SEARCH

SEARCH BY CITATION

Keywords:

  • breast imaging;
  • DCIS;
  • double reading;
  • mammography.

Abstract

Introduction

Ductal carcinoma in situ (DCIS) is often only mammographically evident as microcalcification. Although the overall percentage of screening cases with histologically proven DCIS microcalcification is small, the clinical relevance of missing this finding is significant. The current guidelines in Australia for breast screening departments are for double reading of mammograms to reduce both perceptive and interpretative error.

Methods

This retrospective study identified patients from a state screening program with histologically proven DCIS whose mammograms showed microcalcification. The initial double reader results were documented according to the 5-point grading scale of BreastScreen Tasmania, and discrepancies between readers were noted. Mammographic factors such as breast density, lesion location, morphology, distribution, size and presence on previous imaging were assessed for significant influence on inter-reader discrepancy. Histological evidence of invasion and grade of malignancy were also analysed.

Results

Of 65 identified cases, 29 (45%) showed that one of the two readers had not flagged the microcalcification on the report. Analyses revealed no significant difference in reader discrepancy with any of the analysed factors including breast density, size of microcalcification or presence on previous imaging. Twenty-five of 29 (86%) cases of discrepancy were perceptive.

Conclusion

Breast screening reading for microcalcification is poorly correlated to mammographic or histological features. The majority of errors were perceptive rather than interpretative. Double reading is advocated as standard practice to reduce perceptive error.