Distinct Vulnerability Profiles Mapped to Screening Gaps in Pregnancy
A large French cohort of pregnant women has drawn a sharper line between social vulnerability and cervical cancer prevention. In this retrospective study of 18,253 women who delivered at a single tertiary maternity unit, six distinct vulnerability profiles emerged from 15 indicators spanning administrative, psychological, and violence-related domains. Those profiles were not evenly distributed in their impact. Some were associated with modestly higher odds of cervical screening inadequacy, while others carried a far steeper penalty, especially among women facing migration-related barriers or exposure to violence.
The cohort covered deliveries between May 1, 2019, and May 1, 2025, and included pregnant women 25 years or older. Median age was 32.3 years, and median body mass index was 25.9. Just 56.1% were up to date with screening, leaving a substantial share outside recommended cervical prevention pathways. That gap is not an abstract public health metric. It is the practical consequence of missed contact with a health system that depends on repeated engagement, trust, language access, and continuity of care.
Migration Barriers Carried the Sharpest Screening Penalty
Among the six profiles, the most disadvantaged group was women with migrant or linguistic barriers. Compared with the low-vulnerability reference profile, this group had an adjusted odds ratio of 3.40 for screening inadequacy, with a 95% confidence interval from 3.16 to 3.66. No other profile approached that magnitude. The result points to a layered barrier structure in which administrative navigation, communication, and prior access to prevention services can fail at the same time.
The remaining profiles also showed worse screening adherence than the reference group. Women in the substance use and psychological vulnerability profile had an adjusted odds ratio of 1.33, while those in the psychological violence profile had an adjusted odds ratio of 2.01. The sexual violence profile carried an adjusted odds ratio of 2.08, and the physical violence profile had an adjusted odds ratio of 2.54. Each confidence interval stayed clearly above 1, indicating a consistent excess risk across the vulnerability spectrum.
The spread across these estimates matters clinically. It suggests that a single screening reminder or a generic outreach campaign is unlikely to work equally well for all socially vulnerable pregnant women. A woman who faces administrative barriers and language discordance encounters a different obstacle course than a woman living with coercion or trauma. Both may miss screening, but for different reasons, and prevention systems that treat them as interchangeable will miss the point.
Violence Exposure Linked to Missed Screening and More Detected Lesions
The most clinically unsettling pattern came from the violence-exposed profiles. Women in the psychological, sexual, and physical violence groups not only had higher odds of screening inadequacy, they also had higher proportions of detected intraepithelial lesions despite lower participation in screening. The same pattern appeared in the substance use profile. In practical terms, fewer women in these groups were screened, yet among those who were screened, abnormal findings still emerged often enough to signal a real burden of disease.
This is where prevention and diagnosis collide. Lower uptake does not imply lower need. When women with repeated trauma or substance use are screened less often, the system is not simply seeing less disease; it may be missing disease that is already present. The study did not quantify progression beyond lesion detection, but the finding is enough to show that social vulnerability can shape both who enters screening and who leaves it with a clinically meaningful result.
Administrative, Psychological, and Violence-Related Factors Did Not Behave the Same Way
The investigators used multiple component analysis and hierarchical clustering on principal components to reduce 15 social indicators into six profiles. That approach is important because vulnerability is rarely singular. Administrative insecurity, psychological distress, and violence exposure often overlap, but they do not always travel together. The clustering approach allowed the study to compare combinations of risk rather than isolate one social factor at a time.
Adjusted analyses accounted for maternal age and body mass index. Even after that adjustment, every vulnerable profile carried higher odds of screening inadequacy than the low-vulnerability reference group. The result argues against the idea that poor screening participation is explained by age or body size alone. In this cohort, social position itself remained tightly linked to whether preventive cervical care had been delivered on time.
The clinic-level implication is straightforward. Screening programs that wait for routine engagement will continue to miss the women at highest risk of falling behind. Migration barriers call for interpretation support, administrative simplification, and access pathways that do not assume fluent navigation of the health system. Violence-exposed women require a different kind of response, one that recognizes trauma, safety concerns, and the likelihood that a pelvic examination may not be a neutral encounter.
Why Pregnancy Is a High-Value Window for Prevention
Pregnancy offers repeated contact with care, which makes it a valuable moment for catching overdue screening. The fact that more than 40% of this cohort was not up to date despite that contact suggests a stubborn gap in preventive delivery. Prenatal care can surface unmet needs, but only if systems are designed to notice them and respond with the right intervention at the right time.
The findings also carry relevance beyond France. Organized screening programs in high-income countries are often judged by population-level participation rates, yet those averages can conceal deeply uneven access. A program can look successful on paper while still leaving women with language barriers, unstable lives, or violence exposure outside its protective reach.
This study does not establish causation, and it does not test a specific intervention. It does, however, identify where the faults in the system are deepest. Migration-related barriers were associated with the greatest disadvantage, and violence-exposed women were the clearest example of a group facing both lower screening and greater evidence of disease. Any future response will need to be profile-specific rather than one-size-fits-all, because the pathways into missed screening are not the same for every vulnerable patient.