gms | German Medical Science

54. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

07. bis 10.09.2009, Essen

Histopathology of Core Biopsies of the Female Breast: Interobserver Variability and its Determinants. Preliminary Results of the DIOS Study

Meeting Abstract

  • Andreas Stang - Institut für Klinische Epidemiologie, Medizinische Fakultät, Martin-Luther-Universität Halle-Wittenberg, Halle
  • Alexander Kluttig - Institut für Medizinische Epidemiologie, Biometrie und Informatik, Universität Halle, Halle
  • Pietro Trocchi - Institut für Klinische Epidemiologie, Medizinische Fakultät, Martin-Luther-Universität Halle-Wittenberg, Halle
  • Kathirn Ruschke - Zentrum für Radiologie, Universitätsklinik und Poliklinik für diagnostische Radiologie, Halle
  • Andrea Schmidt-Pokrzywniak - Institut für Klinische Epidemiologie, Medizinische Fakultät, Martin-Luther-Universität Halle-Wittenberg, Halle
  • Hans-Jürgen Holzhausen - Institut für Pathologie, Medizinische Fakultät, Martin-Luther-Universität Halle-Wittenberg, Halle
  • Thomas Löning - Albertinen-Pathologie, Hamburg
  • Jörg Buchmann - Institut für Pathologie, Städtisches Krankenhaus Martha-Maria Halle-Dölau, Halle
  • Anke Heinig - Zentrum für Radiologie, Universitätsklinik und Poliklinik für diagnostische Radiologie, Halle
  • Christoph Thomssen - Zentrum für Frauenheilkunde und Geburtshilfe, Universitätsklinik und Poliklinik für Gynäkologie, Halle
  • Tillmann Lantzsch - Klinik für Frauenheilkunde und Geburtshilfe, Krankenhaus St. Elisabeth und St. Barbara, Halle
  • Thomas Decker - Gerhard-Domagk-Institut für Pathologie, Universität Münster, Münster
  • Steffen Hauptmann - Institut für Pathologie, Medizinische Fakultät, Martin-Luther-Universität Halle-Wittenberg, Halle
  • Werner Böcker - Gerhard-Domagk-Institut für Pathologie, Universität Münster, Münster

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 54. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds). Essen, 07.-10.09.2009. Düsseldorf: German Medical Science GMS Publishing House; 2009. Doc09gmds043

DOI: 10.3205/09gmds043, URN: urn:nbn:de:0183-09gmds0435

Published: September 2, 2009

© 2009 Stang et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

Introduction: According to European guidelines, the histology of breast core biopsies is categorized by the B-categorization (B1-B5). If histologies are categorized as B3-B5, women usually undergo open surgery of the breast. The aim of this study is to estimate the interobserver variability of the B-categorization and to study potential determinants.

Methods: We recruited consecutive women undergoing core biopsies of the breast at University of Halle. We documented radiological findings including BIRDAS-assessment categories (BIRADS: 0=incomplete, 1=negative, 2=benign finding(s), 3=probably benign, 4=suspicious abnormality, 5=highly suggestive of malignancy, 6=known biopsy-provden malignancy), clinical findings and gynecological histories in a standardized way. Histological slides of the biopsies were assessed by two independent experienced pathologists in a blinded fashion. We calculated observed and chance-corrected agreements (unweighted kappa) and 95% confidence intervals (95%CI).

Results: From April 2006 through July 2008, 765 consecutive women were included. 112 women (15%) had a history of breast cancer, 142 women (19%) without a history of breast cancer had breast symptoms, 83 women (11%) had pathological breast images in the organized screening program, and 428 women(56%) without a history of breast cancer and without symptoms had pathological images (either ultrasound, MRT, or mammography) detected outside of the screening program. The observed agreement of B-categorization (B1-B5) between pathologists was 0.86 (95%CI: 0.83-0.88). The kappa was 0.76 (95%CI: 0.72-0.80). Kappa depended on the BIRADS assessment: BIRADS 2-3: kappa=0.49 (95%CI: 0.34-0.65), BIRADS 4: 0.71 (95%CI: 0.65-0.76), and BIRADS 5: 0.84 (95%CI: 0.64-1.00).

Discussion: Overall, the reliability of the histopathological assessment of breast core biopsies is high. We provide evidence that this agreement is substantially influenced by the pre-biopsy BIRADS-based ratings of the images. Agreement is especially low for lesions rated as benign or probably benign according to BIRADS. The distinction between B1 and B2 was especially problematic when MRT-based biopsies were rated.


References

1.
Acknowledgment: Funded by a Federal grant of BMBF, grant number 01ZP0507