Demographic and Screening-to-Diagnosis Cascade Indicators for Risk Stratification in an Organized Cervical Cancer...

To assess an organized cervical cancer screening-to-diagnosis cascade in Yunnan Province, China, and to examine the associations of demographic factors, HPV genotyping, and cytology results with the detection of cervical precancer and cancer.

This retrospective analysis included 82,141 women who participated in a government-supported free screening program between January 2023 and July 2025. The primary endpoint was the final integrated diagnosis based on cytology, colposcopy, and histopathology, with histopathology serving as the diagnostic gold standard. Associations between candidate variables and the outcome were assessed using chi-square tests and multivariable logistic regression, and adjusted odds ratios (aORs) with 95% confidence intervals (95% CIs) were reported. Uniform Manifold Approximation and Projection (UMAP) was used for multidimensional visualization of high-risk clustering patterns.

Among the participants, 1678 (2.04%) were classified as positive for cervical cancer or precancerous lesions. In the descriptive analysis, positivity rates showed significant non-linear variation across age groups, with higher proportions in women aged < 25 and ≥65 years, and were also higher among women with lower educational attainment and in some ethnic minority groups, including Hani and Dai women. In multivariable analysis, age remained significant overall, although only women aged 35-44 years had higher adjusted odds than those aged 45-54 years, while ethnicity was not significant overall except for a higher risk observed among Dai women. HPV positivity showed the strongest association with the outcome, and risk increased markedly across cytology categories from NILM to HSIL. UMAP visualization further showed that high-risk cases formed compact clusters enriched for HPV16/18 positivity and high-grade cytology.

In this organized screening program, HPV status and cytology were strongly associated with the detection of cervical precancer and cancer. Educational level and some demographic factors were also associated with variation in screening outcomes after adjustment, although these associations were more modest than those observed in the descriptive analysis. Demographic indicators such as age and educational level may be useful for identifying underserved groups and informing targeted referral and follow-up strategies within organized screening programmes.