P. A. Trott, R.R. Millis, Nick Ryley, J. Coyne, S Humphreys, Clive A. Wells, J. Lowe, John P. Sloane, J. M. Theaker, C. L. Brown, N S Dallimore, D.J. Scott, J.M. Sloan, J. Nottingham, J. D. Davies, C.W. Elston, D. Eakins, Ian O. Ellis, R. Ellman, H. D. Zakhour, J.O'd. McGee, Thomas J. Anderson, and D. Lawrence
The aim of the scheme was to determine consistency of histopathological reporting in the United Kingdom National Breast Screening Programme. This external quality assessment scheme involved 51 sets of 12 slides which were circulated to 186–251 pathologists at intervals of 6 months for 3 years. Participants recorded their diagnoses on standard reporting forms, which were submitted to the U.K. National Cancer Screening Evaluation Unit for analysis. A high level of consistency was achieved in diagnosing major categories of breast disease including invasive carcinoma and the important borderline lesions, radial scar and ductal carcinoma in situ (DCIS), the latter exceeding a national target set prior to the onset of the scheme. Atypical hyperplasia (AH) was reported with much less consistency although, where it was the majority opinion, over 86% of diagnoses were of benign disorders and only 14% were of DCIS. Inconsistency was encountered in subtyping and measuring DCIS, the former apparently due to current uncertainties about classification and the latter to poor circumscription, variation in size in different sections and merging with zones of AH. Reporting prognostic features of invasive carcinomas was variable. Measurement of size was achieved with adequate consistency except in a small number of very poorly circumscribed tumours. Grading and subtyping were inconsistent although the latter was not specifically tested and will be the subject of future study. Members of the National Coordinating Group achieved greater uniformity than the remainder of the participants in all diagnostic categories, but both groups experienced similar types of problem. Our findings suggest that participation in the scheme improves diagnostic consistency. In conclusion, consistency in diagnosing invasive carcinoma and radial scar is excellent, and good in DCIS, but improvements are desirable in diagnosing atypical hyperplasia, classifying DCIS and reporting certain prognostic features of invasive rumours. Such improvements will require further research, the development of improved diagnostic criteria and the dissemination of clearer guidelines.