Discordance between uterine cervical cytology and biopsy: results and etiologies of a one-year audit.

To investigate the etiologies of discrepancies between cervicovaginal smear and corresponding cervical biopsy results, a total of 15,474 cervicovaginal smears were sampled in a one-year period. Among these, 427 patients were diagnosed with atypical squamous cells of undetermined significance (ASCUS), dysplasia, or malignancy. The screen positive rate was 2.8%. All of the positive cases had histologic follow-up. Forty-nine of the 427 patients had a discrepancy of at least two grades (the grades are divided to negative, ASCUS, mild dysplasia, moderate dysplasia, severe dysplasia and invasive carcinoma), between the cytologic and histologic diagnoses. The discrepancy rate was 11.5%. Ten of these discrepant cases had poorly-preserved slides or a not definitely final diagnosis. A total of 39 cases (79.6%) of discrepancy were reviewed in this study. In thirty (77%) of the 39 discrepant cases, the errors were cytologic and in 9 cases (23%) the errors were histologic. Cytologic error was the major cause of cytohistologic discrepancy. The etiologies of cytohistologic discrepancy included: cytologic interpretation error, 17 cases (44%); cytologic sampling error, 10 cases (25%); biopsy sampling error, 6 cases (15%); cytologic screen error, 3 cases (8%); and biopsy interpretation error, 3 cases (8%). The major etiology of cytohistologic discordances was cytologic interpretation error. In this retrospective study, we determined the etiologies of cytohistologic discrepancies. This information can be useful for improving diagnostic accuracy and the quality of patient care.