Clinical and Virological Profiles Associated with CINTEC^® PLUS Positivity: A Data-Driven Clustering and Modeling...

Area of Science:

• Gynecologic Oncology
• Medical Diagnostics
• Machine Learning in Healthcare

Background:

• The diagnostic accuracy of CINtec® PLUS for cervical cancer screening can be affected by patient characteristics and risk factors.
• Understanding these influences is crucial for accurate interpretation and patient management.

Purpose of the Study:

• To evaluate and model risk factors associated with CINtec® PLUS test positivity.
• To assess the predictive performance of machine learning models for cervical intraepithelial neoplasia (CIN) 2/3.

Main Methods:

• Retrospective analysis of medical data from 134 patients with human papillomavirus (HPV) infection who underwent CINtec® PLUS testing.
• Application of a gradient boosting classifier and XGBoost model to predict CINtec® PLUS positivity and CIN2/3 outcomes based on clinical risk factors.

Main Results:

• The gradient boosting model achieved 75% precision and 0.77 AUC for predicting CINtec® PLUS positivity, with body mass index and age as key predictors.
• High-grade squamous intraepithelial lesion (HSIL), atypical squamous cells of undetermined significance (ASC-US), and high-risk HPV strains increased positive test likelihood.
• The XGBoost model showed strong performance for CIN2 prediction (0.90 AUC) but limited specificity for CIN3 prediction (0.58 AUC).

Conclusions:

• Patient characteristics and risk factors significantly influence CINtec® PLUS positivity rates.
• These factors must be carefully considered for appropriate management strategies in cervical cancer screening.