Association between social vulnerability profiles, prenatal care use and pregnancy outcomes

Abstract

Abstract Background Evaluating social vulnerability is a challenging task. Indeed, former studies demonstrated an association between geographical social deprivation indicators, administrative indicators, and poor pregnancy outcomes. Objective To evaluate the association between social vulnerability profiles, prenatal care use (PCU) and poor pregnancy outcomes (Preterm birth (PTB: <37 gestational weeks (GW)), small for gestational age (SGA), stillbirth, medical abortion, and late miscarriage). Methods Retrospective single center study between January 2020 and December 2021. A total of 7643 women who delivered a singleton after 14 GW in a tertiary care maternity unit were included. Multiple component analysis (MCA) was used to assess the associations between the following social vulnerabilities: social isolation, poor or insecure housing conditions, not work-related household income, absence of standard health insurance, recent immigration, linguistic barrier, history of violence, severe dependency, psychologic vulnerability, addictions, and psychiatric disease. Hierarchical clustering on principal component (HCPC) from the MCA was used to classify patients into similar social vulnerability profiles. Associations between social vulnerability profiles and poor pregnancy outcomes were tested using multiple logistic regression or Poisson regression when appropriate. Results The HCPC analysis revealed 5 different social vulnerability profiles. Profile 1 included the lowest rates of vulnerability and was used as a reference. After adjustment for maternal characteristics and medical factors, profiles 2 to 5 were independently associated with inadequate PCU (highest risk for profile 5, aOR = 3.14, 95%CI[2.33–4.18]), PTB (highest risk for profile 2, aOR = 4.64, 95%CI[3.80–5.66]) and SGA status (highest risk for profile 5, aOR = 1.60, 95%CI[1.20–2.10]). Profile 2 was the only profile associated with late miscarriage (adjusted incidence rate ratio (aIRR) = 7.39, 95%CI[4.17–13.19]). Profiles 2 and 4 were independently associated with stillbirth (highest association for profile 2 (aIRR = 10.9, 95%CI[6.11–19.99]) and medical abortion (highest association for profile 2 (aIRR = 12.65, 95%CI[5.96–28.49]). Conclusions This study unveiled 5 clinically relevant social vulnerability profiles with different risk levels of inadequate PCU and poor pregnancy outcomes. A personalized patient management according to their profile could offer better pregnancy management and reduce adverse outcomes.