging.

Unresolved

months1

Order a diagnostic mammogram if none in last 6 Follow all radiological recommendations Consider additional imaging Obtain consultations and/or additional tests

Suspected Cyst (Fluid)

Suspected Solid

Not a suspected Cyst or Solid

Aspirate to total resolution

(No fluid)

Re-examine in 1 month

Not Bloody And No Residual Mass exists Examine at 4-8 weeks Mass Does NOT Reoccur

Bloody OR Indeterminate cytology OR Residual Mass

Exists

Persistent Residual Mass

Mass Resolves

Biopsy (regardless of mammogram results)

Benign Consider complete excision

Mass Reoccurs

SCPMG Breast Cancer Member Satisfaction Survey findings suggest that the time frame from suspicion to diagnosis should not exceed 14 days.

Malignant or Indeterminate Pathology

Surgical Follow-up

Document mammogram and pathology findings

Document surgical outcomes and other treatment, if applicable

Inform all providers and patient of findings and treatment results

Inform all of appropriate follow-up including next screening date (mammogram and CBE)

©2003 Kaiser Permanente Medical Care Program

Version: January 28, 2003 Next review: January 2004

—For use within Kaiser Permanente only—

Breast Care Management Algorithm Spontaneous Nipple Discharge (other than lactating state) (without mass, if mass see page 5)

Evaluate and document the characteristics of discharge Single Duct may be cancer. Multiple Duct is not as worrisome. Order diagnostic bilateral mammogram if none in the last 6 months and patient is 35 years or older. Follow all radiological recommendations1 Unilateral

Bilateral

Non-Bloody Discharge

Clear, yellow, green, gray or milky.

Bloody

Discharge

Rule Out Mechanical stimulation Rule Out medication2

Consider additional imaging

SCPMG Breast Cancer Member Satisfaction Survey findings suggest that the time frame from suspicion to diagnosis should not exceed 14 days.

Surgical

Follow-up

•

If the patient is under 35, If the patient is under 35, consider consultation consider consultation with the appropriate with the appropriate specialist before ordering specialist before ordering any breast imaging. any breast imaging. (2) Medications that may (2) Medications that may cause discharge include: cause discharge include: Phenothiazine, Phenothiazine, Reserpine, Estrogen, and Reserpine, Estrogen, and Opiates. Opiates.

Endocrine Evaluation (prolactin/TSH)

Normal

Prolactin/

TSH

Annotations to Algorithm Annotations to Algorithm

Abnormal

Prolactin/

TSH

•

Endocrine

Follow-up

Document mammogram and pathology findings

Document surgical outcomes and other treatment, if applicable

Inform all providers and patient of findings and treatment results

Inform all of appropriate follow-up including next screening date (mammogram and CBE)

©2003 Kaiser Permanente Medical Care Program

Version: January 28, 2003 Next review: January 2004

—For use within Kaiser Permanente only—

Breast Care Management Algorithm

Breast Pain (without mass, if mass see page 5)

Non- Cyclical

Cyclical Resolved

Not resolved

Reexamination

Ob/Gyn

after 2 cycles1

Evaluate medication (e.g., hormones, birth control) R/O pregnancy Differential diagnosis (Breast disorder, chest wall pain (bone, muscle, joint), pleural disease, cardiac disease, GI disease) Order a diagnostic bilateral mammogram if none in last 6 months and patient is 35 or older. Follow all radiological recommendations 2

Bilateral

Unilateral

Consult

Focal

Annotations to Algorithm Annotations to Algorithm

(1) 2 cycles if pre(1) 2 cycles if premenopausal or 2 months if menopausal or 2 months if post-menopausal post-menopausal (2) If the patient is under (2) If the patient is under 35, consider consultation 35, consider consultation with the appropriate with the appropriate specialist before ordering specialist before ordering any breast imaging. any breast imaging. (3) Symptom (3) Symptom management and management and treatment may include: treatment may include: For mild to moderate For mild to moderate cyclical pain there are no cyclical pain there are no evidence-based evidence-based treatments available. treatments available. Non-evidence based, nonNon-evidence based, nonhormonal therapy hormonal therapy recommended for 4-6 recommended for 4-6 months includes: months includes: reassurance, caffeine reassurance, caffeine reduction, supportive wellreduction, supportive wellfitting bra, vitamin E, B6, fitting bra, vitamin E, B6, Evening Primrose oil, or Evening Primrose oil, or over-the-counter over-the-counter analgesics. analgesics. Moderate to severe Moderate to severe cyclical pain: Possible cyclical pain: Possible hormone therapy. hormone therapy.

Global

Symptom Management3 Follow-up after 2 cycles1

Resolves

Persistent Pain

> 35

< 352

Bilateral mammogram if none in last 6 months

Negative Bilateral Mammogram Symptom Management

Abnormal Bilateral Mammogram Follow all radiologic recommendations

Follow-up after 2 cycles1

Immediate referral to appropriate specialists

3

Consider additional tests and consultations Persistent Pain

SCPMG Breast Cancer Member Satisfaction Survey findings suggest that the time frame from suspicion to diagnosis should not exceed 14 days.

Pain Resolves

Symptom Management3 and follow-up PRN Consider additional imaging studies, biopsy, and consultations

Document mammogram and pathology findings Document surgical outcomes and other treatment, if applicable Inform all providers and patient of findings and treatment results Inform all of appropriate follow-up including next screening date (mammogram and CBE)

©2003 Kaiser Permanente Medical Care Program

Version: January 28, 2003 Next review: January 2004

—For use within Kaiser Permanente only—

III. Information Related to Clinical Practice Guidelines for Breast Cancer Screening

This section provides contact information for your regional representative who can give you specific information regarding your regions Clinical Practice Guidelines for Breast Cancer Screening.

Internal Web Site (if available)

Contact for Breast Cancer Screening Information

ADDRESS Phone/Fax/E-mail

Sue Jane Fox Prevention Specialist II Kaiser Permanente, Colorado

Department of Prevention 10350 East Dakota Avenue Denver, CO 80231 303-344-7256 303-344-7721 Fax [email protected]

Breast Cancer Screening

Adrienne Mims, MD, Chief of Prevention and Health Promotion Or John Zetzsche Director of Population Based Care Kaiser Permanente, Georgia

9 Piedmont Center 3495 Piedmont Road NE Atlanta, GA 30305-1736 404-364-7289 404-364-4798 [email protected]

Clinical Practice and Prevention Guidelines Use Bookmarks under Adult Medicine to locate Breast Cancer Screening Guidelines

Charles Sakamoto, MD Chief of Family Practice

KP Mililani Clinic 95-660 Lanikuhana Avenue Mililani, HI 96789 808-432-4254 808-625-4242 [email protected]

Or

404-364-7297 404-364-4798 [email protected]

Preventive Care Services for Adults

Jean Kumamoto, Director of Health Education Kaiser Permanente Hawaii

KP Honolulu Clinic 1010 Pensacola Street Honolulu, HI 96814 808-432-2260 808-432-2249 [email protected]

Hanadi Shamkhani, MD Director for Practitioner Performance Review and Oversite Kaiser Permanente, Mid-Atlantic Susan Kutner, MD Department of Surgery Breast Cancer Care Committee Permanente Medical Group, Northern California

KPMG, 2101 E. Jefferson Street, 2 West, Rockville, MD 20852 301-816-6624 [email protected]

Preventive Care of Asymptomatic Adults

KP, 280 Hospital Parkway San Jose, CA 95119 408-972-6017 Tie-line: 8-440-6017 [email protected]

Screening Neoplastic Diseases, Breast Cancer

Version 2.0 October, 2003 -5-

Contact for Breast Cancer Screening Information Nancy Stevens, PhD Prevention Systems Manager Kaiser Permanente, Northwest

Internal Web Site (if available)

ADDRESS Phone/Fax/E-mail

500 NE Multnomah Suite 100 Portland Oregon 97232 503-813-3828 503- 813-2669 Fax [email protected]

Group Health

Breast Cancer Screening

Breast Cancer Screening Guideline

Wendy Dahar Health Outcomes Coordinator Kaiser Permanente, Ohio

Division of Clinical Innovation 5410 Lancaster Street Brooklyn Heights, OH 44131 216- 778-6088 216-778-6090 [email protected]

Preventive Health Recommendations for Screening

Susan Chen, RN, MSN Clinical Services Kaiser Permanente, Southern California

KP, 393 E. Walnut Street, 7th Floor Pasadena, CA 91101 626-405-2554 Tie-line:8- 335-2554 626-405-6991 Fax [email protected]

Breast Cancer Screening

Version 2.0 October, 2003 -6-

IV. New Technology Report

This section provides information regarding new technologies used to diagnose breast cancer. •

Full field digital mammography (3-2002, 12-2002)

•

Computer-assisted detection mammography (3-2002, 12-2002)

•

Random epithelial cell cytology evaluation (3-2002, 12-2002)

•

MRI

•

PET Scan

Version 2.0 October, 2003 -7-

New Technology

Breast Cancer Screening

Summary of Technology Assessments March 18, 2002

Full-Field Digital Mammography (FFDM)

FFDM is proposed as an alternative to screen-film mammography (SFM). SFM, the conventional form of mammography, uses film as the medium for image acquisition, storage and viewing. SFM is neither perfectly sensitive nor highly specific. Dense breast tissue and diffuse involvement of the breast with tumor tends to reduce the sensitivity of screening mammography. Approximately 10–20% of breast cancers that are detected by self-breast examination or physical examination are not visible by screen-film mammography. (Baker 1982) In addition, when lesions are detected by mammography and biopsy is recommended by experienced radiologists, only 5–40% of lesions prove to be malignant. (Sickles 1991). A major limitation of screen-film mammography is the film itself. The film serves as the medium of image acquisition, storage and display. Once a screen-film mammogram is obtained, it cannot be significantly altered. Contrast loss due to film underexposure, especially of dense glandular tissues, cannot be regained through film display. Radiologists cannot manipulate the image directly. Improvements in image display involve acquiring more images with magnification or focal compression (and thus exposing the patient to more radiation), or looking at the images with a hot light or magnifying glass. Breast cancer is often quite similar in X-ray absorption to surrounding normal dense breast tissue.

Digital mammography uses solid state digital detectors and has the potential to improve breast cancer detection and breast lesion characterization . The distinguishing characteristic of FFDM is digital image acquisition, without the need for film. The processes of acquisition, storage and display can be separated and individually optimized. The digital images can be processed by a computer and displayed in multiple formats, on film or on a monitor. Since the steps of image acquisition and display are separated, each can be optimized. In addition, image storage, transmission and retrieval can be improved. Computer-aided diagnosis (CAD) software to assist the radiologist in interpreting the images can also be utilized to provide a second opinion and improve diagnostic accuracy. Furthermore, digital mammograms may provide the best estimate of mammographic density, a predictor of breast cancer risk. (Boyd et al. 1998). Two FDA systems have received PMA approval by the FDA: GE Senorgraphe 2000 system (1/28/00) and Fischer Imaging Corportation Senoscan System (9/25/01). This BCBSA TEC assessment (March 2002) reviews the evidence comparing FFDM and SFM for use in two settings: the screening population and the population of patients referred for diagnostic mammography based on initial suspicious findings. Recall rates, biopsy rates and cancer detection are of primary importance in this review. The assessment compared the results of screen-film mammography (SFM) and full-field digital mammography (FFDM), performed on either the same group or two groups of patients. Additionally, studies had to report on cancer detection based on histologic results for at least some of the patients in the sample. For the screening population, the assessment concluded that there is insufficient evidence to permit conclusions about the effects of full-field digital mammography (FFDM) relative to screen-film mammography (SFM) in screening for breast cancer. Two main studies comprise the currently available data, both of which used the GE system. The first study, performed in two U.S. centers, was reported in a journal article and a conference slide presentation. Target enrollment in this study is 15,000 patients, but the most recent update is for 4,521 patients. The recall rate for FFDM (11.8%) is significantly lower than that for SFM (15.0%). The biopsy rate is also significantly lower for FFDM compared with SFM (11.8% versus 14.3%). FFDM detected 27

Version 2.0 October, 2003 -8-

of 51 cancers while SFM detected 34 cancers. The positive predictive values for both modalities was 3.4%. The sensitivities, 52.9% for FFDM and 66.7% for SFM, were not significantly different in this interim analysis. If the magnitude of this difference in sensitivity remains at this level through subject accrual increases, FFDM could be shown to significantly inferior to SFM in cancer detection. This underscores the need to reserve conclusions until final reporting of the study.

The second study, conducted in Oslo, Norway, has been reported in a proceedings paper and two conference abstracts. Statistical test results are lacking for all comparisons between FFDM and SFM. Unlike the findings of the U.S. study, recall rates in the Norwegian study were higher for FFDM than for SFM. However, recall rates varied from in all three data sources, although 2 of them were based on the same number of patients (3,683 in the largest update). The number of cancers detected by FFDM was 23, compared with 27 for SFM, a nonsignificant difference. Sensitivities was 74.2% for FFDM and 87.1% for SFM. It is uncertain whether this study was powered to detect a difference in sensitivities as large as 13 percentage points. A third study using the GE system, presented only as a conference abstract, does not clearly state whether patients presented for screening or diagnostic mammography, or both. Separate groups of patients underwent FFDM (n=1,849) and SFM (n=5,415). It is also unclear whether the cancer detection rates for FFDM and SFM differed significantly. A fourth study using the GE system is being conducted by the American College of Radiology Investigative Network (ACRIN), with a target of 49,500 patients. No reports on this large study are available yet. A single journal article, presenting interim results for less than half of study target enrollment, has been published on use of FFDM in screening. All other sources of data come from a proceedings paper or conference abstracts. Additional detailed publications for large numbers of patients are needed before conclusions can be reached about the detection of breast cancer by FFDM relative to SFM in a screening population. Diagnostic Population. Patients with findings suspicious for breast cancer based on initial SFM or breast physical examination are referred for additional work-up and comprise the diagnostic mammography population. Four studies are available on the GE system in this population: one journal article on 692 patients, a proceedings paper on 625 patients, one conference abstract on 55 patients and another abstract on 100 patients. The published article reported relatively low inter-test agreement between FFDM and SFM (a kappa value as low as 0.20). A total of 16 cancers were detected by SFM and 13 were detected by FFDM. Sensitivities were 72.7% for FFDM and 88.9% for SFM. The study appeared to lack sufficient power to detect a difference in sensitivities as large as 16 percentage points. The proceedings paper reported sensitivities of 68.2% for FFDM and 69.5% for SFM, indicating that FFDM was not inferior to SFM, however the number of cancers detected was not specified. This paper, reported in a low degree of detail does not resolve concerns raised in the only published paper about what may be a substantially lower sensitivity for FFDM. The two other GE studies have only been reported as conference abstracts and do not provide sufficient detail to permit meaningful analysis of methods and results. Thus, conclusions about the performance of FFDM relative to SFM in a diagnostic mammography population await additional published reports for large patient samples. Evidence on the Fischer FFDM system is limited to a single conference abstract of 247 patients. This source did not mention results of statistical tests comparing FFDM and SFM. Sensitivities were reported to be 66% for FFDM and 74% for SFM, while specificities were 67% for FFDM and 60% for SFM. The area under the ROC curve was 0.715 for FFDM and 0.765 for SFM. A journal article described results for 22 patients imaged with the Fuji digital luminescence mammography (DLM) device. Area under the ROC curve was 0.781 for DLM, compared with 0.732 for SFM. These results are clearly inadequate to permit conclusions about relative performance of these digital mammography systems versus SFM.

Version 2.0 October, 2003 -9-

Computer-Assisted Detection (CAD) , also called computer-aided detection, CAD systems for mammography digitize mammogram films, analyze them for patterns that may represent microcalcifications or masses, and provide a marked image (either printed copy or on a display monitor) that is compared to the original mammogram. CAD systems are not designed to replace the original mammograms or the radiologist(s) reading the film but has been proposed as an alternative to double reading. Designed as a tool for the screening environment, the value of this technology depends on the ability to diagnose and treat early disease. You did not receive an assessment prior to today on this technology. However, SCPMG’s TAG unit has addressed this technology. We will forward the final assessment to you as soon as it is available but I was told that I could discuss the draft assessment with you today. Several types of computer-assisted detection devices are FDA-approved for marketing in the United States. , the ImageChecker M1000 System produced by R2 Technologies, Inc. (Los Altos, CA), the CADx Medical System’s Second Look system, and ISSI’s MammoReader are FDA-approved for marketing in the United States. The MAMMEX TR system is being tested at the University of Southern California, and study results were recently reported in abstract form. This system has been installed in Switzerland.

The assessment didn’t results in a a precise determination of the sensitivity or specificity of CAD mammography because the available evidence is lacking. The NTAT recommends that the NTDST carefully review the estimates of the marginal increase in sensitivity along with the associated number of expected “call backs” and incorporate operational considerations in the evaluation of CAD mammography. The NTDST, with the support of the NTAT and the key technology clinical committees, should determine if CAD technology is appropriate for deployment in Kaiser Permanente, Southern California. The available published evidence does not permit definitive conclusions regarding the sensitivity and specificity of CAD. In addition, the evidence does not permit a determination of the effect of CAD screening mammography on intermediate or long-term health outcomes. However, because CAD is used as a “second read” and identifies locations of suspected mammographic abnormalities, the sensitivity is expected to be the same or higher compared to a single reading. The false positive rate is also expected to be higher. The literature demonstrates that the effect of CAD on sensitivity and specificity is dependent on the experience and skill of the radiologist with most benefit accruing to radiologists less experienced in reading mammograms. In addition, sensitivity and specificity are likely to be related to the size of the tumor, type of tumor and age of the patient. Moreover, the ratio of extra workups to extra cancers is sensitive to small changes in sensitivity and specificity and these are quite variable from radiologist to radiologist and from study to study. As calculated from the average increase in sensitivity reported in the literature (5 retrospective studies with an enriched case selection and variability in radiologist experience), the marginal increase in sensitivity for CAD plus radiologist compared to radiologist alone is 6.5%. A large prospective study (Freer and Ulissey, 2001) found a 19.5% increase in cancers detected when using CAD. A decrease in specificity of 1.1% can be calculated from two of the small retrospective studies (Brem and Schoonjans, 2001; Thurfjell et al., 1998). Freer and Ulissey found a 1.2% increase in recall rate. This could result in an estimated 12/1000 women being called back for further medical and/or surgical follow up to determine whether or not the highlighted abnormalities identified on CAD and warranted as actionable by the radiologist are undetected breast cancers. These numbers may not be applicable to all SCPMG radiologists and/or may vary among departments or among radiologists within a department. In Kaiser Permanente Southern California, it is estimated that CAD mammography would generate in the range of 100-125 callbacks per 10,000 mammograms. Of these 100-125 callbacks, it is estimated that 2-5 additional breast cancers would be detected.

Version 2.0 October, 2003 - 10 -

PMG expert opinion is divided. All of the experts contacted agree that CAD mammography is an emerging technology, and several physicians see it as a useful adjunct to radiologist-read mammography because of its demonstrated ability to pick up “missed” cancers. Others have concerns about the specificity and the cost of the device, and question whether it would be worth the investment.

SCPMG Thomas Mahon, MD (Diagnostic Imaging, LAMC, 8-363-2374) Ideally we would all perform double reading of mammograms, however, few if any of the S.C. Kaiser Radiology Departments have sufficient Radiologists to do this. Double reading increases cancer detection rates by 4 to 15%. Therefore, is CAD a useful and viable alternative to double reading? As you point out there are no studies comparing these two techniques. I agree the studies published so far often have potential bias and have an "enriched" case selection, however, they do seem to suggest that CAD increases detection rate and in the study by Freer of 12,860 patients the patient recall rate did not significantly increase. I think the paper by Jiang in Radiology, September 2001, is important because he states that with CAD there was a two-thirds decrease in substantial disagreements between different radiologists assessing the same group of mammograms. The author thus suggests CAD may result in an improvement in both accuracy and consistency in image interpretation. The main potential disadvantages of CAD would be if the recall rate was to significantly increase and this is not suggested by the studies, nor by the users I talked to in KP in Northern California. Another potential problem would be if use of CAD were to increase the time spent by Radiologists reading each mammogram, and again the users I talked to did not report this. The false positive rate is inconvenient, but if it does not affect either the recall rate or the time spent interpreting a mammogram, then it is not a significant issue. A false negative rate would only be significant if the radiologist using CAD were to miss a cancer that he or she without CAD would diagnose. By that I mean, were CAD to miss a large spiculated mass, then this lesion should be easily seen by the Radiologist. Whereas it is more important that CAD should detect a subtle group of calcifications in a dense breast that could be easily missed by the Radiologist interpreting without CAD. I think it is important that CAD detect clinically relevant cancers, and this is suggested by the paper by Burhenne et all, where of the 8 cancers seen by CAD, 3 were invasive cancers. As you know missed breast cancers is a very important source of litigation and the average award in this area settled by Kaiser, I understand has now reached $700,000. It would be of interest to know if CAD could successfully decrease successful litigation. There are a number of ways it could potentially do this. Firstly, it may decrease the rate of missed cancers, particularly in the important demographic of younger women with dense breasts that can obscure calcifications. Secondly, that by using CAD we can say that the mammograms are in effect second read. Thirdly, that lesion detectability by CAD might establish a reasonable standard of detectability. The current status of CAD appears similar to that in the early 1990's when Stereotactic breast biopsy was introduced. Stereotactic biopsy has proven to be a great success and has literally revolutionized the way we manage breast lesions and yet to this day the studies in the literature remain relatively small, with potential bias. To this day the only Radiology modality that I am aware of that has been tested with large randomized studies is Mammography, and the interpretation of the results of these studies remains probably the single most controversial issues in all of Medicine. CAD will be standardly available when we go over to Digital Mammography, but this is still several years down the road and I think we need to consider CAD before then. The expense of CAD is an issue. If CAD were to decrease medical malpractice claims then that could go towards paying for it. TPMG KM Tan, MD, Chair, KPNC Imaging Clinical Technology Committee. (as of March, 2001) A FEW YEARS AGO, KPNC HAD THREE BETA SITES FOR R2 IMAGECHECKER: REDWOOD CITY, SOUTH SAN FRANCISCO, AND SAN FRANCISCO. FOLLOWING THE TRIAL, THE VENDOR OFFERED THE EQUIPMENT TO EACH SITE FOR APPROXIMATELY $75,000, DISCOUNTED FROM THE RETAIL PRICE OF APPROXIMATELY $170,000. DR. TAN’S IMAGING COMMITTEE WOULD NOT APPROVE REGIONAL FUNDING. REDWOOD CITY AND SAN FRANCISCO CHOSE TO FUND THE EQUIPMENT OUT OF MEDICAL CENTER FUNDS; SOUTH SAN FRANCISCO DID NOT BUY THE R2 IMAGECHECKER. TPMG RADIOLOGY CONSIDERS THE R2 IMAGECHECKER A TECHNOLOGY Version 2.0 October, 2003 - 11 -

THAT MAY BE NICE TO HAVE, BUT A LUXURY. EVEN THE TWO TPMG RADIOLOGISTS, SHERRY BUTLER, MD, SAN FRANCISCO, AND DEBORAH KASS, MD, REDWOOD CITY, WHO EMPLOY THIS TECHNOLOGY WITH THE GREATEST FREQUENCY, HAVE STATED TO DR. KAM THAT R2 IMAGECHECKER IS NOT COMMUNITY STANDARD. DR. TAN ESTIMATES THAT, GIVEN THAT THE UNITED STATES HAS MORE THAN 10,000 BREAST IMAGING PRACTICES, A NATIONWIDE TOTAL OF 140 TO 180 R2 IMAGECHECKER SYSTEMS IN USE MEANS THAT PERHAPS 1% OF MAMMOGRAMS IN THE UNITED STATES ARE NOW READ WITH THE ASSISTANCE OF R2 IMAGECHECKER. DR. KASS HAS INFORMED DR. TAN THAT, IN HER SEVERAL YEARS OF EXPERIENCE WITH THE R2 IMAGE CHECKER, NONE OF THE CASES SHE SENT TO BIOPSY ON THE BASIS OF R2 IMAGECHECKER’S FLAGGING AREAS OF INTEREST TURNED OUT TO BE CANCER. ON PATIENTS WHO HAVE BEEN DIAGNOSED WITH BREAST CANCER AFTER A MAMMOGRAM USING THE R2 IMAGECHECKER, DR. KASS HAS PROCESSED THOSE PATIENTS’ PRIOR MAMMOGRAMS READ WITHOUT R2 IMAGECHECKER. IN A FEW CASES, R2 IMAGECHECKER HAS FLAGGED POTENTIAL AREAS OF CONCERN ON THE PRIOR MAMMOGRAMS, SO THERE IS SOME VALUE TO THE TECHNOLOGY. DR. KASS TOLD DR. TAN THAT SHE DOES FEEL THAT R2 IMAGECHECKER TECHNOLOGY WILL EVENTUALLY BE INCORPORATED INTO MAMMOGRAPHY MACHINES AND BECOME PART OF THE COMMUNITY STANDARD IN THE FUTURE. DR. TAN’S IMAGING COMMITTEE HAS NO CURRENT PLANS TO PURCHASE R2 IMAGECHECKER SYSTEMS FOR THE REGION. Deborah A. Kass, MD (Radiology, Redwood City Medical Center, 8-424-2315) (December 2001) THE VALUE OF CAD MAMMOGRAPHY HAS BEEN DOCUMENTED WITH A VERY LARGE NUMBER OF CASES OBTAINED FROM MULTIPLE MEDICAL CENTERS AND MULTIPLE DIFFERENT MQSACERTIFIED RADIOLOGISTS. EXTENSIVE DATA WAS PROVIDED TO THE FDA, LEADING TO APPROVAL OF THE IMAGECHECKER. STATISTICALLY, IN A HOSPITAL WITH 14,000 SCREENING MAMMOGRAMS PER YEAR, THE DATA PREDICT THAT 4-5 OVERSIGHT ERRORS MIGHT BE PREVENTED. THIS COULD POTENTIALLY RESULT IN DETECTION OF BREAST CANCER IN ITS MORE EASILY TREATABLE EARLY STAGE POTENTIALLY LEADING TO SIGNIFICANT COST SAVINGS. THERE IS A VERY RAPID LEARNING CURE AND IMAGECHECKER HAS NOT SLOWED DOWN THE WORKFLOW. IT BECOMES VERY AUTOMATIC, AND IS VALUED AS A “SAFETY NET “ BY THE RADIOLOGISTS WHO USE IT. THEY FIND THAT IF THEY ARE GETTING TIRED, IMAGECHECKER PROMPTS THEM SO THAT THEY BECOME MORE ALERT. FORMAL DATA COLLECTION SHOWS THAT THE ADDITIONAL PATIENTS IDENTIFIED AND CALLED BACK DO NOT RESULT IN A STATISTICALLY SIGNIFICANT INCREASE IN WORK UPS WHEN COMPARED WITH THE BASELINE VARIATION. IN THE BAY AREA, IMAGECHECKER IS USED ONLY AT SAN FRANCISCO AND REDWOOD CITY. THEY RECENTLY HAD A “NEAR MISS” WITH DMHC, WHEN A MEMBER WHO HAD HAD A NEGATIVE MAMMOGRAM AT ANOTHER FACILITY ATTEMPTED TO HAVE HER NEGATIVE FILMS SENT TO REDWOOD CITY TO HAVE THEM PUT THROUGH IMAGECHECKER. WHEN THIS WAS DENIED, THE MEMBER CONTACTED DMHC. DR. KASS WAS ABLE TO INTERVENE AND READ THE FILMS IN THIS INSTANCE, AVOIDING A POTENTIAL VIOLATION FOR DENIAL OF CARE. LATELY HER FACILITY HAS HAD 100 REQUESTS FROM OTHER NCAL FACILITIES. ONE PHYSICIAN EVEN CIRCUMVENTED THE SYSTEM BY GIVING A PATIENT A NEW MAMMOGRAPHY REFERRAL SO SHE COULD HAVE A SECOND MAMMOGRAM DONE AT REDWOOD CITY. THESE ISSUES HAVE LED NCAL TO THE CONCLUSION THAT THEY NEED A REGION-WIDE POLICY, WHICH THEY ARE IN THE PROCESS OF DEVELOPING. JOHN REGO, MD, CHAIR (SAN FRANCISCO-BASED), CHIEFS OF RADIOLOGY. (MARCH, 2001) WE HAVE FOUND THE R2 IMAGE CHECKER A VALUABLE ADJUNCT TO OUR PRACTICE. IT IS MY FEELING THAT IT ADDS CONSISTENCY TO ALL OF OUR READINGS. ITS MAJOR STRENGTH IS THAT IT ACT AS A SECOND PAIR OF EYES OR A DOUBLE READ WHICH IS SOMETHING WE ARE NOT RESOURCED TO DO. IT IS MY FEELING THAT ALL WOULD BENEFIT FROM THIS TECHNOLOGY BUT ESPECIALLY THOSE WHO HAVE AVERAGE ABILITY. THE EXPERT MAMMOGRAPHER BENEFITS LESS FROM THE TECHNOLOGY. SINCE MOST OF US ARE IN THE FORMER CLASS, I THINK OVERALL WE WOULD ALL BENEFIT. I THINK FROM A FINANCIAL OR MEDICAL LEGAL ASPECT OR EVEN FROM A HEALTH BENEFIT ASPECT IT IS VERY TOUGH TO PROVE THE BENEFIT OF THE TECHNOLOGY SINCE YIELD OF CANCER ON MAMMOGRAPHY IS LESS THAN ONE PERCENT ANYWAY. I THINK AS AN ORGANIZATION THERE ARE MANY OTHER PRIORITIES WE SHOULD BE ADDRESSING WITH OUR MONEY THAN TO INVEST IN THIS TECHNOLOGY FOR ALL OTHER FACILITIES. I WOULD DEFINITELY

Version 2.0 October, 2003 - 12 -

FEEL DIFFERENT IF DOUBLE READING WERE MANDATED AND THIS TECHNOLOGY WOULD SUBSTITUTE FOR THAT.

Random Epithelial Cell Cytology Evaluation. IN BREAST CANCER RISK ASSESSMENT AND HIGH-RISK PATIENT MANAGEMENT This discussion will include three ductal cytology techniques: random periareolar fine-need aspiration (FNA), nipple aspiration, and ductal lavage. The assessment will review the value of these techniques in the evaluation of risk and high-risk patient management. The BCBSA TEC assessment (March 2002) will be the basis of this discussion. •

random breast epithelial cell cytology for breast cancer risk assessment and patient management in women who are already at increased risk of breast cancer does not meet the Blue Cross and Blue Shield Association Technology Evaluation Center (TEC) criteria.

[Direction to Staff: Because the concept of increasing detection of breast cancer is compelling, the Assessment needs to clearly and carefully spell out the reasons why this technology doesn’t meet criteria. Specifically discussed were the unknowns of whether the technology would increase unnecessary biopsies, potential adverse effects of the procedure, impact on future monitoring/diagnostic procedures, and the unknown implications of a negative cytology result. Also, the data in the cited studies do not derive from a patient population in whom the test would be used; and there are concerns as to the reproducibility of the test.

use of epithelial cell cytology in breast cancer risk assessment and high-risk patient management Random breast epithelial cell cytology for breast cancer risk assessment and patient management does NOT meet criteria.

Women identified as being at a higher risk of breast cancer than the general population, whether by Gail risk assessment, family history, genetic analysis, or a history of breast cancer, must make choices regarding prevention or frequency of surveillance. The high-risk population represents a continuum of risk, and women at the upper end of the continuum may make different choices than those at the lower end. Thus, additional information derived from intermediate biomarkers associated with later cancer development could revise risk estimates and influence management choices. Women originally at a lower risk who are positive for an additional risk factor and at a higher risk level as a result may be more likely to select beneficial preventive treatment. Conversely, women negative for an additional risk factor who may still benefit from preventive treatment may be more likely to avoid such treatment. Potential biomarkers are hyperplasia and hyperplasia with atypia identified by cytologic analysis of random epithelial cell specimens. Case-control and prospective cohort studies of benign breast disease have shown an increased risk of breast cancer in women with a history of ductal hyperplasia by surgical biopsy, particularly in those with atypical ductal hyperplasia. Family history of breast cancer combined with atypical hyperplasia results in a higher risk than either factor alone. There is no indication for biopsy in asymptomatic high-risk women with no evidence of a mammographic lesion or a palpable mass. However, various techniques are available for obtaining random samples of ductal epithelial cells. These include:

Version 2.0 October, 2003 - 13 -

• • •

Random periareolar fine-needle aspiration (FNA): the needle is positioned near the areolar margin and breast tissue is probed deeply in several locations to sample the terminal lobular duct unit. Nipple aspiration: Aspiration of ductal fluid from the nipple using a nonpowered breast pump similar to those used to express milk in lactating women. The procedure may be assisted by breast compression. Ductal lavage: A microcatheter is inserted into the natural nipple opening of the individual mammary ducts that have been identified by nipple aspiration as yielding fluid. Saline is then infused and diluted ductal fluid withdrawn.

This Assessment evaluates whether random epithelial cell cytology evaluation on samples obtained by nipple aspiration, ductal lavage, or random periareolar FNA can assist in risk assessment and patient management and improve outcomes for women at high risk of breast cancer. Outcomes of interest are incidence of breast cancer, and morbidity and mortality from breast cancer in high-risk women whose management is influenced by ductal epithelial cell cytology results. The female adult, non-pregnant, non-lactating breast secretes fluid into the breast ductal system; this fluid also contains shed ductal epithelial cells. Such fluid can be obtained via nipple aspiration. Ductal lavage enhances ductal fluid collection by first infusing saline solution to increase the fluid and cellular volume. Either technique can be performed during an office visit by non-physician, trained health care personnel. Both are limited in that not all women yield fluid, and in successful procedures, only fluid-yielding ducts are sampled. Random periareolar FNA is a more invasive procedure, performed by a physician, that attempts to randomly sample ductal epithelial cells from all quadrants of the breast. Nipple Aspiration. Ductal fluid can be aspirated from the nipple using a syringe or a pump connected to a small cup that fits over the nipple; the procedure may be assisted by breast compression. Only those ducts that yield fluid can be studied. The nipple aspirate fluid can be used for cytologic and biochemical studies. Patient factors that are associated with increased ability to obtain nipple aspirate fluid include an age of 30–50 years, early onset of menarche, non-Asian ethnicity, and prior parity and/or lactation (Petrakis 1993). Random Periareolar Fine Needle Aspiration. In the absence of a mammographic lesion or palpable mass, the needle is positioned near the areolar margin. Tissue is probed deeply to sample the terminal lobular duct unit. Typically, 8–10 aspirations are performed per breast, sampling from all 4 quadrants of the breast, and aspirates are pooled for each breast for cytologic and biochemical studies. Ductal Lavage. The technique of ductal lavage involves several steps. First, fluid-yielding mammary ducts are identified using nipple aspiration. Next a microcatheter is inserted into the natural nipple opening of the individual mammary ducts. Only ducts that initially yielded nipple fluid are studied. Saline is then infused and ductal fluid withdrawn. The fluid is then analyzed for cytologic or biochemical abnormalities. In women with a history of breast cancer who received mastectomy or lumpectomy plus radiation, only the contralateral breast is examined. In cancer patients who received lumpectomy alone, and in all other high-risk patients without evidence of disease, both breasts are examined. Ductal endoscopy or ductoscopy. Fiberoptic ductoscopy is an emerging technique that allows direct visual access of the ductal system of the breast through nipple orifice cannulation and exploration. Saline can also be injected and extracted for examination of ductal lavage cytology. To date, this technique has been used in pilot studies of women with spontaneous nipple discharge. The technology was developed in Japan and has been studied primarily in Asia. As no studies of ductoscopy have focused on random examinations of high-risk women, ductoscopy will not be evaluated in this Assessment.

Breast duct fluid aspirators are similar to nonpowered breast pumps used to express milk in lactating women. TM They include the InDuct Breast Aspirator manufactured by Cytyc Health Corporation (originally Windy Hill Technology, renamed Pro-Duct Health, now a wholly owned subsidiary of Cytyc Corporation) and the Mammary Aspiration Specimen Cytology Test (MASCT) manufactured by Nastech Pharmaceutical Company, Inc.

Version 2.0 October, 2003 - 14 -

The InDuctTM Breast MicroCatheter is manufactured by Cytyc Health Corporation for performing ductal lavage and is similar to catheters originally designed for galactography. Each of these devices has received U.S. Food and Drug Administration (FDA) approval as substantially equivalent to devices previously cleared for marketing (i.e., 510(k) approval). Fine-needle aspiration is a surgical procedure and, as such, is not subject to regulation by the FDA. ALTHOUGH NOT EXTENSIVE, THE EXISTING EVIDENCE SUGGESTS THAT CYTOLOGIC ATYPIA IS ASSOCIATED WITH SIGNIFICANTLY INCREASED RISK OF BREAST CANCER, AND IS INDEPENDENT OF OTHER RISK FACTORS SUCH AS FAMILY HISTORY. THIS IS TRUE WHETHER SAMPLING IS BY NIPPLE ASPIRATION OR BY RANDOM FNA AND IS LIKELY TO BE TRUE WHEN SAMPLES ARE OBTAINED BY DUCTAL LAVAGE, ALTHOUGH DIRECT EVIDENCE IS NOT YET AVAILABLE. HOWEVER, THE PREDICTIVE SIGNIFICANCE OF A NEGATIVE CYTOLOGY RESULT BY ANY SAMPLING METHOD IS UNKNOWN. NO STUDIES ADDRESS THE CONSISTENCY OF CYTOLOGIC ANALYSIS AND THE EFFECT ON RISK ESTIMATES. NOR HAVE ANY QUANTITATIVE METHODS OF EVALUATING CYTOLOGY BEEN APPLIED IN THE STUDIES OF RISK ESTIMATION. FINALLY, THERE IS INSUFFICIENT EVIDENCE TO SHOW THAT CYTOLOGIC HYPERPLASIA WITH ATYPIA IS REVERSIBLE WITH PREVENTION INTERVENTIONS KNOWN TO DECREASE CANCER INCIDENCE; NOR DO ANY STUDIES DIRECTLY ADDRESS A POSSIBLE ASSOCIATION OF REDUCED CANCER INCIDENCE WITH PREVENTIVE TREATMENT OF HIGH-RISK WOMEN WITH CYTOLOGIC ATYPIA. ALTHOUGH THE BREAST CANCER PREVENTION TRIAL P-1 INDICATES A GREATER BENEFIT FROM TAMOXIFEN TREATMENT IN WOMEN WITH A HISTORY OF ATYPICAL HYPERPLASIA THAN IN WOMEN WITHOUT THIS BIOPSY RESULT, IT IS NOT CLEAR THAT WOMEN WITH A CURRENT CYTOLOGIC RESULT OF ATYPIA REPRESENT A SUFFICIENTLY SIMILAR POPULATION SUCH THAT THE SAME BENEFIT CAN BE PREDICTED. IT IS ALSO NOT CLEAR THAT WOMEN WITH BENIGN CYTOLOGY RESULTS WOULD NOT BENEFIT FROM TAMOXIFEN. GIVEN THE LOW UPTAKE OF TAMOXIFEN THERAPY IN WOMEN WHO ARE AT INCREASED RISK (PORT ET AL. 2001), IT IS POSSIBLE TO SPECULATE THAT KNOWLEDGE OF ADDED RISK AS A RESULT OF CYTOLOGIC ANALYSIS MAY INFLUENCE THOSE AT GREATEST RISK AND MOST LIKELY TO BENEFIT, TO CHOOSE TAMOXIFEN THERAPY. HOWEVER, WHETHER OR NOT A CYTOLOGIC RESULT OF ATYPIA DOES INFLUENCE PATIENT MANAGEMENT AND IMPROVES PATIENT OUTCOMES IS NOT KNOWN. IT IS ALSO POSSIBLE THAT KNOWLEDGE OF A NEGATIVE CYTOLOGIC ANALYSIS RESULT WOULD INFLUENCE PATIENTS TO AVOID TAMOXIFEN WHEN THEY MIGHT OTHERWISE BENEFIT. HOWEVER, THERE IS NO EVIDENCE FOR SUCH POTENTIAL HARM. THUS, THE EVIDENCE IS INSUFFICIENT TO SUPPORT THE USE OF CYTOLOGIC HYPERPLASIA WITH ATYPIA AS A CLINICALLY USEFUL INTERMEDIATE BIOMARKER OUTSIDE OF CLINICAL TRIALS AT THIS TIME. THE EXISTING EVIDENCE IS OF HIGH CLINICAL INTEREST BUT FURTHER FOLLOW-UP STUDIES OF RISK AND TRIALS OF INTERVENTION IN WOMEN WITH THIS MARKER WILL BE NEEDED. Effect on risk assessment. Only a few studies evaluate the effect of cytologic hyperplasia with atypia on cancer risk in women who are already at increased of breast cancer. However, the results of these studies consistently indicate significantly increased risk whether sampling is by nipple aspiration or by random FNA. Ductal lavage has been compared to nipple aspiration and is more likely to obtain an evaluable sample with much larger numbers of epithelial cells; no prospective data evaluates risk when sampling is by ductal lavage. Presumably, risk estimates using ductal lavage cytology would be similar to those of nipple aspiration. However, the single study comparing ductal lavage to nipple aspiration cytology also indicates greater prevalence of atypia by ductal lavage; the effect of this on risk estimates is unknown. Also unknown is the predictive value of the absence of cytologic hyperplasia with atypia, regardless of sampling method. Effect on patient outcomes. There are no studies that directly compare routine surveillance vs. routine surveillance plus epithelial cell cytology analysis in the follow-up of high-risk women for the detection of longterm clinical outcomes. No studies compared the outcomes of patients whose management was determined Version 2.0 October, 2003 - 15 -

by the results of routine surveillance vs. routine surveillance plus epithelial cell cytology analysis. No studies have used ductal lavage, nipple aspiration, or random periareolar FNA to influence patient management in the population of interest. There is some indirect evidence from the NSABP Breast Cancer Prevention Trial (P-1), which enrolled women at high risk and randomly assigned them to placebo or tamoxifen for 5 years. Women with a history of atypical hyperplasia who received tamoxifen had a risk ratio for subsequent breast cancer of 0.14 (0.03–0.47) compared to those who received placebo over a median follow-up time of 54.6 months. Thus, high-risk women with a history of atypical hyperplasia benefited to a greater degree than the study population as a whole (RR 0.51; 95% CI, 0.39–0.66). It was noted, however, that the number of women in this subgroup was small, and that this was only one of five subgroups examined. Women without a history of atypical hyperplasia who received tamoxifen also benefited with a risk ratio of 0.56 (95% CI, 0.42–0.73). Thus, lack of a history of atypical hyperplasia does not preclude improved outcomes with tamoxifen treatment. These trial results cannot address whether or not participants, particularly those with a negative history, had cytologic evidence of hyperplasia or atypical hyperplasia at the time of enrollment. It is possible that some of the women who were negative for a history of atypical hyperplasia would have been positive at study entry by random cytology, and may have accounted for at least part of the benefit in this subgroup. Nevertheless, it cannot be ruled out that women with no detectable atypical hyperplasia may still benefit from tamoxifen treatment.

TPMG Pascal Fuchshuber, MD, Thoracic Surgery, Walnut Creek. I believe this is an excellent review of the topic and me being a strong advocate of evidence based medicine I would agree with the conclusions. However, I believe that Kaiser is particularly well positioned to participate in developing its own trial regarding the value of breast cytology in this context. We have to have better evidence to fully support its application in the routine clinical practice.

Emerging Technology NMP-66 Matritech, Inc. has developed a blood test for NMP-66, a protein, that may detect breast cancer. This test is being described as a a test similar to the PSA test for prostate cancer. It is recognized that even if early research shows the test has promise (current study in 1,100 women), additional research will be necessary to show it works. The test is being touted as a possible “back-up” to a mammogram.

Version 2.0 October, 2003 - 16 -

New Technology

Breast Cancer Screening

Summary of Assessments on Select New Technologies December 2002

In March 2002, the Interregional Breast Cancer Leaders requested a review of new technologies used to screen for breast cancer. Evidence summaries for full-field digital mammography (FFDM), computer-assisted detection (CAD) mammography, and random epithelial cell cytology were provided. The purpose of this document is to update the interregional group on these technologies, based on evidence reviews created since March 2002. (See Appendix A for additional references to evidence reviews for other breast cancer technologies.)

FULL-FIELD DIGITAL MAMMOGRAPHY (FFDM) In May 2002, KP’s Interregional New Technologies Committee (INTC) discussed full-field digital mammography, using the technology assessment from the Blue Cross Blue Shield Association (BCBSA) as the primary evidence review. (This BCBSA review was summarized in the previous new technology summary, March 2002.) The INTC concluded that there was insufficient evidence to determine whether full-field digital mammography (FFDM) is a medically appropriate diagnostic option for any patient. FFDM has not been demonstrated to be superior to screen-film mammography (SFM) applied in the screening or diagnostic setting. The evidence on use of FFDM in screening for breast cancer include interim data for 4,521 of 15,000 patients in the Lewin study. These preliminary results suggest a trend toward lower cancer detection with FFDM. A second study from Norway (Skaane) shows lower sensitivity for FFDM, but no statistical test results are available. Lewin consistently reports a lower recall rate for FFDM (11.5% and 11.8% versus 13.8 and 15% for SFM). Skaane observed higher recall rates for FFDM (reported by different investigators in the same study as 13.3%, 4.6% and 11.4% for FFDM versus 10%, 3.5% and 7.1% for SFM). The INTCnoted that the American College of Radiology Investigative Network (ACRIN) is conducting a study with a target enrollment of 49,500 subjects in the United States and Canada. The ACRIN study should have sufficient power to detect very small differences in cancer detection between SFM and FFDM, if any exist. Nineteen facilities will participate. Digital mammography systems from GE Medical Systems, Fischer Imaging, Hologic/Lorad, and Fuji Medical Systems will be tested in the study. Results will not be available until 2005 or later. In addition, the INTC noted that if FFDM is shown to be clinically equivalent to SFM, there may be operational reasons to consider FFDM including throughput, efficiency of FFDM versus SFM when used with CAD, and teleradiology advances. ---------------

_______________________ Questions concerning this paper can be addressed to Robin Cisneros (Director of Medical Technology Assessment, The Permanente Federation) at (510)271-5863 or at [email protected]

Version 2.0 October, 2003 - 17 -

A technology assessment vendor, ECRI, completed a TARGET Fact Sheet on FFDM in May of 2002. No additional evidence was available for this review than for the BCBSA TEC assessment. The report noted the following: °

The published evidence evaluating the effectiveness of full-field digital mammography is sparse, although there are numerous studies on performance characteristics and physics. There is a forthcoming comparative study on digital and film mammography for detecting microcalcifications (Fischer et al. 2002) and data from the abstract was reviewed in the report. Since both the BCBSA TEC and the ECRI reviews were completed, the Fischer U et al study was published in European Radiology 2002 Nov;12(11):2679-83. ( See Appendix B for the abstract for this study.)

°

Compared to film mammography, FFDM reduces film costs and other disadvantages associated with film storage and use, such as film degradation. It also can lower the radiation dosage to the patient and reduce the need for retakes. Additional advantages over conventional film mammography include manipulation of digital images to enhance detection of suspicious lesions and easy availability of digital data files for such applications as telemammography and computer-assisted detection. Initial studies and user reports also suggest that digital mammography offers advantages in imaging the dense breast. More clinical trials are necessary to determine whether full-field digital mammography using dedicated workstations results in improved detection of earlier-stage cancers and cancers in dense breasts and decreased false-positive biopsy rates.

°

Cost-effectiveness will ultimately determine whether full-field digital mammography technology is adopted, since hospitals must justify their purchase based on exam volume and patient population. The acquisition cost of digital mammography systems is significantly higher than film mammography systems, and high-resolution monitors used for soft-copy interpretation are expensive. Reimbursements for screening mammography are low and may not be sufficient to offset the increased expenditures for the new technology. Initial studies suggest that digital mammography may be able to reduce recall rates, and possibly necessary follow-up procedures, which will lower costs. Some facilities have reported substantial decreases in examination time and retakes, which will increase patient throughput and lower costs. Additional larger studies in screening populations are necessary to determine whether the cost benefits of using digital mammography relative to screening mammography outweigh the high acquisition costs. ----------------

There has also been another publication from Lewin et al which appears in the America Journal of Roentgenol 2002 Sep;179(3):631-7. (See Appendix B for abstract ). This publication updates the data available from this prospective trial to now include 6,736 of 15,000 examinations. This data is discussed in the report from CCOHTA, October 2002 (summarized below). ---------------The Canadian Coordinating Office for Health Technology Assessment (CCOHTA) published a Technology Report (Issue 30) in October 2002, Digital Mammography versus Film-Screen Mammography: Technical, Clinical and Economic Assessments. (This report is available upon request.) The scope of the report was “digital mammography systems for breast cancer detection” and the terminology used in the report is as follows: ° Film-screen mammography (FSM) – current standard of x-ray examination of the breast ° Digital mammography (DM) – digital capture of images through two different technologies: o Digital radiography mammography (DR-M) – a direct system; x-ray information is directly converted into a digital image. o Computed radiography mammography (CR-M) is an indirect system; x-ray information captured on a detector plate, from which a digital image is created. The report offered the following overall conclusion:

Version 2.0 October, 2003 - 18 -

DR-M has significantly higher annualized costs than either FDSM or CR-M. Potential clinical benefits (improved diagnostic accuracy, shorter examination time, lower radiation dose) for patients, institutions and payers have not been demonstrated in a clinical setting. The ability to detect cancer is comparable for DR-M and FSM. (There is no sufficient data on clinical effectiveness of CR-M.) Assuming that DR-M and CR-M are, at best clinically equivalent to FSM, the minimum-cost system is preferred; therefore, conventional FSM is preferable to DM at this time. An excerpt from the CCOHTA executive summary on the review of clinical effectiveness: Limited evidence indicated that DM might result in fewer women being recalled for additional workup after a screening examination. DM was not found to be superior or inferior to FSM for screening, based on receiver operating curve analysis and cancer detection rate. There was a large variability in interpretation of images. Studies also showed that DM is better for visualizing subcutaneous structures, while FSM is limited in detecting microcalcifications in dense breast tissue. Although computed radiography would be expected to have performance similar to digital radiography, there was limited direct evidence of similarities or differences.

COMPUTER-ASSISTED DETECTION (CAD) MAMMOGRAPHY In May 2002, KP’s Interregional New Technologies Committee (INTC) discussed computer-assisted detection (CAD) mammography, using an assessment from SCPMG’s Medical Technology Assessment unit as the primary evidence review. The INTC concluded that CAD is a medically appropriate diagnostic option when used to supplement a radiologist’s initial reading of a mammogram. CAD has demonstrated that it has the potential to be equivalent in cancer detection to double reading of mammograms in some operational settings depending on the experience of the radiologists and the volume of mammograms. CAD for mammography has not been demonstrated to be superior to mammogram interpretation by two radiologists or by a single expert mammographer. CAD has also not been shown to improve health outcomes of breast cancer in any operational setting. Excerpts from the INTC’s discussion follow: ° There is no published evidence from controlled trials comparing the benefits and harms of CAD to conventional mammography. Since CAD is used as a “second read” and identifies locations of suspected mammographic abnormalities, the sensitivity is expected to be the same or higher compared to a single reading. The recall rate is also expected to be higher. In addition, the sensitivity and specificity are likely to be related to the size of the tumor, type of tumor and age of the patient. Current data from five retrospective studies demonstrate the marginal increase in sensitivity for CAD plus radiologist compared to radiologist alone to be 4%. Dr. Adcock noted that the improvement in sensitivity will vary depending on the experience and skill of the radiologist reading the initial mammogram and the volume of mammograms the radiologists reads on a daily basis. In addition, the evidence suggests that CAD mammography is more likely to detect missed microcalcifications rather than malignant masses, calling to question to clinical significance of the discovered disease. There is no evidence that CAD screening mammography improves health outcomes. -----------Another external vendor of technology assessment, HAYES, Inc., updated an evidence review of CAD mammography in July 2002. This update did not yield evidence that would lead HAYES to change its rating of “C” from 1999. (The definition of the “C” Rating is: Investigational and/or experimental. The data on this procedure are promising but inconclusive regarding safety and/or efficacy. The 2002 update stated there is no clear medical consensus regarding its safety and/or efficacy.) Although HAYES rates the overall quality of the evidence as “good”, they found the efficacy and patient selection criteria unchanged and no long-term follow-up data available. ------------In October of 2002, the Blue Cross Blue Shield Association (BCBSA) completed a technology assessment on CAD mammography. Following review and discussion, the Medical Advisory Panel voted that:

Version 2.0 October, 2003 - 19 -

•

use of computer-assisted detection (CAD) devices after initial radiographic interpretation as a quality adjunct to single-reader mammography meets the Blue Cross and Blue Shield Technology Evaluation Center (TEC) criteria for patients having film-screen mammography;

•

use of CAD devices in patients without a prior independent interpretation of a mammogram does not meet criteria;

•

use of CAD devices as an adjunct to single-reader mammography does not meet the TEC criteria for patients having full-field digital mammography. Note: Technology Evaluation Center (TEC) criteria. 1. The technology must have final approval from the appropriate government regulatory bodies. 2. The scientific evidence must permit conclusions concerning the effect of the technology on health outcomes. 3. The technology must improve the net health outcome; and 4. The technology must be as beneficial as any established alternatives. 5. The improvement must be attainable outside the investigational settings.

------------A third party vendor, ECRI, published a Windows on Medical TechnologyTM report on CAD mammography in October of 2002. ECRI concluded there is insufficient evidence that CAD improves patient outcomes or lowers breast cancer mortality. While cancers detected with CAD are earlier in stage than cancers detected without CAD, over diagnosis, lead-time bias, and length bias all detract from the apparent benefits of mammography and other screening tests. Furthermore, the effect of earlier diagnosis is reduced by the fact that treatment is ineffective for many patients. These phenomena all contribute to the ongoing controversy about the net benefits of screening mammography after all the benefits and harms of screening and subsequent tests and treatments are accounted for. Ultimately, screening mammography and CAD are best evaluated in outcomes-oriented clinical trials. But it will be years before such trials on CAD can be completed. Additional excerpts from the WINDOW report follow: °

CAD systems are designed to reduce the number of cancers missed in diagnostic imaging examinations. Because this is a relatively new technology and clinical follow-up of screened patients takes time, clinical trials measuring both sensitivity and specificity of CAD in screening mammography have not yet been reported in the literature.

°

One large trial has found that CAD with the ImageChecker system increases the sensitivity of screening mammography without decreasing the PPV. Because CAD systems are intended to increase the number of examinations called positive, they will increase both the number of cancers detected and the number of women who must undergo follow-up procedures unnecessarily because of a false-positive mammogram result. But if the PPV of screening mammography without CAD is clinically acceptable (19 false positives per cancer detected), the increased false positives with CAD will be an acceptable price to pay for increased sensitivity. There is no evidence available for analysis of the effectiveness of other CAD systems, nor is there evidence directly comparing the effectiveness of CAD with the effectiveness of double reading and alternative methods of increasing the sensitivity of screening mammography. -------------

Group Health Cooperative discussed the R2 Image Checker in October of 2002. This review found two studies of reasonable size and design and critically reviewed them. (Both of these studies were included in the BCBSA TEC assessment and ECRI Windows reports from October of 2002.) Freer TQ, Ulissey MJ. Screening mammography with computer-aided detection: Prospective study of 12,860 patients in a community breast center. Radiol 2001; 220: 781-786. Brem RF, Schoonjans JM. Radiologist detection of microcalcifications with and without computer-aided detection: A comparative study. Clin Radiol 2001; 56: 150-154. This review acknowledged the following:

Version 2.0 October, 2003 - 20 -

There is evidence from a large, prospective study (Freer) that, when radiologists interpret mammograms aided by the R2 Imagechecker, they can detect additional cancers. In the Freer study of over 12,000 mammograms, 8 out of 49 malignancies (16%) were first identified by the R2 Imagechecker. This led to the identification of an additional 0.6 cancers per 1,000 women screened (from 3.2 to 3.8 cancers per 1,000) compared to mammogram interpretation by the radiologist-only. The clinical significance of this degree of increase needs to be considered. Brem did not find a statistically significant difference in the sensitivity of mammography reading by radiologists alone versus reading by radiologists aided by the R2 Imagechecker, but the study may have been underpowered or may not have found a difference because the radiologists were highly experienced. Additional large clinical studies of the R2 Imagechecker would strengthen the evidence. There is a study on the use of R2 ImageChecker among GHC radiologists forthcoming. In summary, The R2 Image Checker passed all of GHC’s Medical Technology Diagnostic Test Evaluation Criteria.

RANDOM EPITHELIAL CELL CYTOLOGY EVALUATION In May 2002, KP’s Interregional New Technologies Committee (INTC) discussed random epithelial cell cytology techniques. The INTC concluded that there is insufficient evidence to determine whether random epithelial cell cytology evaluation is a medically appropriate diagnostic option to: • evaluate breast cancer risk and contribute to high-risk patient management • screen patients for breast cancer • follow-up previously diagnosed breast cancer patients or •

monitor response to chemoprevention treatment.

The INTC used a technology assessment from BCBSA as its primary evidence review. (This assessment was summarized for the interregional breast cancer leaders in the March 2002 summary.) The BCBSA TEC assessment from March of 2002 considered the following methods of ductal cytology: ductal lavage, nipple aspiration, and random periareolar fine-needle aspiration (FNA). The assessment reviewed the value of these techniques in risk assessment and high-risk patient management. Excerpts from the INTC discussion follow: There is some evidence that patients already at increased risk of breast cancer with cytologic hyperplasia with atypia (sampling by nipple aspiration or random FNA), are at significantly increased risk. There is no prospective data that evaluates breast cancer risk based on epithelial cytology obtained by ductal lavage. No studies have used ductal lavage, nipple aspiration, or random periareolar FNA to influence treatment decisions in high-risk women. Only a few studies evaluate the effect of cytologic hyperplasia with atypia on cancer risk in women who are already at increased of breast cancer. However, the results of these studies consistently indicate significantly increased risk whether sampling is by nipple aspiration or by random FNA. Ductal lavage has been compared to nipple aspiration and is more likely to obtain an evaluable sample with much larger numbers of epithelial cells; no prospective data evaluates risk when sampling is by ductal lavage. Presumably, risk estimates using ductal lavage cytology would be similar to those of nipple aspiration. However, the single study comparing ductal lavage to nipple aspiration cytology also indicates greater prevalence of atypia by ductal lavage; the effect of this on risk estimates is unknown. Also unknown is the predictive value of the absence of cytologic hyperplasia with atypia, regardless of sampling method. There are no studies that directly compare routine surveillance vs. routine surveillance plus epithelial cell cytology analysis in the follow-up of high-risk women for the detection of long-term clinical outcomes. No studies compared the outcomes of patients whose management was determined by the results of routine surveillance vs. routine surveillance plus epithelial cell cytology analysis. No studies have used ductal lavage, nipple aspiration, or random periareolar FNA to influence patient management in the population of interest. There is some indirect evidence (NSABP Breast Cancer Prevention Trial, P-1), that women with a history of atypical hyperplasia who received tamoxifen had a reduced risk ratio for subsequent breast cancer compared to those who received placebo. Thus, high-risk women with a history of atypical hyperplasia benefited to a greater degree than the study population as a whole. However, women without a history of atypical hyperplasia who received tamoxifen also benefited. Thus, lack of a history of atypical hyperplasia does not preclude

Version 2.0 October, 2003 - 21 -

improved outcomes with tamoxifen treatment. It is possible that some of the women who were negative for a history of atypical hyperplasia would have been positive at study entry by random cytology, and may have accounted for at least part of the benefit in this subgroup. Nevertheless, it cannot be ruled out that women with no detectable atypical hyperplasia may still benefit from tamoxifen treatment. The net benefit of tamoxifen treatment varies with patient age and Gail risk. If the presence of atypical cytology substantially changed Gail risk, as does a history of atypical hyperplasia on surgical biopsy, then cytologic analysis of random duct fluid or FNA specimens could be used to revise Gail risk and influence patient management. However, there is insufficient evidence to show either how Gail risk would be modified, or whether the use of cytology results to change patient management influences net health outcomes. ------------------A third party vendor, HAYES, Inc. published a report on ductal lavage in July of 2002. This review also noted a paucity of evidence on ductal lavage. All studies found in this review were included in the BCBSA TEC assessment from March of 2002. The review concluded that the limited published evidence regarding ductal lavage suggests that ductal lavage is feasible and well tolerated, yields a greater number of breast duct epithelial cells for cytological analysis than nipple aspiration, is less invasive than fine-needle aspiration, and is capable of detecting abnormal cells in nipple aspirate fluid specimens from individual breast ducts. However, there are no published data on the diagnostic sensitivity or specificity of ductal lavage, on the clinical significance of the presence of atypical cells in ductal lavage specimens, on whether the demonstration of no atypical or malignant cells correlates with a decreased risk for breast cancer, or on the use of either positive or negative test results in patient management. Although previous studies have demonstrated a relationship between the presence of atypical epithelial cells in the breast ducts and an increased risk for breast cancer, particularly in women with a strong family history of the disease or those carrying gene mutations, whether the results of ductal lavage can be used to aid in the estimation of a woman’s risk of breast cancer and, therefore, to guide clinical decisionmaking regarding preventive or therapeutic strategies has not yet been adequately determined. There is insufficient evidence from well-designed trials showing that this technique reduces breast cancer morbidity and mortality or otherwise improves health outcomes. Definitive conclusions regarding the role of ductal lavage in the diagnosis and management of breast cancers and precancers require data from well-designed clinical trials employing strict patient inclusion criteria and long-term follow-up and through the refinement of riskassessment models incorporating ductal lavage results. ------------------TM A technology assessment vendor, ECRI, completed a Windows on Medical Technology report on ductal lavage and nipple aspiration in August of 2002. No studies were discussed that were not previously discussed in the BCBSA TEC report of March 2002. The conclusions of this review follow:

Sampling ductal epithelial breast cells using ductal lavage or nipple aspiration has been suggested as a way to identify women at high risk of developing breast cancer. The relative risk of developing breast cancer after a diagnosis of atypical hyperplasia by nipple aspiration was reported by one study of more than 8,000 women to be 2.1 (95% confidence interval, 1.1 to 3.9) relative to women from whom fluid cannot be obtained by nipple aspiration. Nipple aspiration was well-tolerated and only mildly uncomfortable, with no reported side effects or morbidity. No information about the ability of ductal lavage to predict the development of breast cancer was identified. ECRI evaluated the evidence regarding correlation between the two methods and found poor agreement. Thus, despite the apparent similarities between the two screening tests, the results of studies of nipple aspiration cannot be extended to ductal lavage. Patients reported that ductal lavage was somewhat painful (median VAS score of 2.4 on a scale of 0 to 10). The procedure carried a slight risk of breast infection. Side effects of bruising, swelling, and pain were also reported. The available evidence suggests that nipple aspiration is not a very useful method of identifying women at high risk of developing breast cancer. Available data are insufficient to draw any conclusions about ductal lavage.

Version 2.0 October, 2003 - 22 -

The review noted the following ongoing trials: A study sponsored by the National Cancer Institute is currently recruiting patients. This trial is enrolling patients carrying BRCA1 or BRCA2 mutations and will explore methods to screen for early breast cancer. Patients will be followed for three years and will regularly be evaluated by ductal lavage, mammography, physical examination, MRI, and positron emission tomography scanning. Cytyc Health Corporation is sponsoring the Ductal Lavage Outcomes Tracking System (DLOTS). Cytyc states that “DLOTS is a prospective, multicenter, observational outcomes database designed to collect data on the demographics, lavage procedures, cytology results, and surgical pathology results of patients at risk for developing breast cancer.” The purpose is to assess the performance of ductal lavage procedures among practicing physicians and to assist physicians in patient follow-up and management. DLOTS plans to correlate ductal lavage cytology results with pathology results in patients who undergo a surgical biopsy or tumor excision and to give clinical feedback to practicing physicians using the procedure.

Version 2.0 October, 2003 - 23 -

APPENDIX A: ADDITIONAL ASSESSMENTS ON TECHNOLOGIES RELATED TO BREAST CANCER

°

March 18, 2002 FULL-FIELD DIGITAL MAMMOGRAPHY (FFDM), COMPUTER-ASSISTED DETECTION (CAD) MAMMOGRAPHY, AND RANDOM EPITHELIAL CELL CYTOLOGY

°

BREAST BRACHYTHERAPY (INTC 10/02, BCBSA 10/02, BS of CA 10/02, HAYES 4/00 updated 9/02, CCOHTA on Mammosite 3/02)

°

CRYOABLATION FOR TREATMENT OF BREAST FIBROADENOMAS (HAYES 6/02)

°

DUCTAL LAVAGE (HAYES 7/02)

°

GENETIC SUSCEPTABILITY TO BREAST CANCER (HAYES 7/97 updated 3/02)

°

HIGH DOSE CHEMOTHERAPY WITH AUTOLOGOUS STEM CELL SUPPORT FOR THE TREATMENT OF BREAST CANCER (ICSI 4/02, HAYES 5/00 updated 9/02)

°

IMAGE-GUIDED, VACUUM-ASSISTED BREAST BIOPSY (ECRI TARGET, 7/02 and HAYES 11/99 updated 11/02)

°

LYMPHATIC MAPPING WITH SENTINAL LYMPH NODE BIPSY FOR BREAST CANCER (ICSI 7/02)

°

PROPHYLACTIC MEDICAL TREATMENTS FOR BREAST CANCER (HAYES 5/99 updated 7/02)

°

SCREENING MAMMOGRAPHY FOR WOMEN AT AVERAGE RISK FOR BREAST CANCER (HAYES 2/02)

°

SENTINAL NODE BIOPSY (HAYES 3/01 updated 11/02)

°

STEREOTACTIC BREAST BIOPSY (ECRI TARGET 8/02 and HAYES UPDATE 9/02)

°

TAXOL (HAYES 10/00 updated 7/02)

°

THERMOGRAPHY (HAYES 7/97 updated 3/02)

If you are interested in any of these reviews, please contact Robin Cisneros (Director of Medical Technology Assessment, The Permanente Federation) at (510)271-5863 or at [email protected].

Version 2.0 October, 2003 - 24 -

APPENDIX B: KEY ABSTRACTS

Comparative study in patients with microcalcifications: full-field digital mammography versus screen-

film mammography. European Radiology 2002 Nov; 12(11): 2679-83.

Fischer U, Baum F, Obenauer S, Luftner-Nagel S, Von Heyden D, Vosshenrich R, Grabbe E. Department

of Radiology, Georg-August-University of Goettingen, Robert-Koch-Strasse 40, 37075 Goettingen, Germany,

[email protected] OBJECTIVE: The goal of this prospective study was to compare a full-field digital mammography system (FFDM) to a conventional screen-film mammography system (SFM) for the detection and characterization of microcalcifications. SUBJECTS AND METHODS: Fifty-five patients with 57 isolated microcalcification clusters were examined using a FFDM system (Senographe 2000D, GE Medical Systems, Milwaukee, Wis.) and a SFM system (Senographe DMR, GE Medical Systems, Milwaukee, Wis.). A conventional screen-film mammogram and a digital contact mammogram were obtained of each cluster. The image quality and the number of calcification particles were evaluated, and a characterization (BI-RADS 1-5) of microcalcifications was given by four experienced readers. RESULTS: Histopathology revealed 16 benign lesions (sclerosing adenosis, dysplasia, hamartoma, radial scar) in 15 patients and 21 malignant tumors (in situ carcinoma, invasive carcinoma) in 20 patients. Twenty patients had benign changes verified by long-term follow-up. Image quality of FFDM was assessed as superior to SFM in more than 50% of the cases. The FFDM showed more calcifications in 41% of all cases. Sensitivity and specificity for FFDM vs SFM were 95.2 vs 91.9% and 41.4 vs 39.3%, respectively. Moreover, FFDM demonstrated a higher diagnostic accuracy (deviation: 0.86 BI-RADS steps) compared with FSM (deviation 0.93 BI-RADS steps).

CONCLUSIONS: The FFDM system with a 100- micro m pixel size provides better image quality than SFM in

patients with mammographic microcalcifications. The FFDM has a higher sensitivity and a higher reliability in

characterizing microcalcifications.

PMID: 12386757 [PubMed - in process]

Clinical comparison of full-field digital mammography and screen-film mammography for detection of

breast cancer. America Journal of Roentgenol 2002 Sep;179(3):631-7. .

Lewin JM, D'Orsi CJ, Hendrick RE, Moss LJ, Isaacs PK, Karellas A, Cutter GR.

University of Colorado Health Sciences Center, 4200 E. 9th Ave., Mail Stop F724, Denver, CO 80262, USA. OBJECTIVE: The purpose of this work is to compare full-field digital mammography and screen-film mammography for the detection of breast cancer in a screening population. SUBJECTS AND METHODS: Full-field digital mammography was performed in addition to screen-film mammography in 6736 examinations of women 40 years old and older presenting for screening mammography at either of two institutions. Two views of each breast were acquired with each technique. The digital and screen-film mammograms were each interpreted independently. In addition to a clinical assessment, each finding was assigned a probability of malignancy for use in receiver operating characteristic analysis. In cases in which the digital and screen-film interpretations differed, a side-by-side analysis was performed to determine the reasons for the discrepancy. With few exceptions, findings detected on either technique were evaluated with additional imaging and, if warranted, biopsy. RESULTS: Additional evaluation was recommended on at least one technique in 1467 cases. These additional evaluations led to 181 biopsies and the detection of 42 cancers. Nine cancers were detected only on digital mammography, 15 were detected only on screen-film mammography, and 18 were detected on both. The difference in cancer detection is not statistically significant (p > 0.1). Digital mammography resulted in fewer recalls than did screenfilm mammography (799 vs 1007, p < 0.001). The difference between the receiver operating characteristic curve area for digital (0.74) and screen-film (0.80) mammography was not significant (p > 0.1). Reasons for discrepant interpretations of cancer were approximately equally distributed among those relating to lesion conspicuity, lesion appearance, and interpretation. CONCLUSION: No significant difference in cancer detection was observed between digital mammography and screen-film mammography. Digital mammography resulted in fewer recalls than did screen-film mammography. PMID: 12185042 [PubMed - indexed for MEDLINE]

Version 2.0 October, 2003 - 25 -

MRI for Screening Women at

High Risk of Breast Cancer

Joanne Schottinger MD

Medical Technology Assessment

SCPMG

Presentation Date: July 20, 2003

Next Review:

July 20, 2005

TECHNOLOGY

MRI uses powerful magnetic fields to create images. Contrast agents

(typically Gadolinium) are employed to enhance or decrease signal

in areas of increased vascular permeability.

Most studies employed a high-field MRI system (1.0-1.5 Tesla) and a

dedicated breast coil.

The MRI is typically performed in the prone position with the

breasts hanging through a cut out in the table.

MRI compatible needles and localization equipment are available to

perform biopsies.

POPULATION

Health problem - Breast Cancer

Population - Women at high risk of breast cancer, either

because they carry a known mutation in BRCA 1 or 2, or

because of a strong family history of premenopausal

breast or ovarian cancer, and who have a negative

mammogram examination

HEALTH OUTCOMES

•Mortality from breast cancer

•Morbidity from breast cancer

•Unnecessary procedures - followup exams or biopsies

•Anxiety

•Allergic reaction to contrast

INTC REVIEW 1/03

•3 studies with limited followup showed MRI can identify

cancers in women at high risk with negative mammograms

•Relatively high false positive rates (up to 14%) and high

recall rates (up to 30%) for incidental enhancing foci

•Lack of standardized protocols and interpretation criteria,

resulting in lack of reproducibility

INTC CONCLUSION:

There is insufficient evidence to determine whether

MRI is a medically appropriate option for screening

patients at high risk of breast cancer.

EVIDENCE REVIEW

3 new abstracts presented at ASCO meeting 6/03

Kuhl -Bonn- update - 5 year results of a prospective trial of 462

women at high risk based on genetic testing or pedigree analysis

Annual CBE, Mammo, U/S, and MRI

Median age 39 (range 18-65)

51 breast cas identified in 45 patients; 2/51 invisible on screens

but found at prophylactic surgery

13/51 (25%) cancers were palpable on CBE

EVIDENCE REVIEW

U/S

MRI

MAMMO Sensitivity 42.8%

47.0%

96.1%

Specificity

94.3%

88.4%

95.1%*

PPV

38.0%

17.5%

56.9%*

NPV

95.1%

96.9%

99.7%

Recall Rate

5.2%

7.3%

EVIDENCE REVIEW

Robson - MSKCC -

53 women documented BRCA 1,2 mutations, with negative

mammograms received 115 breast MRIs over 3.5 yr period

90 were screening; 25 interval followup exams

Further study, short followup MRI, or biopsy advised in

35/90 (39.9%) screening studies

12 biopsies revealed 2 (16.7%) cancers - both DCIS

EVIDENCE REVIEW

Kriege - prospective, multicenter Dutch trial of women with

a lifetime risk of >15% developing ca breast.

Q 6 mo CBE and Q year mammo and MRI

1848 total women, 294 with BRCA 1,2 mutations

Mean age 41

30 breast tumors detected, including 6 DCIS

Sensitivity Specificity

CBE

16% 97%

Mammo 36% 95%

MRI

71%

88%

RECOMMENDATION

•There is evidence that MRI can detect additional cancers of the

breast in women at high risk who have a negative screening

clinical exam and mammogram.

• There are high recall rates for incidental foci as well as

a high false positive rate resulting in additional

radiology tests and biopsies.

•Lack of standardized protocols limits reproducibility of results.

•There are no studies showing improved outcomes.

PET SCAN FOR THE EVALUATION

OF BREAST LESIONS AND

AXILLARY LYMPH NODES

Joanne Schottinger MD

Medical Technology Assessment

SCPMG

Presentation Date: July 20, 2003

Next Review:

July 20, 2005

TECHNOLOGY

n

PET (Positron Emission Tomography) is a unique

imaging modality that reveals metabolic activity in

addition to anatomic information.

n

Organic compounds labeled with positron emitting

radioisotopes are used as the tracers. FDG, a glucose

analog, is most commonly used in cancer imaging.

n

Tumor cells show increased utilization of glucose.

TECHNOLOGY

n

PET scan is a non-invasive test.

n

Possible morbidity includes:

In theory, small risk carcinogenesis from exposure to

gamma radiation

Anxiety, delay in getting definitive therapy

Important morbidity would be the consequences of an

incorrect result, in these cases, particularly a false

negative test.

n

n n

POPULATION

n

Women with an abnormal mammogram suspected of

having a breast malignancy.

n

Standard procedures:

Diagnostic mammogram, Ultrasound

Image guided biopsies

Ultrasound, stereotactic

Core needle biopsies

Open surgical biopsies

BACKGROUND

n

Screening mammography improves early detection of

breast cancer, but in studies 70 - 90% of lesions on

mammography referred for biopsy are benign.

n

Biopsies involve anxiety, discomfort, and may produce

scarring that interferes with future screening.

n

PET imaging has been proposed as a test that may

evaluate lesions further to avoid unnecessary biopsies.

EVIDENCE

n

Literature review by California Technology

Assessment Forum and meta -analysis of studies

performed by ECRI are the basis of this review.

n

16 studies with a total of 661 patients are included in

the analysis.

n

No studies compared management or outcomes of

women who had PET scans vs. those who did not.

EVIDENCE

n

Studies included both dedicated PET scanners as well

as modified gamma cameras and SPECT systems.

n

Quality of studies was difficult to ascertain because

necessary information was omitted to determine if

there was verification bias in 11; if the PET readers

were blinded to the reference standard in 8.

n

The prevalence of disease was 75%, much higher than

the general population; spectrum bias also evident in

that the mean tumor size was large (2-4 cm).

EVIDENCE

n

n

Lacquement (1999) reviewed 688 biopsies BIRADS 3 Category - 3% malignancy

BIRADS 4 Category - 23% malignancy

BIRADS 5 Category - 92% malignancy

The evidence in this meta-analysis cannot be

generalized to the population with BIRADS

3,4 lesions where the prevalence of disease is

much lower than 75%.

EVIDENCE

n n

n n n

Sensitivity 89.7% (95% CI 86.4-92.2%)

Specificity 81.0% (95% CI 75.7-85.3%)

With prevalence of 75%:

PPV 93.4%

NPV 72.2%

SUMMARY

n

n

Women who test + on PET need to have biopsy

done anyway; did not avoid a procedure; slight

delay to biopsy in getting PET unlikely to be

harmful.

Women who test - on PET still have 28%

chance of having malignancy; informed of this

risk, most prudent women would still opt for a

biopsy to avoid risk of delayed diagnosis of

cancer with possible worse prognosis.

POPULATION

n

Women with biopsy + breast cancer , no evidence of

distant mets, no palpable lymph nodes

n

Standard therapy:

Axillary lymph node dissection (ALND)

Define prognosis

May improve local control (but not survival)

Determines aggressiveness of adjuvant therapy

Determines role for radiotherapy (>4 LN)

PROBLEM

n

n

n

ALND carries risk of morbidity with chronic

lymphedema reported in up to 8-25%, possible pain,

wound complications, restricted shoulder ROM

Sentinel lymph node biopsy has been proposed as a

procedure to predict the presence of axillary node

disease and spare women full ALND if the sentinel n

node is negative. SN biopsy has much lower morbidity.

Recent meta-analyses have reported sensitivities of

91% and 95%.

PET scan has been proposed as another procedure

that, if negative, would spare women the need for

ALND.

EVIDENCE

n

Literature review by California Technology

Assessment Forum and meta-analysis by ECRI are the

basis of this review.

n

19 studies of 965 patients were reviewed in the analysis.

Only 6 studies with 249 patients clearly address women

without palpable axillary lymph nodes.

n

Many different PET protocols were used.

EVIDENCE

n

No studies compared outcomes or management of

women who received PET scans vs. those who were not

scanned.

n

No studies assessed whether PET scan could determine

the extent of involvement of nodes.

EVIDENCE

n

All studies (includes palpable nodes)

Sensitivity 79% (95% CI 70 - 84%)

Specificity 89% (95% CI 83 - 93%)

n

6 studies without palpable axillary nodes

Sensitivity 74% (95% CI 56 - 94%)

Specificity 91% (95% CI 86 - 95%)

n

Prevalence of disease in the 6 studies was 35%,

with a NPV of 87%

SUMMARY

n

Women who test + PET would still undergo ALND, so

derive no benefit from testing.

n

Women who test - PET still have a 13% chance of

having malignancy in their lymph nodes.

V. Breast Cancer Tracking System

Computer Tracking Systems for Breast Care Important Note: Work is currently underway in the Care Management Institute to identify and

develop functionality to support population care management as part of the KP HealthConnect

implementation. One of the components includes tracking capabilities which could be used for

breast care issues as well as chronic care needs. Please contact [email protected] for

additional information concerning this work.

BACKGROUND In 2002, an Interregional Breast Cancer Leaders (IRBCL) task force, convened by the Permanente Federation, met to address enhancing the care process for women who present with breast complaints. During that time, the IRBCL recognized computer prompting and tracking systems as one of the key systematic processes to ensure that members receive consistent, timely, high-quality breast care and requested an inventory of systems. Kaiser Permanente’s regions have computer tracking systems in place to prompt and track routine care as well as to help reduce delays in diagnosis and treatment by tracking care from routine screening to disposition . These systems may be stand-alone or rely on existing clinical systems/databases. While there are some core functions present in most systems, the features and functions of systems vary as do their deployment of staff to enter data and coordinate patient care. This paper will provide general information about the key functionality of a comprehensive breast care tracking system, by including all the functions currently available in several systems. A list of clinical and IT contacts along with some general information about each system (where available) is attached in Tables 1 and 2 to facilitate communication between Regions concerning implementations of tracking systems or enhancements to existing systems.

BENEFITS OF BREAST CARE TRACKING SYSTEMS Breast care tracking systems provide several benefits to the organizations. • Improved quality and member satisfaction with breast health care due to: ⇒ Prompt, reliable access to care ⇒ Improved coordination of care between departments ⇒ Reduced delays in diagnosis and treatment ⇒ Facilitated, systematic approach for communication of results • Easier compliance with the Mammography Quality Standards Reauthorization Act of 1998 (MQSRA) regulations • Reductions in litigation due to delay or failure to diagnose

KEY FUNCTIONS/REQUIREMENTS A comprehensive breast care computer tracking system will support the complete breast care

processes including self-risk appraisal, routine screening, diagnosis, treatment and subsequent

follow-up . This paper addresses four major care events:

Version 2.0 October, 2003 - 26 -

• • • •

Routine Screenings Abnormal Exams Abnormal Mammograms Abnormal Breast Biopsies

In addition, these systems will provide key reporting data to meet requirement of the Mammography Quality Standards Reauthorization Act of 1998 (MSQA) regulations. Depending of the capabilities of these systems, tracking may be more automatic with information extracted from other systems, or may require manual input of specific information which is collected from the member or derived from reports, clinical records or other medical information. As such, the FTEs will vary depending on manual requirements and their specific role in patient outreach and care coordination. Given this tradeoff between automation and staff, cost comparison of these systems is difficult. Routine Screening The primary function of the routine screening tracking system is to identify and prompt women due for routine mammograms and track compliance. The majority of regions utilize current computer systems for identification of patients requiring routine mammograms. Colorado, Hawaii, NCAL, SCAL, Ohio, and KPNW have (will have) automated systems which provide inreach for routine prevention, including mammograms. These systems remind both patients and providers that a routine mammogram is due when a member presents for care. This information can be available for online viewing and may also print notices for patients when they are in clinic. In addition to these preventive alert systems, there are also general outreach systems in place. Hawaii has a mammogram reminder system which sends notifications for routine screening and Northern California is implementing a new Reports on Request System (ROAR system), an internet based system, which will generate information about screening rates as well as files which can be used for outreach. These systems are not solely devoted to breast care; however, include breast care as one of their focuses and play a key role in routine screening. Two Regions, Colorado and KPNW and Group Health Cooperative (GHC) have systems which are specifically designed for in reach and outreach for breast care. These systems may target to women of screening age, or a specific subset of at risk members who are overdue for these preventive services or identified as higher risk individuals. These systems operate based on specific guidelines for screening timeframes, education, and follow up set by the Program or within each individual Region. Routine screening functionality includes the ability to:

• Provide prevention alerts for clinicians and members during clinic visit • Collect and maintain specific risk data for member populations (via risk surveys and pathology information)

• Identify members by gender, age (via membership data) and/or risk (family and personal history or prior medical event) who are due for routine mammograms

• Check appointment/radiology records for occurrence of mammogram based on specific timeframes • Generate automatic notification tools (lists, computer alerts, outreach letters) • Track and identify patients requiring educational materials

Version 2.0 October, 2003 - 27 -

• Perform updates to identify members (via lists, alerts or automatic notification) who have not received care to receive additional prompts based on timeframes

• Provide members with written notification of results Colorado’s Breast Screening System tracks activity for all women ages 40 – 75. When a female member turns 40 and 50, or when new members between 40-49 and 50-65 years of age join the program, a set of Regional educational materials specifically developed for each of these two age groups are sent. Each time a member has a mammogram, she completes a questionnaire with information about three key risk factors. This information is then manually entered in the database to determine whether a member will receive an annual or biannual reminder. GHC’s sends a four-page risk survey to a member when she turns 40 or joins the GHC after age 40. This survey is scanned and stored in their Breast Screening Recruitment and Reminder System. This information, as well as updates scanned from new surveys a member completes when she presents for a mammogram, updates provided by patient and provider phone calls to a special 1-800 number and emails, and pathology data are used to determine the screening schedule for each member. KPNW’s Safety Net program builds files from available data bases to specifically target members who have not received a screening mammogram in more than two years and tend to be resistant to customary inreach and outreach efforts for preventive services. Letters are sent to a subset of the Safety Net population – first-time safety net patients. Available Outcomes Information:

• In 2000, 80% of Colorado women age 50 – 69 had evidence of a mammogram in the past 2 years compared to a national average of 75% (HEDIS). Abnormal Exam/Mammograms Systems which track abnormal exams, abnormal mammograms and abnormal biopsies can play a critical role in timely diagnosis and treatment of breast cancer. Computer system functionality for tracking abnormal mammograms includes:

• Capture of abnormal/suspicious findings or complaint • Identify members who require outreach for scheduling next appointment • Check appointment/radiology/surgery records for occurrence of follow-up mammogram, surgery • • • • •

consult or biopsy within designated time period Identify members who have not received recommended follow up care and/or require follow up call, notices/certified letters Generate notification tools (lists, computer alerts, notices, letters) for outreach Provide information for timely callback concerning results and next steps Revises system-generated routine preventive alerts with revised timeframes for procedures and revert to routine schedule when care completed Provide point of service reminders to physicians concerning member’s next care step

Version 2.0 October, 2003 - 28 -

The trigger for initiating tracking for abnormal mammograms may be generated automatically by a system, or could be triggered by a report which results in manual data identification and initiation of tracking. KPNW TRAK system begins with a report generated by radiology and the creation of an event by the staff. The basis of the KPNCAL Breast Cancer Tracking System (BCTS) abnormal mammography event is the use of standard alert codes, assigned by radiologists and entered into the system by transcriptionists as a mandatory field. These codes are the sole indicator of a patient’s tracking eligibility in the BCTS and are required for all mammograms and breast ultrasounds. KPNCAL’s alert codes further break down BIRAD 3 into three sub-groups which dictate varying follow up mammogram timeframes of 3, 6 and 12 months. BIRAD 4 and BIRAD 5 require follow up biopsies within two weeks. With this detailed information, the system is able to track specific care paths for each member. Six of the Regionally-supported tracking systems follow abnormal mammograms and abnormal biopsies. Available Outcome Information: • In Colorado, this system has been successful at getting 94% of women with abnormal

mammograms in for timely follow up.

• For 2001, KPNCAL achieved a 95% rate of return (within one year) for mammograms which recommended a 6-month follow up mammogram. BCTS intervened in 9% of these appointments to bring patients. • For the 5% who did not return in 2001, the BCTS generated certified letters for the 4% of

patients no longer covered by KP and the 1% who were active, but chose not to return.

Abnormal Breast Biopsies Abnormal biopsies include diagnoses of invasive breast cancer, ductal carcinoma in situ (DCIS) and pathology results of increased risk for developing breast cancer, i.e. atypical hyperplasia and lobular carcinoma in situ (LCIS). These systems perform many of the same functions as those for abnormal mammograms:

• Capture pathology information via risk code • Provide information for timely callback concerning results and next steps • Check appointment records for occurrence of surgery appointments, oncology appointments, • • • •

mammogram and clinical follow up based on risk code Identify members who have not received recommended follow up care and/or require follow up call, notices/certified letters Generate alerts for physicians for outreach Revises system-generated routine preventive alerts with revised timeframes for procedures and revert to routine schedule when care completed Provide point of service reminders to physicians concerning member’s next care step

The KPNCAL system tracks three types of abnormal breast biopsy diagnoses: invasive breast cancer, ductal carcinoma in situ (DCIS) and pathologies indicating a higher risk for breast cancer, such as lobular carcinoma in situ (LCIS) atypical ductal hyperplasia, and atypical lobular hyperplasia. Their BCTS is programmed to recognize these pathology codes for automatic data entry.

Version 2.0 October, 2003 - 29 -

Available Outcome Information: • Since its inception in 1995, BCTS has tracked a total of 19,233 abnormal biopsies and 109,374 abnormal mammograms. During this time, BCTS intervened to ensure that care was received in 8,486 cases. As a result of these BCTS interventions, 175 patients were diagnosed with a new or recurrent breast cancer or increased risk diagnosis. (data current as of March 2003.)

Outcome Statistics for MQSRA In addition to ensuring that members receive timely care, these systems also provide information to meet the 1998 Mammography Quality Standards Reauthorization Act (MQSRA) including:

• Number and type of mammographic procedures by physician and facility • Correlation of pathology results with the mammograms that recommended surgical • • • • •

intervention True and false negatives and positives Specificity and sensitivity and cancer detection rates Positive predictive values Time from mammography to biopsy Time from biopsy to surgery

Staff The more a system is connected to existing computer information, the less data entry is required. At the same time, during the abnormal mammogram and abnormal biopsy tracking phase, additional staff may be required to enter additional information to the tracking system, to review and evaluate clinical information and medical records, as well as perform outreach and coordination functions at the clinic level. Staff may be centralized and/ or located in medical centers. Colorado and KPNW staffs are centrally located. KPNCAL’s master’s prepared registered nurses and support staff is located at a centralized location and coordinate with department contacts throughout the Region in medicine, OB-GYN, surgery, radiology and pathology. These local contacts are chosen based on an interest to be a point person, a clinical ability for chart review, knowledge of the appointment system and communication and facilitation skills. Georgia’s tracking system relies on LPNs who serve as Facility, Radiology and Surgery Breast Health Coordinators to monitor and track patient care. Their database supports the work of these coordinators.

Please submit additions, updates, and question via email to [email protected] or phone 510-271-5773

Version 2.0 October, 2003 - 30 -

Table 1 1. REGION

Computer Tracking Systems for Breast Care 2. SYSTEM NAME 3. PLATFORM

4. CLINICALCONTA CT

Contacts and General Information

6. # OF FTE’S

7. LINKAGES

1.0 Supervisor (FTE not dedicated solely to this system)

Membership database; Radiology system, appointment system

8. SCOPE (see Table 2)

9. NOTES 10. ADDITIONAL INFO

5. IT CONTACT(S) Colorado

Georgia

BREAST SCREENING SYSTEM

Sue Jane Fox Prevention Specialist 303-344-7256

IBM Mainframe with internet access

Mike Bodily 303-338-3185 Laurence Mapp 303-338-3126

BREAST HEALTH TRACKING DATABASE

Alison Phipps RT(R)(M) Supervisor of Breast Health and Ultrasound Services 770-677-5889

Sybase System

4.0 Women Health Technicians who work from a centralized location 5.0 Breast Health (LPNs) Coordinators

Membership (demographics) database.

Routine Screening Abnormal Mammo

Abnormal Mammo Abnormal Biopsy

MQSRA Support

Farida Bambot 404-943-7726

Used in Denver/Boulder only; Information on family history, personal history and prior diagnosis of atypical hyperplasia is collected and maintained to prompt for more frequent screening. In 2003, inreach prompting will be added. Colorado Description of Breast Cancer Screening and Computer Tracking System

System uses a database, populated by Breast Health Coordinators, to maintain information. Tracking and follow up done by Breast Health Coordinators in Medicine, Radiology and Surgery Department Georgia Breast Health Tracking Powerpoint, Alison Phipps, RT(R) (M), Supervisor of Breast Health and Ultrasound

Group Health Cooperative (IGP)

Group Health Cooperative (IGP)

Breast Screening Recrutiment and Reminder (BSRR)

Susan Carol Bradford, Mangaer, Cancer Management Programs (206) 326-3472

C++, Unix, Sybase

Jim Jackson (206-448-5091)

Mammogram Tracking System (IDX Mam)

Jo Ellen Callahan Manager, Breast Cancer Screening Centers (206) 326-3158

2.0 FTE Operations 0.5 FTE manager, projects 0.1 FTE ISD 5.8 FTE’s RN’s 1.5 FTE admin

IDX, IDX-Mam, Membership and Billing clinical demographics,

Risking Correspondence for Routine Screening

BSRR, CIS Membership and Billing

Routine Screening Abnormal Mammo Abnormal Biopsy MQSA report

.2 FTE ISD Beth Grissom 206-901-6536 Version 2.0 October, 2003 - 31 -

Risk-based, centralized program serving over 100,000 patients; sends letters recommending individual screening program; send reminders when due for screening mammogram. (copy of questionnaire available upon request.)

RN’s in centralized program specialize in tracking and contacting patients

Table 1 (cont) Computer Tracking Systems for Breast Care 1. REGION

2. SYSTEM NAME 3. PLATFORM

Northwest

TRAK

4. CLINICALCONTA CT

6. # OF FTE’S

7. LINKAGES

2.0 FTE

Membership databases; EpicCare; Results reporting; KARE (appointment system)

Contacts and General Information 8. SCOPE (see Table 2)

9. NOTES 10. ADDITIONAL INFO

5. IT CONTACT(S) DEC/Vax II; MUMPS

Bev Battaglia Tumor Registry 503-813-3828 Elaine Hall 503-778-2442

SAFETY NET

A program which builds files

Northern California

BREAST CANCER TRACKING SYSTEM (BCTS)

Nancy Stevens, PhD Prevention Systems Mgr. 503-813-3828

Not Available

Membership databases; EpicCare; Results reporting; KARE (appointment system)

Not a dedicated resource

Elaine Hall 503-778-2442 Mary Callahan Supervising RN Coordinator 510-752-7934

4.0 FTEs RN 3.0 Research Assts. 2.0 Secretaries

Jeff A Hart 510-752-7993 Shirley Palur Director of Radiology 216-362-2586 Bob Szczepanik 216-749-8393

Facility Contacts

DEC/Vax II; MUMPS

Ohio

MAMMOGRAM TRACKING SYSTEM Encounter system in the mainframe

Not available

Radiology, appointments, laboratory, member demographics, pathology, provider database Radiology Information System (RIS)

Abnormal Mammo Abnormal Biopsy

MQSRA Support

Routine Screening

Abnormal Mammo Abnormal Biopsy MQSRA Support

Routine Screening Abnormal Mammo MQSRA Support

Used for other abnormal x-rays in addition to mammograms; Original event is entered manually by Tumor Registry staff. This included entry of protocol which dictates time frame for follow-up. TRAK automatically runs check for follow-up on weekly basis. If pathology report indicates cancer, the incident is entered in the tumor registry and followed per American College of Surgeon guidelines. KPNW Detailed Description of TRAK Letters sent to subset; Health Maintenance Reminders used for inreach. Used for both breast and cervical cancer screening; specifically geared for women 52-69 who have not had a screening mammogram in greater than 2 years. KPNW Detailed Description of Safety Net A Breast Cancer Tracking System, Mary Callahan, RN, MPH; Judith Sanderson, RN, MPA, The Permanente Journal, Fall 2000/Volume 4, No. 4 In mid-2003, BCTS data will transfer into a new webbased Population and Condition Tracking System (PACTS). Please contact Linda Cliff at 510-625-5524. Encounter system links with RIS; tracks abnormal mammograms by code assigned by Radiologist. There is a need for a new system when AMR is implemented. KP Ohio Description of Mammography Tracking System

Version 2.0 October, 2003 - 32 -

Table 2. Computer Tracking Systems for Breast Care Functionality Table

1

CO

GA

GHC BSSR

X

ROUTINE SCREENING

GHC IDX Mam

X

Provide prevention alerts for clinicians and members during 4 clinic visit

• •

Collect and maintain specific risk data for member populations Identify members by demographic/risk information

NW

OH

NCAL

MARS

PHP

TRAK

EpicCare

X

3

X

2 PAS-

•

NW Safety Net

X X

X

X

X

X

• • ●

• •

Check appointment/radiology records for occurrence of mammogram based on specific timeframes

X

Generate automatic notification tools (letters, lists, computer alerts)

X

X

X

Track and identify patients requiring educational materials

X

X X

X

Perform updates to identify and notify members who have not received care Provide written notification of results

X

ABNORMAL EXAM or MAMMOGRAM • • • • • • • •

X

X

Capture of abnormal/suspicious findings Identify members who requiring outreach for scheduling appts. Check appointment/radiology records for occurrence of follow up mammogram, surgery consult or biopsy within designated time period Identify members who have not received recommended follow up care and/or require follow up call, notices/certified letters Generate notification tools (lists, computer alerts, letters) for outreach Provide information for timely callback concerning results and next steps Revises system-generated routine preventive alerts with revised timeframes for procedures and revert to routine schedule when care completed

X

X

X

X X X

X

X X

2003 Upgrade

X

X X X

X X

X X

X X

X

X

X

X

X

X

X X

X

X

X

X

X

X X

X

X X

Version 2.0 October, 2003 - 33 -

X

X X

X

X

Provide point of service reminders to physicians concerning

X

ROAR system

Report to

X

If

2003

member’s next care step

Version 2.0 October, 2003 - 34 -

MD for signoff

overdue

CO ABNORMAL BIOPSIES

• • • • • • •

Capture of pathology information via risk code Provide information for timely callback concerning results and next steps Check appointment records for occurrence of surgery appointments, oncology appointments, mammogram and clinical follow up based on risk code Identify members who have not received recommended follow up care and/or require follow up call, notices/certified letters Generate alerts for physicians for outreach Revises system-generated routine preventive alerts with revised timeframes for procedures and revert to routine schedule when care completed Provide point of service reminders to physicians concerning member’s next care step

OUTCOME STATISTICS FOR MSQRA 5

GA X X X

GHC BSSR

GHC IDX Mam

NW Safety Net

NW

OH

NCAL

X X

X X

TRAK

X Manually

3

Manually

X

X

X For radiology generated referrals

X X X

X X X

X X

X

X

1

X

X

X

Given the differences in functionality of systems and staff responsibilities, this chart should not be used to infer robustness of the system. These functionality tables are current as of March/April 2003. If a box is not checked in the column, it does not infer that the function is not performed in the Region, only that the function is not provided by this computer system. 2 Colorado’s PAS system is under development and will be available 2003/early 2004 3 KPNW links biopsies to mammograms, therefore, no separate functionality is listed for abnormal biopsies 4 This functionality is not part of the system described but available through other systems. SCAL (POINT) and Hawaii (TKNI) also have prevention alert systems in place. 5 Systems vary in ranges of outcome statistics provided.

Version 2.0 October, 2003 - 35 -

X

Kaiser Permanente

Breast Health

Tracking

Alison Phipps, RT(R)(M)

Supervisor of Breast

Health & Ultrasound

Services

Georgia Region

Presentation Date: September 2003 1/20/2004

Next Review: September 2005

Why did Breast Health Tracking Start ?

1/20/2004

Mammography Quality Standards Act of 1992 ( Public Law 102-539)

MQSA was enacted because of growing public concern: • Breast Cancer

•Quality of mammography services to detect early breast cancer

1/20/2004

Mammography Quality Standards Act of 1992 ( Public Law 102-539)

A set of federal guidelines to ensure minimum uniform quality standards in mammography

• Equipment • Personnel • QA/QC • Communication of results

• Medical outcomes audit

1/20/2004

Medical Outcomes Audit 900.12(f)(1) General Requirements

•Each facility must have a system to track positive mammograms and a process to correlate the findings with biopsy results. •Positive mammograms are those with final assessment categories of “Suspicious” or “Highly suggestive of malignancy.”

1/20/2004

Kaiser Permanente Breast Health Coordinating and Tracking Department

1/20/2004

Caring for the patients and treating the disease

History of Breast Health Tracking The Breast Health Tracking department was officially established in April 2000. Prior to the inception of the current model, the tracking of positive breast images were tracked by radiology. Clinical symptoms (I.e.. Abnormal CBE) were tracked by the health care team. Due to extensive membership growth and increasing federal guidelines, the need for a structured system for Breast Health Tracking came into existence

1/20/2004

Why is Breast Health Tracking Needed?

* Systematic approach for communication of results * Ensure compliance with MQSA regulations * Ensure the reliability, clarity and accuracy of interpretation * Reduce possibility of litigation due to delay of failure to diagnose

1/20/2004

•Provide Medical, Surgical, OB/Gyn and Radiology department tracking

What sets us apart from the other tracking departments ?

•Track positive breast imaging studies, abnormal clinical breast exams and patients with strong family history •Coordinate member’s follow-up care for Medical, Surgical, OB/Gyn and Radiology departments •Provide member education on Breast Cancer Awareness •Provides training to health care teams on Breast Health Tracking • Notify members of mammography results •Manage database with highly trained nurses

1/20/2004

Divisions of Breast

Breast Health Coordinators

Coordinators

There are three divisions of breast health coordinators presently available to perform appropriate tracking needs of patients and providers. * Medical BH Coordinator * Radiol ogy BH Coordinator * Surgery BH Coordinator

1/20/2004

Facility Breast Health Coordinators

Coordinators * Coordinates from the aspect of the Health Care Team * Communicates between patients and providers * Ensure providers orders are appropriately followed * Schedules patients being “called back” for additional imaging

* Mobilizes non-compliant patients * Ensure patients are given and comply with surgical referral 1/20/2004

Radiology Breast Health Coordinators

Coordinators * Coordinates from the aspect of the Radiology department

* Monitors all patients assigned a BIRAD code of 3,4, & 5 * Schedules patient for “short term follow up” appointments

* Maintains Radiology Pathology log

1/20/2004

Surgery Breast Health Coordinators

Coordinators * Coordinates from the aspect of the Surgery department * Monitors surgery patient from initial consult, until released by surgeon back to primary care. * Monitor and calculates Medical Outcomes Audit

1/20/2004

Database reports to maintain the standard of care

Due Report: to notify coordinator when reminder letters require mailing. No Results Report: to check for results on studies ordered. No Due Date Report: notify coordinator to check patients outstanding in the system.

1/20/2004

1. Practitioner completes the “Breast Health Tracking Worksheet” (Pink sheet)

Two key steps toward successful breast health tracking

1/20/2004

•Reason for Tracking •Patient History •Follow-up instructions

2. Lead RN or assigned staff Forward worksheet to assigned Breast Health Coordinator

Breast Health Tracking

HCT-OB/GYN

Today’s Date:___________ Medical Office:_________

Breast Health Tracking Worksheet HCT-OB/GYN (sample)

Reason for Tracking: �Abnormal CBE �Cyst Aspiration �Bloody

�Clear

�Non-Resolution �Resolution �Strong Family History Breast Imaging Studies Ordered �Mammography �Ultrasound �Schedule follow-up visit: �_______ weeks �_______ months �Refer patients to:_______________

1/20/2004

____________________Provider Name

Breast Health Coordinators

1/20/2004

Sharon Donahue, LPN

Crescent Centre

770-496-3527

•Surgery

•Alpharetta

•Gwinnett

•Eastside Medical

Bonita Skillman, LPN

Crescent Centre

770-496-3518

•Crescent Center

•Townpark

•Glenlake

•Cumberland

Breast Health Coordinators

1/20/2004

Audryan Graham, LPN

Crescent Centre.

770-496-3529

•Southwood •Panola •Cascade Lavoris Johnson, LPN Crescent Centre 770-496-3530 •Radiology

Moncia Alexander

Southwood

770-603-3703

•Radiology

Breast Health Coordinators

1/20/2004

Department of Breast Health Tracking is centralized at the Crescent Centre Medical Office.

What happens next ?

The coordinators enter the information provided into the system and the tracking process begins. It’s as simple as that !!

1/20/2004

Caring for the patients and treating the disease

• Complete the Worksheet

Essential steps to remember

1/20/2004

• Forward the information

to Breast Health Tracking

For additional information

or assistance, contact:

Alison Phipps, RT(R)(M)

Supervisor of Breast

Health & Ultrasound

Services

Crescent Center

770-677-5889

1/20/2004

MECHANICS OF THE TRAK SYSTEM

Introduction

The TRAK System is a computerized tracking system used to monitor

the care of patients with an abnormal or high risk condition related to

cancer in an effort to reduce any delays in diagnosis. Currently the

system enters all patients with suspicious findings on x-rays,

ultrasounds, CT scans, hemoccult positive tests, cytology reports

excluding pap smears and pap smears for which no follow-up is recorded

within six months. Patients are followed in this system until a

definitive resolution of the condition is made.

Although seen as an extension of the Tumor Registry because of its use

of Registry personnel, TRAK is a totally separate system. Developed as a

manual system in l976, the TRAK System was automated in l984 and

converted to MUMPS language in l986. It runs on the DEC VAX-II, has

on-line inquiry and update and includes an automated link to the

membership data base. Regular reporting requirements are handled through

automated batch processes while ad hoc reporting and inquiry are

supported by a separate database tool known as l032. It does not

interface with the Tumor Registry.

Input - Initial Entry

Based upon identified laboratory and radiology reports that are

sent directly to the Registry, the health record number is entered to

determine if the patient is known to the system. If not, a new Event

Registry Chart Request Worksheet is created and the MMPS file is checked

to automatically enter name, date of birth and sex. This Worksheet

contains the following information:

Event No.

Date Code

Each suspicious/abnormal study is entered as a separate event in order to follow each event to resolution. Thus, one patient may have more than one event being followed. The date the test was performed. Using the CA Relative Value Code, each test is coded and a brief description entered.

Provider

The ordering provider number is entered.

Protocol

Depending upon the type of test, the protocol being used and the individual recommendation of the radiologist reading the film, a code identifying the recommended time for follow-up to have occurred is entered. The protocol codes are: A-I Y-I,II

(red dot) (yellow dot)

C-2

(hemoccult +)

In other words, these codes identify the time frame within which the patient should have received further evaluation or treatment. Comment:

Short description of abnormality and recommended follow-up.

F/U:

A specific date based upon the protocol and the date of the test used to identify when activity regarding this patient should be reassessed. A Friday is always used.

Input-Subsequent

Each Monday every Event Registry Chart Request Worksheet with a

follow up date prior to the previous Friday, is printed.

EpicCare(computerized medical record), Results Reporting and the KARE

System is checked to determine if the patient has been back in contact.

Appt. scheduled Date and notation of Appt. entered.

with logical

provider for F/U date modified up to one month after the Appt.

problem: or procedure to be sure results available in

EpicCare.

Appt. previously scheduled: New abnormal

test result:

EpicCare reviewed, date of visit, code of visit or

procedure type and comment entered.

Enter health record number, determine if this is

follow-up of previously entered event. If yes

date, procedure code and comment entered.

F/U date modified depending upon recommendations or

protocol.

No follow-up indicated:

Upon EpicCare review, if no follow-up has occurred

an EpicCare alert note is sent to the provider

requesting information about what follow up is

planned for the patient. An alert code and a F/U

date of two weeks is entered. If there is no

response from the provider by that date, a second

alert is sent.

Persistent lack of follow-up:

Case referred to Quality Management (KSMC,LK,

Salem) or Medical-Legal (BKMC area)

This process continues until resolution is reached as defined by a

definitive pathology report or other reason as noted in EpicCare. The

resolution code is entered and no further follow up is required.

Simultaneously, all malignant pathology results are entered into

the Tumor Registry for continued monitoring.

JLoewen 6/91

edited BBattaglia 9/93

edited BBattaglia 1/02

doc#12032

Kaiser Permanente Ohio: I. Key Contact: Shirley Palur, Director of Radiology (216-362-2586) (information confirmed with Shirley Palur on 11/05/02)

II. Computerized Mammogram tracking: • All women who have a mammogram receive written notification of their

results

• Tracking of women who require a magnification view or ultrasound • Mammograms that recommend surgical consult and/or biopsy • Mammograms recommending a 6 or 12 month follow-up mammogram are tracked for that appointment • Point of service reminders to physician for women currently being tracked to remind physician of next action for member III. Abnormal Mammography tracking: • Tracking of abnormal mammograms to determine that patient is called within 24 hours, and scheduled for diagnostic follow-up (magnification view or ultrasound) • Tracking to determine that magnification views and ultrasounds are performed within 5 days of patient being notified. • Tracking to determine that surgery appointment is scheduled within 5 days of patient leaving Radiology department • Tracking time from abnormal mammogram to biopsy and pathology report available to the Radiology Department surgeon, and F/U surgery appointment IV. Additional service: Radiology departments where mammography is performed in the region are provided with monthly reports indicating number of mammograms read by each radiologist, false negatives, and follow-up information including pathology result, of all mammograms recommended for surgical consult and/or biopsy. A monthly report provides statistics of each Radiologist, the facility, and the region for a quarterly outcome audit/facility, reviewed semi-annually by the Radiologists. These reports satisfy state licensing and federal requirements of the Mammography Quality Standards Reauthorization Act (MQSRA) of 1998. V. Learnings: • Currently data for time from mammogram to biopsy is being used for quality improvement activity. The quality Improvement activity to track the abnormal mamm event to biopsy is no longer being tracked. Years of tracking indicate significant improvement and only the factors controlled by the Radiology Department will be tracked (diagnostic procedures scheduled and performed within the guidelines; pathology report available to Radiology Department).

PACTS = Population and Condition Tracking System

PACTS is a web-based clinical application developed to provide Kaiser NCAL PCPs, specialists, and departments with a tool for identifying, tracking, and providing outreach to patients with specific or "at risk" medical conditions such as abdominal aortic aneurysms, prostate cancer (abnormal PSA results), breast cancer (abnormal mammograms/biopsies) and cervical cancer; PACT was developed to provide a "safety net" with which to ensure appropriate and timely clinical followup care. PACTS works by comparing nightly downloaded clinical data from various mainframe legacy systems to defined events and timeframes to ascertain patient compliance with agreed upon care paths. Patients who are determined to be out of compliance with a care path are placed on alarm lists accessible by Personal Care Providers (PCPs), specialists, and other responsible staff members. Alarms not acted upon are escalated via email to assigned facility administrators. Anticipated benefits of PACTS include: improved quality of care decreased late or lost-to-follow-up cases reduced medical-legal exposure improved clinician and patient satisfaction. Implementation of PACTS in Northern California is anticipated in mid-2003. Contact: Linda Cliff The Technology Group PACTS Business Project Manager 510.625.5524 8-428-5524 (tieline) [email protected]

VI. Sample Presentations, Additional Resources and Regional CME Contact information Sample presentations and other resources describing issues related to breast cancer and risk management information are listed in this section. Continuing Medical Education contact information is listed below. Individuals are invited to submit presentations and resources for review and posting on this website for program-wide use. Please submit these documents to [email protected].

Presentation 1: Daniel J. Lee, MD., F.A.C. S., The Southeast Permanente Medical Group, Regional Program on

Breast Health,

Presentation: “Issues in Breast Cancer” presented March 20, 2002.

Contents: • Statistics/Risk Factors • Types/Treatment/Follow-up • Recurrent/Metastatic Disease • Management of the Axilla • Sentinel Lymph Node Biopsy Presentation 2: Robert van der Meer, MD, MBA, Chief of Risk Management, The Southeast Permanente Medical

Group, Regional Program on Breast Health, Presentation: “Failure to Diagnose Breast Cancer:

A Risk Perspective” presented March 20, 2002.

Contents: • Legal Reality/Trends • Median Medical Malpractice Awards/Resources • Costs of Delay • Case Profiles Direct questions related to the presentations listed above to Dr. Robert van der Meer Presentation 3: Stefanie D. Kolpak, MD, Department of General Surgery, The Colorado Permanente Group, “

“Assessment and Treatment of Breast Complaints” presented February 5, 2003 as part of CME

program.

Contents: • Case Studies/Assessment and Treatment for • Breast Lump • Breast Pain • Nipple Discharge • Breast Infections

Version 2.0 October, 2003 - 36 -

Issues in Breast Cancer

Daniel J. Lee, M.D., F.A.C.S. The Southeast Permanente Group Regional Program on Breast Health Presentation Date: March 20, 2002

Next Review: March 2004

Issues in Breast Cancer

Statistics/Risk Factors Types/Treatment/Follow-up Recurrence/Metastatic Disease Management of the Axilla Sentinel Lymph Node Biopsy

Making the Diagnosis

• Lifetime risk: 1/8 • 205,000 new cases of breast cancer will be diagnosed in the U.S. in 2002 • 40,000 will die from breast cancer in 2002 • #1 cause of cancer deaths in women 20-59 yo • Number of newly diagnosed cases expected to increase to 420,000/year in 20 years

Risk Factors • Average woman has a 2.5 % chance of developing breast cancer between the ages of 35 and 55 • AGE: 1/2 of a woman’s lifetime risk occurs after age 65

Incidence Breast Cancer

Risk Factors for Breast Cancer Relative Risk 1.7 - 136

Major Factor Radiation Exposure Family History

1.2 - 10 20

Genetic Mutation LCIS ADH Prior Breast Cancer

7 - 12

4-5

2-4

Risk Factors for Breast Cancer Minor Factor Early Menarche Late Menopause Live Birth after 30

Relative Risk 1.5 - 2.0 1.1 - 1.9 2 - 3

Nulliparous Estrogen Replacement EtOH Use Post-menopause Obesity 3

1 - 2.72 1-2

Assessing Risk Level • Gail Model – current age, age at menarche, age 1st live birth, # of previous bx’s, # relatives with breast cancer

• Claus Model – family history, age, # relatives with breast cancer, type/age realatives with breast cancer

Options for High Risk Patients • Prophylactic Mastectomy • Close Surveillance • Tamoxifen 50% reduction in incidence of breast cancer after five years of use (pre & post menopausal)

Breast Cancer • In-situ – ductal – lobular

• Invasive – Infiltrating – Locally advanced – Inflammatory

In-situ Carcinoma of the Breast • DCIS (ductal carcinoma in-situ) – intraductal carcinoma – noninvasive ductal carcinoma

• LCIS (lobular carcinoma in-situ)

DCIS “A group of neoplasms arising from the ductal epithelium confined by the basement membrane and therefore not invading the surrounding stroma. As such, it is largely a maligancy confined to the breast itself with little or no risk of regional or distant spread.”

DCIS • Much more frequently seen on biopsies today – 1985: 7% – 1995: 14% – Today: 25-30%

• Presentation: – Mass – Abnormal calcifications

DCIS: Natural History • Prevalence: 9% of autopsy specimens • High grade: progression to invasive carcinoma in 5-10 years • Small studies of untreated groups: 25% progression to invasive within 10 years

• A precursor lesion

DCIS: Subtypes • Papillary/Micropapillary • Cribriform • Solid • Comedo (Necrosis)

DCIS with calcifications

DCIS

DCIS with necrosis

DCIS with necrosis and calcification

DCIS: Treatment • Simple Mastectomy – high grade, large tumor, multicentric

• Breast Conservation – Lumpectomy with/without breast irradiation – 5-25% recurrence rate

• Tamoxifen – Decrease recurrence in treated breast by 47 %

DCIS: Breast Conservation • Van Nuys Prognostic Index – Point values (1,2,3) assigned based on: • Size/extent of DCIS lesion • Grade of lesion • Margin status

– 8 year disease free survival: • 3,4 (low): 97% • 5,6,7 (intermediate): 77% • 8,9 (high): 20%

DCIS: Radiation - Yes or No?

• No radiation therapy probably fine for low Van Nuys index – small area of DCIS – well differentiated – wide clear margins

DCIS: Lymph Node Dissection?

• Not necessary: about 1% positive LN’s • Role of sentinel lymph node biopsy Only if part of a peer reviewed trial

DCIS: Recurrence

• 50% recurrence is DCIS Treatment: WLE +/- XRT, simple mastectomy

• 50% recurrence is invasive breast cancer Treatment: As with any invasive breast cancer

DCIS: Follow-up • Lifetime follow-up • Mammography/PE every 6 months in treated breast for 2 to 3 years • Bilateral mammography yearly

LCIS • 17% incidence of subsequent invasive carcinoma in either breast (0.5-1 % per year) – usually infiltrating ductal

• 50-90% have LCIS in the other breast • Presentation: Found on biopsy, no typical mammographic findings, “thickening” • Treatment: – Bilateral simple mastectomy – close clinical observation

Invasive Breast Cancer: Pathology

1. 2. 3. 4. 5.

Infiltrating ductal (75%) Infiltrating lobular (5-10%) Medullary (5-7%) Mucinous or Colloid (3%) Tubular (2%)

Infiltrating Ductal Carcinoma

Infiltrating Ductal Carcinoma

Metastatic Breast Cancer to lymph node

Invasive Breast Cancer: Pre-treatment workup • Rule out metastatic disease – History and Physical Exam – CXR, LFT’s – Bone scan low yield if normal alkaline phosphatase

TNM Classification • T: tumor size T1: 2 - 5 cm

T3: > 5 cm direct extension

• N: lymph node status T4: N0: no nodes N1: mobile axillary

N2: matted/fixed N3: internal mammary

• M: metastasis M0: no mets

M1: distant mets

Breast Cancer Staging • Stage I

87% 5-yr survival

– T1N0

• Stage II a

78%

“

68%

“

51%

“

20%

“

– T1N1, T2 N0

• Stage II b – T2N1, T3 N0

• Stage III – T1N2, T2N2, T3N1,2

• Stage IV – any M1

Early Stage Breast Cancer

Stage I and II

• T 1 or 2, N 0 or 1 • Treatment Options (NSABP B-06, Veronisi)

– Breast conservation, axillary treatment, breast irradiation – Modified radical mastectomy

• Need adequate removal of the primary tumor with pathologic negative margins • Individualize for the patient

Early Stage Breast Cancer Stage I and II 1992 NIH Consensus Conference: “breast conservation treatment is an appropriate method of primary therapy for the majority of women with stage I and II breast cancer and is preferable because it provides the surgical equivalent to total mastectomy and axillary dissection while preserving the breast”

Locally Advanced Breast Cancer • T 3, N 2 or 3 • T 4, any N • Large • Extensive regional lymph node involvement

Locally Advanced Breast Cancer Historical Perspectives • 1940’s, surgery alone – 53% local failure – 0% 5 year survival

• Radiation – 10-30% 5 year survival

• Combined – Better local control but no better survival

Locally Advanced Breast Cancer Treatment • Obviously a systemic disease at time of diagnosis • 1970’s: Added chemotherapy • Treatment of Choice – Induction – Surgery – Adjuvant Chemotherapy/XRT

Inflammatory Breast Cancer • • • •

Rare, virulent form of advanced breast cancer 1-6% of all breast cancer erythema/warmth/edema of the breast distinguish from infection – rapid onset of symptoms

• Diagnosis via breast skin biopsy • Multimodality treatment

Invasive Breast Cancer Follow-up • Aware of long-term risk of recurrence/metastatic disease • most recurrence within 5 years, some >20 • 65-85% recurrence identified on H&P • PE every 4 months for two years, every 6 months for the next three years then yearly • Mammography at least yearly

Invasive Breast Cancer Local Recurrence • 5-10% 10 year recurrence in the conserved breast • Salvage mastectomy: 50-65% 5 year survival • Chest wall recurrence after mastectomy – usually within 2-3 years after surgery

Invasive Breast Cancer Metastatic Disease

• Median survival after diagnosis: 2 years • Treatment: palliative

Management of the Axilla • Contributes little to overall survival • Local Control – 40% patients with clinically negative nodes have histologic involvement of axilla – 18% patients without axillary treatment will develop axillary disease

• Staging – Nodal status is a predictor of outcome • 0 nodes involved: 70% 10-year survival • 1-3 nodes involved: 40% 10-year survival • >4 nodes involved: 60% of claims for delay of diagnosis are from premenopausal women. >70% of total indemnity claims are from premenopausal women. >30% of claims are from women under the age of 40. ~40% of the total indemnity claims go to women under the age of 40.

THE PATIENT F

The most common patient is peri- or premenopausal, presenting with a self identified lesion or abnormality

THE PATIENT The ages ranged from 23-84. F Median age was 45. F 6.5% were in their 20’s. F Although women under the age of 40 only make up ~7% of breast cancer cases, they account for 40-50% of cases alleging delay of diagnosis of breast cancer. F 6% of patients were pregnant. F

THE PATIENT 17%

9%

46%

9%

16%

MASS / NO PAIN MASS & PAIN PAIN / NO MASS OTHER NO SYMPTOMS

THE PATIENT PRESENTING SYMPTOMS PA M N O A M I SS SY A N/ SS N /N M PT O & OM O TH PA P O A A IN IN SS ER MS

17% 9% 9% 16% 46% 0%

10% 20% 30% 40% 50%

THE DELAY F

The average alleged delay in diagnosis was 14.6 months.

PIAA PIAA- -1995 1995

THE COST OF DELAY >54 MO. (7 CASES)

$437,786

48- 53 MO. (4 CASES) 42-47 MO. (1 CASE) 36-41 MO.(9CASES) 30-35 MO. (10 CASES) 24-29 MO. (34 CASES 18-23 MO. (60 CASES) 12-17 MO. (113 CASES) 6-11 MO.(164 CASES) 35

It‘s important to differentiate the patient‘s actual concern. • —A“ œ that her symptoms are due to a sinister underlying pathology

• —B“ œ that she is sufficiently affected by the pain and wants specific treatment

It‘s almost always due to —A“, where reassurance may be sufficient therapy.

Cyclic breast pain without masses • Mild to moderate œ œ œ œ œ œ

reassure supportive bra like a sports bra NSAIDS decrease caffeine oil of primrose* → long chain FA (γ linolenic acid) vitamin E

• Moderate to severe œ oral contraceptive or cessation of ERT *only agent actually shown by a double blind study to significantly decrease cyclic mastalgia compared to placebo

Mondor’s Disease → • Superficial thrombophlebitis of lateral thoracic vein • Exam reveals linear pain and palpable cord, most prominent in upper aspect of breast • Causes œ œ œ œ

50% no underlying pathology 30% local trauma or surgery 8% inflammatory process 12% carcinoma

• So these women require good exams, mammograms, and follow-up for resolution. • Treatment: œ NSAIDS œ warm compresses œ support bra

Case # 3 • 73 year old woman comes into your office with a chief complaint of fluid draining from her right breast

Where do you start? • What specific things do you want to know? • What are the important risks factors? • Why is her family history important (or is it)?

On physical exam... • What specifics are you looking for? • Do you need to perform guaiac test or send for cytology? • What if you can‘t reproduce discharge?

Next step in management... • • • •

What do you do if it‘s bloody? What if it‘s not? What if you feel a mass? Does she need a mammogram? What about a galactogram? • What sorts of things can cause bloody discharge?

Nipple Discharge • History œ œ œ œ œ

breast cancer risk factors? duration? spontaneous? associated lump? recent manipulation?

• Exam œ bilateral exam with attention to nipple • single versus multiple ducts • bloody versus milky/blue/green • unilateral versus bilateral • review mammogram

Evaluation of Fluid • Non-bloody

• Blood or guaiac +

œ ? Medication cause • phenothiazines • estrogens • opiates

œ check labs • TSH • prolactin

œ treat accordingly • surgical referral unnecessary

œ mammogram œ surgical referral

What happens at surgical referral? • Review mammograms and history • Full breast exam œ œ œ œ œ

number of ducts involved? easy or difficult to express fluid? grossly bloody or guaiac+ both breasts involved? any palpable masses? • mass should be worked up first with FNA or core biopsy

œ responsible quadrant?

Bloody Nipple Discharge • Fewer than 10% of patients with a bloody

discharge will have an underlying cancer

• Varies with age (% malignant)

œ < 40 (3%)

œ 40-60 (10%)

œ >60 (32%)

Surgical Options→ • Selective Duct Excision œ responsible duct cannulated œ duct access via a circumareolar incision œ duct removed from insertion on nipple out laterally

• Central Duct Excision œ used if can‘t cannulate duct or identify a responsible quadrant œ larger circumareolar incision and removal of all ductal elements at nipple • problems with sensory loss, nipple necrosis, cosmesis

Case # 4 • 28 year old woman comes into your office 3 weeks post partum with redness and pain in her left breast

Evaluation of this patient includes... • What aspects of her history, both past and recent, would you like to know? • Does it matter if she‘s breastfeeding or not? • Do you need to know her breast cancer risks factors? Why or why not? • What personal history is important?

On physical exam... • What aspects of the physical exam are concerning for abscess? For cancer? • Are there non-operative options for abscesses? What findings mandate open drainage? • What complication can develop for open drainage of a lactating woman? • When/if would you recommend cessation of nursing?

Lactating versus non-lactating • How do approaches differ? • Is follow up any different? • Is the presence of palpable axillary nodes significant? • What about inflammatory cancer?

Breast Infections • History œ œ œ œ œ œ

duration? prior abscesses? lactating? systemic symptoms? breast cancer risk factors? nipple piercing or tattoo?

• Exam œ clearly examine breast for extent of erythema, fluctuance, skin thinning, tenderness, etc.

Treatment: warm packs tylenol/ NSAIDS antibiotics

----------------------------------------------------------------------------------------

If resolved: ⇓ Resume regular screening guidelines

If resolving:

If no better or worse: ⇓ ⇓ Continue Ultrasound antibiotics, and surgical re-examine to referral for document abscess. resolution.

What happens at the surgery appointment? • Full exam and review of most recent mammograms and ultrasound. • FNA and/or ultrasound to r/o abscess • Biopsy any persistent masses. • Repeated aspirations for abscesses with operative drainage reserved for failures.

A Word About Lactational Mastitis • < 2% incidence • typically presents 2-4 weeks post-partum • therapy with compresses, ibuprofen, and antibiotics the same • it‘s ok to continue breastfeeding œ in fact this can improve symptoms and drainage

• open drainage should be avoided œ milk fistulas can develop and are difficult to treat

Inflammatory Breast Cancer • < 2% of all breast cancers œ quite uncommon

• initially appears as tenderness œ a mass may be present but initially overlooked by the patient

• inflammatory response of the breast with enlargement, induration, and edema of the ENTIRE breast œ erythema most pronounced in the dependent portion of the breast œ most patients will have palpable axillary nodes

Refer patient to surgery for evaluation and skin biopsy. These patients are NOT curable at presentation and chemotherapy is only option and results are dismal. Rarely will the woman require a mastectomy for control of local disease (ulceration, bleeding, etc.)

Paget‘s Disease • Often mistaken for dermatitis of areola • < 2% of all breast cancers œ Paget cells are not invasive, they may indicate underlying invasive component.

• Progression: œ burning, itchy nipple œ nipple erythema and edema œ erosion and crusting of nipple and areola

• Reasonable course of therapy: œ change bra or washing soap œ steroid cream for 2-3 weeks œ surgical referral for evaluation if no resolution

Treatment of Paget‘s Disease • Review mammogram for other suspicious findings. • Since disease is in the nipple-areolar complex, mastectomy is usually necessary. • Therapy is total mastectomy with evaluation of axillary lymph nodes reserved for those patients with invasive cancer

Additional Resources Clinical Forms • A sample breast evaluation clinic record (adapted from Walnut Creek, CA) • A sample mammography requisition (provided by Ohio region)

Procedures • A biopsy result call back procedures (provided by Hawaii) • Best practices for Telephone Outreach for Mammography screening (provided by Northern California) • Breast Screening Program (provided by Colorado) • Safety Net System (provided by KP Northwest)

Education • A brochure detailing a 7.25 hour CME program: “Breast Health Update” (sponsored by Colorado) • A list of patient education resources available on intranet (PKC) and internet • Continuing Medical Education Information Evidence Reviews • Evidence Synopsis Template: Breast Problem Management Clinic Evaluation KP-Health Connect Clinical Content Development • Fibrocystic Breast Disease Mini Evidence Review

Version 2.0 October, 2003 - 37 -

SAMPLE BREAST EVALUATION RECORD

REFERRAL: Age:

BP:

TEMP:

S: PRESENT COMPLAINT: G____ P____ A___

LNMP:

PERSONAL Hx OF BREAST CANCER:

HORMONE STATUS: MEDS:

F.H. OF BREAST CANCER or OTHER CANCERS (Relationship, Ages, DX): GYN SURGERY:

DATE:

SURGERY FINDINGS: PREVIOUS BREAST SURGERY

R:

L:

DATE:

SURGERY FINDINGS:

PREVIOUS MAMMOGRAM NO___ YES___

DATE:

MAMMOGRAM FINDINGS:

PREVIOUS US Findings:

Mass: R: ______ L: ______

O:

Size:

Character:

Contour:

NIPPLE DISCHARGE: SKIN CHANGES

NO___ YES___

NO___ YES___

ADENOPATHY NO___ YES___ A:

CLINICAL DIAGNOSIS:

P:

DISPOSITION:

R

L

Version 2.0 October, 2003 - 38 -

___ MAMMO/US:

___TRACKING:

___ BREAST ASPIRATED: DATE

___RETURN VISIT: DATE

___ NEEDLE BIOPSY: DATE

___DO SELF EXAM EACH MONTH AND REPORT CHANGES ( AS PER INSTRUCTIONS)

___ SURGICAL BIOPSY: DATE

___ HANDOUT

___ REFERRAL:

___ OTHER

Version 2.0 October, 2003 - 39 -

RADIOLOGY DEPARTMENT MAMMOGRAM REQUISITION

REFERRING PROVIDER’S I.D. #

PROVIDER’S PAGER #

DATE REFERRED:

FACILITY:

DOES THIS PATIENTS HAVE BREAST IMPLANTS? [

] NO [

] YES => Requires 2 appointment slots when scheduled

[

] SCREENING MAMMOGRAM (ONLY if Patient has NO lump, focal pain or discharge, NO prior breast cancer)

[

] DIAGNOSTIC CLINICAL MAMMOGRAM (You MUST check one or more of the following): [ ] PRIOR BREAST CANCER ] Right [ ] Left Date(s): ______________________ [ ] NO CURRENT SYMPTOMS ] CURRENT SYMPTOMS (MUST Indicate Abnormality and [

Location Below)

[ ]

[

ABNORMAL CLINICAL BREAST EXAM or SYMPTOMS (MUST Indicate Abnormality and Location Below) [ [ ] DISCHARGE [ ] LUMP (Must identify location) [ ] FOCAL PAIN (Must identify locat ion) [ ] OTHER (Describe, including location): _______________________________________________ Location (Circle Right or Left to indicate which breast): [ ] Upper Outer Quadrant Right Left ] Upper Midline (12 o’clock) Right Left [ ] Upper Inner Quadrant Right Left Right Left [ ] Retro-areolar [ Lower Outer Quadrant Right Left ] Lower Midline (6 o’clock) Right Left [ ] [ ] Lower Inner Quadrant Right Left 6 MONTH [FOLLOW-UP of prior abnormal mammogram (Code 7) ONE YEAR FOLLOW-UP of prior abnormal mammogram (Code 8)

[ ] [ ] ] STEREOTACTIC BIOPSY

[

] NEEDLE LOCALIZATION

L.M.P. ___________________________ PREGNANT: [ ] YES [ ] NO Mammographer’s Initials

[

] Right Breast

[

] Left Breast

[

] Bilateral

DATE OF LAST MAMMOGRAM _______________________________ BWC:

[ ] YES [ ] NO 8 X 10

EMPLOYEE: 10 X 12

[ ] YES [ ] NO

FOR MAMMOGRAPHER’S USE ONLY:

Version 2.0 October, 2003 - 40 -

0THER

] NO

Family history of breast cancer Sister [

] NO

Hormonal Replacement Therapy [ ____________ ] Prior Breast Cancer: ] Mastectomy Bilateral/Date_______________ ] Lumpectomy [ Bilateral/Date_______________ [ ] Radiation Therapy [ Bilateral/Date_______________ [ [[ ] Prior Biopsy

] YES ⇒ Family Member (please circle):

[ [

Daughter Year Started ___________ Year Stopped

] YES ⇒ If Yes,

[

[

] Right/Date_______________

[

Mother

]

Left/Date_______________

[

] Right/Date_______________

[

] Left/Date_______________

[

] Right/Date_______________

[

] Left/Date_______________

[

] ] ]

] Right/Date_______________ [ Left/Date_______________ ] ] Bilateral/Date_______________ [ [ 3 years in Radiology Dept, attach the report; if bx. not performed in Radiology Dept., Note: If bx. performed within past attach the Pathology report. [ ] Reduction Mammoplasty Bilateral/Date_______________ [

] Right/Date_______________

[

] Left/Date_______________

Additional Comments: (Revised: January/03)

Version 2.0 October, 2003 - 41 -

[

]

Biopsy Results Call Back Process (Hawaii Regional Process)

This process applies to breast biopsies done in the Diagnostic Imaging Department (stereotactic or U/S guided breast biopsies)

1.

Pt. has biopsy appointment

2.

At time of biopsy, pt. is given a results phone call appt. date and time a. This results phone call appt. will be 48 – 72 hours after biopsy is done

3.

Pt. is phoned at time of appt. and given biopsy results and follow-up information

4.

If biopsy result is negative, pt. is given follow-up information a. For stereotactic bx., follow-up in 6 months with diagnostic mammogram on breast that was biopsied, return to normal screening mammo at 1 yr. b.

5.

For U/S guided biopsy, there is no set guideline so we usually f/u in 1 year with a screening mammo bilat.

If biopsy is positive, surgical consultation appt. date and time is obtained and patient is phoned with information that surgical consultation is needed. a. Patient is NOT told over the phone that they have breast cancer. Pt. is told that the biopsy results have shown some changes and that the surgeon will be talking to them about removing a few more cells from around that area. b.

Much reassurance is given with this phone call and the surgical consultation appt. is usually given within 1-3 working days.

Version 2.0 October, 2003 - 43 -

Breast Care Task Force

Mammography Outreach Recommendations Operational Best Practices

July 2003

Version 2.0 October, 2003 - 44 -

Summary This document was developed by the Mammography Subgroup of the Breast Care Task Force to outline operational Best Practices for conducting Mammography Screening Outreach to ensure that women at the highest age related risk for breast cancer receive appropriate screening. The recommendations included were derived from research on mammography screening, including outreach practices currently in use throughout the Region, literature review, and clinical and operational expertise contributed by members of a multidisciplinary workgroup. The recommendations for mammography outreach detailed in this document address the following steps: 1. Regularly obtaining up-to-date patient outreach lists by using Reports on Automated Request (ROAR). 2. Calling patients on the outreach list to schedule mammogram appointments. 3. Reserving mammogram appointments to ensure appointment supply for outreach patients. The members of the BCTF Mammography Subgroup are: Lee Davidson, MD Chief, Radiology, Diablo Service Area Lou Fehrenbacher, MD Chief, Oncology, Vallejo Jennifer Horsma Quality and Operations Support Laurie Miller, MD Physician, OB/GYN, San Francisco Jennifer Slovis, MD Assistant Chief, Internal Medicine, Oakland Carol Somkin, PhD Division of Research, Regional Offices, Oakland Dave Spring, MD Physician, Radiology, Oakland Laurie Weisberg, MD Physician, Oncology, South San Francisco

Version 2.0 October, 2003 - 45 -

We would like to thank the Stockton Kaiser facility for considerable contributions to this work.

Contact Jennifer Horsma at 8-428-6893 or Linda Satterfield at 8-428-3078 with questions regarding this document. Purpose The intent of this document is to assist medical centers in conducting mammography outreach in the most efficient and effective manner by providing:

1. A flexible recommendation format to enable customization to individual facility operations 2. A framework from which to structure a Mammography Screening Program 3. Supporting tools and resources

Background The practices listed in this document have been shown to be the most effective in improving metrics for Mammography Screening Quality Goals. The mammography screening Quality Goal was determined by the TPMG Board of Directors and is similar to the NCQA HEDIS requirement for mammogram screening. The objective of this goal is to ensure that women at the highest age related risk for breast cancer, between the ages of 50-75 years of age, receive a mammogram at least once every two years. While we wish to ensure that all our members receive optimal care, we particularly wish to encourage female members between the ages of 50-75 years to comply at least with these minimum requirements. Kaiser’s Clinical Practice Guidelines for breast cancer screening are more comprehensive than the basic Quality Goals mammography screening requirement, providing guidelines for optimal breast care health. Regional Clinical Practice Guidelines for breast cancer screening in average risk patients* are as follows: ‹

Clinical breast exam is recommended every 1-2 years starting at age 20 for women of average risk.

‹

Mammography is recommended every 1-2 years for average-risk women age 40 -74.

‹

Mammography is at patient/provider discretion for all women over age 75.

*Average risk applies to patients with none of the risk factors listed below: • Personal history of breast cancer • Family history of first-degree relative (mother, sister, daughter) whose onset of breast cancer was before age 50 • Breast biopsy showing atypical hyperplasia or lobular carcinoma-in-situ (LCIS) Version 2.0 October, 2003 - 46 -

How to Use this Document The intended audience for this document is anyone responsible for designing, managing, or coordinating mammography screening outreach. This document is meant to assist in the development or revision of a mammography screening outreach program. Begin by comparing the checklist (on page 5) with your facility’s current practices to ensure that the most comprehensive approach for Outreach is used.

Version 2.0 October, 2003 - 47 -

Table of Contents:

Topic

Page

I.

Checklist: Basic Criteria in Effective Mammography Outreach Programs

5

II.

Recommendations Regarding Mammography Outreach Operational Steps

6-7

General recommendations pertaining to the operational steps involved with a Mammography Outreach program are outlined including the rationale supporting the recommendations. Recommendations are divided among the following four operational steps: 1.

Requesting ROAR Mammography Outreach List

6 6-7

2.

Mammography Outreach

7

3.

Mammogram Appointment Scheduling

7

4. III.

Monitoring of Mammography Outreach

Recommended Mammography Outreach Operational Model

8-9

A graphical representation of the recommended Mammography Outreach model is depicted on page 8, with supporting descriptive documentation found on page 9. 1.

Mammography Outreach Model 2.

IV.

8

Key Departments and Staff Functions In Outreach Model

Documentation: Mammography Outreach Tools and Resources

9 11-17

The following tools and resources support the operational steps involved with mammography outreach efforts. 11-12

1.

Outreach Call Script

2.

Breast Imaging Consultation Form

3.

Outreach Outcome Log

4.

Outreach Support: Who to Contact, When, and How

13 14 15

5.

Patient Letters

Version 2.0 October, 2003 - 48 -

16-17

Checklist: Basic Criteria for Effective Mammography Outreach Programs Based on a Regional survey, the most effective Mammography Screening Outreach Programs incorporate the following criteria. If you currently have a Mammography Screening program in place at your facility, use the following criteria as a checklist. If your current program fulfills all of the following criteria and you feel that your current program is effective in supporting achievement of Mammography Screening Quality Goals, then revision of your program is probably unnecessary. However, if one or more of the criteria below are not addressed by your current program, target these areas for improvement.

ß Clearly defined personnel roles to support mammography outreach efforts:

Do the individuals involved with mammography outreach know what they have to do and why? Are all the pieces of mammography outreach covered and clearly assigned? ß Accountability for outreach efforts:

Do the individuals involved with mammography outreach know how they are supposed to fulfill their role and the timeline for completion of their responsibilities? ß Designated outreach resources:

Do all parties concerned understand that mammography outreach is a designated part of the participant’s job description? ß Continuous phone outreach efforts:

Are outreach calls conducted on a daily basis over the course of the year? (This evens out outreach resources and supports consistent demand for mammography appointments to assist Radiology in supply planning.) ß Centralized outreach efforts:

Are outreach efforts coordinated at the Medical Center level? (This centralizes outreach resources.) Does one department coordinate outreach? (This helps avoid duplication of outreach efforts.) ß Designated mammogram appointments to ensure appointment supply for outreach

patients: Does Radiology reserve special appointments that are only bookable for patients contacted through outreach efforts? ß Consistent outreach counseling:

Do all outreach callers follow the same outreach call script? ß Documentation of outreach attempts:

Is a record kept of attempts to contact the patient and the outcome? ß Continuous feedback and monitoring of outreach efforts and results: Version 2.0 October, 2003 - 49 -

Does someone oversee the success of the outreach program, taking the lead to trouble-shoot and address problems as they arise? ß Routine communication among key personnel involved with Mammography

Outreach: Do all individuals involved with mammography outreach communicate regularly? Mammography Outreach Operational Steps Operational Step 1. Requesting Mammograph y Outreach Patient Lists

Recommendation

Rationale

Request ROAR Mammography Outreach Lists monthly.

Monthly requests result in a shorter list of “new” patients and support continuous outreach efforts (advantages are listed below).

Parameters: women 50-75yo without mammogram in past 22 months.

Permits enough time for the patient to schedule and have her mammogram within a 24mo (two year) period.

(ROAR can be used to request patient outreach lists for women of any age within any time period.) 2. Outreach

Coordinate outreach efforts at the Medical Center Level (i.e., one Call Center Manager per MC).

Coordination of outreach efforts at the Medical Center centralizes outreach resources. Most Medical Centers have the ability to schedule mammograms at their satellite facilities.

Design systems to support continuous patient outreach: conduct daily outreach calls at a rate to reach all new patients on the outreach list for that month, within the month.

Continuous outreach efforts even out the resources devoted to outreach (i.e. time per week devoted to outreach calls), and provide Radiology with a consistent demand for mammograms which helps Radiology manage its appointment supply.

(Example: 200 new patients on the Outreach list divided by 20 workdays in the month equals 10 outreach calls per day.) Contact patients by telephone (where possible).

Data from a Regional survey of mammography outreach practices indicates that phone outreach is more effective than mail outreach, resulting in approximately 30% acceptance of mammogram screening versus