BCSC
Breast Cancer Surveillance Consortium: Working together to advance breast cancer research
 
Home   |   Data   |   Statistics   |   Tools   |   Collaborations   |   Work with Us   |   Publications   |   About   |   Links

A Breast Pathology Quality Improvement Project

Grant #: U57
PI Name: Carney P
Title: A Breast Pathology Quality Improvement Project
Institute: New Hampshire Division of Public Health and Centers for Disease Control and Prevention


The overarching goal of this study is to assess and improve breast pathology diagnostic agreement. In 1991, Rosai reported that even for highly respected surgical pathologists with considerable experience in breast pathology, interobserver diagnostic variability for borderline epithelial lesions was unacceptably high (Rosai, 1991). In a similar, follow-up study, Schnitt reported that the diagnostic reproducibility of proliferative breast lesions was improved (complete agreement in 58% of cases) if all participants adhered to previously agreed upon standardized diagnostic criteria (Schnitt, 1992). Few studies have focused on the diagnostic reproducibility of routinely practicing pathologists without a special interest or expertise in diagnostic breast pathology. An Italian study showed good overall diagnostic agreement but cases of potentially harmful diagnostic errors were also evident. This Italian study, unlike most, did control for the technical quality of the histological sections, but the cases selected for review were known to present diagnostic problems (Bianchi, 1994). In the present study, the slides evaluated represent randomly selected cases encountered in community practice.

The investigators propose to undertake three study aims as part of this investigation: 1) to work with New Hampshire pathologists to achieve statewide standardization of breast pathology report content; 2) to conduct a statewide evaluation of overall diagnostic agreement for breast pathology; and 3) to assess diagnostic agreement according to the type of ductal carcinoma in situ classification systems used. By conducting these studies, they may identify how interpretive performance could be improved.