A systematic review of psychosocial screening tools for identification of postpartum mental health status beyond depressive symptoms

Abstract

Background:

Mental health is a key driver of maternal health, particularly in the postpartum period. Psychiatric screening may improve linkage to care and outcomes. Limited evidence exists beyond screening for depression.

Objective:

The goal of this systematic review was to characterize the effects of systematic psychiatric and psychosocial screening on the identification, referral for, and resolution of postpartum mental health symptoms beyond depressive symptoms.

Methods:

We identified studies that screened for psychiatric or psychosocial symptoms during pregnancy or postpartum and assessed at least one result of the screening in the postpartum year (e.g., diagnosis, linkage to care, resolution of symptoms). We searched six bibliographic databases on October 25, 2022 and again on December 15, 2023. Results were screened using Covidence. Two reviewers assessed each title/abstract and full text. We conducted a quality assessment using a modified version of the EPHPP (Effective Public Health Practice Quality Assessment Tool for Quantitative Studies).

Results:

Of 5,758 initially eligible studies, we selected 47 which fit our criteria. The most common symptom type was anxiety (25), followed by life course trauma (7), general psychosocial screening (4), and posttraumatic stress disorder (6). Studies primarily assessed diagnostic accuracy, followed by referral to care as outcomes. We deemed most studies of low quality (80.8%), primarily due to a lack of an unscreened control group for assessing screening effectiveness.

Conclusions:

Postpartum screening is feasible and accurate for many psychiatric symptoms beyond depression. It is unclear whether it is effective in linking to care or improving postpartum outcomes.

Introduction

Behavioral and substance use disorders are an increasingly significant contributor to maternal deaths in the United States (Joseph et al., 2021; Trost et al., 2021). Often undiagnosed and untreated, postpartum psychiatric disorders are common and result in a range of negative consequences for postpartum women, infants, and families (Chin et al., 2022; Johannsen et al., 2020; Rogers et al., 2020). An estimated 20% of postpartum women experience clinically-relevant levels of mental illness symptoms in the postpartum period with common categories including depression (estimated postpartum prevalence 13%), anxiety (~prevalence 10%), posttraumatic stress disorder (~prevalence 4–6%), thoughts of self-harm (~prevalence 2–5%) and psychosis (~prevalence 0.01%) (Dekel et al., 2020; Meltzer-Brody et al., 2018; Palladino et al., 2020; Yildiz et al., 2017). Postpartum mental illness may represent the exacerbation or re-emergence of pre-existing mental illness or incident mental illness, potentially triggered by the physiological and social changes of pregnancy, childbirth, and postpartum. One commonly proposed intervention to address postpartum mental health is expansion of screening for mental illness to a greater number of settings and morbidities.

There are many barriers to receiving care for postpartum psychiatric disorders, beginning with identifying a need (Rouhi et al., 2019). An estimated 30% of women do not attend the recommended postpartum follow-up visit in the United States, when screening for mood disorders is part of the standard of care (Attanasio et al., 2022). Among those who do attend the postpartum visit, 20% report that screening for depressive symptoms did not occur (Interrante et al., 2022; Stanhope & Kramer, 2021). If screening does occur, accessing timely and appropriate care may be challenging. A 2023 analysis of Medicaid-enrolled women in Michigan showed that only 19.8% percent of women who screened positive for depression or anxiety symptoms at the postpartum visit received a diagnosis of depression or anxiety (Hall et al., 2023). Women of marginalized racial/ethnic identities may experience a greater burden of postpartum mental illness (Cannon & Nasrallah, 2019; Heck, 2021). However, evidence suggests that, in the presence of depression symptoms, Black women were 80% less likely than white women to receive a diagnosis at their postpartum visit (Hall et al., 2023)

Systematic screening for postpartum psychiatric symptoms may improve the identification of potential problems, diagnosis, and referral (Byatt et al., 2020). The American College of Obstetricians and Gynecologists recommends screening for perinatal mood disorders (anxiety and depression) at least once during the perinatal period (ACOG Committee on Practice Bulletins, 2018). Further, additional studies show that mood disorder screening is feasible across a variety of settings including obstetric and pediatric well-visits, home visiting programs, and community organizations (Bhat et al., 2022). While evidence demonstrates that systematic screening for depressive symptoms improves detection, referral, treatment, and resolution of depressive symptoms (O’Connor, Rossom, Henninger, Groom, & Burda, 2016), information regarding non-mood disorder screening is limited. Expanding screening beyond depressive symptoms may better identify the range of experienced mental illness during the postpartum period.

There have been at least six prior systematic reviews on postpartum depression screening (Bhat et al., 2022; Felice et al., 2018; Ford et al., 2017; O’Connor, Rossom, Henninger, Groom, & Burda, 2016; van der Zee-van den Berg et al., 2017; Waqas et al., 2022); however only three reviews included another psychiatric morbidity, which in all cases was anxiety (Bhat et al., 2022; Ford et al., 2017; Waqas et al., 2022). Key gaps remain in the evidence base regarding postpartum psychiatric and psychosocial screening. First, prior reviews focused primarily on the detection of depression, and missed other potential psychiatric morbidities, including posttraumatic stress disorder (PTSD), suicide ideation, psychotic disorders, or anxiety, which may occur or be exacerbated in the postpartum period. Second, no prior reviews included information on the resolution of anxiety or other non-depressive psychiatric symptoms. We build on prior work to systematically review evidence for screening for psychiatric morbidities outside of depressive symptoms and psychosocial wellbeing in the postpartum period. Symptoms beyond depression, including anxiety, suicide ideation, PTSD, and psychotic disorders, are associated with poor outcomes in the postpartum period and may disproportionately burden disadvantaged populations, suggesting an urgent need for intervention in this area (Chin et al., 2022; Johannsen et al., 2020; Rogers et al., 2020).

Additional screening for perinatal mental illness has been proposed as an intervention to address the considerable burden of poor maternal mental health (American College of Obstetricians and Gynecologists, 2023). Barriers to screening for health systems and providers may limit its uptake, particularly in the absence of information on clinical utility (Bayrampour et al., 2018). These barriers include time burden, potential stigma, and a scarcity of resources to treat identified mental illness (Bayrampour et al., 2018). Understanding the effectiveness of screening for postpartum psychiatric symptoms can inform clinical practice and improve outcomes for postpartum women and other birthing people. The overarching goal of this systematic review is to characterize the effects of systematic psychiatric and psychosocial screening on the identification, referral for, and resolution of postpartum mental health symptoms.

METHODS

The current review was registered with PROSPERO (CRD42022367321) and was organized and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement for systematic reviews (Page et al., 2021).

Inclusion/Exclusion Criteria

This study sought to determine if screening for postpartum psychiatric symptoms in different contexts and across diverse settings can inform clinical practice and improve outcomes for postpartum women and other birthing people. Studies were included that had 1) people who had given birth in the last 12 months (delivery to 12 months); 2) systematic screening for psychiatric symptoms or psychosocial wellbeing beyond depressive symptoms during pregnancy or the postpartum year; 3) outcomes that included detection of psychiatric disorder (mood disorder, anxiety disorder, posttraumatic stress disorder, psychotic disorder, suicide ideation/attempt); referral (to specialty care, GP, OBGYN, support group); access to treatment (counseling, therapy, support group, medication); psychologic symptoms (depression, anxiety, posttraumatic stress disorder, psychotic disorder, suicide ideation, self-rated mental health, parenting stress, sleep) and overall wellbeing (self-rated health, quality of life, functioning, readmission); and 4) an empirical study design or clinical trial with published results. We use the term “woman” throughout the manuscript for simplicity and due to the gendered experiences of motherhood that contribute to postpartum psychiatric disorders. However, we note that not all people who give birth identify as women and that gender minorities may experience an excess burden of psychiatric disorders (Gedzyk-Nieman & McMillian-Bohler, 2022). We note that we included depressive symptoms as a potential outcome of screening if the study screened for a symptom beyond depression.

Inclusion criteria were peer-reviewed articles published between January 1, 2016 and December 31, 2023 in English or Spanish languages. The date was restricted to articles published after January 1 2016 since this was the year that the US Preventative Task Force recommended universal postpartum depression screening (O’Connor, Rossom, Henninger, Groom, Burda, et al., 2016). We recognize that these guidelines would not apply to global settings; however, we use this start date for consistency. The World Health Organization did not recommend universal postpartum depression screening until 2022 (Behl, 2023; World Health Organization, 2022). We excluded studies that (1) were qualitative only, (2) only included screening for depression, (3) did not assess systematic screening, (4) screened for physical symptoms only, and/or (5) did not include an outcome assessment during the postpartum year.

Literature Search Strategy

A comprehensive literature search strategy was developed and conducted by an experienced medical librarian with input from the research team to identify relevant articles. The searches combined controlled vocabulary supplemented with keywords related to the concepts of postpartum (e.g., postnatal, perinatal), screening (e.g., assessment, questionnaire), mood disorders (e.g., mental health, psychological well-being) and detection (e.g., diagnosis, identification). The draft strategies were peer-reviewed by another medical librarian and retested.

Six bibliographic databases were searched: Cochrane Central Register of Controlled Trials (CENTRAL), CINAHL (EBSCO), Embase.com, MEDLINE/PubMed, APA PsycInfo (EBSCO) and Web of Science Core Collection. Searches were initially undertaken on October 25, 2022 and re-run unchanged December 15, 2023. The search strategy for MEDLINE via PubMed may be found in Figure 1. Full search strategies for each database may be found in Appendix A.

Figure 1.

Search strategy for MEDLINE via PubMed

Article review and screening

Two independent reviewers reviewed each title/abstract. Two independent reviewers then reviewed full texts. Conflicts were resolved by consensus or a third reviewer.

Data extraction and synthesis

Following the identification of the final set of articles, we extracted information from each article on the population, setting, screening tool, study design, and results of screening. We then conducted a quality assessment on each included article using a modified version of the EPHPP (Effective Public Health Practice Quality Assessment Tool for Quantitative Studies) (Effective Public Health Practice Project, 1998) with questions specific to screening tools adapted from the Standards for Reporting Diagnostic accuracy studies (STARD) checklist (Cohen et al., 2016). The quality assessment tool is available in Appendix B. We considered aspects of study design (overall and presence of an unscreened control group), analysis (any confounding control), and validity of measurement tools. Measurement tools were considered strong if they had been previously validated or were validated within the study itself in reference to a gold-standard criteria (e.g., physician diagnosis). Two members of the study team completed the quality assessment tool independently and met to come to a consensus regarding any disagreement.

A priori, we planned to conduct quantitative (meta-analysis) synthesis on any screening-outcome relationship with at least five articles and qualitative synthesis on the remaining identified articles. We summarize the results across outcome categories and psychiatric symptom types.

Results

A total of 11,688 citations from the databases were uploaded to EndNote X20 (Endnote 20, n.d.), and 4,137 duplicates were manually excluded. This left 7,551 initially eligible studies, which were uploaded to Covidence systematic review software (Covidence Systematic Review Software, n.d.). Covidence identified 1,794 additional duplicates, leaving 5,757 records for title and abstract eligibility screening, which was performed independently by at least two members of the study team according to the inclusion/exclusion criteria. Consensus was reached by discussion between the two reviewers. Of these records, 5,517 were excluded for irrelevancy, leaving 241 eligible for full-text review. Of these, 194 were excluded, leaving 46 for data extraction and synthesis. The primary reasons for exclusion were screening exclusively for depressive symptoms (n = 81 excluded), measurement of symptoms was within the context of a research study only (n = 65 excluded), or the study did not assess outcomes within 0–12 months postpartum (n = 41 excluded). Further details about the review and selection processes are summarized in the flow diagram in Figure 2. All included studies are summarized in Table 1.

Figure 2.

PRISMA Flow diagram presenting studies identified and excluded at each stage.

Table 1.

List of included studies (n = 46), published 2016–2023

Lead Author	Year	Population	Symptoms Assessed	Country	Screening Timing	Sample Size
Arora	2023	postpartum women who reported a traumatic birth experience	posttraumatic stress	United States	Varied	59
Ashby	2016	pregnant women enrolled in social services at a pediatric outpatient who were adolescent parents	anxiety, trauma, and bipolar symptoms	United States	2 weeks to 12 months postpartum	894
Ayers	2018	postpartum women recruited from OBGYN outpatient, ≥ 18 years	posttraumatic stress	UK	Not reported	1018
Bachani	2022	pregnant persons women from COVID-19 maternity ward for delivery and ≥ 18 years	anxiety	India	< 1 week postpartum and after 6 to 8 weeks postpartum	243
Barrett	2023	pregnant and postpartum people recruited from an outpatient mental health clinic	emotional regulation difficulties	Canada	Varied	237
Bauer	2017	postpartum women recruited from a pediatric outpatient clinic and were suspected of having postpartum depression	anxiety	United States	3 to 15 months postpartum	73
Caropreso	2020	pregnant and postpartum women from outpatient psychiatric/mental health clinic	anxiety, OCD	Canada	Not reported	52
Cauli	2018	pregnant women receiving outpatient care	anxiety, suicidal ideation	Italy	2 months to 1 year postpartum	318
Chrzan-Detkos	2021	postpartum women receiving care in a breastfeeding clinic	anxiety, insomnia, social dysfunction	Poland	6 months postpartum	193
Cole	2018	postpartum women with a infant born with a fetal anomaly current in the NICU	posttraumatic stress	United States	1 to 3 days postpartum	725
Craemer	2023	pregnant and postpartum women seeking care in outpatient OBGYN clinics	depression, anxiety, posttraumatic stress disorder, substance use disorder, and suicidality	United States	First trimester of pregnancy or at the postpartum visit	717
Dikmen-Yildiz	2017	pregnant and postpartum women recruited from OBGYN inpatient, ≥18 years	anxiety, posttraumatic stress	Turkey	6 months postpartum	829
Dodge	2019	postpartum women recruited from community setting	general psychosocial	United States	6 months postpartum	936
Fairbrother	2016	postpartum women recruited into the study from OBGYN outpatient	anxiety	Canada	6 to 8 weeks postpartum	347
Fairbrother	2019	postpartum women recruited into the study from OBGYN outpatient	anxiety (OCD)	Canada	6 to 8 weeks postpartum	310
Fairbrother	2021	pregnant women from hospital and community-based recruitment in rural areas	anxiety (OCD)	Canada	Third trimester of pregnancy	574
Grunberg	2022	postpartum mothers of singleton babies who were in the NICU for 5–14 days, ≥ 18 years	posttraumatic stress	United States	Not reported	378
Guille	2021	postpartum women recruited from OBGYN outpatient	general psychosocial	United States	Not reported	3535
Harvey	2018	pregnant and postpartum women from outpatient psychiatric/mental health clinic	coping	Australia	Not reported	770
Highet	2019	postpartum women in Maternal and Child Health outpatient clinics	general psychosocial	Australia	4 to 6 weeks postpartum	144
Inness	2023(a)	pregnant and postpartum women recruited either from an outpatient mental health clinic or community advertisements	worry	Canada	0 to 6 months postpartum	351
Inness	2023(b)	pregnant and postpartum women recruited either from an outpatient mental health clinic or community advertisements	anxiety, worry	Canada	Not reported	214
Johnson	2021	pregnant and postpartum women from OBGYN outpatient	anxiety	United States	Between 8 weeks gestation postpartum	373
Kohlhoff	2022	pregnant women recruited from OBGYN outpatient	stress exposure	Australia	9 weeks and 10 months postpartum	485
Koukopoulos	2021	pregnant women recruited from OBGYN outpatient	anxiety, bipolar	Italy	Third trimester and 6 months postpartum	289
Lefever-Rhizal	2023	pregnant and postpartum women seeking care in midwifery clinics	posttraumatic stress	United States	Not reported	99
Makino	2020	pregnant women in outpatient psychiatric/mental health clinic with normal eating behavior	eating disorder symptoms	Japan	Not reported	24
Marley	2017	pregnant and postpartum women recruited from OBGYN outpatient ≥ 16 years	anxiety	Australia	Not reported	97
McCabe-Beane	2018	postpartum women recruited from Level IV NICU, ≥ 18	anxiety	United States	2 weeks postpartum	200
Menke	2019	pregnant and postpartum women from outpatient psychiatric/mental health clinic, ≥ 18 years	anxiety, PTSD, sleep quality	United States	Not reported	578
Millan	2021, 2023	pregnant and postpartum women receiving mental health services	bipolar disorder	United States	Varied - at least twice prenatally and once following delivery	1510
Moreyra	2021	postpartum women with an infant hospitalized in the NICU for at least 2 weeks.	anxiety, posttraumatic stress	United States	2 weeks post NICU admission	158
Moya	2022	postpartum women who had recently given birth	anxiety	Malawi	1 to 12 months postpartum	120
Nagle	2022	postpartum women from OBGYN inpatient during delivery hospitalization	perceived birth trauma	Ireland	6 to 12 weeks postpartum	1154
Okamoto	2022	pregnant women recruited from OBGYN outpatient during prenatal check-up	general psychosocial	Japan	first trimester	7462
Pereira	2022	pregnant women recruited from OBGYN inpatient and ≥ 18 years old without medical conditions	Anxiety (OCD)	Portugal	3 to 6 months postpartum	212
Reilly	2022	pregnant or postpartum women recruited from OBGYN outpatient hospital	general psychosocial	Australia	Second trimester to 3 months postpartum	2919
Schwartz	2021	postpartum women recruited from OBGYN inpatient clinic ≥ 18	anxiety	Canada	2 weeks postpartum	344
Slade	2019	postpartum women recruited from an OBGYN outpatient clinic ≥ 16 years who had given birth to a live baby	perceived birth trauma	UK	6 to 12 weeks postpartum	2414
Smith-Nielsen	2021	postpartum women recruited from community setting ≥ 18 years	anxiety	Denmark	2 months postpartum	762
Toler	2018	postpartum women	anxiety	United States	10 weeks postpartum	318
Torres-Gimenez	2021	postpartum women recruited from outpatient psychiatric/mental health clinic who had given birth to a live baby	bonding disorders	Spain	4 to 6 weeks postpartum	156
Vengadavaradan	2019	postpartum women ≥ 18	bonding disorders	India	4 weeks to 6 months postpartum	250
Voegtline	2021	pregnant women recruited from OBGYN outpatient clinic with pre-existing mood and anxiety disorders	anxiety	United States	Varied - first trimester to 6 months postpartum	295
Waqas	2021	pregnant and postpartum mothers of children up to 3 years old	depression, anxiety	Pakistan	Not reported	2492

Abbreviations: OCD- obsessive compulsive disorder, PTSD – post-traumatic stress disorder

Overall, there was no screening/outcome combination with sufficient studies to conduct meta-analysis on the relationship. The most common type of screening tool was for anxiety symptoms (25 studies), over half of which focused on diagnostic accuracy (17/25; Table 2). Depending on the study and screening tool, screening for anxiety resulted in modest to excellent diagnostic accuracy (area under the curve: 0.734–0.967) (Fairbrother et al., 2019, 2021; Inness, Furtado, et al., 2023a; Inness, McCabe, et al., 2023b; Koukopoulos et al., 2021; Marley et al., 2017; Moya et al., 2022; Pereira et al., 2022; Reilly et al., 2022; Smith-Nielsen et al., 2021; Voegtline et al., 2021; Waqas et al., 2022). Further, studies which compared cases identified using an anxiety screening tool to those with depression screening alone found that an additional 5–15% of individuals screening positive would not have been identified using a depression screening tool alone (McCabe-Beane et al., 2018; Toler et al., 2018). Five studies measured care uptake following anxiety symptom screening without a comparison group, showing that 14 – 47% of individuals screened sought mental health care (Ashby et al., 2016; Bauer et al., 2017; Guille et al., 2021; Moreyra et al., 2021; Schwartz et al., 2021). One study compared care uptake following anxiety symptoms to historic controls (Johnson et al., 2021). The researchers showed that after implementing systematic anxiety screening in a health system, the proportion of perinatal patients completing at least one mental health visit increased from 7.2% to 15.0% (Johnson et al., 2021). Three studies showed that anxiety symptoms tended to decline over the postpartum period, though they could not attribute the decline to screening or care (Bachani et al., 2022; Caropreso et al., 2020; Chrzan-Dętkoś et al., 2021). Further, postpartum anxiety diagnoses were more often prevalent (from ongoing anxiety) rather than incident during the postpartum period (Fairbrother et al., 2016).

Table 2.

Summary of study results by symptoms screened for and postpartum outcome assessed, n = 46 included studies

Symptom Type	Na	Outcome Assessed				Qualitative Summary of Results
		Diagnostic Accuracy	Care Use	Mental Health	Other
Anxietyb	25	16	6	4	0	Screening tools for perinatal anxiety have modest (~0.75) to excellent (0.96) accuracy compared to gold standard diagnostic tools. Many individuals screen positive for postpartum depression as well with 5–15% screening positive only for postpartum anxiety. In one study, fewer than half of individuals screening positive reported successfully attending a mental health visit.
Posttraumatic stress	7	2	4	1	0	PTSD screeners demonstrate high discriminatory ability (AUC: 0.93 (0.87, 0.99)). Between 6–33% of individuals screened for PTSD received a referral for mental health care; in a low-risk population, 16% asked for help with PTSD symptoms. Instituting PTSD screening in standard care increased the proportion of patients who disclosed trauma and had a documented care plan.
Stressors (current or life course)	4	1	1	2	0	Screening for life course and birth trauma is predictive of postpartum PTSD incidence and symptom severity. Screening for life course stressors may also identify individuals with unmet need for mental health care.
Comprehensive/risk factor screeners	5	1	3	2	1	Comprehensive psychosocial screeners (including life stressors, social support, pregnancy intention, and mental health history) were predictive of postpartum incident mental health problems. In the context of a comprehensive home visiting program, they were associated with improved parenting and reduced CPS investigations.
Bipolar	2	1	1			A screening tool for bipolar disorder displayed modest diagnostic accuracy (AUC 0.75) and 41% of individuals screened were referred for treatment, 47% of whom attended treatment.
Sleep quality and insomnia	2			2		Poor sleep quality was correlated with more severe depression and anxiety in postpartum people. Sleep symptoms improved over time in the postpartum period.
Bonding	2	2				Validated screening scales exist for bonding disorders in several languages.
Coping	1					Increase in perceived coping following a home visiting intervention including screening for coping, depression, and anxiety.
Emotional regulation	1	1				An emotional regulation screener was able to accurately identify current diagnoses of anxiety, depression, and trauma-related disorders.

^aNumbers are not exclusive; studies could screen for more than one type of symptom.

^bInclusive of 5 studies focused specifically on obsessive-compulsive disorder

Four identified studies presented screening for life course or birth trauma (Kohlhoff et al., 2022; Menke et al., 2019; Nagle et al., 2022; Slade et al., 2020). Life course trauma modestly predicted postpartum incident mental health problems and PTSD symptom severity (Kohlhoff et al., 2022; Nagle et al., 2022). Five studies considered comprehensive psychosocial screeners (not focused on a single symptom profile) (Cauli et al., 2019; Craemer et al., 2023; Dodge et al., 2019; Highet et al., 2019; Okamoto et al., 2022), one of which was modestly predictive (AUC: 0.63) of incident mental health diagnoses in the postpartum period. We include one study in this category that examined the clinical utility of Computerized Adaptive Testing for Mental Health (CATMH), which screens for multiple potential diagnoses (Craemer et al., 2023). The study found that of the 10.7% of people screening positive for one or more ongoing mental health disorders, 35.1% received a referral to treatment (Craemer et al., 2023).

Seven studies examined PTSD symptom screeners (Arora et al., 2023, 2023; Ayers et al., 2018; Cole et al., 2018; Dikmen-Yildiz et al., 2018; Grunberg et al., 2022; Lefever‐Rhizal et al., 2023; Moreyra et al., 2021), two of which demonstrated high accuracy (Sensitivity: 81–92%, Specificity: 90–76%) (Arora et al., 2023; Dikmen-Yildiz et al., 2018). Rates of referral following screening for PTSD ranged from 6–33% with the highest referral rates in NICU parents (Arora et al., 2023; Ayers et al., 2018; Cole et al., 2018; Dikmen-Yildiz et al., 2018; Grunberg et al., 2022). One study in midwifery care found that PTSD screening resulted in the creation of a care plan in 24% of screened participants (Lefever‐Rhizal et al., 2023). Additional identified studies screened for bipolar disorder (2) (Ashby et al., 2016; Millan et al., 2023), bonding disorders (2) (Torres-Giménez et al., 2021; Vengadavaradan et al., 2019), emotional regulation (1) (Barrett et al., 2023), eating disorder symptoms (1) (Makino et al., 2020), and coping (1) (Harvey et al., 2018).

None of the identified studies met our criteria for strong global quality (Table 3). Overall, most included studies had strong screening tools (82.6%) and validly measured outcomes (strong: 76.1%) The primary weakness was a lack of an unscreened comparison group in nearly all studies (97.9%) and no control for confounding (97.8% received poor scores in this area). A full summary of the quality assessment results is available in Appendix D.

Table 3.

Results of quality assessment using a modified version of the Effective Public Health Practice Quality Assessment Tool for Quantitative Studies with added questions specific to screening tools adapted from the Standards for Reporting Diagnostic accuracy studies, n = 46 included studies

Global Ratingb	% (n)
Strong	0 (0)
Moderate	17.4 (8)
Weak	82.6 (38)
Study Design
Good	2.2 (1)
Fair	60.9 (28)
Poor	37 (17)
Selection Bias
Good	32.6 (15)
Fair	15.2 (7)
Poor	52.2 (24)
Confounding
Good	2.2 (1)
Fair	4.3 (2)
Poor	93.5 (43)
Screening tool accuracy
Good	82.6 (38)
Fair	4.3 (2)
Poor	13 (6)
Information bias- Outcome
Good	76.1 (35)
Fair	13 (6)
Poor	10.9 (5)
Analysis
Good	47.8 (22)
Fair	10.9 (5)
Poor	2.2 (1)
N/A	39.1 (18)

^bStrong = no weak/poor ratings in any area; Moderate = 1 weak/poor rating; Weak = 2 or more weak/poor ratings

Discussion

Mental health concerns represent a growing contributor to maternal deaths in the United States and worldwide (Mangla et al., 2019; Trost et al., 2021). Despite this, our review shows that we lack high-quality evidence on the impact of screening for perinatal mental health symptoms beyond depressive symptoms. Most identified studies focused on anxiety symptoms. These demonstrated the existence of accurate screening tools for a variety of settings and populations and the ability of anxiety screening to identify unmet needs that would not have been captured through depression screening. We identified few studies for other symptom types (posttraumatic stress, obsessive-compulsive disorder, and bipolar). Regardless of the symptoms studied, most studies focused on diagnostic accuracy and few studies compared care use or health outcomes between screened to unscreened individuals. This gap limits our ability to make inferences about the effectiveness of perinatal screening to improve outcomes. Further research is needed to test whether screening of low-risk pregnant and postpartum women is an effective intervention to prevent negative sequelae of postpartum mental illness.

Similar to previous studies on mood disorder screening (Bhat et al., 2022; Felice et al., 2018; O’Connor, Rossom, Henninger, Groom, Burda, et al., 2016), our review supports the feasibility and accuracy of screening for anxiety and PTSD and, to a lesser extent, obsessive compulsive-disorder, bonding disorders, and bipolar disorder. Our review identified limited data on the effectiveness of screening on referral, treatment, and resolution of psychiatric symptoms. Without this information, busy clinical practices and health systems may be slow to adopt standardized screening beyond what is already recommended. Indeed, despite recommendations for prenatal depression screening at least twice prior to delivery (American College of Obstetricians and Gynecologists, 2023), many patients (10–50%) are not screened during pregnancy, with clear disparities based on race, geographic location (within the United States), and provider type (Eakley & Lyndon, 2024). Further evidence is needed on the effectiveness of screening for psychiatric symptoms on improving quality of life and preventing adverse outcomes (e.g., readmission, severe maternal morbidity or mortality, Child Protective Services contact). Notably, several of the included studies showed that even among women who screened positive, half or fewer received a referral and a similarly low proportion reported attending a subsequent mental health visit (Craemer et al., 2023; Grunberg et al., 2022). Screening for a range of psychiatric disorders appears to be accurate and feasible; future research and practice must focus on identifying ways to increase the provision of referrals and linkage to needed care following a positive screen.

Despite the feasibility of perinatal screening, obstetric care providers and patients describe barriers to completing screening in clinical settings. Obstetric care providers describe time constraints as a barrier to screening for postpartum depression (Bayrampour et al., 2018). Further, being familiar with an illness (depression or psychosis) increases the likelihood of screening, and few obstetric care providers report screening for rarer mental illnesses (e.g., psychosis) (Leddy et al., 2010). Evidence from trauma screening suggests that clinicians are most willing to screen when robust mental health resources are available within their health system (Rariden et al., 2021). While the majority of patients tend to describe perinatal psychiatric screening as acceptable, those who do not find screening acceptable appear to be at the highest risk of psychiatric disorders (e.g., those with a history of mental illness and/or those with a history of trauma) (Yapp et al., 2019). Some postpartum women describe reluctance to discuss mental health symptoms with providers and exhibit stigma towards sharing mental health symptoms (Foulkes, 2011; Kingston et al., 2015; Xu et al., 2021). The Trauma Informed Care framework may offer a framework to support health systems seeking to expand screening for psychiatric symptoms and potentially trauma through models of patient support, non-judgement, and individually-tailored approaches (Lewis-O’Connor et al., 2019).

The current review should be interpreted considering its limitations. While we did not exclude articles based on language, our search string was only in English, and we may have missed relevant articles published in other languages. Second, nearly all identified articles lacked a comparator group. We include those articles to provide information about the accuracy, feasibility, and acceptability of screening in diverse populations. However, this limits our ability to attribute outcomes to the screening itself. Third, we were unable to describe differences in screening accuracy or acceptability by race/ethnicity or language due to the limited number of articles that presented results stratified by these characteristics. This is an area that may be fruitful for future research. Fourth, we were unable to complete a meta-analysis due to the low numbers of articles with similar exposures (screening tools) and outcomes.

Implications for Policy and Practice

Given patient and provider barriers to psychiatric symptom screening, health systems may consider other innovative options to improve maternal mental health outcomes. One option that has received limited attention is the relative benefit of more general psychosocial risk screening versus screening specific to individual psychiatric disorders. In our review, we identified four studies that assessed the impact of a general psychosocial risk screening tool on postpartum wellbeing (Cauli et al., 2019; Dodge et al., 2019; Highet et al., 2019; Okamoto et al., 2022). The value of these tools may be to identify individuals who warrant additional screening or monitoring, to refer to social services when appropriate, and, when appropriately tailored, to provide an acceptable screening option that captures risk for a range of psychiatric disorders (Barrios et al., 2021). Another option may be to increase referrals to prevention and support programs such as perinatal home visiting, doulas, and Centering Pregnancy, which may prevent the development or exacerbation of maternal mental illness in the postpartum period. Most identified studies in the current review focused on clinical settings and clinician-led screening. However, moving screening away from clinicians and the clinical settings may be feasible and acceptable. We identified three studies that examined screening in a community or home-visiting setting. Two of these found that screening in home settings could be effective at identifying probable anxiety disorder in the postpartum period (Smith-Nielsen et al., 2021; Waqas et al., 2022). A third showed that a comprehensive home visiting program, which included screening and referral for mental illness, improved maternal postpartum mental health and decreased rates of childhood maltreatment compared to the standard of care (Dodge et al., 2019). Community and home visiting strategies should be considered further to examine whether they effectively link people with care and reduce barriers to mental health treatment. Another study showed that phone-based screening for anxiety, compared to in-clinic screening, increased screen-positive and care use rates among Black people during the postpartum period (Guille et al., 2021).

Further research may consider novel, large-scale studies (e.g., multi-site) to evaluate the effectiveness of different types of systematic perinatal screening for psychiatric symptoms. As shown in our review, the available evidence to date is largely based on single-arm studies without comparison groups. Rigorous study designs can provide actionable policy-relevant evidence. Implementation science approaches and quasi-experimental designs, including a step-wedge design or even pre/post historic controls, may allow for universal implementation of screening while providing evidence on its effectiveness (Dadich et al., 2021; Lefever‐Rhizal et al., 2023; Miller et al., 2020). Given the relative rareness of some severe mental illnesses (e.g., psychosis), novel, large designs are necessary to inform guidance regarding whether screening is an effective prevention measure. There is some evidence that postpartum psychiatric symptoms present differently across populations; for example, one study showed that Black postpartum women demonstrated a different constellation of posttraumatic stress disorder symptoms than white postpartum women (Thomas et al., 2021). Low-resourced Black women show particularly high rates of PTSD and depression in pregnancy, which often goes untreated and carries over into the postpartum period (Powers et al., 2020). Two studies identified in our review demonstrated the diagnostic accuracy of locally tailored methods of screening for perinatal mood disorders (specifically, the Kimberly Mum’s Mood Disorder screening tool in Australia for use with a specific indigenous group and a home-visiting model for rural Pakistan) (Marley et al., 2017; Waqas et al., 2022). Further research is needed to understand how locally tailored tools compare to standard (e.g., EPDS) screening tools and whether they improve diagnosis, referral, treatment, and resolution for postpartum mental illness.

Conclusions

Maternal mental illness is a major contributor to maternal mortality and morbidity (Trost et al., 2021). Innovative, multi-pronged strategies to prevent, identify, and resolve maternal mental illness are needed. Screening for postpartum psychiatric symptoms may be used as part of a wider strategy to address maternal mental health. Yet screening without improved referral and linkage to treatment and prevention is unlikely to solve the maternal mental health crisis.