Surgical Pathology—Second Reviews, Institutional Reviews, Audits, and Correlations: What's Out There? Error or Diagnostic Variation?

Abstract

Abstract Context.—A variety of methodologies have been used to report error rates in surgical pathology within the peer-reviewed medical literature. The media has selectively and superficially reported these error rates creating a climate of disinformation between physicians and the public. Objectives.—To review the medical literature on diagnostic error in surgical pathology and summarize and compare these data with selected reports in the print and broadcast media. Design.—A search of the medical literature from the National Library of Medicine database using the heading “Error and Pathology Diagnosis.” Results.—Three thousand nine hundred ninety-two citations were found, of which 83 directly measured in some manner errors in surgical and cytopathology. Major error rates ranged from 1.5% to 5.7% globally for institutional consults. Error rates were less, 0.26% to 1.2% for global in-house prospective review and 4.0% for in-house and retrospective blinded review. Error rates also varied by anatomic site: skin, institutional consult, 1.4%; prostate, institutional consult, 0.5%; and thyroid, institutional consult, 7.0%. Error rates reported in citations used by the Wall Street Journal were as follows: prostate, Gleason score changed by 1 point, 44% and resultant change in treatment for prostate cancer, 10%; for breast, altered lumpectomy or mastectomy plan, 8%; and diagnosis changed for thyroid lesions, 18%. Errors in second opinion on breast lesions (single pathologist author for the study) fall within the range of global reviews. Errors for second opinions on prostate cancer were principally 81% upgrades in Gleason score for prostate core needle biopsies. However, this resulted in an upgrade of patient risk category in only 10.8% of patients. Data for the article on change in diagnosis of thyroid lesions were incomplete. There appeared to be 3 significant diagnostic errors (4.5%). Conclusions.—Pathology is not immune to the power of the media to create concern about accuracy of diagnosis in surgical pathology and cytopathology. Detailed analysis of the medical literature cited by the media determines that painting the big picture and hitting the highlights can be profoundly misleading.