RESUMO
BACKGROUND: To the authors' knowledge, published studies reporting on the performance of the FocalPoint GS (FPGS) imaging system have yielded conflicting results to date. However, the results of the MAVARIC study indicated that the FPGS "No Further Review" (NFR) aspect of the technology demonstrated potential and warranted further investigation. The current validation study was performed prior to implementing the NFR slide reporting technology within the routine cervical screening program in Wales, United Kingdom. METHODS: A total of 45,317 SurePath liquid-based cytology cervical screening samples were submitted for FPGS scanning by 4 Welsh laboratories between 2006 and 2011. The current study (Computer Assisted Evaluation, Screening and Reporting [CAESAR]), reports on a comparison between slides categorized as NFR (8130 slides) and slides manually screened as negative (93,473 slides). Both interventions had a subsequent negative quality control screen. RESULTS: The histological outcome rates of cervical intraepithelial neoplasia 2 (CIN-2) (high-grade squamous intraepithelial lesion or worse [HSIL+]) at 2 years and subsequently 3 years after an FPGS NFR result versus a manually screened negative result were compared. Significantly fewer cases were detected in the NFR cohort compared with the manually screened cohort (P = .043 at 2 years and P = .027 at 3 years). When these cases were subcategorized as cancers and precancers, the interval cancer prevalence between NFR and manually screened samples at 2 years and 3 years was similar; however, the interval precancer prevalence for FPGS NFR was significantly lower (P = .023 at 2 years and P = .026 at 3 years) at approximately one-half that of manual screening. CONCLUSIONS: The negative predictive potential of the FPGS NFR technology is higher than that of manual screening, and the technology has quality/throughput benefits to support and enhance a laboratory cervical screening service.
