Multiple Dimensions of Social Vulnerability and Cervical Cancer Screening Uptake
Julie Haury, Jean Baptiste Chanier, Bruno Renevier, Geoffroy Canlorbe, Simon CrequitImportance
Despite organized screening programs, persistent social inequalities undermine cervical cancer screening effectiveness in high-income countries, disproportionately affecting the most socially vulnerable women.
Objectives
To identify distinct social vulnerability profiles among pregnant women and examine their associations with cervical cancer screening adherence and intraepithelial lesion detection.
Design, Setting, and Participants
This retrospective cohort study assessed all pregnant women 25 years or older who delivered at Centre Hospitalier Intercommunal André-Grégoire, a single tertiary care maternity unit in France, between May 1, 2019, and May 1, 2025.
Exposures
Social vulnerability was assessed across 15 indicators spanning administrative, psychological, and violence-related domains. Multiple component analysis and hierarchical clustering on principal components identified 6 distinct vulnerability profiles. Multiple logistic regression examined associations between profiles and cervical screening inadequacy (no cervical sample within the last 3 years, including those never screened) and squamous intraepithelial lesion detection, adjusting for maternal age and body mass index.
Main Outcomes and Measures
Cervical screening inadequacy (primary outcome) and low- and high-grade squamous intraepithelial lesion detection (secondary outcome).
Results
Among 18 253 women (median [IQR] age, 32.3 [29.1-36.0] years; median [IQR] body mass index, 25.9 [22.7-29.4]), 10 248 (56.1%) were up to date with screening. Six vulnerability profiles were identified. Compared with the low-vulnerability reference profile, all profiles showed increased odds of screening inadequacy: substance use and psychological vulnerability (adjusted odds ratio [AOR], 1.33; 95% CI, 1.20-1.48), migrant or linguistic barrier (AOR, 3.40; 95% CI, 3.16-3.66), psychological violence (AOR, 2.01; 95% CI, 1.75-2.32), sexual violence (AOR, 2.08; 95% CI, 1.74-2.48), and physical violence (AOR, 2.54; 95% CI, 2.19-2.95). Violence-exposed and substance use profiles also had significantly higher proportions of women with a detected intraepithelial lesion despite reduced screening participation.
Conclusions and Relevance
In this cohort study of pregnant women, distinct social vulnerability profiles were differentially associated with cervical cancer screening inadequacy; women with migration-related barriers experienced the greatest disadvantage, and violence-exposed women showed both reduced screening and elevated disease burden. These findings suggest profile-specific interventions could help reduce cervical cancer screening inequalities among socially vulnerable populations.