Does use of the AutoPap assisted primary screener improve cytologic diagnosis?

To correlate the ranked review slides by the AutoPap Assisted Primary Screener with the final cytologic diagnosis and to assess whether the ranking of slides improves diagnostic accuracy.

A total of 5,865 consecutive conventional and suitable cervical/vaginal smears, including high-risk (HR) cases, were processed by the AutoPap System. All slides designated Review were manually screened by two cytotechnologists using the Ranked Review Report. All abnormal findings and reactive/reparative changes were referred to two attending cytopathologists for a final diagnosis. After screening, the Review slides called Within Normal Limits (WNL) were selected according to rank for a further Quality Control (QC) review by the supervisor. All HR cases not selected by AutoPap for QC review were also rescreened. The final diagnoses of the possibly abnormal/reactive/reparative referred slides were recorded and distributed in five ranks according to the Ranked Review Report assignment.

Of 5,865 slides, 5,120 (87%) qualified for scanning. The AutoPap System designated 3,840 (75%) slides for manual review. One thousand three hundred forty-five slides were assigned for QC review. After elimination of nonqualified slides, 763 remained. Rescreening detected 1 high grade squamous intraepithelial lesion (HSIL), 2 low grade squamous intraepithelial lesions (LSIL), 5 atypical squamous cells of undetermined significance (ASCUS) and 5 Benign Reactive Changes (BRC). QC of 364 HR cases revealed 1 LSIL, 2 ASCUS and 3 BRC. Ultimately, 313 (8.1%) smears were diagnosed as ASCUS or a more severe abnormality, 262 (6.8%) as reactive/reparative changes, 3,259 (84.8%) as WNL and 6 (0.1%) as unsatisfactory. Of 313 abnormal slides, 181 were ranked by the system in the 1st rank, 61 in the 2nd, 38 in the 3rd, 19 in the 4th and 14 in the 5th. No HSIL or more severe lesions occurred in the fourth and fifth ranks.

Our study validated the claim by the manufacturer that a significant epithelial abnormality is more likely to be present if a high score is assigned to a slide. The preliminary results support use of the AutoPap Assisted Primary Screener to improve cytologic diagnosis.