Screening parameters for ThinPrep and conventional gynecologic cytology via automated monitoring

To compare diagnostic discrepancies and screening parameters between conventional (CP) and ThinPrep (TP) (Cytyc Corporation, Boxborough, Massachusetts, U.S.A.) cervicovaginal samples using Pathfinder (Neopath, Redmond, Washington, U.S.A.).

Pathfinder tracked average screening time, percent slide coverage and percent overlap of viewing fields for CP and TP. False negative rate (FNR) was determined by rescreening 10% of random and high-risk negative cases. CP and TP FNR with Pathfinder were compared to control groups without Pathfinder.

A total of 46,393 Pathfinder cases were evaluated (43,354 CP, 3,039 TP) as compared to 62,981 without Pathfinder (60,307 CP, 2,674 TP). FNR was calculated for 12,983 negatives. Using Pathfinder resulted in a significant reduction in FNR for CP atypical squamous cells of undetermined significance and atypical glandular cells of undetermined significance cases. No decrease in FNR was observed for CP squamous intraepithelial lesions or for TP cases. TP slides were screened 66 seconds faster on average than CP. With electronic feedback, mean percent slide coverage and percent overlap were similar between CP and TP cases. Without feedback, coverage dropped and overlap increased slightly for both CP and TP. Technologists screened faster with feedback, saving an average of 50 seconds on CP and 41 seconds on TP.

Pathfinder significantly reduced FNR for CP but not TP. Technologists screened TP significantly faster than CP while maintaining similar coverage and overlap. Pathfinder feedback itself may decrease screening time.