Related Experiment Videos
Detection of false-negative Pap smears using the PAPNET system.
G M Troni1, I Cipparrone, M P Cariaggi
1Department of Cytopathology, Center for Cancer Study and Prevention, Florence, Italy.
Tumori
|February 24, 2001
Summary
Computer-assisted Pap smear reading using PAPNET technology slightly improved detection of cervical cancer precursors but at a significantly higher cost. This approach may identify a few additional cases but is not cost-effective for screening programs.
Area of Science:
- Gynecologic Oncology
- Digital Pathology
- Screening Technologies
Background:
- False-negative diagnoses in cervical cancer screening are a significant challenge.
- Computer-assisted methods, like neural network technology, are explored to improve early detection.
Purpose of the Study:
- To evaluate the performance and cost-effectiveness of computer-assisted (PAPNET) versus conventional manual Pap smear reading.
- To compare sensitivity, specificity, and cost per detected lesion.
Main Methods:
- Independent interpretation of 1,000 routine smears (including 81 known false-negatives) by two readers using both conventional and PAPNET-assisted methods.
- Analysis of sensitivity for false-negative smears, specificity, and cost-outcome.
Main Results:
- PAPNET-assisted reading demonstrated a minor increase in sensitivity for one reader.
- The cost per detected lesion was $7,543, and the cost per additional detected lesion was $25,748, including PAPNET costs.
Conclusions:
- PAPNET-assisted screening can detect a limited number of additional cervical intraepithelial neoplasia grade 2+ (CIN2+) lesions compared to conventional methods.
- The high cost associated with PAPNET makes its current cost-outcome ratio unfavorable for widespread screening implementation.