Home > For Researchers > All Cancers > All Cancers Overall Summary: Cancer Cases and Validation > Breast Cancer Validation
breast cancer Validation (DR7)
Confirmation of Breast Cancer Diagnosis
Self-reports are the main source of initial information collected on breast cancer and other health conditions in the Sister Study. Participants who develop breast cancer report their diagnosis through telephone calls, e-mails, correspondence with the Sister Study Helpdesk, or regularly scheduled follow-ups (Annual Health Updates or Detailed Follow-Up Questionnaires). Once a study participant reports a breast cancer diagnosis, she is asked to provide a copy of her pathology report, if available.
Approximately six months after the reported date of diagnosis, participants are asked to complete a brief questionnaire (Breast Cancer Follow-Up Form, BCFF) and to permit contact of their health care provider for retrieval of relevant medical records and pathology reports. Medical records are abstracted onto standard forms (Breast Cancer Medical Report Form, and when applicable, Stage IV Breast Cancer Medical Report Form) to verify self-reported breast cancer diagnosis information, collect information needed for staging, and obtain additional diagnostic information. Information abstracted includes:
- Tumor characteristics
- Hormone receptor status
- Lymph node involvement
- Metastasis
- Treatments (surgery, chemotherapy, radiation, biological therapy, hormonal treatment, clinical trial enrollment)
- Genetic testing
Concordance of Self-Reported and Medical Record-Abstracted Breast Cancer Characteristics
Data Release Information: DR 7.0:
(Requires Adobe Acrobat - Download Here) 
Links to other Data Releases, including detailed analysis of concordance in DR 4.0
As of release 7.0, the Sister Study has medical records for 82.3% of women that reported a breast cancer diagnosis (3,512 total incident breast cancer events). Among those who have provided medical records, the Positive Predictive Value (PPV) of a self-reported breast cancer is 99.3%. Concordance PPVs for self-reported subtypes and characteristics can be found below.
| Self-Reported Breast Cancer* | % Confirmed by Medical Records (PPV) |
|---|---|
Breast Cancer (any type) |
99.3% |
* All breast cancers reported to the Sister Study.
| Self-Reported Breast Cancer Type by Invasiveness | % Confirmed by Medical Records (PPV) |
|---|---|
|
Invasive (any)
|
99.3%
|
|
In Situ (only)
|
64.6%
67.3% |
| Self-Reported Type by Ductal or Lobular Status†‡ | % Confirmed by Medical Records (PPV) |
|---|---|
|
Ductal (Any)
|
98.5% 96.3% |
Both Ductal/Lobular or Mixed |
39.7% |
|
Lobular (Any)
|
73.6% 74.6% |
† Excludes n = 5 cases not of ductal or lobular origin (phyllodes tumors).
‡ Special subtypes of inflammatory or Paget’s were assigned accordingly to ductal or lobular categories.
Hormone Receptor Characteristics among those with Medically Confirmed Invasive Cancers
[This excludes those that: a) did not complete the Breast Cancer Follow-Up Form (self-report of hormone receptor characteristics) and b) are missing hormone receptor data in medical records].
| Self-Report Type | % Confirmed by Medical Records (PPV) |
|---|---|
|
Estrogen Receptor (ER)
|
99.3% |
|
Progesterone Receptor (PR)
|
98.5% |
|
Human Epidermal Growth Factor Receptor 2 (HER2)
|
64.6% |
*Further excluded 2 women from concordance results who have medical confirmation of breast cancer but disavowed (refuted) their original self-report of breast cancer.
Other Data Releases:
Breast Cancer Validation: Data Release 3
Breast Cancer Validation: Data Release 4
For more information on concordance between breast cancer self-report and medical record(s) (DR 4.0):
- D'Aloisio AA, Nichols HB, Hodgson ME, Deming-Halverson SL, Sandler DP. Validity of self-reported breast cancer characteristics in a nationwide cohort of women with a family history of breast cancer. BMC Cancer. 2017 Oct 23;17:692.
For full text (citation above), please see the link below. Note: Manuscript pdfs are not 508 compliant however, compliant versions can be obtained from the respective journals, accessed via the link in the citation.
|