Pregnant Black Women and Emergency Department Utilization: Assessing Self-Reported Receipt of Prenatal Counseling

Abstract

There is limited literature on emergency department (ED) use among pregnant women. In this manuscript, we examined the associations between prenatal counseling with the use of the ED during pregnancy. In our cohort of Black women in the Metro Detroit area, we found that approximately 70.5% of the women had an ED visit at some point during pregnancy. In unadjusted models of prevalence ratios (PRs), we found women reporting receipt of prenatal counseling regarding fetal movement, what to do about baby’s movement slowing down, and smoking (but not what to do about smoking) were at statistically significantly greater risk of ED utilization during pregnancy. Adjustment for confounders slightly weakened the associations for counseling about baby’s movement or smoking, so that the associations were no longer statistically significant. These findings call for further research on ED utilization among this population, especially differentiating urgent vs. non-urgent use of the ED during pregnancy.

There is a high rate of emergency department (ED) use in the US. The CDC estimated 130 million ED visits in the US in 2018 (Centers for Disease Control [CDC], 2018). Reports on the utilization of the ED for care among pregnant women are limited to date. Data collected in the U.S. on ED visits in the National Hospital Ambulatory Medical Care Survey from December 2016 – December 2017 showed that problems related to pregnancy were among the top 10 reasons for ED visits in women 15–65 years old, which includes visits for non-urgent indications (Rui & Kang, 2017). A non-urgent ED visit is defined as a visit for a condition for which delaying care by several hours would not increase the likelihood of an adverse outcome (Uscher-Pines et al., 2013). Results from a cross-sectional study of women delivering over a 2-month period at a high-volume maternity hospital in Rhode Island showed that 84% of pregnant women received care in the ED during their pregnancy (Kilfoyle et al., 2017).

Outside of pregnancy, reliance on the ED for non-urgent care has long been a concern of providers and policy makers in terms of health outcomes as well as health care spending (Uscher-Pines et al., 2013). Overall, almost 37% of adult ED visits are considered non-urgent (Uscher-Pines et al., 2013). In one of the few published studies in pregnant women, the ratio of non-urgent ED visits appeared similar to the ratio in adult ED visits generally: slightly more than one-third (35.6%) of pregnant women had at least one visit to the ED that was nonurgent (Kilfoyle et al., 2017). Outside the U.S., a German study found 56.3% of pregnant patients at a single institution utilized emergency department services for non-urgent reasons (Thangarajah et al., 2018). Non-urgent care may be better provided in an outpatient setting by an established obstetric provider (Kilfoyle et al., 2017). Continuity of care, an important element of prenatal care as it is with primary care, is disrupted by ED visits. Furthermore, in contrast with ED providers, prenatal care providers have specialized professional training in caring for pregnant and postpartum women.

Risk factors for ED use, especially for non-urgent ED use, have also been little examined in pregnant women in the published literature. Theoretically, pregnant women should have few financial barriers to accessing regular and timely prenatal care for non-urgent problems outside of an ED setting. However, despite decades-long expansions of Medicaid to pregnant women, utilization of timely prenatal care continues to be a problem, especially for minority women (Gennaro et al., 2016). A cross-sectional secondary analysis done at two Midwestern hospitals reported a high rate of delayed initiation of prenatal care in their sample, even though their population had access to federally funded insurance (Malik et al., 2017). Prenatal care is a possible area where efforts to ensure equitable and just access could impact ED utilization among pregnant women.

Prenatal care was originally developed as a surveillance system for pregnant women to improve pregnancy and neonatal outcomes (Carter et al., 2016). The American Congress of Obstetrics and Gynecology (ACOG) outlines an evidence-based prenatal schedule, which is a series of one-on-one visits with a patient’s obstetric care provider intended to prevent poor perinatal outcomes and provide education to women throughout their pregnancy (ACOG, 2018). The American College of Nurse-Midwives (ACNM) and the Association of Women’s Health, Obstetrics and Neonatal Nurses (AWHONN) similarly follow these ACOG guidelines. The study in Rhode Island explored prenatal care experiences of the women. Despite most of the women reporting receiving verbal and written information, only a small proportion (3.9%) could accurately identify signs of labor or pregnancy complications. While a small group, women who had at least one non-urgent ED visit were more likely to have correct knowledge compared to women without any non-urgent ED visits (9.6 versus 0.7%, p = 0.001) (Kilfoyle et al., 2017). This study suggests that providing education to mothers may increase ED care seeking and that interventions may need to go beyond basic education and consider the contexts in which the information is relayed, as this could relate to the clinical environment in which the information is presented and not the mother herself.

More needs to be known about utilization of the ED among pregnant women, including the reasons for the visits and factors associated with the use of the ED in pregnancy. If rates of ED use are as high as they appear, identifying risk factors for ED visits is a necessary first step to develop strategies to decrease such care seeking and provide the needed appropriate support to pregnant women. Decreasing ED use during pregnancy is important because excess utilization can lead to unnecessary diagnostic and treatment services, resulting in excess cost (Magriples et al., 2008). In this manuscript, we estimated the rate of overall ED use in pregnancy by conducting a secondary analysis within a cohort of births to Black women recruited for a preterm birth study. We were interested to examine the associations between prenatal counseling with the use of ED during pregnancy among a sample of Black women from Detroit and the surrounding suburbs, as counseling during prenatal care visits can be an important tool to prevent adverse maternal and infant outcomes in pregnant women (Alibekova et al., 2013). Some of the topics are particularly salient (e.g., baby’s movement slowing down). Furthermore, the limited literature suggests that prenatal care content is both modifiable and may relate to care utilization. This information could be used as a first step to analyzing the impact of pregnancy education on ED care seeking.

Methods

Design

Data were derived from the Life-course Influences on Fetal Environments (LIFE) study, a retrospective cohort study of 1383 African American mothers in the Detroit metropolitan area. Inclusion criteria were as follows: self-identification of Black or African American, 18–45 years old, and new birth of a singleton baby. Women were excluded if they did not understand English, had serious cognitive deficits or significant mental illness, or were currently incarcerated. No women were noted to be ineligible due to mental illness.

Participants

Women were recruited for the study from a Detroit suburban hospital from June 2009 through December 2011 and were approached approximately 48 hours after giving birth.

Data collection

An interview was conducted in private, during which subjects were asked about experiences during pregnancy as well as lifetime exposures and past experiences. A total of 1410 women were consented to participate, which represented 71% of all eligible women approached for the study. Data for analysis were derived from in person interviews and medical record abstractions. The interview provided data on sociodemographic factors, prenatal care counseling, and health behaviors. The medical record was abstracted to determine whether patients had major pre-existing conditions (asthma, hypertension, diabetes, heart disease, thyroid disease). Adequacy of utilization of prenatal care was computed based on the Kotelchuck index using data from the medical record abstractions (Kotelchuck, 1994). Adequacy was classified by the number of actual prenatal visits compared to the expected number of visits in the time period between initiation of prenatal care and delivery. The expected number of visits is adjusted for gestational age when care began and the gestational age upon delivery (Kotelchuck, 1994). Adequate care is defined as a ratio of actual to expected prenatal care visits equal to 80–109%.

ED care was determined from the medical record abstraction. Any ED visits were compared to no ED visits during pregnancy. Available medical record data were insufficient to determine the reason for each visit, as well as to differentiate urgent from non-urgent ED care.

Measures

The primary predictors of interest in this analysis related to prenatal care content as reported by the mother. Women were asked about prenatal counseling in several domains, specifically what patients reported being told about what might happen and what they should do if a problem arose. This was measured by asking women if a specific topic was discussed (“Since you have been pregnant, has anyone at the prenatal clinic talked to about any of the following topics?”, and if so, if women were told what to do about the problem (“Did they tell you what to do about it?). These were presented as closed-ended questions (yes/no), and each woman interpreted for herself what the questions meant. Eight topics were included in the interview with two not included in analyses because of the very limited variability. Nearly all women reported being told about two topics and what to do (eating; sleeping) so these two topics were not analyzed as outcomes. The remaining six counseling domains analyzed included existing chronic health problems, early contractions/labor pain, the baby’s movement slowing down, smoking, alcohol, and prescription drug use. These topics are ones which providers typically have clear guidance about how to discuss actions with patients. For example, if baby’s movement slows down, women are typically told to seek immediate medical care. If women smoke, they are generally told to stop smoking or referred to smoking cessation programs.

Covariates were evaluated as potential confounders if they met the criteria for confounding (associated with both predictor and outcome and were not in the causal pathway). Inclusion of covariates that are not true confounders not only reduces power but can distort associations in either direction (Szklo & Nieto, 2006). Potential confounders included adequacy of prenatal care utilization, education as a proxy for low health literacy (Kutner et al., 2006), and smoking. Women who have inadequate care may be less likely to receive prenatal care counseling and more likely to visit the ED. Individuals with low health literacy utilize more ED care compared to those with higher health literacy (Schumacher et al., 2013). In pregnant women with low health literacy, even if the women received adequate counseling from their providers, they may not be able to understand how to apply the counseling they received. Finally, smoking can correlate with a woman’s general health behavior and receptiveness to counseling from their physicians. Smokers perceive less health risks of smoking compared to nonsmokers, implying that smokers may not be as receptive to health guidance by their providers if they do not think they are at a health risk (Upadhyay et al., 2019). A woman who is a smoker may not be as receptive to following the guidance of her provider, who has likely advised her about smoking cessation in the past. This may be indicative of her receptiveness and awareness of prenatal counseling during pregnancy.

Statistical analysis

Odds ratios (OR) were computed to compare proportions of any ED visit by predictor with chi square tests of statistical significance in bivariate analysis. This was an exploratory step in the analysis; with a high prevalence outcome, such as ED visits here, ORs will overestimate the effect. For each prenatal care counseling domain, logistic regression was used to compute ORs and test for statistical significance after addition of potential confounders. When the OR was statistically significant (p<0.05), we computed the more conservative prevalence ratio. Confounders were included in the final models if the regression coefficient (for the prevalence ratio) changed more than 5 percent as evidence of confounding. Analysis was conducted with SPSS version 25 and R Studio (https://rdrr.io/cran/prLogistic/man/prLogisticDelta.html). The study was IRB approved at Wayne State University and Ascension Providence (clinical site).

Confounders

Potential confounders include the woman’s perception of receiving adequate prenatal care, education level and smoking status. If the woman determines that she is receiving adequate prenatal care, she is likely receiving prenatal counseling about what to do in certain emergencies compared to those who feel they are not receiving adequate prenatal care. Another confounding variable in this population is the education level of the participants. According to the NIH, those with low health literacy utilize more healthcare services, including the emergency department, compared to those with higher health literacy. A reason for low health literacy can be explained by limited educational opportunity, more specifically receiving a high school education or lower (Kutner et al., 2006). In pregnant women with low health literacy, even if the women received adequate counseling from their providers, they may not be able to understand how to apply the counseling they received. Additionally, smoking/nonsmoking status can correlate with a woman’s general health behavior and receptiveness to counseling from their physicians. A woman who is a smoker may not be as receptive to following the guidance of her provider, who has likely advised her about smoking cessation in the past. This may be indicative of her receptiveness to prenatal counseling during pregnancy (Francis et al., 2019).

Results

Approximately 70.5% (n= 994) of the women had an ED visit at some point during their pregnancy. Table 1 describes the characteristics of the cohort. Per eligibility, all women were Black. Nearly all women were insured but most were insured with Medicaid. Most (71.9%) had at least some college education. Nearly half of our sample received adequate or more prenatal care. Prenatal care counseling was reported in all domains, ranging from 51.5% in the least frequent domain of existing chronic health problems to 90.1% in the most frequently reported domain for discussion.

Table 1.

Participant Characteristics, LIFE Study 2009–2011, metro Detroit Michigan

Variable	% (N)
Race
Black or African American	100% (1410)
Age (years)
≤18	3.3 (47)
19–24	36.3 (511)
≥25	60.4 (852)
Marital Status
Single	45.9 (645)
Married or Cohabitating	54.1 (761)
Level of Education
High School Graduate or Less	16.5 (232)
Some College, Technical Training, or Associates Degree	72.3 (1020)
College Graduate	5.2 (73)
Graduate School	6.0 (85)
Family’s Current Financial Situation
Very poor, not enough to get by	1.4 (20)
Barely enough to get by	6.5 (92)
Enough to get by with no extras	50.5 (711)
More than enough or well to do	41.4 (583)
Insured	99.5 (1394)
Medicaid	57.8 (810)
1+ Prior Births	57.6 (812)
Employment
Part time (< 40 hours/week)	36.3 (306)
40 hours/week	46.2 (390)
>40 hours/week	17.5 (148)
Unemployed	40.1 (566)
Financial Assistance#	62.9 (887)
1 other source of support	27.9 (394)
2 or more other sources of support	35.1 (493)
History of Preterm Birth	11.1 (156)
Smoking
Prior to pregnancy	11.5 (162)
During pregnancy	5.7 (79)
History of Substance Use%	46.0 (648)
Utilization of healthcare^
Less than adequate prenatal care	51.2 (722)
Adequate or more prenatal care	48.8 (688)
ED Visit During Pregnancy	70.5 (994)

^#Other sources of income included supplemental social security income (SSI), cash welfare, unemployment, food stamps, alimony, women infant children nutrition program, other (self reported)

^%substance defined as marijuana, hashish, pot, grass, cocaine, crack, coke, heroin, methadone, amphetamines, LSD, others

^{^}using adequacy of prenatal care index which characterizes prenatal care utilization on the dimensions of adequacy of prenatal care initiation and adequacy of services receives (Kotelchuck, 1994)

In unadjusted analyses (Table 2), women reporting receipt of prenatal counseling regarding fetal movement slowing down (odds ratio (OR), 95% confidence interval (CI): 1.55; 1.08, 2.24, p<0.05) and smoking (OR, 95% CI: 1.42, 1.08, 1.86, p<0.05) showed significantly greater odds of ED utilization during pregnancy. For women who reported being told what to do regarding baby’s movement slowing down, there was also a larger and significant impact on ED use (OR, 95% CI: 2.23,1.24, 3.98, p<0.01). For those associations that were not significant predictors of ED usage, the magnitude of the estimated ORs was quite small (odds ratios less than 1.2 if increased risk, greater than 0.8 for decreased risk). For the statistically significant associations with ED use, we computed unadjusted and adjusted prevalence ratios (PRs) as ORs overestimate effects when the outcome is not rare (ED utilization ~30 percent is not rare). PRs indeed were smaller in magnitude than ORs for all topics. However, those variables representing topics for which the OR was statistically significant also had PRs that were significant in unadjusted analyses. None of the covariates examined as a potential confounder were determined to confound the association between discussing the baby’s movement slowing down and emergency department visits. Confounders were present for the associations between emergency department use and what to do about the baby movement slowing down (age, smoking in second half of pregnancy) and discussing smoking (age). Adjusted PRs and 95% CIs are included in Table 2. For “what to do about the baby movement slowing down” and “smoking,” adjustment for the confounders was deemed necessary (changing the coefficients by >5%). While the PR was little changed for either variable, the lower confidence interval no longer excluded 1 (=0.99) and the associations were no longer statistically significant.

Table 2:

Prenatal Counseling and ED visits during pregnancy

Prenatal Counseling†	≥ 1 emergency department visit during pregnancy†† % (N)	Unadjusted Odds Ratio (OR) 95% Confidence Interval (CI)	Unadjusted Prevalence Ratio (PR) 95% CI	Adjusted PR (95% CI)
Prenatal provider discussed....
Existing chronic health problems? (n=1397)
Yes (720)	72.1% (519)	1.15 (0.92, 1.10)	1.04 (0.97, 1.12)
No (677)	69.9% (468)
If discussed (n=711), Explained what to do?
Yes (555)	73.3% (407)	1.22 (0.83, 1.80)	1.06 (0.94, 1.19)
No (156)	69.2% (108)
Early contractions/labor pains? (n=1400)
Yes (1111)	71.2% (791)	1.11 (0.84, 1.48)	1.03 (0.95, 1.13)
No (289)	68.9% (199)
If discussed (1107), Explained what to do?
Yes (1055)	71.4% (760)	1.25 (0.65, 2.40)	1.07 (0.86, 1.33)
No (42)	66.7% (28)
Baby’s movement slowing down? (n=1400)
Yes (1261)	71.6% (903)	1.55 (1.08, 2.24)*	1.16 (1.01, 1.32)*	(no confounding)
No (139)	61.9% (86)
If discussed (n=1257) Explained what to do?
Yes (1209)	72.5% (876)	2.23 (1.24, 3.98)**	1.34 (1.03, 1.74)*	1.32 (0.99, 1.77)§
No (48)	54.2% (26)
Smoking? (n=1393)
Yes (1089)	72.3% (787)	1.42 (1.08, 1.86)*	1.12 (1.02, 1.22)*	1.10 (0.99, 1.21)§§
No (304)	64.8% (197)
If discussed Explained what to do? (n=1081)
Yes (664)	72.7% (483)	1.04 (0.79, 1.37)	1.01 (0.94, 1.09)
No (417)	71.9% (300)
Alcohol use? (n=1390)
Yes (1069)	71.9% (769)	1.30 (0.995, 1.70)	1.08 (0.99, 1.18)
No (321)	66.4% (213)
If discussed Explained what to do? (n=1064)
Yes (690)	72.8% (502)	1.08 (0.,82, 1.43)	1.02 (0.95, 1.11)
No (374)	71.1% (266)
Using prescription drugs? (n=1398)
Yes (1134)	71.3% (808)	1.14 (0.85, 1.52)	1.04 (0.95, 1.14)
No (264)	68.5% (181)
(If discussed) Explained what to do? (n=1129)
Yes (1076)	71.2% (766)	0.80 (0.42, 1.52)	0.94 (0.81, 1.10)
No (53)	75.5% (40)

^*p<0.05,

^**p<0.01

^†Interview data

^††Medical records data

^§Adjusted for maternal age and smoking in second half of pregnancy

^§§Adjusted for maternal age

Discussion

More than two thirds of the women in our study of Black women reported at least one visit to an ED during their pregnancy. Black women are a population of particular interest as they have the highest rates of adverse maternal and infant health outcomes in the U.S (Callaghan, 2012). Vladatiu et al found that high pregnancy-related emergency care utilization was associated with Black race (Vladutiu et al., 2019). The ED utilization in our study is comparable but slightly lower than the rate in the published study on ED use in pregnancy of primarily Hispanic and white women (maternity hospital in Rhode Island), and slightly higher than the rate in the published study on ED use in pregnant Medicaid recipients in North Carolina (57.5%) (Kilfoyle et al., 2017; Vladutiu et al, 2019).

In this study, we examined how prenatal counseling relates to the rate of use of the emergency department in Black pregnant women in Metro Detroit. While several domains of counseling were studied, the interview did not specifically capture counseling on the signs and symptoms that should warrant a visit to the ED during pregnancy. Prenatal counseling increased ED usage across all topics covered, though not all were statistically significant. Prenatal clinic providers discussing the topic of baby’s movement slowing down, discussing what to do if baby’s movement slows down, and discussing smoking all increased risk of ED use but associations for two of the topics were weakened by adjustment for confounders albeit remaining close to significant. Our results are consistent with those of the other few publications on ED visits in pregnancy. Kilfolye et al. reported that women who had one or more non-urgent ED visits were more likely to correctly know the signs compared to women who had no non-urgent ED visits. On the other hand, Magriples et al. reported that women with inadequate prenatal care use had a higher total number of additional unscheduled visits during pregnancy and were more likely to utilize the emergency department (Magriples et al., 2008). Our results and those of these two additional studies highlight prenatal counseling as an important area of focus to address appropriate ED usage in pregnancy.

Does counseling specifically direct patients to visit the ED? For example, when women notice baby’s movement slowing down or when they experience labor pains, are women who are counseled on these topics more likely to seek out emergency care? If the goal of prenatal counseling on these topics is to direct pregnant women to the ED, then comprehensive and effective prenatal counseling must be delivered to all women. Based on our data, this does not appear to be the case. In our study, though about half of women received adequate or more prenatal care, that did not translate to effective prenatal counseling, as many women answered in the negative as to whether their prenatal provider discussed various topics. For example, approximately half of women reported that the prenatal care provider did not discuss any existing chronic health problems. Patients’ uptake of information must also be considered. The patients might have some recollection that a topic was discussed and therefore answering affirmatively to the questions, but yet be unable to translate the education into practice. Standardized counseling from the perspective of the provider is of little benefit to patients if they do not recall receiving the information and do not understand the counseling they receive. The continued efforts over the past several decades in the U.S. to increase utilization of timely prenatal care may need to expand to promote high quality prenatal care, including effective prenatal counseling.

Our study was unable to differentiate between urgent and non-urgent ED visits. It is important to identify risk factors that contribute to non-urgent ED visits, as non-urgent care provided in the ED is usually not in the best interest of the patient and could be better provided in an outpatient setting by an established provider (Kilfoyle et al., 2017). If the goal of the counseling these women received is to decrease ED visits, counseling may need to be more comprehensive and nuanced. Accessibility to providers outside of daytime clinic hours may also be a factor. Regular and timely care for prenatal patients often occurs only during clinical hours. For patients in this study, we do not know if patients sought ED use during clinical or off-hours, and data was not collected regarding whether or not these patients were guided to seek emergency care by their providers or if they went without the knowledge of their prenatal care provider. In a study of pregnant women coming for non-urgent reasons to a specialized women’s emergency care facility, the majority of women had access to a usual source of medical care (84%) but they reported that the source of care was not always accessible or available to them (Matteson et al., 2008). This can be a challenge, because when women access emergency care outside of their provider’s network, the provider has difficulty obtaining information from these outside sites, relying on patients to provide the relevant and correct information. Neither our study nor the Kilfoyle publication addressed accessibility of the prenatal providers.

There are a few important limitations to consider in this study. We were unable to differentiate between urgent and non-urgent ED visits with the available medical data abstracted from this study. As this was a secondary analysis, we were unable to return to the records for a more detailed review outside the original study purpose and consent. Therefore, as discussed above, some of the visits might be absolutely warranted. As we could only identify few prior publications on ED use in pregnancy, our findings contribute to the field even with this limitation.

Other limitations relate to the prenatal counseling measurement. First, the specific content of the prenatal counseling is not known. Topics may have been mentioned briefly or without much detail. No information is available as to the goal of counseling. Perhaps increasing ED use was intentional, and this study’s results capture that prenatal counseling works as intended. A second related limitation is the dependence on self-report of the prenatal counseling topics. Recall bias is a systematic error. Asking women post birth what they were educated about in a prenatal visit is inherently challenging. If a patient forgets the discussion of these topics, if the provider used different verbiage, or if the woman did not fully comprehend the information received, she may report that the prenatal provider did not discuss these topics. However, asking women offers at least some perspective. Furthermore, even if they recall incorrectly, it is their recall on which they may have based their ED seeking behavior. One could argue that if the patient reports that the provider did not discuss the topic even if they did, the topic then effectively was “not” discussed.

However, in studies of congenital malformations, specifically designed to evaluate the possibility of recall bias related to reporting, little evidence has in fact been found of recall bias in mother’s exposure reports (Klemetti & Saxon, 1967; Mackenzie & Lippman, 1989; Zierler & Rothman, 1985). In one such study designed to evaluate the impact of recall bias, Khoury et al. found that use of affected controls did not change etiologic inferences derived from using normal controls. They conclude that “concerns about recall bias are overrated in birth defects studies and that the use of normal controls is acceptable unless evidence of substantial recall bias exists” (Khoury et al, 1994). Several other studies of pregnancy are consistent with this conclusion (Buka et al, 2000; Infante-Rivard & Jacques, 2000; Mackenzie & Lippman 1989). Therefore, we view recall bias as an unlikely explanation for the findings. Furthermore, only some of the topics were associated, whereas all or more of the topics would be expected to be associated with ED use if recall is differential.

Prenatal care is a potential area of focus to address factors that influence ED use in pregnancy. It presents an opportunity to educate women on appropriate and not appropriate use of the ED during pregnancy. As such, it is vitally important to define what is appropriate and what is not appropriate use. Counseling should provide effective education to women so they can determine when they should be seeking urgent care in the ED or visiting their outpatient provider for non-urgent care.

This study was a secondary analysis of an important group of interest, specifically pregnant black women. As such, it presents as an exploratory paper on a topic that has been neglected in research on prenatal care and birth outcomes.

We end this paper with a call to action for both clinicians and researchers. There is still much to be learned regarding who uses the ED during pregnancy, what conditions are associated with high rates of ED use during pregnancy, especially non-urgent visits, and finally the role of prenatal care providers to decrease inappropriate utilization.

For future direction on this topic, some points of interest include: updating and standardizing prenatal counseling for all pregnant women, determining whether decreasing ED usage is an appropriate goal for pregnant patients, directly counseling women about ED usage and how to avoid non-urgent use, determining pregnant women’s care seeking and if they are contacting their providers before seeking emergency care, and improving resources for pregnant women outside of the ED and after clinic hours, such as an emergency hotline to act as an intermediary in addressing patient concerns.