Diagnostic discrepancies and their clinical impact in a neuropathology referral practice

Open Access
Publisher Website
Google Scholar
Abstract
BACKGROUND During the course of their neuropathology practice, the authors received cases to review in consultation. In some cases, the patients came to the authors' hospital for therapy; in others, the primary pathologists requested a consultation. Because changes in diagnosis might significantly alter patient management, protocol entry, care costs, or the potential for physician liability, the authors determined the frequency and degrees of their disagreements with the original diagnoses submitted to them. METHODS The authors reviewed the first 500 brain or spinal cord biopsy cases that were submitted to their neuropathology consultation service for a second opinion in 1995. Disagreements were coded into 10 categories, but were grouped for this analysis as follows: serious (having immediate significance for therapy or intervention), less serious but potentially substantial (calling for a change in type or grade of glioma), minor (adding or deleting information), and those in which the authors made the first diagnosis themselves. RESULTS There was some degree of disagreement between the original and review diagnoses in 214 (42.8%) of the 500 cases. Disagreements were counted as serious in 44 cases (8.8%), less serious but substantial in 96 cases (19.2%), and minor in 50 cases (10.0%); the authors made the first diagnosis in 24 cases (4.8%). CONCLUSIONS Clinically important diagnostic errors that can affect immediate patient care decisions occur in a substantial number of brain and spinal cord biopsy cases. Thus, seeking expert neuropathology consultation is prudent and cost‐effective for pathologists who are less experienced with these types of cases. Cost savings in case management might result from confirmation of diagnosis before definitive therapy is administered to the patients. The rates of discrepancy between original diagnoses and second opinions in other subspecialties of pathology should be examined. [See editorial on pages 665‐7, this issue.] Cancer 1997; 79:796‐803. © 1997 American Cancer Society.