Depression and anxiety among high-risk obstetric inpatients

Abstract

Objective

To assess the following among women hospitalized antenatally due to high-risk pregnancies: (1) rates of depression symptoms and anxiety symptoms, (2) changes in depression symptoms and anxiety symptoms and, (3) rates of mental health treatment.

Methods

Sixty-two participants hospitalized for high-risk obstetrical complications completed the Edinburgh Postnatal Depression Scale (EPDS), Generalized Anxiety Disorder 7-item scale (GAD-7) and Short-Form 12 weekly until delivery or discharge, and once postpartum.

Results

Average length of total hospital stay was 8.3±7.6days for women who completed an initial admission survey (n=62) and 16.3±8.9 (n=34), 25.4±10.2 (n=17) and 35±10.9 days (n=9) for those who completed 2, 3 and 4 surveys, respectively. EPDS was ≥10 in 27% (n=17) and GAD-7 was ≥10 in 13% (n=8) of participants at initial survey. Mean anxiety (4.2±6.5 vs. 5.2±5.1, p=.011) and depression (4.4±5.6 vs. 6.9±4.8, p=.011) scores were lower postpartum compared to initial survey. Past mental health diagnosis predicted depression symptoms [odds ratio (OR)=4.54; 95% confidence interval (CI) 1.91–7.17] and anxiety symptoms (OR=5.95; 95% CI 3.04–8.86) atinitial survey; however, 21% (n=10) with no diagnostic history had EPDS ≥10. Five percent (n=3) received mental health treatment during pregnancy.

Conclusion

Hospitalized high-risk obstetrical patients may commonly experience depression symptoms and/or anxiety symptoms and not receive treatment. A history of mental health treatment or diagnosis was associated with depression symptoms or anxiety symptoms in pregnancy. Of women with an EPDS ≥10, >50% did not report a past mental health diagnosis.

1. Introduction

Antenatal depression and anxiety occur in approximately 13% and up to 21.7% of women, respectively [1]. Rates of antenatal depression among women hospitalized for obstetrical risk can be as high as 19% [2]. Depression and/or anxiety during pregnancy have been associated with poor maternal health behaviors, including tobacco use [3] and poor maternal weight gain [4], and adverse birth outcomes, including preterm labor and preterm delivery [5,6]. Anxiety and/or depression during pregnancy may also adversely affect infant [7] and child development [8,9].

High-risk pregnancies can exacerbate depression and anxiety [10], and hospitalization can further increase the stress of a high-risk pregnancy [11]. Women hospitalized for high-risk pregnancies may therefore be at increased risk of depression and the subsequent adverse neonatal outcomes [2,12,13]. Although women may have access to psychiatric professionals in the hospital setting, psychiatric consultation referral rates in inpatient obstetric settings can be as low as 0.3% [14].

To date, no study in the United States has examined depression, anxiety, quality of life and rates of mental health treatment over the course of hospitalization among women admitted due to high-risk pregnancy. In order to better understand the impact of obstetric hospitalization on women's mental health, we assessed the following among women admitted antenatally for high-risk pregnancies: (1) rates of depression symptoms and anxiety symptoms, (2) change in depression symptoms and anxiety symptoms and quality of life throughout hospitalization and, (3) rates of mental health treatment.

2. Materials and methods

Participants were recruited from the inpatient antenatal service at a large tertiary care facility in an academic medical center in Central Massachusetts. Women on the antenatal service are all admitted to private rooms in the maternity center. During the study period, routine mental health assessments were not being conducted on admission or during hospitalization; however, psychiatric consultation was available when requested by clinical care teams. Group meetings of antepartum patients were not conducted by the antenatal service. The maternity center has general visitation hours of 11 a.m. until 8 p.m.; however, antenatal patients have relatively unrestricted access to 1 or more visitors 24 h per day, in addition to telephonic and wireless internet access. The service has a wide catchment area, and many hospitalized patients do not live proximate to the hospital. Accommodations, including a sleeping area, are made for one support person to remain with admitted patients.

During the 17-month study period, 82 women were deemed eligible and invited to participate in the study. Several study staff were involved in inviting women to participate; three were research assistants and two were obstetrics and gynecology resident physicians. Eligibility criteria for the study were determined based on information obtained from the medical record. Participants were considered eligible if they were (1) pregnant, (2) over 15 years of age, (3) English speaking and (4) admitted to the antenatal service due to high-risk obstetrical complications between June 2011 and October 2012. Prisoners and women with a cognitive disorder that would preclude them from participating in the informed consent process or from completing questionnaires were excluded. This study was approved by the University of Massachusetts Medical School Institutional Review Board.

The Edinburgh Postnatal Depression Scale (EPDS) is a validated, self-administered 10-item screening questionnaire that is widely used to assess depression during pregnancy and the postpartum period [15]. The intensity of depression is rated for the preceding 7 days by answering 10 multiple-choice items. Each item is scored on a 4-point scale for a total score range of 0–30, with higher scores reflecting a greater severity of symptoms [15]. The minimum cutoff used to indicate possible depression ranges from 9 to 13 in relevant literature. l. [15] reported a sensitivity of 86% and a specificity of 78% using a cutoff of ≥9 for “possible” depression and ≥12 for “probable” depression. In a systematic review of 37 validation studies of the EPDS for postpartum depression (PPD), Gibson et al. [16] noted that a cutoff score of 9/10 provides “strong evidence” for ruling out minor and/or major depression and that a cutoff of 12/13 provides “convincing diagnostic evidence” for minor/major/depression because the accuracy of diagnosing PPD increases. Gibson and colleagues recommended that investigators/clinicians consider whether it is more important to rule in cases of PPD or rule out cases that do not have PPD to determine the most useful cutoff point. Consistent with many studies in the peer reviewed literature [17–23], we used a cutoff of 10 to ensure that we captured most or all women with scores requiring further assessment for depression.

The Generalized Anxiety Disorder-7 (GAD-7) is a self-administered seven-item tool that is used to screen for and assess the severity of generalized anxiety disorder (GAD) [24,25]. GAD is evaluated for the preceding 2 weeks by scoring seven questions on a 4-point scale for a total score range of 0–21, with higher scores reflecting a greater severity of symptoms. A GAD-7 score > 10 is recommended as a cutoff point for identifying GAD. Scores of 5, 10 and 15 have been interpreted as indicative of mild, moderate and severe anxiety, respectively [24,25].

The 12-Item Short-Form Health Survey (SF-12), a shorter version of the Medical Outcomes Study 36-Item Short-Form Health Survey, is a set of self-administered quality-of-life measures. The 12 items provide a sampling of eight health concepts and their various operational definitions by evaluating how participants feel, what they are able to do and how they evaluate their health. Physical health is predicted by two items or scales: Physical Functioning and Role Physical. Mental health is also predicted by two items or scales: Role Emotional and Mental Health Scales. The Physical Component Summary (PCS) detects differences based on physical criteria (such as the severity of diabetes). The Mental Component Summary (MCS) detects differences using mental criteria (such as impact of depression or anxiety symptoms) [26]. The PCS and MCS scores have a range of 0 to 100 witha mean±standard deviation (S.D.) of 50±10. Scores above and below 50 are above and below the mean, respectively, based on the 1998 general US population [27]. Higher scores indicate better physical and mental health. Scores greater than 50 represent above average health status. Conversely, scores of 40 represent functioning at a level lower than 84% of the population (1 S.D.), and scores less than 30 represent functioning at a level lower than approximately 98% of the population (two S.D.s) [27].

We also ascertained basic demographics, psychiatric history, current and past mental health treatment history, and interest in a support group or other activities during hospitalization via a 21-item initial questionnaire developed by study staff. The questionnaire asked about current and past mental health treatment via medication treatment with a mental health provider, medication treatment with a non-mental health care provider, psychotherapy, counseling, case management, utilization of community supports such as support groups or church groups, and/or involvement with Department of Children and Families. We also reviewed the medical record to capture past mental health history and current or past mental health treatment that participants may not have disclosed on the questionnaire. The latter half of the questionnaire (five questions), along with the EPDS, GAD-7 and SF-12, was repeated weekly for participants with extended hospital stays in order to track changes in these measures throughout hospital stay.

After informed consent was obtained, participants completed the aforementioned surveys. Baseline and weekly surveys were administered and completed by subjects. Initial surveys were administered an average of 7 days after admission (median 4 days). Women who were hospitalized for 1 week or less only completed surveys for baseline measures. Those participants who remained in the hospital for > 1week were asked to repeat the standardized questionnaires and follow-up questionnaire on a weekly basis. Women who delivered during the hospitalization in which they signed the study informed consent were asked to repeat the questionnaires postpartum. When possible, participants who were discharged and readmitted for delivery were asked to complete the questionnaires postpartum. Women who were readmitted during their pregnancy were invited to continue to participate in the study and complete the surveys weekly. Data obtained during readmissions were entered as subsequent surveys.

If participants had EPDS ≥12 and/or GAD-7 ≥10, they were contacted by the principal investigator (N.B.) or her delegate (another psychiatrist) to discuss the results of the screening and to offer the participant mental health services if not already connected to mental health treatment. Participants who remained in the hospital for a prolonged time period were offered psychiatric consultation while in the hospital. If a participant scored positive on the suicide screening question (#10) of the EPDS, the principal investigator (N.B.) or her delegate (another psychiatrist) was contacted to further assess the subject's safety. If it was determined that there was a risk of danger to self or others, the patient was placed on 1:1 observation for safety, and a psychiatric consultation was obtained.

Descriptive statistics are displayed as mean±S.D. for continuous variables and count (percentage) for categorical variables. Mixed-effects regression models were fitted to the repeated longitudinal survey assessments. In these longitudinal analyses, patients were modeled as random effects, and time was coded as the number of weeks of hospitalization. Due to decreasing sample sizes with length of hospital admission, we did not include surveys past week 5 in the analyses. Changes in ratings of mood, anxiety and quality of life were modeled over two time periods: first, we examined available rating values obtained only antenatally, and then we examined rating values obtained both antenatally and during the postpartum assessment. To identify individual patient characteristics that were associated with depression symptoms or anxiety symptoms upon admission, we fit a series of models (one for each clinical and demographic variable of interest) using linear regression for mean rating scores (EPDS and GAD-7 total scores) and logistic regression for the risk of clinically elevated scores (EPDS ≥10 and GAD-7 ≥10). These models were not fit with any covariates and used only the data collected at the baseline admission visit (results of all models are available upon request). Statistical significance was determined at p≤.05.

3. Results

After reviewing the medical charts of all women admitted to the high-risk obstetrics service during the 17-month study period, 82 women were deemed eligible and invited to participate in the study. Sixty-three women consented to participate in the study and 62 completed all the questionnaires throughout their hospital stay. The overall study participation rate was 76%. We did not observe any trends between patients who chose to participate and those who did not. Sixty-two women completed the initial survey. The number of women who completed subsequent surveys decreased with time due to participants being discharged from the hospital (see Table 2). In addition, two women completed six surveys, and one woman completed seven surveys. Each survey time point was separated by a 1-week period. One subject was discharged and readmitted during her participation in the study. The length of hospital stay ranged from 1 to 99 days. The decrease in sample size over the hospital stay was due to participants' being discharged from hospital (either delivered or undelivered), not loss to follow-up. The attrition rate was 1.6% as only one participant was lost to follow-up.

Table 2.

Change in survey ratings over time.^a

	Admission
	Survey 1 (n=62), mean±S.D.	Survey 2 (n=34), mean±S.D.	Survey 3 (n=17), mean±S.D.	Survey 4 (n=9), mean±S.D.	Postpartum (n=10), mean±S.D.	w/PPb, p-value
Length of stay	8.3±7.6	16.3±8.9	24.8±10.1	33.8±11.1	28.1±17.9
EPDS	6.9±4.8	6.1±5.4	4.8±4.4	4.0±4.0	4.4±5.6	.011
GAD-7	5.2±5.1	4.2±4.8	3.0±3.6	1.9±2.7	4.2±6.5	.011
SF-12 — PCS	26.7±8.5	23.6±7.4	22.3±4.9	23.2±5.3	25.6±6.8	.175
SF-12 — MCS	48.2±6.4	45.9±7.9	47.1±6.9	43.8±6.9	47.1±6.3	.281

^aSample size decrease was due to participants being discharged from hospital, not attrition.

^bp-Values from mixed regression model trend tests for the impact of time first including the postpartum measure (w/PP). All available surveys from every time point were included.

The demographic and clinical composition of the included analytic sample (n=62) is described in Table 1. Only one participant, who was 17 years old, was < 18 years of age. Participants were of a diverse socioeconomic background; however, a majority was white or married. The mean gestational age on admission was 27.7±4.1. Twenty-four participants (39%) were still pregnant when they were discharged from the hospital, and 38 participants (61%) were postpartum on discharge. All participants had live births. Fourteen participants (23%) reported having been previously diagnosed with a mental health condition by a mental health provider. Seventeen participants (27%) reported having received prior mental health treatment, but only three participants (5%) were receiving mental health treatment during pregnancy or at the time of the initial or survey.

Table 1.

Demographics.

Characteristic	N (%)
Age, mean±S.D.	30.3±6.6
Parity, mean±S.D.	1±1
Gravidity, mean±S.D.	2.4±1.5 (0-9)
Race, n (%)
White	41 (66)
African American	10 (16)
Asian	1 (2)
Other	10 (16)
Ethnicity, n (%)
Latino	11 (18)
Not Latino	60 (82)
Education level, n (%)
HS or less	19 (30)
College (2 or 4 years)	27 (42)
Graduate study	10 (16)
Income level, n (%)
<40,000	21 (38)
40,000–80,000	15 (23)
>80,000	18 (33)
Relationship status, n (%)
Single	16 (25)
Married	39 (63)
Engaged	7 (11)
Prior admissions, n (%)	8 (13)
Obstetrical diagnosis, n (%)
Premature dilation, incompetent cervix	22 (35)
Preterm labor/contractions	7 (11)
Premature rupture of membranes	15 (24)
Vaginal bleeding/placenta previa	7 (11)
Preeclampsia	4 (6)
Mental health history, n (%)
Positive diagnostic history	14 (27)
Prior mental health treatment	17 (27)
Mental health treatment in pregnancy	3 (5)
Gestational age, mean±S.D.
At admission	27.7±4.1
At delivery	33.9±4.2
Days between admission and delivery	42.6±38.4

We found a statistically significant decrease in EPDS and GAD-7 survey scores throughout the course of the hospitalization whether considering time up to delivery or including the postpartum survey (Table 2). There were no statistically significant changes in the SF-12 PCS and MCS throughout the hospitalization, whether considering time up to delivery or including the postpartum survey. The mean PCS scores were very low among all participants throughout the hospital stay; they were more than 2 S.D.s below the mean reported for a normal nonobstetric population [27].

We next did a bivariate analysis to evaluate all the clinical and demographic patient characteristics listed in Table 1 as potential predictors of elevated symptom ratings at admission. The two statistically significant (p< .05) predictors are listed in Table 3. The mean EPDS scores were higher in women with a history of mental health diagnosis. The mean GAD-7 scores were higher in women with either a history of mental health treatment or diagnosis. Change in EPDS score over time was not impacted by a history of mental health treatment (p=.6) or diagnosis (p=.6) prior to delivery. Similarly, GAD-7 ratings over time were not associated with history of mental health treatment (p=.6) or diagnosis (p=.7) prior to delivery.

Table 3.

Variables associated with ratings of anxiety and depression symptoms upon admission.^a

	n	EPDS		GAD-7
		Mean±S.D.	p	Mean±S.D.	p
Obstetrical diagnosis
Premature dilation, incompetent cervix	20	7.2±5.5	.68	5.1±5.9	.59
Preterm labor/contractions	7	8.6±6.2		7.0±7.6
Premature rupture of membranes	15	7.1±4.9		5.7±5.2
Vaginal bleeding/placenta previa	7	7.0±3.3		5.9±2.9
Other	12	5.3±3.4		3.1±2.6
History of mental health treatment
Yes	17	8.6±4.0	.09	8.5±6.1	.001
No	45	6.3±5.0		3.9±3.9
History mental health diagnosis
Yes	14	9.9±4.5	.001	9.5±5.9	.0001
No	38	5.3±4.1		3.6±4.1

^aUnivariate results from linear regression models (one for each clinical and demographic variable of interest listed in Table 1) of EPDS and GAD-7 summary scores.

Based on admission surveys, 27% (n=18) of participants had symptoms of depression (EPDS≥10) and 13% (n=8) had symptoms of anxiety (GAD-7≥10). No participants scored positive on the EPDS suicide screening question (question #10). History of mental health diagnosis predicted depression symptoms [odds ratio (OR)=4.54; 95% confidence interval (CI) 1.91–7.17] and anxiety symptoms (OR= 5.95; 95% CI 3.04–8.86) at initial survey. Women who reported a history of mental health treatment were more likely to have scored ≥10 on the EPDS at the time of the initial survey compared to women who did not report a history of mental health treatment [47% (n=8) vs. 20% (n=9), p=.04]. Women who reported having received a mental health diagnosis were also more likely to have scored ≥10 on the EPDS at the time of the initial admission survey compared with those who did not report a mental health diagnosis [50% (n=7) vs.16% (n=6), p=.01]. Fourteen women (23%) reported having received a mental health diagnosis from a mental health provider, and 48 (77%) reported they had not received a mental health diagnosis. Of the 48 participants with no diagnostic history, 21% (n=10) had an EPDS ≥10. Therefore, more than half of the women with an EPDS ≥10 reported that they did not have a history of a mental health diagnosis. Similarly, women with a history of mental health treatment or history of mental health diagnosis were more likely than those without a mental health diagnosis to have scored ≥10 on the GAD-7 at initial admission survey [35% (n=6) vs. 5% (n=2) p=.002].

Three participants (5%) received mental health treatment during their pregnancy; this includes outpatient treatment received before their obstetric hospitalization. Seventy-seven percent of women (n= 47) reported that they would or maybe would benefit from a supportive psychotherapy group during their hospitalization. Fifty-seven percent (n=8) of participants with a history of a mental health diagnosis anticipated that they would benefit from a supportive psychotherapy group, compared with 21% (n=8) of participants without a prior mental health diagnosis (p=.03).

4. Discussion

Despite the negative impact of depression and anxiety on maternal and infant outcomes, there is a dearth of studies examining rates of mental health treatment and symptoms of depression, anxiety and quality of life over time in high-risk obstetrical inpatients. Our study suggests that high-risk obstetrical inpatients commonly experience depression (26.9%) symptoms and anxiety (12.6%) symptoms, yet few are receiving mental health assessment and/or treatment (5%).

We found rates of depression symptoms and anxiety symptoms that were higher than previously reported in both nonhospitalized [28–30] and hospitalized [12] pregnant women, and lower than rates previously reported by Brandon et al. [2]. Our finding that depression symptoms and anxiety symptoms decreased throughout the course of the hospitalization is consistent with previous studies [13]. Prior authors have postulated that this is due to the relief of stress and anxiety secondary to the birth of a live infant [13].

Women with a history of mental health treatment or diagnosis were more likely than those without a mental health history to experience symptoms of depression and anxiety during pregnancy. However, women without a history of mental health treatment or diagnosis also commonly experienced depression symptoms and were more likely to experience depression symptoms than anxiety symptoms. Prior history of mental health diagnosis or treatment is a risk factor for antenatal depression [31]. Thorough assessment of past psychiatric history during obstetrical care could help identify women at higher risk for antenatal depression symptoms and anxiety symptoms. Our results also suggest, however, that a lack of past mental health treatment or diagnosis may be falsely reassuring in this study population because a surprisingly large percentage of women without a mental health history (21%) experienced depression symptoms and/or anxiety symptoms. It is possible that this finding is a reflection of the highly stressful and uncertain environment of pregnancy complications.

Despite the high rates of symptoms of anxiety symptoms and depression symptoms, as well as previous mental health diagnoses and treatment, a surprisingly low number of participants were in mental health treatment. One particularly surprising finding was that 23% of women reported having received a mental health diagnosis from a mental health provider and 27% of participants reported having received mental health treatment. This may be because women received a diagnosis from a non-mental health provider or perceived that they were treated without having received a diagnosis, or being educated about their diagnosis. Regardless, this represents a missed opportunity to further assess women in the obstetrical inpatient setting to determine whether treatment(s) may be indicated. Clinicians should be vigilant and/or aware of the possibility of depression and anxiety symptoms in obstetrically hospitalized women. Identification of women who may have depression or be at risk of developing depression could facilitate prevention, assessment and treatment. This should also be considered during hospital discharge planning and follow-up to ensure that women receive the mental health assessment and/or treatment they may need.

Depression has been noted to prolong length of hospital stay is nonobstetric [32,33] and obstetric populations [34]; however, it is unclear whether hospitalized women with high obstetric risk are more likely to experience depression [2,12]. Thus, there is a need for studies that examine change in depression symptoms over the course of hospitalization [13]. We did not find an increase in the severity of depression symptoms or anxiety symptoms with longer hospital stays, suggesting that further research is needed to clarify whether the duration of hospitalization increases psychiatric symptoms in high-risk obstetric populations.

Obstetric inpatients may experience a lower quality of life than those who receive care in outpatient obstetrics settings [35]. We did not find any significant effect of length of hospital stay on quality of life as measured by the SF-12. One possible explanation is that hospitalization itself, rather than length of hospitalization, impacts quality of life. We found that the mean PCS scores throughout the hospital stay were remarkably low among all participants; however, the mean MCS scores were near the normal range. PCS scores may have been low because participants were experiencing high-risk pregnancies requiring hospitalization and likely bed rest.

The majority of inpatient high-risk obstetric patients (77%) reported that they may benefit from a supportive psychotherapy group. Group-based psychotherapy interventions such as supportive psychotherapy have substantially less evidence to support their efficacy in pregnant women than individual psychotherapy interventions, such as interpersonal psychotherapy or cognitive-based therapy [36]. However, in inpatient obstetric settings, where access to individual psychotherapy is often extremely limited, a weekly group-based supportive psychotherapy may be more feasible. Supportive psychotherapy groups may also provide an additional support option, particularly for women who may not have an EPDS ≥10, or who may not meet the criteria for depression upon further assessment, yet may still benefit from mental health support. Depressive symptoms can improve over the course of the pregnancy and in the postpartum period when women receive treatment [37,38]. Further research is needed in this specific setting and, in general, to determine whether interventions such as supportive psychotherapy could effectively prevent progression to depression among at-risk pregnant women with subsyndromal depression symptoms [38,39].

While the small sample size limits the generalizability of our data, the relatively high participation rate (76%) increases the confidence with which we can generalize our results, particularly to the obstetric inpatient population at our institution. However, we have limited data on trends over longer periods of hospitalization due to the small sample size and lengths of the participants' hospitalization. Although we identified symptoms of depression and anxiety, we did not conduct a Structured Clinical Interview for DSM Disorders or Mini International Neuropsychiatric Interview to determine whether women met the diagnostic criteria for an affective or anxiety disorder. The GAD-7 is also used to evaluate symptoms of anxiety/GAD over the preceding 2 weeks, which may have impacted our weekly assessments and thus findings. Because we used a low EPDS threshold, women with depressive symptoms may have had subsyndromal depression symptoms and anxiety symptoms. While EPDS scores ≥10 indicate that further assessment is warranted, it is important to note that the EPDS is not a diagnostic tool; thus, we are not able to determine whether participants with an EPDS ≥10 needed treatment, without a mental health assessment. In addition, we had a large proportion of white and married women which may not be representative of the general population in the city where our study took place. Further research is needed to determine (1) the factors that place obstetrically hospitalized women at risk for unrecognized depression, (2) the natural course of subsyndromal depression in hospitalized pregnant women and (3) the optimal approach for detecting and addressing both subsyndromal and major depression among hospitalized and newly discharged obstetric patients.

Antenatal anxiety and depression increase the risk of PPD [40] and negatively impact infant [7–9,41–43] and child development [44]. Therefore, it is critical to detect, assess and address depression and anxiety as part of obstetric care. While thorough assessment of past psychiatric history could help identify women at higher risk for antenatal depression and anxiety, a substantial proportion of women with elevated EPDS scores did not have such a history. Most women, regardless of whether they had an EPDS or GAD-7 ≥10, or a mental health history, had not participated in a mental health assessment and were not in mental health treatment. This suggests that clinicians should have a low threshold to suspect and screen for depression and anxiety symptoms. Future interventions and work could focus on leveraging the inpatient obstetric setting, which inherently involves frequent patient–provider interactions, as an opportunity to detect, assess and address women's mental health concerns.