Reviewing histologic diagnosis of lymphoma: comparison of original and review diagnoses in 269 cases.

By analyzing the pathology records of patients at Emory University Hospital, Atlanta, who had a diagnosis of suspected or proved malignant lymphoma (ML) based on examination of a biopsy specimen obtained elsewhere, the rates of agreement between original and review diagnoses were determined. Ninety-six percent of the cases submitted as ML were accepted as ML on review. In 96% of those referred to us as non-Hodgkin's lymphoma (NHL), this diagnosis was confirmed. All cases submitted as Hodgkin's disease (HD) were accepted as ML, but 13% were reclassified as NHL. Based on the review diagnoses, expressed according to the Lukes-Butler (HD) or Rappaport (NHL) classifications, 69% of the HD cases and 29% of the NHL cases bore original diagnoses that agreed. Nonuse and incorrect use of standard classifications contributed greatly to these discrepancies. Pathology review is strongly recommended for all cases of ML, to reduce diagnostic error where possible and to promote the use of histopathologic terminology that bears on clinical decisions.