Young investigator challenge: Comparison of 2 different methods of manual slide screening in semiautomated gynecologic cytology.

Abstract

Negative gynecologic cytology cases (ie, those diagnosed as negative for intraepithelial lesion or malignancy) are manually reviewed by 2 methods using semiautomated screening: 1) immediate full slide review (FSR) after fields-of-view analysis (FOV) (FOV + FSR), and 2) quality-assurance/high-risk, quintile-directed full manual review (FMR). Data supporting current guidelines were limited. The authors investigated FMR, FOV + FSR, and the review process in general. Gynecologic cytology cases from 2009 to 2014 at Massachusetts General Hospital were analyzed. The data comprised 93,169 patients, 194,656 specimens, and 49,979 human papillomavirus (HPV) tests. In patients who underwent FMR, the epithelial cell abnormality (ECA) rate was correlated with the HPV-positive rate (correlation coefficient [r(2) ] = 0.82; Y = 0.19X + 0.02), and both rates decreased with age. For patients who underwent FOV + FSR, the ECA rate was also related to the HPV-positive rate (r(2) = 0.86; Y = 0.39X + 0.11), and both rates decreased with age. The FMR group had similar HPV-positive rates compared to the FOV + FSR group (2%-52% vs 9%-68%, respectively). HPV-positive patients had a higher risk of ECA than HPV-negative patients (40% vs 8%; P < .0001). Currently, manual review allocates resources inefficiently, to older, HPV-negative patients as compared to younger, HPV-positive patients. Additionally, too many patients who have HPV-positive specimens with ECA proceed from primary screening to FMR. This imbalance can be observed in the slope of the line comparing the FMR ECA rate with the HPV-positive rate (which might serve as a surrogate marker for adequacy of FOV + FSR screening). Therefore, laboratories should try to reduce the number of HPV-positive cases that reach FMR. Current review procedures devote too few resources to younger patients. FMR should be assigned according to HPV status and age. The slope of the line comparing the FMR ECA rate with the FMR HPV-positive rate might be a useful surrogate marker of screening accuracy.