Baseline inaccuracy rates for the comparison of cervical biopsy to loop electrosurgical excision histopathologic diagnoses.

Abstract

The purpose of this study was to quantify the baseline inaccuracies in histopathologic diagnosis between colposcopically directed cervical biopsies and loop electrosurgical excision procedure specimens. A retrospective review was made of cervical biopsies and loop electrosurgical excision procedure pairs (n = 448) that were performed at the University of Arizona between 1992 and 2000. Agreement was assessed by kappa statistics and Spearman rho coefficients. Subgroup analysis was performed (n = 139) to assess pairs with significant inaccuracy between biopsy and loop electrosurgical excision procedure, when these were performed within 28 days of each other. Eighty-one percent agreement was found between cervical biopsy and loop electrosurgical excision procedure histopathologic diagnoses, when compared to within 1 degree (kappa, 0.73; P <.0001). A discrepancy that would be attributed to regression or progression was found: mild biopsy (regression, none by definition; progression, 4%), moderate biopsy (regression, 9%; progression, 0%), and severe biopsy (regression, 26%; progression, 0%). Even when colposcopically directed cervical biopsy correlates strongly with loop electrosurgical excision procedure histopathologic diagnosis, there is a definable percentage of patients for whom there is a discrepancy between these 2 diagnoses. This discrepancy represents an inherent inaccuracy of the diagnostic testing modality and should not be interpreted as progression or regression of disease.