Community Health Worker-Delivered Interventions for Perinatal Mental Health in the United States: A Scoping Review

Abstract

Background:

Maternal health in the United States is in crisis, and perinatal mental health conditions are emerging as a critical contributor to maternal morbidity and mortality. Perinatal mood and anxiety disorders (PMADs) encompass a range of mental health disorders experienced during pregnancy and 1 year after delivery. Expanding the perinatal mental health workforce to include paraprofessionals could help address a national shortage of mental health professionals.

Objective:

This scoping review aims to investigate the literature on perinatal mental health interventions delivered by community health workers (CHWs) or paraprofessionals in the United States.

Methods:

Studies on interventions addressing perinatal mental health were included if delivered by CHWs or paraprofessionals in the United States. Seven databases were searched via EBSCO through August 29, 2025. No publication year limits were applied. Results were exported to Zotero for deduplication, Catchii for screening, and Excel for data extraction and analysis. Risk of bias was not assessed.

Results:

After removing 4148 duplicates from the 7524 records identified, 3376 records were screened, and 13 studies met the inclusion criteria. CHW-delivered interventions for PMADs in the United States varied in design, content, and implementation. Most were initiated during the antenatal period, extended postnatally, and targeted primary prevention of depression symptoms. Ten studies evaluated outcomes among predominantly low-income, Latina, or Black mothers in urban environments. Interventions enhanced social support, cognitive, emotional, behavioral, parenting, and practical skills. Intervention results varied. Most studies reported positive effects: all nonexperimental designs found statistically significant improvements in depressive symptoms, and many experimental designs demonstrated statistically significant improvements in depressive symptoms among women receiving adequate intervention dosage.

Conclusions:

This review expands knowledge of perinatal mental health interventions delivered by CHWs and health paraprofessionals in the United States, including information about populations served, CHWs, intervention characteristics, and outcomes. Gaps in evidence and recommendations for future research are presented.

Introduction

In the United States, perinatal mental health conditions are emerging as a critical contributor to maternal morbidity and mortality. ¹ Mental health conditions, including suicide and substance use disorders, were a leading cause of pregnancy-related death in 2021, ² accounting for 22.5% of deaths, based on the most recent reporting of detailed causes of death. ³ Among deaths due to mental health conditions, which are considered preventable, nearly two-thirds (63%) occurred during the late postpartum period (days 43–365 postpartum). ³ These deaths represent only a small fraction of those suffering from perinatal mood and anxiety disorders (PMADs). PMADs encompass a range of mental health disorders, but most commonly include anxiety and depression, which manifest during pregnancy and 1 year after delivery,^4–7 although these conditions may persist beyond 1 year postpartum.^8,9 Several interrelated factors increase risk for PMADs, including biological factors (e.g., genetic and hormonal), psychosocial and interpersonal factors (e.g., stress, social support, and relationship quality), and pregnancy, fetal, or neonatal complications.^10–12 Racial and ethnic minoritized women and those with challenging social determinants of health (e.g., economic insecurity, food insecurity, discrimination, and lack of stable housing) are more likely to experience untreated PMADs.^7,13–21 These disparities can be driven by inadequate or inconsistent screening and diagnostic practices, which are exacerbated by inequitable access to appropriate treatment and care.^7,13–20

Undiagnosed and untreated PMADs can have lasting consequences across the life course, including suicide or self-harm, adverse pregnancy outcomes, and child health and behavioral problems.^22,23 Timely and systematic screening and referral during the perinatal period are critical to improving mental health outcomes and reducing disparities in care. Although the American College of Obstetricians and Gynecologists recommends screening at 6 weeks postpartum, ²³ perinatal providers often lack training in mental health and face other administrative and billing challenges that limit integration of services. ²⁴ Additionally, interaction with perinatal providers tends to diminish in the postpartum period when many mental health conditions emerge.^25,26 Workforce shortages for both perinatal and mental health professionals only exacerbate these issues.^22,23,27 Therefore, addressing infrastructure and service delivery-related challenges requires innovative solutions that complement and extend beyond clinical care.^22,27–30

One promising strategy to overcome limitations in clinical care systems and improve mental health is the integration of paraprofessionals, or lay practitioners (hereafter referred to as community health workers [CHWs]), into the health workforce. CHWs typically share similar life experiences, cultural knowledge, and language as the community they serve. Although they are typically considered nonspecialized health workers or peer support specialists, they receive targeted, standardized training to provide culturally appropriate health services to their communities. ³¹ For example, CHWs are often equipped to support health promotion, increase self-efficacy, support continuity of care through case management and treatment adherence, and connect communities to important resources to improve long-term well-being and prevent poor health outcomes, thereby addressing broader social determinants of health that may contribute to disparities in health care access and outcomes.^31,32 Particularly within resource-constrained health systems, CHWs can alleviate burdens on clinicians, serving as frontline workers for less intensive needs by connecting communities to clinicians when higher levels of care are needed (a task-shifting, stepped-care approach).^31,32

Evidence shows CHWs have been successfully integrated into interventions for physical health in various settings, including diabetes, hypertension, and HIV^33–38; however, their integration into providing mental health interventions, especially for perinatal populations in the United States, is limited. ³⁹ Most systematic or scoping reviews have captured interventions delivered by CHWs in non-U.S. countries or are aimed at training nonspecialist health workers to provide mental health services in general.^40–43 Studies in low-resourced settings outside the United States show that CHWs can effectively deliver perinatal mental health interventions.^40,41,44,45

Despite their potential impacts on improved health outcomes,^{33–38,40,41,44,45} CHWs remain significantly underutilized in the U.S. health workforce.^35,46–48 In the face of numerous barriers to equitable health and health care, and rising perinatal mental health concerns, CHWs may play a key role in delivering perinatal mental health interventions, but there is no synthesized literature detailing the components of these interventions and implementation strategies. Accordingly, it is critical to understand how CHWs and other paraprofessionals are used to support perinatal mental health, which can inform their potential impact on maternal outcomes within the unique U.S. health system. Therefore, our objective is to summarize the literature on studies evaluating perinatal mental health interventions delivered by CHWs or other paraprofessionals in the United States. This review aims to consolidate knowledge on alternative models that reimagine perinatal mental health care by identifying the populations served, outcomes assessed, key intervention components, CHW characteristics, theoretical frameworks, delivery strategies, and remaining knowledge gaps.

Methods

This scoping review followed the JBI Manual for Scoping Reviews ⁴⁹ and the Preferred Reporting Items for Systematic reviews and Meta-Analyses Extension for Scoping Reviews guidelines. ⁵⁰

Protocol and Registration

The methodology for this scoping review was determined in advance and registered on the Open Science Framework. ⁵¹

Eligibility Criteria

Guided by the providers–concept–context (PCC) framework, ⁴⁹ eligible studies were empirical investigations evaluating perinatal mental health interventions delivered by CHWs or comparable paraprofessionals without any formal mental health training (provider). Eligible interventions focused on the prevention, identification, or treatment of mental health disorders during pregnancy and up to 1 year postpartum (concept). Studies were included if conducted in the United States and involved participants aged 18 years or older who were pregnant or within 12 months postpartum (context). Randomized clinical trials, quasiexperimental studies, or pre–post nonexperimental evaluations were eligible, while single-subject designs, study protocols, gray literature, and unpublished studies were excluded. Only studies published in English were considered, with no restrictions on publication year.

Information Sources

Using the EBSCOhost interface, a public health librarian (NT) conducted systematic searches across seven health science and psychology databases up to August 29, 2025: CINAHL, Family and Society Studies Worldwide, Health Source: Nursing/Academic Edition, MEDLINE, PsycINFO, Psychology and Behavioral Sciences Collection, and SocINDEX. Additionally, the reference lists of all included sources of evidence were reviewed to identify any additional eligible studies.

Search Strategy

We developed a search strategy based on the PCC framework to ensure comprehensive coverage of relevant studies. Database searches combined text words and controlled vocabulary terms for provider (e.g., community health workers, health educators, and lay counselors), perinatal mental health concept (e.g., perinatal, postpartum in combination with depression, anxiety, and psychological distress), and intervention-related context (e.g., home visit, program, telehealth, and mHealth). Proximity operators and truncation symbols were applied to capture variations in phrasing and terminology. Complete search terms and full search strings are provided in the Supplementary Data S1 and S2.

Selection of Sources of Evidence

All citations identified through database searches were imported into the citation management tool Zotero (https://www.zotero.org/) for duplicate removal and initial screening by a librarian (NT) to exclude clearly irrelevant records (e.g., editorials, conference abstracts, and news). The remaining records were exported to Catchii (https://catchii.org/) for collaborative screening. Reviewer pairs (MET, MM, APM, JRJ, LMM, AP, DG, and SEH) independently screened titles, abstracts, and full-texts against the predefined eligibility criteria. Discrepancies were discussed and resolved in weekly meetings to ensure consistency in study selection.

Data Extraction and Coding Procedures

A standardized Excel form was developed for data extraction, and data extraction was conducted in pairs: one reviewer extracted the data, and another verified it. Extracted information included population characteristics, targeted PMADs, intervention components, theoretical frameworks, CHW characteristics, and intervention outcomes. Reviewer pairs met to reach consensus and resolve discrepancies, with a third reviewer consulted as needed. Coding questions or challenges were addressed in biweekly meetings led by the second (MET) and senior author (MM).

Critical Appraisal

Consistent with established scoping review guidelines, no formal critical appraisal of individual sources was conducted. ⁵⁰ However, study design, sample size, and methodological rigor were considered when interpreting the overall strength and consistency of the evidence base.

Synthesis of Results

To address the research objectives, we followed narrative synthesis recommendations for scoping reviews, employing tabulation, textual descriptions, comparison, pattern identification, and vote counting to synthesize findings. ⁵² Tabulation and narrative summaries were used to describe study populations, settings, theoretical frameworks, PMADs assessed and their measurement; intervention timing, duration, components, and CHW-delivered services; and CHW education, training, and implementation supports such as supervision and fidelity monitoring. Textual descriptions summarized each study’s key features, including populations, PMADs, intervention content, theoretical frameworks, CHW training, intervention delivery, and key findings to support the initial synthesis of data on these topics.

Results

Selection of Sources of Evidence

The database search identified 7524 records, of which 4148 were duplicates. Titles and abstracts of the remaining 3376 records were screened, and ultimately 51 studies were selected for full-text review. Of these, 11 studies^53–63 met all eligibility criteria for inclusion. Two additional studies^64,65 identified through a hand search of related systematic review articles met the eligibility criteria for inclusion, resulting in a total of 13 publications^53–65 included in the final review.

Studies were excluded at the full-text stage if they failed to meet all inclusion criteria. The most common reasons for exclusion were that studies did not focus on a U.S.-based population, did not evaluate interventions delivered by CHWs or paraprofessionals, or were exclusively qualitative. The study selection process is summarized in Figure 1.

FIG. 1.

PRISMA flow diagram of study selection process. Records were identified through database searching (CINAHL, Family and Society Studies Worldwide, Health Source: Nursing/Academic Edition, MEDLINE, PsycINFO, Psychology and Behavioral Sciences Collection, and SocINDEX) and through review articles and reference lists. After removal of duplicates (n = 4148), 3376 records were screened by title and abstract. Fifty-one full-text articles were assessed for eligibility, and 13 studies met the inclusion criteria. PRISMA, Preferred Reporting Items for Systematic reviews and Meta-Analyses.

Characteristics of Sources of Evidence

Study populations

Interventions were studied across a variety of populations. Nine were conducted in urban settings,^{53–57,61–64} eight study populations were low-income or predominantly low-income,^{57–59,61–65} and three focused primarily on first-time mothers.^54,58,60 Most study populations were predominantly racial or ethnic minorities,^55–65 with six of these studies specifically focusing on Hispanic populations^{55–57,59,61,64} in the United States, including Mexican-American, Central American, and first-generation communities (Table 1).

Table 1.

Study Characteristics

Lead author, date	Study population	Sample size (N)	Study design	Setting	Theoretical approach of intervention	Primary outcomes	Analytic methods	Evidence of change
Barkin et al., 2021 53	Postpartum mothers (mean age: 33 years), from a range of racial and socioeconomic backgrounds, living in an urban environment	156 [402 enrolled]	Pretest–posttest, single group	Intervention delivered at home of client	Maternal functioning model, attachment theory	Depression (PHQ-9) at program completion	Paired t test with complete case sample	Yes
Collado et al., 2025 55	Latina mothers (18+ years of age) living in urban or rural environments	126 [141 enrolled]	Pretest–posttest, single group	Intervention delivered in various locations with the client	Behavioral activation	Depression (PHQ-9), anxiety (GED-7)	Hierarchical linear modeling adjusted for baseline values among the full intention-to-treat sample	Yes
Robidoux et al., 2023 59	Majority first generation, low-income, postpartum Latina mothers (18+ years of age) living in a periurban environment	1846 visits	Pretest–posttest, single group	Intervention delivered at community clinic	Health care access	Screening and referral rates for postpartum depression (EPDS)	Univariate statistics and chi-squared analysis	Yes
Sampson et al., 2024 61	Low-income Latina mothers (18–44 years of age) up to 12 months postpartum living in mostly urban environments	74 [96 enrolled]	Pretest–posttest, single group	Intervention delivered at home of client	Problem-solving therapy and coaching	Depressive symptoms (EPDS and PHQ-9) at various time points	Paired sample t tests, Cohen’s d estimated effect size. ANCOVA estimated effects in subgroups (nativity, Spanish speaking), controlled for baseline level of outcome	Yes
Sperlich et al., 2022 62	Pregnant, mostly African American, low-education, low-income women with a history of trauma living in an urban environment	38 [56 enrolled]	Pretest–posttest, single group	Intervention delivered at home of client	Trauma and PTSD-specific psychoeducational model	Depressive symptoms (EPDS), PTSD symptoms (PCL-5) at program completion	Paired-sample t tests with pooled estimates of posttest scores from intention-to-treat sample; McNemar’s change test for small samples for categorized scores in per-protocol sample	Yes
Mundorf et al., 2018 58	Low-income, majority African-American, first time mothers (18+ years of age), living in a periurban environment	102 [141 enrolled]	Quasiexperimental, 2 arm	Intervention delivered at homes of clients and at community center	Social support model	Postpartum depression (EPDS) at 6 months postpartum	Independent t tests; linear correlation measured using Pearson’s coefficient or Spearman’s Rho; adjusted multiple regression	Yes
Tandon et al., 2022 65	Pregnant, low-income, predominantly racial minority women (mean age: 29 years) enrolled in a home visiting living across one U.S. state	1017 [1229 enrolled]	Quasiexperimental, 2 arm	Intervention delivered at home of client	Cognitive behavioral theory, attachment theory	Depressive symptoms at 6 months postpartum (BDI)	Linear mixed effects model controlling for baseline values and unadjusted logistic regression to examine dosage effects using intention to treat sample	Mixed
Gordon et al., 1999 54	First time mothers (18+ years) with uncomplicated pregnancies living in an urban environment	314 [478 enrolled]	Randomized control trial, 2 arm	Hospital: doula led care for women during labor and delivery	Psychosocial support model	Depression (MHI) at 4–6 weeks postpartum	Chi-square and Fisher’s exact test	No
Keller et al., 2014 56	Sedentary, overweight to obese, (BMI 25–35) postpartum Latina mothers (18–40 years) living in an urban environment	139 [177 enrolled]	Randomized control trial, 2 arm	Intervention delivered at community center	Social support model, healthy lifestyle	Postpartum depression (EPDS) at 12 months postpartum	ANOVA	No
Kieffer et al., 2013 57	Pregnant or early postpartum low-income Mexican-American mothers (18+ years of age) living in an urban environment	275 [278 enrolled]	Randomized control trial, 2 arm	Intervention delivered at homes of clients and at community center	Social support, healthy lifestyle	Depressive symptoms (CES-D) at 6 weeks postpartum	Unadjusted linear mixed models using intention-to-treat sample; adjusted linear mixed model, subgroup analysis of Spanish-speaking participants	Mixed
Le et al., 2011 64	Pregnant, low-income, Central-American mothers (18–35 years of age) living in an urban environment	217 [220 enrolled]	Randomized control trial, 2 arm	Intervention delivered at the community center	Cognitive behavioral theory, health management of reality model	Depressive symptoms (BDI) and major depressive episodes at 4 and 12 months postpartum	Adjusted regression analysis using intention-to-treat sample; Fisher’s exact test and complier average causal effects methods	Mixed
Samankasikorn et al., 2016 60	Pregnant and postpartum majority Hispanic teenagers (mean age 17) living in rural and urban environments	150 [186 enrolled]	Randomized controlled trial, 2-arm	Intervention delivered at home of client	Peer support model, social support model	Postpartum depression (EPDS) at 3 months postpartum	t-tests, ANOVA, and Kruskal–Wallis tests	No
Tandon et al., 2021 63	Pregnant women 16+ years of age, predominantly racial minority, enrolled in a home visiting program across 7 US states	824 [874 enrolled]	Cluster-randomized noninferiority trial	Intervention delivered at community center	Cognitive behavioral theory, attachment theory	Depressive symptoms at 24 weeks postpartum (QIDS-SR16)	Adjusted linear mixed effects models	Mixed

ANOVA, analysis of variance; ANCOVA, analysis of covariance; BMI, body mass index; PTSD, posttraumatic stress disorder.

Theoretical frameworks and models

A range of theoretical frameworks and models were applied to design the interventions, with most studies^{53–58,60,61} emphasizing the role of social support in the mental health and well-being of pregnant and postpartum women. Specific frameworks and models included the social support model,^56–58,60 attachment theory,^53,63,65 maternal functioning model, ⁵³ behavioral activation, ⁵⁵ cognitive behavioral theory,^63–65 healthy management of reality model, ⁶⁴ peer-support model, ⁶⁰ problem-solving therapy and coaching, ⁶¹ health care access, ⁵⁹ psychosocial, ⁵⁴ and psychoeducational ⁶² information. Additionally, most studies implied the application of community-based participatory research ⁶⁶ (CBPR) principles in the design and development of their intervention, and a number described the processes they followed to engage and incorporated community perspectives and needs.^{55–57,61,65}

Intervention characteristics, including timing, duration, and components

The study authors’ descriptions of the intervention’s characteristics, timing, duration, and components varied considerably in detail and completeness. Of the 13 studies, 2 interventions took place exclusively during pregnancy^54,62; 7 interventions spanned both pregnancy and the postpartum period^{55,57,58,60,61,64,65}; 3 were limited to the postpartum period^53,56,59; and 1 during labor and delivery. ⁵⁴ The duration of the interventions also varied substantially, ranging from a screening session ⁵⁹ to as long as 19.5 months. ⁶⁰ Most interventions, however, lasted longer than 6 months^{53,55,58,60,61,64,65} (Table 2).

Table 2.

Intervention Timing, Duration, and Components

Study	Intervention time relative to pregnancy	Intervention duration and dosage	Intervention components (as provided by CHW/PPs)
			Screening and referral?	Health education	Delivery modalities
Barkin et. al., 2021 53	Postpartum: 0–12 months	Approx. 10 months: weekly, 1-h home visits after enrollment until child’s first birthday	No	Yes, practical skills-based teaching/mentoring: modeling, nurturing, and supportive relationships, normalizing/universalizing concerns, encouragement to explore and reflect, resource sharing, problem-solving	Home visits
Collado et al., 2025 55	Pregnancy and postpartum: first trimester to 6 months postpartum	Approx. 15 months: six, 1-h, peer mentoring sessions during pregnancy, two, 1-h, meetings postpartum (optional)	No	Yes, 8 sessions of skills-based cognitive behavioral and behavioral activation content to address depressed mood, anxiety, and stress	Individual meetings with mentor
Robidoux et. al., 2023 59	Postpartum: days after delivery to 6 months postpartum	<1 day: varied by case; mothers screened at 5 well-child visits between birth and 6 months; longer if referred	Yes	Yes, but limited. Overview of postpartum depression provided during repeated PPD screening with patients attending the clinic for well-child visits (newborn, 1, 2, 4, and 6 months)	Individual sessions, if referred, follow up with phone calls
Sampson et. al., 2024 61	Pregnancy and postpartum: birth to 12 months postpartum	Approx. 12 months: five 1-h sessions	No	Yes, mental health and stressor management	Home visits and individual sessions
Sperlich et. al., 2022 62	Pregnancy: enrolled at any point during pregnancy	Approx. 10 weeks: sessions (30–60 min in length) once per week for up to 10 weeks	Yes	Yes, 10 modules on managing emotions, relationships, and trauma reactions in a range of contexts	Home visits and individual sessions
Mundorf et. al., 2018 58	Pregnancy and Postpartum: first trimester through 6 months postpartum	Approx. 12–14 months: at least 5 in-person sessions, mobile health technology, and phone calls	No	Yes, content not described by the authors	In-person meetings, some at home, plus phone calls and texts
Tandon et al., 2022 65	Pregnancy and postpartum: enrolled before 32 weeks of gestation until 6 months postpartum	Approx. 8 months: weekly home visits	No	Yes, cognitive behavioral content to teach mothers skills and behaviors that lead to positive mood states, manage stress and depressive symptoms; information and skills on attachment theory to promote strong parent–child relationships, plus existing home visiting curriculum content. CHW given autonomy to adapt content and delivery from 11 standard sessions to fit home visiting format	Home visits
Gordon et. al., 1999 54	At the time of parturition	Approx. 2–3 days: during labor and postdelivery hospital stay	No	Yes, during labor and delivery. Information on treatment options provided such as guidance on pain medication, epidurals, c-section, use of oxytocin, as well as postdelivery guidance on breastfeeding	Individual care primarily during labor and delivery
Keller et. al., 2014 56	Postpartum: 0–3 months	Approx. 12 weeks: weekly sessions	No	Yes, primarily on the importance of physical activity for overall health, diet, and women’s health issues	Group sessions plus phone calls and texts
Kieffer et. al., 2013 57	Pregnancy and postpartum: <20 weeks of gestation to 6 weeks	Approx. 26 weeks: Two home visits and 9 group meetings during pregnancy; 2 home visits and 1 group meeting postpartum	No	Yes, information, discussion, and activities aimed at empowering women to develop knowledge and skills and reduce social and environmental barriers to healthy eating and regular exercise. And information on pregnancy, childbirth, postpartum period, identifying and managing stress, and symptoms of depression	Home visits and group sessions, activity days
Le et. al., 2011 64	Pregnancy and postpartum: second trimester through 3 months postpartum	Approx. 9 months: eight, 2-hour, group sessions during pregnancy, 3 individual sessions postpartum	No	Yes, psychoeducational and CBT content teaching women mood-regulation skills, local social services	Group sessions and individual sessions
Samankasikorn et. al., 2016 60	Pregnancy and postpartum: first trimester to 12 months postpartum	Approx. 19.5 months: From 6 weeks pregnant through 1 year postpartum	Yes	Yes, adequate pregnancy spacing, improved psychosocial well-being, completion of education, and child outcomes	Individual sessions and home visits
Tandon et. al., 2021 63	Pregnancy: enrolled before 33 weeks of gestation	Approx. 6 weeks: 6 weekly sessions (average session = 86 min long) plus usual home visiting services	No	Yes, CBT content included managing emotions and mood, engaging in pleasant activities, identifying helpful/unhelpful thought patterns, strategies for reframing, promoting positive social interactions, assertive communication, stress management plus existing home visiting curriculum content	Small group sessions (1:3 cohorts)

CBT, cognitive behavioral therapy; CHW, community health worker; PPD, postpartum depression.

CHWs carried out a range of tasks aligned with the design and goals of each intervention. The number of participants served by each CHW ranged widely (1–300+) and was not reported in several studies. Three studies’ interventions included a screening and referral component.^59–61 All interventions included health education content delivered by CHWs. All but two interventions^54,59 used structured modules covering a broad range of topics related to maternal mental health and well-being, delivered over multiple sessions (Table 2).

Delivery modalities also varied considerably across studies. All interventions involved at least one face-to-face session, supplemented in some cases with telephone or text contact.^55,56,58,59 Seven interventions were delivered through a combination of home visits and community center sessions, with some sessions conducted in participants’ homes and others held at community centers,^{53,57,58,60–62,65} whereas three interventions were conducted solely at community centers^55,63,64 and two at a medical facility^54,59 (Table 2). One study reported transitioning sessions to online video conferencing due to the COVID-19 pandemic. ⁵⁵

In most studies, CHWs delivered the intervention after receiving intervention-specific training; however, the duration and rigor of training varied widely. Descriptions of training content and processes were limited in all studies, with four studies^56,57,60,64 providing no information (Table 3). Five interventions^55–59 were delivered by CHWs recruited from the same communities as participants, sharing cultural, linguistic, and social backgrounds. Across studies, it was often unclear if CHWs were compensated for their work. In 5 of the 13 studies, CHWs are noted as being employed by a nonprofit or faith-based organization or the research team.^{60,61,63–65}

Table 3.

Intervention Delivery and Implementation Fidelity

Study	CHW education and background	CHW training and supervision			Fidelity monitoring of intervention?
		CHW training: duration and description	Refresher training?	Supervision?
Barkin et. al., 2021	Trained volunteers	Five 2-h training sessions: content included orientation to the program, communication and honing listening skills, problem-solving and cultural awareness, roles and boundaries, postpartum depression, and anxiety	Yes	Yes	Yes
Collado et al., 2025 55	Trained, community-based, Spanish-speaking paid volunteers	2-day workshop, mock mentoring sessions, and a weekly 2-h consultation meeting with supervisor for remainder of the program	Not reported	Yes	Yes
Robidoux et. al., 2023 59	Trained community based, Spanish-speaking community recruited volunteers	CHWs and clinic staff trained by project coordinator on how to administer and score EPDS, and follow an algorithm to determine possible treatment options, including crisis plan for severe cases of PPD	Yes	Yes	Yes
Sampson et. al., 2024 61	Spanish-speaking CHW employed by local FBOs	1-day: information on symptoms and assessment of PPD, how culture influences perceptions and identification of PPD, motivational interviewing, project data collection; included role play/practice	Yes, training recordings available for review	Not reported	Yes
Sperlich et. al., 2022 62	Trained CHWs with in-depth understanding of the community’s culture, language	3-day training on the SMC’s foundational research, the effects of trauma, trauma-informed care, and intervention materials. Project-specific training was also provided to the agency as a whole and to supervisors	Not reported	Yes	Not reported
Mundorf et. al., 2018 58	Trained volunteers	Trained using a competency-driven curriculum, and 2 modules on reproductive and sexual health and core public health topics	Not reported	Not reported	Yes
Tandon et al., 2022 65	Trained home visiting paraprofessionals employed by agency, 60% with GED or bachelor’s degree	Onboarding training on the content and implementation logistics of the health start home visiting model, plus training on a range of maternal and health topics	Yes, ongoing training	Not reported	Yes
Gordon et. al., 1999 54	Doulas	All doulas attended approved doula training programs in the community, served as doulas for at least two births; attended a half-day orientation to research protocol, hospital’s labor and delivery protocol	Not reported	Yes	Yes
Keller et. al., 2014 56	Trained community-based, Spanish-speaking volunteers	Not reported	Not reported	Yes	Yes
Kieffer et. al., 2013 57	Trained community-based, Spanish-speaking volunteers	Not reported	Not reported	Yes	Yes
Le et. al., 2011 64	Trained post-bachelors Spanish-speaking research staff	Not reported	Not reported	Yes	Yes
Samankasikorn et. al., 2016 60	CHW, some bilingual	Not reported	Not reported	Yes	Not reported
Tandon et. al., 2021 63	Trained home visiting paraprofessional employed by the agency	1–1.5-day training led by principal investigator, plus previous training from employing agency on home visiting	Not reported	Yes	Yes

CHW, community health worker; PPD, postpartum depression.

Study designs and outcomes

Of the 13 studies, 8 used experimental designs, 2were quasiexperimental,^58,65 6 included randomized control trials,^{54,56,57,60,63,64} and 5 used nonexperimental, single-group, pretest–posttest designs^{53,55,59,61,62} to evaluate the interventions under study. Only one of the eight studies that used an experimental design reported a statistically significant difference in depressive symptoms among women who received the intervention as compared with those who did not. ⁵⁸ Four of the eight studies using experimental designs reported mixed findings among participants: those who received an adequate dose of the intervention (as defined by the study authors) experienced statistically significant improvements in PMAD outcomes compared with those who did not.^57,63–65 Three of the studies in the experimental design group, all evaluating the mothers and babies (MB) intervention, showed the greatest amount of change in depression among women who exhibited mild depressive symptoms at baseline (i.e., secondary prevention rather than primary prevention). In particular, the Tandon et al.’s ⁶³ study demonstrated similar reductions in depressive symptoms regardless of whether paraprofessionals or mental health professionals delivered the intervention. All five studies^{53,55,59,61,62} that used a pretest/posttest design reported improvements in participants’ PMAD-related outcomes.

Although this scoping review included all PMADs (e.g., depression, posttraumatic stress disorder [PTSD], anxiety, and bipolar), few studies evaluated CHW-led interventions to address mental health conditions other than depression during the perinatal period. All assessed depression outcomes, two also measured changes in anxiety,^55,58 and one assessed changes in PTSD symptoms. ⁶¹

In alignment with the theoretical frameworks guiding the interventions, five studies assessed social support as a secondary outcome.^{53,56,60,63,65} Additional secondary outcomes were examined across three domains: (1) perinatal emotional and psychological factors—maternal function, ⁵³ self-esteem,^54,60 self-efficacy, ⁶¹ anger, ⁶² birth perceptions, ⁵⁴ mood regulation,^63,65 reframing of negative thoughts,^63,65 engagement in positive activities,^55,63,65 positive mood, ⁵⁸ mother’s adverse childhood experiences, ⁶² prenatal psychological distress, ⁵⁸ perceived stress,^55,65 and perception of neighborhood environment and safety ⁵⁶ ; (2) postpartum physical health—delivery-related complications, ⁵⁸ general health, ⁵⁸ body mass index, ⁵⁶ tissue inflammation, ⁵⁶ and energy intake ⁵⁶ ; and (3) postpartum health care access—postpartum depression (PPD) referral rate. ⁵⁹

Discussion

PMADs significantly contribute to maternal morbidity and mortality in the United States. ⁶⁷ Racial and ethnic minoritized women, immigrant women, low-income women, and those living in under-resourced urban or rural settings are disproportionately affected yet often face barriers to timely and effective care. ⁷ In this scoping review of 13 studies, CHW-delivered interventions hold promise as equity-promoting strategies because they are rooted in community trust, cultural and linguistic alignment, and increase accessibility beyond traditional medical settings. However, the current evidence base in the United States is small and fragmented, leaving important questions unanswered about how these models can be scaled and adapted to advance health equity in perinatal mental health care.

Summary of Findings

This scoping review adds to existing reviews assessing CHW-staffed interventions for health^{41,42,68–70} and is the first to focus solely on CHW-delivered interventions for PMADs in the United States. Results of this limited but growing body of evidence suggest that CHW-delivered interventions for PMADs in the United States vary widely in design, content, and implementation. Overall, these interventions are typically initiated during the antenatal period and extend into the postnatal period, targeting primary prevention of depression symptoms among maternal populations with multiple, intersecting risk factors in the United States. Ten of the studies were conducted among predominantly low-income, Latina, or Black mothers.^55–65 Interventions generally aimed to improve PMAD outcomes by enhancing social support, teaching cognitive, emotional, behavioral regulatory skills, and building parenting and practical skills. They were delivered mostly by volunteer CHWs recruited from the same community as intervention participants, who were then trained^{53–55,58,59,61–63,65} and supervised^{55,57,59,62,63} to meet with participants in person to deliver health education content, build and practice skills, and provide social support. Three studies^63–65 implemented the MB intervention, a cognitive behavioral therapy intervention that provides health education and skills to support healthy mood management and reduce stress and depression, using different delivery strategies and formats.

The outcomes of interventions vary, but there is evidence of success in some contexts and populations. Notably, among studies using nonexperimental designs, all reported statistically significant improvement among participants, and these results should be interpreted carefully, with consideration of the intervention under study and the time points at which PMAD-related outcomes were assessed.^{53,55,59,61,62} Among studies using experimental designs, one reported statistically significant improvements among intervention participants overall, ⁵⁸ with other studies finding statistically significant improvements among participants who received an adequate dosage of the intervention.^57,63–65 Further research is needed to better understand intervention effects, taking into account intervention retention challenges and underlying population characteristics, which may bias evaluation results.

Gaps in Evidence

Scope of interventions

As most of the reviewed interventions focused on primary prevention of PMADs, particularly depression symptoms, it remains unclear whether these interventions would be appropriate as standalone treatments for clinically significant PMADs or whether they would have a synergistic effect when applied alongside clinical care. It is also important to note that only three studies^55,58,62 in this review assessed PMAD outcomes other than depression. It is unclear if these interventions, primarily designed to address depressive symptoms, would be appropriate and helpful for other PMADs. While it is plausible that CHW-delivered interventions developed for depressive symptoms may also improve other PMADs, this has not been empirically tested.

Populations studied

While interventions in this review focused on populations known to be disproportionately at risk of developing PMADs, many high-risk groups in the United States remain understudied, including Native Americans, women living in rural areas, Asian women, fathers,^71,72 and LGBTQ + populations. ⁷³ Additionally, the usefulness of this type of intervention for the prevention or treatment of PMADs in dyads, fathers, or partners of birthing people is understudied. ⁷² Evidence from other countries suggests that CHW models may benefit perinatal populations across socioeconomic and racial/ethnic groups, but this remains untested in the U.S. context. ⁴¹ Only one study ⁵⁸ explored outcomes among mothers with delivery complications or neonatal intensive care unit (NICU) admissions, finding that mothers who experience delivery complications and received the intervention had three time the odds ratio (OR 3.10 confidence interval [CI] 1.26–7.61) of experiencing improvement in depressive symptoms than those without complications, mother’ whose newborn spent any time in the NICU had almost five times the odds (OR 4.96 CI 12.9–19.16) of improvement in depressive symptoms. This study highlights the potential impact for these unique populations. ⁵⁸

Theoretical frameworks and models

Interventions in this review were largely grounded in individual behavior change frameworks and models, yet a critical comorbid factor—substance abuse ⁷⁴ —was notably absent. None of the frameworks or models addressed interpersonal risk factors such as family systems, relationship health, ⁷⁵ or intimate partner violence.^73,74,76,77 Only one study ⁶² incorporated elements from the life course theory, and only one study ⁵⁶ considered other domains of social determinants of health, such as neighborhood safety.

Implementation and delivery

One intervention explicitly addressed the issue of optimal intervention timing, ⁵⁸ leaving a significant gap in the literature on when to intervene to address PMADs. Mundorf et al. ⁵⁸ found no evidence that intervening at a specific prenatal time is key to realizing PPD prevention outcomes. Otherwise, delivery modalities are also relatively unexplored. For example, the relative effectiveness of group versus one-on-one sessions is unknown, as are participants’ preferences for telehealth versus in-person delivery during the postpartum period. Questions remain as to which delivery modality is best for various communities. Training and supervision approaches were inconsistently reported, making it difficult to assess standards for preparation and support.

Strengths and Limitations

Authors of studies included in this review may have sparsely described the interventions under study, particularly intervention resources and inputs such as training, recruitment, and CHW supervision, as well as processes for tailoring and designing the interventions. There may have been variation in extraction across reviewers; however, the first author further evaluated and reconciled this information as necessary to ensure consistency. Minimal detail on these intervention aspects may have limited analysis and assessment. Additionally, some interventions, such as the MB intervention, have been evaluated more frequently than others and are more represented in this review. Finally, our inclusion criteria focused on studies evaluating PMADs, potentially excluding studies of other CHW-delivered interventions with related outcomes.

This scoping review is strengthened by the multidisciplinary composition of our research team. Team members are leading scholars and interventionists in maternal, reproductive, and perinatal health whose work spans epidemiology, community health systems strengthening, mental health and trauma, family and couples’ health, health policy, health services research, and clinical obstetrics. Our combined experience includes conducting research, policy analysis, and developing and implementing community-based or CHW-led MCH interventions. Our vision for this review was to generate actionable information to spur innovation and improve perinatal health care for the well-being of all families in the United States.

Future Directions: Reimagining Perinatal Care through Health Equity

Reimagining perinatal health care to address mental health in the United States requires rethinking current models of care and considering how best to engage communities and families. The human aspect of care may be the missing key, and incorporating compassionate, resourceful CHWs to listen, empower, teach, and guide families to access the types of care or services they need may be the best option; however, evidence gaps persist. ⁴⁸ To advance health equity in perinatal mental health, future research and practice should seek out and include understudied populations, ⁷³ integrate family and relational health, ⁷⁷ and assess the effectiveness of CHW intervention as a standalone treatment for clinically significant PMADs. Applying CBPR principles ⁶⁶ to tailor existing interventions and prioritize research with populations such as Native American, Asian American, LGBTQ+, and rural populations, as well as fathers and partners, will address gaps in knowledge and reduce inequities in access to culturally responsive care within the United States.^73,74

Currently, a lack of evidence prevents a clear understanding of whether CHW interventions can effectively address family relationship outcomes. These outcomes, including relationship satisfaction, parenting stress, coparenting quality, and partner involvement, are central to equity in perinatal care. ⁷² Incorporating family systems perspectives to assess outcomes not only among mothers but also partners and infants may reveal new ways to support and strengthen vulnerable families and catalyze lasting improvements in mental health and well-being outcomes at a population level in the United States. Family-level interventions may more easily incorporate education and skills to improve relationship health and reduce substance use and intimate partner violence, which disproportionately affect communities facing structural inequities.^73,74

CHWs working in tandem with primary, obstetric, pediatric, and mental health systems, as well as communities, may expand access to equitable, high-quality, appropriate, and coordinated care.^14,48 In the United States, evaluation is critically needed to understand how CHW interventions can best integrate within existing workflows and synergize with maternal health care providers and structures. It remains to be understood if CHW interventions improve access to treatment for PMADs for mothers or the caseload of higher-level care providers, such as psychiatrists and counselors. To ensure continued investment in building the mental health workforce, rigorous large-scale evaluations of CHW-delivered interventions are needed, and it is imperative that we apply CBPR principles ⁶⁶ to engage communities and CHWs as codevelopers from the start.

Research is needed to optimize intervention design and delivery and evaluate whether CHWs with different training levels and education deliver interventions with equivalent effectiveness. Research should seek to elucidate the best time to intervene during the perinatal period, what types or combinations of delivery modalities are the best for participants (e.g., group vs. individual, telehealth vs. in-person), barriers to participation, and how to empower families to access this type of care.

Finally, research is needed to ensure fair compensation and support for the CHW workforce. Sustaining this workforce equitably requires formalizing CHW training programs and qualifications, assessing training needs, setting minimum education and continuing education requirements, and defining effective supervision models and standards to ensure quality and effectiveness. ⁴⁸ Addressing these gaps can help reimagine perinatal care in ways that not only reduce PMADs but also build equitable, family-centered systems of support.

Authors’ Contributions

This scoping review was conceptualized by M.E.T., M.M., R.G., L.M.M., A.P.M., and N.T.; Methodology was developed by M.E.T., N.T., M.M., R.G., L.M.M., and A.P.M.; Investigation, including article review and data extraction, was conducted by N.T., M.E.T., M.M., L.M.M., A.P.M., and D.G.; Data curation was conducted by N.T., A.P.M., G.D., and L.M.M.; Project administration: M.M., M.E.T., and A.P.M.; Supervision: M.E.T., M.M., and N.T.; Funding acquisition: M.M. M.E.T., and R.G.; Resources: N.T., M.E.T., and M.M.; visualization: N.T., A.P.M., M.E.T., and L.M.M.; Validation: M.E.T., M.M., R.G., and N.T.; Writing—original draft: M.E.T., M.M., A.P.M., L.M.M., D.G., and N.T.; Writing—review and editing: M.E.T., N.T., M.M., R.G., L.M.M., and A.P.M.