RESUMEN
For quality assurance purposes, the authors correlated aspiration cytology and subsequent tissue findings and examined the reasons for discrepancies. In three months from Jan. to Mar. 1996, 1,383 aspirations were performed, of which 285 (20.6%) had subsequent tissue diagnoses within two months. The aspiration sites were thyroid (483), lymph node (LN) (290), breast (250), soft tissue (154), liver (89), lung (49), salivary gland (26), pancreas (22), gallbladder (3), bone (3), joint (2), adrenal gland (2), scrotum (2), mediastinum (2), omentum (2), oral cavity (1), chest wall (1), and intraabdominal (1) and pelvic cavities (1). A total of 68 discrepancies were identified, and biopsies and smears from these cases were reviewed monthly. In 27 cases (40%), the discrepancy was attributed to sampling error. In five cases (9%), aspiration gave superior results with better sampling and preservation than biopsy. Thirty six cases (53% of discrepant cases) were errors in cytologic diagnosis. We categorized these discrepancies into "A", "B", and "C" ("A": minor disagreement with no affect on patient care, "B": minimal affect on patient care, "C": major affect on patient care), which were 9 (13%), 14 (21%) and 13 (19%) cases, respectively. In thirteen cases of category "C", there were eleven false negative and two false positive diagnoses. Eleven false negative cases included thyroid (3), lymph node (2), breast (2), bone (1), salivary gland (1), lung (1), and liver (1). Three cases of thyroid were papillary carcinomas diagnosed as nodular hyperplasia (1), occasional pleomorphic cells (1), and cystic change (1). Two breast cases of invasive ductal carcinomas were diagnosed as ductal hyperplasia. A malignant lymphoma was diagnosed as reactive hyperplasia and a metastatic carcinoma of LN was diagnosed as tuberculosis. Other cases were malignant tumors of bone, salivary gland, lung, and liver those were misinterpreted as benign lesion or normal. Of two false positive cases, one was nodular hyperplasia of thyroid diagnosed as papillary carcinoma and the other was normal islet cell of pancreas diagnosed as islet cell tumor. A continuous monitoring of laboratory performance is an essential component of the quality control and assurance, and the review of discrepant cases provides useful information for improvement of diagnosis.