Social vulnerability and triage acuity among pregnant people seeking unscheduled hospital care

Introduction

Obstetric triage units, where pregnant patients are evaluated for both obstetric and non-obstetric complaints, function as a form of emergency department. At many institutions, the number of unscheduled hospital visits by pregnant persons exceeds the overall number of births by 20–50 % [1]. Unscheduled hospital visits, particularly for non-obstetric concerns, may reflect health disparities and inadequate access to prenatal care. There is scant literature on whether more socially vulnerable pregnant patients present to triage with higher acuity conditions.

Management algorithms are used by many emergency departments for risk stratification [2]. These triage tools, which are predictive of hospital admission and resource utilization [3], 4], were developed for units serving predominantly non-pregnant patients and have limited applicability in obstetric triage [5]. Standardized, evidence-based obstetric triage tools have been developed to ensure safety, quality, and efficiency. [5], [6], [7]. One such tool is the Maternal-Fetal Triage Index (MFTI), which has high inter-rater reliability, and is endorsed by the Association of Women’s Health, Obstetric and Neonatal Nurses (AWHONN) and the American College of Obstetricians and Gynecologists (ACOG) [8], 9]. Acuity is classified using a 5-level system ranging from priority 1 (highest acuity) to priority 5 (lowest acuity or scheduled services).

Many factors influence whether patients will seek care in emergency departments instead of outpatient clinical settings [10], 11]. Persons belonging to socially marginalized groups, especially those residing in economically disadvantaged areas, are more likely to have missed or delayed appointments, which contribute to poor health and a disproportionate burden of illness [12], 13]. This may, in part, be attributable to reduced access to reliable and affordable transportation, limited availability of nearby healthcare providers, lack of flexible appointment times, and challenges with navigating the healthcare system. The social vulnerability index (SVI) quantifies the effect of non-medical factors in the local environment that influence health outcomes. The Centers for Disease Control and Prevention (CDC) released the SVI tool, which incorporates data from U.S. Census Bureau on socioeconomic status, household composition, minority status, and housing/transportation. Higher SVI scores indicate communities with increased vulnerability.

For this study, our objective was to determine how individual and neighborhood-level social vulnerability are associated with triage acuity among pregnant people seeking unscheduled hospital care. Specifically, we determined how maternal sociodemographic characteristics, and census tract-linked SVI are associated with MFTI score, while adjusting for medical comorbidities. We hypothesized that there would be a positive association between SVI and MFTI acuity score. We also identified factors associated with three or more triage visits during pregnancy.

Materials and methods

Results

Patient and triage visit characteristics

A total of 11,388 pregnant patients were included for analysis (Figure 1). Non-Hispanic White patients constituted the largest race and ethnicity group (30.7 %), followed by Non-Hispanic Black (22.1 %), Asian or Pacific Islander (20.6 %), and Hispanic patients (13.3 %). Most patients had private health insurance (60.7 %) and about half were nulliparous (47.9 %). Most triage visits were classified as prompt or non-urgent (61.5 %, n=7,007), while 35.1 % (n=3,996) were classified as urgent, and 3.4 % (n=385) as stat. Patient and triage visit characteristics are summarized in Table 2. Patients in the stat (MFTI-1) group were more likely to be of advanced maternal age, have higher BMI, and be multiparous. Additionally, patients with public insurance and hypertension were more frequently classified as urgent (MFTI-2) or stat (MFTI-1). These differences suggest that certain sociodemographic and clinical characteristics may be associated with higher triage acuity.

Figure 1:

Study flowchart.

Table 2:

Patient demographics and clinical characteristics by acuity at first triage visit.

Characteristic	MFTI-3-4 (prompt/non-urgent) (n=7,007)	MFTI-2 (urgent) (n=3,996)	MFTI-1 (stat) (n=385)	p-Value
Social vulnerability index (SVI)
Very low	1,561 (22.3)	811 (20.3)	89 (23.1)	0.04
Low	1,298 (18.5)	763 (19.1)	70 (18.2)
Moderate	1,500 (21.4)	825 (20.6)	92 (23.9)
High	1,702 (24.3)	986 (24.7)	92 (23.9)
Very high	946 (13.5)	611 (15.3)	42 (10.9)
Race and ethnicity
Non-Hispanic White	2,191 (31.3)	1,188 (29.7)	119 (30.9)	0.40
Non-Hispanic Black	1,525 (21.8)	916 (22.9)	73 (19.0)
Hispanic	912 (13.0)	548 (13.7)	54 (14.0)
Asian or Pacific Islander	1,467 (20.9)	799 (20.0)	84 (21.8)
Other or multiracial	677 (9.7)	389 (9.7)	42 (10.9)
Declined or unknown	235 (3.4)	156 (3.9)	13 (3.4)
Public health insurance	2,692 (38.6)	1,611 (40.5)	168 (43.6)	0.03
Advanced maternal age	1,646 (23.5)	1,068 (26.7)	104 (27.0)	0.001
Nulliparity	3,548 (50.7)	1,740 (43.6)	171 (44.5)	<0.001
Body mass index (BMI), kg/m2
Normal, 18.5–24.9	2,104 (30.0)	1,186 (29.7)	64 (16.6)	<0.001
Overweight, 25.0–29.9	1,856 (26.5)	1,065 (26.7)	102 (26.5)
Obese, ≥30.0	2,287 (32.6)	1,335 (33.4)	179 (46.5)
Declined or unknown	760 (10.8)	410 (10.3)	40 (10.4)
Diabetes	89 (1.3)	70 (1.8)	9 (2.4)	0.05
Hypertension	425 (6.1)	299 (7.5)	29 (7.5)	<0.001
Gestational age at initial triage visit, weeks
<34 0/7	2,041 (29.2)	1,241 (31.1)	135 (35.2)	<0.001
34 0/7–36 6/7	776 (11.1)	509 (12.7)	59 (15.4)
≥37 0/7	4,180 (59.7)	2,243 (56.2)	190 (49.5)
COVID-19 pandemic era	3,111 (44.4)	1,665 (41.7)	312 (81.0)	<0.001

1. Data are number (percentage). Missing data: nulliparity (n=8), diabetes (n=581), hypertension (n=581), public health insurance (n=46), gestational age at initial triage visit (n=14). SVI: very low, 0.00–0.20; low, 0.21–0.40; moderate, 0.41–0.60; high, 0.61–0.80; very high, 0.81–1.00.

Discussion

In this retrospective cohort study of obstetric triage patients, we found that SVI score, which quantifies the social vulnerability of each census tract, was associated with an increased likelihood of urgent triage visits but not stat visits. There was no significant association between race and ethnicity and the likelihood of either urgent or stat triage visits. The only factors associated with higher odds of a stat visit were presentation during the COVID-19 pandemic era and a history of chronic hypertension.

Based on our review of the literature, few studies have specifically evaluated how sociodemographic and neighborhood characteristics are associated with the acuity of obstetric triage visits. Prior research has demonstrated that healthcare seeking behaviors in pregnancy are influenced by sociodemographic factors and local community characteristics [15], 16]. Predictors of non-urgent emergency department visits during pregnancy include speaking a language other than English and having public or no insurance [17], but predictors of urgent or stat visits in this population are less studied. Unscheduled hospital-based care by pregnant patients may be driven by unmet clinical and psychosocial needs [15]. However, higher utilization of obstetric triage (i.e. more frequent visits) is not correlated with higher acuity visits [18]. Our finding of an association between very high SVI scores and urgent triage visits indicates that triage acuity may be more affected by neighborhood characteristics than other sociodemographic characteristics, at least in the catchment area of the study hospital. In pregnancy, higher SVI scores have been associated with preterm birth [19] and delayed diagnosis of congenital heart defects [20]. Notably, after adjustment for confounding variables, SVI scores have not been associated with stillbirth [21] or severe maternal morbidity [22]. Individual patient characteristics are often more strongly linked to clinical outcomes than neighborhood factors [23]. However, in this study, aside from chronic hypertension, few examined variables were associated with increased triage acuity, with external factors such as high neighborhood SVI and the COVID-19 pandemic being the primary contributors.

Neither SVI nor other predictors were linked to three or more triage visits, while advanced maternal age, presentation during the COVID-19 pandemic, and later gestational age were associated with fewer visits. One potential explanation for the higher odds of multiple triage visits among patients presenting earlier in gestation is that they have more remaining time in pregnancy, naturally allowing for more opportunities to re-present. Alternatively, patients with early pregnancy complications may inherently require more frequent evaluation, or patients with limited access to routine prenatal care may rely more heavily on triage services throughout pregnancy.

During the COVID-19 pandemic, stat obstetric triage visits were more frequent, likely reflecting changes in care-seeking behavior and disruptions in routine prenatal care. Disruptions and delays in routine and nonemergency care were common [24], 25]. Efforts to avoid severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) exposure likely contributed to increased reliance on acute care. Hospital avoidance may have led patients to delay seeking care until their conditions became more severe. At the same time, reduced access to routine outpatient services and the direct effects of COVID-19 infection may have further increased the risk of pregnancy complications. Notably, the overall number of triage visits declined during this period, further supporting the role of hospital avoidance.

The association between high SVI and increased odds of urgent triage visits suggests that patients from more vulnerable communities may face greater barriers to timely prenatal care, potentially leading to more acute presentations. However, the lack of association between SVI and stat visits indicates that the most severe obstetric emergencies are likely driven by individual clinical factors rather than neighborhood-level vulnerabilities. These findings support the need for targeted interventions such as expanding community-based prenatal care access, improving transportation options for high-SVI neighborhoods, and strengthening outreach efforts to promote early recognition and management of pregnancy-related complications. Addressing these systemic barriers could improve maternal health outcomes and reduce disparities in obstetric care access.

Many pregnant patients use obstetric triage units as their primary or sole source of care. In a study by Knight et al., approximately one-third of the study population presented to obstetric triage without prior prenatal care [26]. Identifying high utilizers of obstetric triage and implementing targeted interventions to support their ongoing care could improve pregnancy outcomes. However, determining the most relevant socioeconomic factors influencing triage utilization is challenging and may vary based on local and regional contexts. Identifying communities with limited access to healthcare, transportation, and essential services such as testing and vaccination would allow for more effective resource allocation, ultimately improving maternal morbidity in high-risk populations.

Future research should explore the mechanisms underlying the association between neighborhood-level social vulnerability and urgent obstetric triage visits, including potential delays in prenatal care, transportation barriers, and differences in health-seeking behaviors among individuals in high-SVI areas. Additionally, further studies should assess interventions such as mobile prenatal care units, telemedicine expansion, or enhanced care coordination, can mitigate the increased need for urgent triage visits in vulnerable populations. Given the lack of association between SVI and stat triage visits, future research should focus on identifying individual clinical and behavioral factors that contribute to the most severe obstetric emergencies. Investigating how patient-level and neighborhood-level risk factors interact could provide a more comprehensive understanding of disparities in acute obstetric care utilization. Lastly, given the observed impact of the COVID-19 pandemic on triage acuity, future studies should examine how public health crises influence obstetric care-seeking behaviors and access to prenatal services. Understanding these patterns could help inform healthcare policies and preparedness strategies to ensure equitable access to obstetric care during future emergencies.

Our study has several strengths. The included patient population is diverse. Based on our review of the literature, no prior studies have evaluated the association between neighborhood social vulnerability and obstetric triage acuity. SVI can objectively quantify neighborhood social vulnerability which can then be correlated with acute care utilization. This approach is generalizable and relevant to diverse populations beyond our metropolitan area, making it applicable to other regions and healthcare settings. SVI was a publicly accessible tool until its removal from the CDC website in January 2025 following the presidential inauguration; the authors remain hopeful that it will be reinstated in the future.

This study has several limitations, including its retrospective nature and inclusion of data from only one obstetric triage unit. Patient-level data on income, occupation, employment status, and educational attainment were not consistently available and therefore could not be analyzed. We did not evaluate whether patients had prenatal care, which trimester any such care was established, the frequency of visits, and whether their healthcare provider was affiliated with our institution. We are also unable to evaluate if individuals in this cohort relocated during the study period resulting in disruptions in care.

We identified an association between neighborhood social vulnerability and obstetric triage visit acuity. Patients from areas with a very high SVI score had increased odds of an urgent triage visit but not a stat visit. No racial or ethnic disparities were observed, as urgent and stat visits occurred at similar rates across self-identified groups. The only factors associated with higher odds of a stat triage visit were presentation during the COVID-19 pandemic era and chronic hypertension. Future studies should investigate whether disparities in prenatal care access, transportation barriers, or delayed symptom recognition contribute to the increased odds of urgent obstetric triage visits in high-SVI areas and evaluate targeted community-based interventions to improve maternal health outcomes.