The role of home visiting in addressing postpartum care gaps: Evidence from a large community sample

Abstract

Background

Maternal mortality in the United States has doubled over the last two decades, primarily due to cardiovascular disease, with more than half of deaths occurring postpartum. Despite national calls to action, attendance at postpartum clinic visits remains low. The role of alternative care delivery strategies, such as home visiting, remains insufficiently studied.

Objective

To describe postpartum clinic visit attendance among participants in a large national home visiting program, identify sociodemographic and clinical factors associated with attendance, and examine how timing and duration of enrollment relate to postpartum follow-up.

Design

We conducted a retrospective cohort study of pregnancy episodes among mothers enrolled in the Parents as Teachers (PAT) national home visiting program between July 2016 and September 2024.

Methods

The primary outcome was documented attendance at a postpartum clinic visit within 12 weeks of delivery. Descriptive analyses estimated overall attendance rates and characterized variation by pregnancy risk status, geography, calendar year, and sequential pregnancies. Multilevel logistic regression with random intercepts for participant and site was used to account for repeated pregnancies and clustering. Sensitivity analyses included all pregnancies, including those exiting home visiting prior to 12 weeks postpartum.

Results

Among 46,899 pregnancy episodes from 39,498 participants, median maternal age was 27 years (IQR 18–36). Overall, 30.1% (95% CI 29.6–30.5) attended a postpartum clinic visit within 12 weeks, including 33.2% (95% CI 32.3–34.1) among high-risk pregnancies. Attendance was lower among Black (adjusted OR 0.80 [95% CI 0.75–0.86]), American Indian/Alaska Native (0.61 [0.55–0.68]), rural participants (0.77 [0.73–0.82]), and participants who primarily used the emergency room or urgent care for medical care before pregnancy (0.78 [0.72–0.85]). Attendance washigher among recent immigrants (1.17 [1.05–1.30]), women with high-risk pregnancies (1.13 [1.06, 1.20]),and those enrolled prenatally (1.93 [1.84–2.04]).

Conclusions

Postpartum clinic attendance was low in this national home visiting cohort. Prenatal and sustained engagement in home visiting were associated with higher odds of attendance. Home visiting is a potential complementary care model to extend early care beyond clinical settings and promote sustained engagement in clinical follow-up.

Keywords

postpartum care home visiting maternal health postpartum clinic attendance health disparities community-based interventions retrospective cohort study

1. Introduction

In the United States the age-adjusted maternal mortality rate has more than doubled between 1999 and 2022, primarily due to increases in cardiovascular disease-related mortality.¹ Over one-half of maternal deaths occur in the postpartum period,² making it a critical window to retain women in care and ensure linkage to longitudinal clinical care.^3–5 Recognizing these risks, the American College of Obstetricians and Gynecologists (ACOG) and the American Heart Association (AHA) have called for continuous, individualized postpartum care with cardiovascular risk assessment and coverage through 12 months.^6,7

Despite these national calls to action, many women disengage from clinical care shortly after delivery.⁸ Nationwide, postpartum follow-up between 6-12 weeks postpartum is low across all groups – around 40% overall and as low as 10% in economically disadvantaged populations.^2,9 Even among high-risk pregnancies, fewer than one in four patients see a primary care clinician within six months postpartum,¹⁰ and only half with severe preeclampsia attend a six-week postpartum clinic visit.¹¹ While these statistics highlight major gaps in clinical care, much of the literature on postpartum care utilization derives from small, single-hospital studies, state-level estimates or administrative claims data. In a systematic review of 88 studies examining postpartum clinical follow-up, only five used nationally representative data, and only one examined a community-based population.² Furthermore, these studies had limited data on postpartum visit timing, determinants, or barriers to accessing care. Overall, greater understanding of population-level drivers of postpartum care engagement is needed to improve clinical outcomes.¹²

Given these care gaps, alternative strategies for delivering longitudinal postpartum care have been proposed. A 2025 American Heart Association (AHA) Scientific Statement emphasized extending postpartum care beyond traditional health care settings to improve population health.¹³ International evidence-based home visiting programs, such as Parents as Teachers (PAT), support families during the perinatal and early childhood periods by providing in-home guidance that removes barriers to clinical care, such as transportation and childcare, while also connecting families to community resources that facilitate linkage to care.¹⁴ PAT supports families through personal visits, developmental guidance, family well-being promotion, resource connections, and goal-oriented parent education beginning during pregnancy and extending through the early years of a child's life. PAT reaches populations often underserved by health systems and engages families during pregnancy and early parenthood. However, PAT is not primarily designed to deliver healthcare-focused interventions,¹³ and evidence for its role in strengthening linkage to clinical care and promoting postpartum care engagement remains relatively limited.

To address these evidence gaps, this retrospective cohort study of pregnancy episodes within a national home visiting program aims to: a) describe patterns in postpartum clinical visit attendance across time and geography; b) identify sociodemographic and clinical determinants of attendance; and c) evaluate how enrollment timing and participation across multiple pregnancies are associated with follow-up. By leveraging a large, community-based sample, this study provides new insights into patterns of postpartum clinical follow-up among families engaged in home visiting.

2. Methods

2.1. Study design and setting

This is a retrospective cohort study of pregnancy episodes examining postpartum clinic visit attendance within 12 weeks postpartum among mothers enrolled in the Parents as Teachers (PAT) national home visiting program between July 2016 and September 2024, when there was a consistent and centralized data collection and management system in place. The analytic unit was the pregnancy episode, with some participants contributing more than one pregnancy during the study period. Pregnancies were followed from delivery through 12 weeks postpartum to ascertain visit attendance. Methods are reported according to STROBE guidelines (eMethods 1).¹⁵

2.1.1. Setting

Parents as Teachers (PAT) is a federally recognized, evidence-based home-visiting program, and the role of PAT educators in promoting maternal and child health has been well described.¹⁶ Participation in PAT is voluntary and referral pathways vary across affiliates, including healthcare, education, and community agencies. Information on education and training of home visiting parent educators is available (eMethods 2). Services are delivered through local affiliates supported by a mix of federal, state, local, and philanthropic funding sources, including the Maternal, Infant, and Early Childhood Home Visiting Program (MIECHV),¹⁷ a program with a focus on pregnant and postpartum mother and child wellbeing.

Affiliates follow program standards for home visit frequency based on family stressors, with at least monthly visits recommended when one or more family-level stressors are present (see eMethods 3).

2.1.2. Data source

Data were obtained from Penelope, a HIPAA-compliant data management system linking family-, parent-, and prenatal/postpartum-level data using unique identifiers. The Parents as Teachers National Center (PATNC) provides affiliates with guidance and data quality reports for ongoing accuracy. All records were de-identified, and the study was deemed exempt by Washington University Institutional Review Board.

2.2. Participants and study size

Participants were included if they: (1) were enrolled in PAT between July 2016 and September 2024, and (2) had at least one live birth recorded during that period, with enrollment occurring either prenatally or postpartum. Participants were excluded if they: (1) exited the PAT program prior to 12 weeks postpartum, or (2) did not have a prenatal or postpartum record initiated or modified after 12 weeks postpartum, as these records could not confirm outcome ascertainment. An a priori power calculation was not performed; the analytic sample was determined by available data, comprising 46,899 pregnancy episodes with documented postpartum follow-up through 12 weeks, as detailed in eFigure 1. Given this sample size, the study had >99% power to detect odds ratios as small as 1.10 for key binary exposures at α=0.05, confirming adequate power for the primary analyses. Because continued engagement may be associated with healthcare utilization behaviors, a supplemental sensitivity cohort including pregnancies that exited prior to 12 weeks postpartum was analyzed separately (eTable 1 and eFigure 4).

2.3. Measures

Data were primarily obtained via structured self-report collected by trained PAT parent educators and documented in standardized electronic records with routine quality assurance monitoring by PAT National Center, except for administrative information on funding sources and program records documenting date of PAT enrollment, record initiation and modification, and exit date.

2.3.1. Outcomes

Postpartum clinic visit attendance and timing were assessed by asking if the birthing parent had at least one postpartum visit with a healthcare provider within 12 weeks after pregnancy in agreement with ACOG guidelines on timing of postpartum assessment.¹⁸ Postpartum clinic visit attendance was documented by trained and certified home visitors through the standardized PAT data collection protocol at the time of each visit, based on participant report. Attendance was documented by trained PAT staff in prenatal/postpartum records and categorized as Yes/No. For all participants that responded yes, the date of the clinic visit was recorded.

2.3.2. Variables of interest

Sociodemographic variables analyzed include maternal age at delivery, race, ethnicity, community type (urban, suburban/small town, and rural, based on participant residential zip code), health insurance status, employment status, housing status, highest education level attained, source of funding for home visiting, and family stressors. Health insurance coverage was classified as coverage at or within four weeks of delivery date.

Employment status was categorized into unemployed (not working for pay), full-time (average of ≥ 30 hours/week), or part-time (average of <30 hours/week). Funding sources were classified based on whether home visiting was funded by MIECHV or non-MIECHV, as MIECHV specifically includes education on the importance of postpartum care.

Family stressors present at delivery and related to parenthood and access to postpartum clinical care were recorded,^19,20 inclusive of low household income (eligible for benefits as listed in eMethods 2), single parenthood, multiple children in the household, ,teen parenthood, and recent immigration (within the past five years).

Clinical variables analyzed include location of medical care prior to pregnancy, timing of enrollment in PAT (prenatal or postpartum), presence of high-risk pregnancy, and mode of delivery. Enrollment in home visiting occurring prior to pregnancy or during pregnancy, up to the delivery date, was considered prenatal enrollment. High-risk pregnancy was identified based on participant report that a prenatal care provider had classified the pregnancy as high risk.

2.3.3. Duration of PAT enrollment

Participants enrolled in home visiting over multiple pregnancies were analyzed by birth order to assess changes in postpartum attendance across pregnancy.

2.4. Statistical analysis

All analyses, including data cleaning and linkage, were conducted using R version 5.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Two-sided p-values <0.05 indicated statistical significance. No adjustment was made for multiple testing.

2.4.1. Descriptive analysis

Participant characteristics were evaluated overall and stratified by postpartum clinic visit attendance. Continuous variables, such as maternal age and number of family stressors, were summarized as median (IQR) because data were skewed. Family stressor results represent linear approximations across the observed range. Complete case analysis was applied to demographic factors of age, race, and ethnicity as well as timing of enrollment, since there were no missing data for these variables. Calendar year was evaluated descriptively only given substantial variation during the study period. Trends in postpartum clinic visits across multiple pregnancies were evaluated as proportions with p-values reported. Timing of postpartum clinic visits was described graphically using a histogram of weeks from birth to the first documented clinic visit.

2.4.2. Regression analysis

Multivariable logistic regression was used to evaluate the association of postpartum clinic visit attendance within 12 weeks postpartum (the primary dependent variable) with the sociodemographic and clinical factors described above. Model variable selection was informed by literature on prenatal care engagement, non-pregnancy home visiting, and limited evidence on postpartum care engagement.^21–23

Repeated observations were clustered by individual within mixed-effects models. Results were also adjusted for site-level clustering. Models were evaluated for collinearity, and univariate estimates were compared with multivariable estimates to assess changes in effect size and direction. Results are reported as odds ratios (ORs) with 95% confidence intervals (CIs).

Models were initially adjusted for age alone. A second model additionally adjusted for employment status and health insurance status. For the purposes of regression analysis, employment status was treated as a binary variable (employed versus unemployed). Health insurance coverage was also treated as a binary variable in the regression model (insured versus un-insured), as all documented insurance providers cover at least one comprehensive postpartum clinical visit. The final model adjusted for funding source (MIECHV versus non-MIECHV) as MIECHV has a maternal and child health focus.

2.4.3. Missing data

Multiple imputation by chained equations (MICE) was applied to variables with <15% missing data. Except for insurance type at delivery, all variables at ≤10% missing data. There were no variables with >15% missing data. Quality of imputation was evaluated by comparing observed and imputed category proportions, with acceptable differences defined as <5%. Variables without missing data were excluded from imputation to avoid model instability.

2.4.4. Sensitivity analyses

Sensitivity analyses for descriptive analyses were performed for a) all pregnancies, including those exiting the visits prior to 12 weeks postpartum, and b) all home visits occurring postpartum, including after 12 weeks. All regression analyses were repeated using complete case data as a sensitivity check.

3. Results

During the study period, there were 51,730 recorded live births among women enrolled in the PAT home visiting program. Of these, 4,831 were excluded due to exiting the program prior to 12 weeks postpartum, resulting in a primary analytic sample of 46,899 pregnancies from 39,498 participants. The participant flowchart is shown in eFigure 1. Demographic characteristics for all recorded live births (including those who exited the program prior to 12 weeks) are shown in eTable 1. Participants were enrolled across 44 states and the District of Columbia (eFigure 2). eFigure 3 shows the distribution of postpartum clinic visit timing for all pregnancies. Forty percent of visits occurred between 2–6 weeks postpartum, 47.6% between 6–8 weeks, and 10.2% between 8–12 weeks. The remaining 2.2%, not included in the primary analysis, occurred beyond 12 weeks postpartum, with three visits recorded more than one year after delivery.

Overall, postpartum clinic visit attendance within 12 weeks of delivery was reported for 30.1% (95% CI 29.6%-30.5%) of pregnancies (Figure 1(a)). Attendance remained low even among high-risk pregnancies, at 33.2% (95% CI 32.3%-34.1%) (Figure 1(b)). In the expanded sensitivity cohort including exit from home visiting prior to 12 weeks, attendance was 29.5% overall (95% CI 28.7%–30.7%) (eFigure 4A) and 31.8% (95% CI 69.2% – 71.6%) (eFigure 4B) among high-risk pregnancies. Rates were not significantly different between the primary analytic cohort and the sensitivity cohort (p=0.45). Rates varied widely by geography, with the highest attendance in New York and the lowest attendance in Nevada (Figure 1(c)). State-level estimates are descriptive and do not account for differences in sample size or case-mix. Across time, attendance rates remained relatively stable (35.5–37.9%) through 2021, dropped to 16.8% in 2021, then rose sharply to 44.5% in 2023 and exceeded 50% (53.5%) by 2024 (Figure 2). Rates of prenatal enrollment in home visiting followed a similar trajectory.

Overall demographic and clinical characteristics are presented in Table 1. The median maternal age was 27 years (IQR 18, 36). Participants were predominantly White (60.0%), followed by Black (18.5%), and American Indian/Alaska Native (6.6%). Approximately one third (34.9%) identified as Hispanic or Latino. Community type was evenly distributed, with 22.9% of participants residing in rural areas. Medicaid was the primary insurer at delivery (37.2% of the sample), and 7.5% of participants were uninsured around the time of delivery.

Table 1.

Characteristics of overall study population.

Characteristic	Overall (n=46,899)
Demographic
Age, median (IQR), y	27 (18, 36)
Race, n (%) ^a
White	28,139 (60.0)
Black	8,676 (18.5)
American Indian/Alaska Native	3,095 (6.6)
More than one race	2,157 (4.6)
Asian	985 (2.1)
Other	1,313 (2.8)
Prefer Not to Report	2,533 (5.4)
Ethnicity, n (%)
Not Hispanic or Latino	29,218 (62.3)
Hispanic or Latino	16,368 (34.9)
Prefer not to report	1,313 (2.8)
Community Type, n (%)
Suburban/Small Town	20,448 (43.6)
Urban	15,242 (32.5)
Rural	10,740 (22.9)
Insurance Coverage at Time of Delivery, n (%)
Medicaid	17,446 (37.2)
Tribal/Military	9,989 (21.3)
Private	6,566 (14.0)
No Insurance	3,517 (7.5)
Employment Status, n (%) ^b
Unemployed	21,105 (45.0)
Full-Time	9,849 (21.0)
Part-Time	6,941 (14.8)
Housing Status, n (%) ^b
Rents Housing	20,260 (43.2)
Owns Housing	9,520 (20.3)
Living with Parents	6,472 (13.8)
Undomiciled	1,360 (2.9)
Education Level, n (%) ^b
High School	13,319 (28.4)
Less than High School	8,770 (18.7)
College education	7,504 (16.0)
Other	8,207 (17.5)
Funding Source for Home Visits, n (%)
Non-MIECHV	32,736 (69.8)
MIECHV^c	14,163 (30.2)
Family Stressors
Number of Family Stressors, median (IQR)	2 (0, 4)
Low Household Income	33,627 (71.7)
Single Parent	13,929 (29.7)
Multiple Children	13,695 (29.2)
Teen Parent	6,191 (13.2)
Recent Immigrant	3,564 (7.6)
Clinical Factors
Prenatal Enrollment in Home Visiting, n (%)	11,115 (23.7)
First Trimester	1,767 (15.9)
Second Trimester	4,346 (39.1)
Third Trimester	5,002 (45.0)
High Risk Pregnancy, n (%) ^d	10,177 (21.7)
Delivery Mode, n (%)
Vaginal	31,844 (67.9)
Planned Cesarean	10,552 (22.5)
Unplanned Cesarean	1,876 (4.0)
Prenatal Preventative Care Locations, n (%)
Primary Care Clinic	23,082 (49.2)
Emergency Room/Urgent Care	15,758 (33.6)
No Location	1,126 (2.4)
Other	4,784 (10.2)

^aDenominator for all percentage (%) calculations.

^bTaken at time of first postpartum home visit screening (≤4 weeks postpartum).

^cMIECHV = Maternal, Infant, and Early Childhood Home Visiting Program.

^dBy participant self-report - “Have you been told by your clinical provider that your pregnancy is high-risk?”

Nearly half of participants (45.0%) were unemployed. Most either rented (43.2%) or owned (20.3%) their housing; 2.9% were undomiciled. 44.4% had completed at least a high school diploma or GED. Close to one-third (30.2%) of participants received home visiting funded at least in part by MIECHV.¹⁷

The median number of family stressors was 2 (IQR 0, 4). 71.7% of participants lived in low-income households, 29.2% had multiple children in the household as a stressor, and 7.6% were recent immigrants.

Nearly half (49.2%) reported a primary care clinic as their usual source of care, 12.6% relied primarily on emergency rooms or urgent care centers, and 2.4% reported having no source of care.

Regarding timing of enrollment, 23.7% of participants were enrolled prenatally. Missingness rates are detailed for each variable included and did not differ substantially by postpartum clinic visit attendance (the primary outcome) or funding source (eTable 2).

3.1. Regression analysis

Associations of sociodemographic and clinical factors with postpartum clinic visit by multivariable logistic regression are shown in Table 2. Compared to White participants, Black (adjusted odds ratio [OR] 0.80 [95% CI 0.75–0.86]) and American Indian/Alaska Native (adjusted OR 0.61 [95% CI 0.55–0.68]) participants were significantly less likely to attend a postpartum clinic visit within 12 weeks.

Table 2.

Association of demographic and clinical factors with 12 week postpartum clinic visit attendance.

Variable	Model 1^a		Model 2^b		Model 3^c
Variable	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Race
White	1.00 (ref)	—	1.00 (ref)	—	1.00 (ref)	—
Black	0.86 (0.80, 0.92)	<0.001	0.87 (0.80, 0.95)	0.002	0.80 (0.75, 0.86)	<0.001
AI/AN^d	0.64 (0.58, 0.72)	<0.01	0.66 (0.57, 0.78)	<0.001	0.61 (0.55, 0.68)	<0.001
More than one Race	0.94 (0.84, 1.05)	0.29	1.02 (0.89, 1.18)	0.73	0.92 (0.82, 1.03)	0.14
Asian	1.25 (1.06, 1.47)	0.01	1.38 (1.14, 1.68)	0.01	1.29 (0.97, 1.74)	0.09
Other	0.83 (0.49, 1.39)	0.47	1.03 (0.58, 1.84)	0.92	0.80 (0.47, 1.34)	0.39
Prefer not to report	0.93 (0.83, 1.05)	0.24	0.91 (0.78, 1.05)	0.21	0.97 (0.86, 1.09)	0.56
Ethnicity
Hispanic or Latino	1.00 (ref)	—	1.00 (ref)	—	1.00 (ref)	—
Not Hispanic or Latino	0.87 (0.80, 0.94)	0.04	0.89 (0.77, 1.04)	0.10	0.90 (0.79, 1.03)	0.09
Prefer not to report	0.91 (0.77, 1.06)	0.22	0.93 (0.76, 1.12)	0.43	0.90 (0.77, 1.06)	0.20
Community Type
Urban	1.00 (ref)	—	1.00 (ref)	—	1.00 (ref)	—
Suburban/Small Town	0.93 (0.67, 1.20)	0.48	0.90 (0.69, 1.17)	0.27	0.91 (0.68, 1.18)	0.24
Rural	0.78 (0.74, 0.82)	<0.01	0.78 (0.72, 0.83)	<0.01	0.77 (0.73, 0.82)	<0.01
Education Level
College education	1.00 (ref)	—	1.00 (ref)	—	1.00 (ref)	—
High School	1.11 (0.97, 1.26)	0.13	1.09 (0.93, 1.29)	0.27	1.07 (0.94, 1.23)	0.29
Less than High School	1.04 (0.97, 1.11)	0.28	1.04 (0.95, 1.14)	0.40	1.02 (0.95, 1.09)	0.62
Other	0.98 (0.90, 1.07)	0.64	0.95 (0.85, 1.06)	0.34	0.98 (0.90, 1.06)	0.57
Housing Status
Domiciled	1.00 (ref)	—	1.00 (ref)	—	1.00 (ref)	—
Undomiciled	0.90 (0.81, 1.01)	0.08	0.91 (0.80, 1.04)	0.17	0.93 (0.83, 1.04)	0.22
Number of Family Stressors ^e	1.09 (0.87, 1.32)	0.53	1.09 (0.87, 1.32)	0.53	1.07 (0.85, 1.30)	0.51
Type of Family Stressor
Low Household Income	0.97 (0.88, 1.06)	0.49	1.05 (0.93, 1.19)	0.40	0.99 (0.89, 1.09)	0.77
Single Parent	1.10 (0.89, 1.18)	0.23	1.16 (0.77, 1.26)	0.24	1.11 (0.70, 1.19)	0.38
Multiple Children	1.02 (0.96, 1.09)	0.45	1.02 (0.94, 1.10)	0.67	1.06 (1.00, 1.13)	0.06
Teen Parent	0.97 (0.89, 1.06)	0.55	0.95 (0.85, 1.06)	0.38	0.99 (0.91, 1.08)	0.86
Recent Immigrant	1.21 (1.09, 1.34)	<0.01	1.15 (1.01, 1.31)	0.03	1.17 (1.05, 1.30)	<0.01
Preventative Care Locations
Primary Care Clinic	1.00 (ref)	—	1.00 (ref)	—	1.00 (ref)	—
Emergency Room or Urgent Care	0.78 (0.71, 0.84)	<0.01	0.80 (0.72, 0.89)	<0.01	0.78 (0.72, 0.85)	<0.01
No Location	1.06 (0.91,1.23)	0.46	1.04 (0.90, 1.22)	0.37	0.99 (0.86, 1.15)	0.32
Other	0.90 (0.76, 1.06)	0.21	0.90 (0.72, 1.13)	0.36	0.89 (0.75, 1.05)	0.16
Prenatal Enrollment	1.95 (1.86, 2.06)	<0.01	1.91 (1.79, 2.03)	<0.01	1.93 (1.84, 2.04)	<0.01
High Risk Pregnancy	1.12 (0.95, 1.25)	0.14	1.09 (1.02, 1.17)	0.02	1.13 (1.06, 1.20)	<0.01
Delivery Mode
Vaginal	1.00 (ref)	—	1.00 (ref)	—	1.00 (ref)
Planned Cesarean	0.98 (0.93,1.03)	0.46	0.97 (0.92,1.01)	0.16	0.99 (0.93, 1.06)	0.77
Unplanned Cesarean	1.06 (0.91, 1.23)	0.48	1.08 (0.87, 1.34)	0.47	1.26 (0.93, 1.71)	0.13

^aModel 1: Adjusted for age as well as individual and site-level clustering using mixed methods.

^bModel 2: Adjusted additionally for employment (employed versus un-employed) and insurance status (insured versus un-insured).

^cModel 3: Additionally adjusted for funder source (MIECHV versus non-MIECHV).

^dAI/AN: American Indian/Alaska Native.

^eOR per 1-unit increase.

Living in a rural community was associated with lower attendance (adjusted OR 0.77 [95% CI 0.73–0.82]) compared to urban areas. Housing status and educational attainment were not significantly associated with the primary outcome.

The overall number of family stressors was not significantly associated with clinic visit attendance. However, being a recent immigrant (adjusted OR 1.17 [95% CI 1.05–1.30]) was associated with higher likelihood of attendance.

Participants who primarily used the emergency room or urgent care for medical care before pregnancy were less likely to attend a postpartum clinic visit at 12 weeks (adjusted OR 0.78 [95% CI 0.72–0.85]) compared to those with a primary care clinic. Participants who enrolled in PAT home visiting prenatally were almost two times as likely (adjusted OR 1.93 [95% CI 1.84-2.04]) to attend a postpartum clinic visit. Having a high-risk pregnancy was positively associated with attendance (adjusted OR 1.13 [95% CI 1.06–1.20]).

Sensitivity analyses using complete case data produced results consistent in direction and magnitude with the imputed analyses. Further sensitivity analyses with participants that had less than 3 personal home visits and participants with postpartum visits after 12 weeks also produced results consistent in direction and magnitude with the final sample used for regression analysis.

3.2. Duration of PAT enrollment

Overall, 6,402 participants experienced multiple live births while enrolled in home visiting. Of these, 5,570 had two live births, 703 had three, and 129 had four or more, for a total of 29.4% of all studied live births. As shown in Figure 3, postpartum clinic attendance within 12 weeks was significantly greater when comparing the first to the second birth among participants with two live births (24.6% [95% CI 23.5–25.7] to 33.7% [95% CI 32.5–35.0]; p < 0.05), and across all three births among participants with three live births (22.5% for first birth, 27.9% for second birth, 34.3% for third birth; p_trend < 0.05).

4. Discussion

In this national, community-based sample of postpartum women enrolled in home visiting, only 30% attended a postpartum clinic visit within 12 weeks of delivery. Attendance remained low even among high-risk pregnancies (33%), underscoring the persistent gap in postpartum follow-up despite national calls to action by professional societies.⁷ This effect remained when all pregnancies, including those with incomplete follow-up, were included. Although postpartum attendance remained low, consistent with national concerns regarding postpartum care engagement, the study identified characteristics associated with greater follow-up and highlighted opportunities for community-based support systems to strengthen continuity of care.

In this cohort, notably, 21.7% of pregnancies were identified as high risk, higher than the U.S. national estimate of 6–8%.¹ Several factors were associated with clinic attendance within 12 weeks: non-modifiable factors included race, community type, recent immigrant status, and high-risk pregnancy, while modifiable factors included source of medical care before pregnancy and enrolling in home visiting prenatally. Notably, women who enrolled prenatally in home visiting had higher documented rates of postpartum clinic attendance, a finding supported both in regression analyses and by the upward trend in attendance in the later years of the study. These findings are observational and likely reflect underlying engagement behaviors; individuals who enroll earlier or remain in home visiting may already differ in health-seeking behavior, care navigation, or resource access, which may influence both program participation and postpartum clinic attendance. Nonetheless, these findings reinforce that routine postpartum clinic attendance within 12 weeks remains low among families reached by home visiting, even among those engaged in longitudinal support services.

4.1. Generalizability of the study sample

This is one of the largest studies to utilize a community-based, national sample rather than a single-center or claims database.^2,22,24 Prior studies report wide variation in postpartum attendance, from 10% in Medicaid analyses up to 50% in the Pregnancy Risk Assessment Monitoring System (PRAMS).^2,25 Attendance in this study fell between these estimates at 30% overall, although attendance reached 54% by 2024. Although the primary outcome was based on quality-verified documentation by home visitors rather than clinical verification through an electronic medical record, similar estimates have been observed in prior analyses using comparable geographic and programmatic data.^26–28 In contrast, higher attendance estimates reported in surveillance systems such as PRAMS likely reflect differences in both population examined and outcome definition. PRAMS captures self-reported receipt of any postpartum checkup rather than attendance within a specific clinical timeframe, and lower rates of postpartum visit attendance in PAT compared with PRAMS may reflect the greater socioeconomic burden among PAT-enrolled families, as PRAMS data represent a broader population not matched to this study sample.

Participants represented all U.S. regions, with relatively even distribution of insurance coverage and a sizable rural population. Rural populations remain underrepresented in postpartum care research, despite well-documented disparities in mortality, adverse outcomes such as gestational diabetes, and higher emergency use postpartum.^29,30 The reach of home visiting into these settings highlights its value as a data source for populations that are often invisible in clinic-based datasets, particularly in maternity care deserts,³⁰ areas lacking obstetrical health services. Participation in PAT is voluntary and varies across affiliates in referral processes and documentation practices. Therefore, findings may be generalizable to families engaged in similar home visiting mechanisms but may not extend to all postpartum populations.

4.2. Sociodemographic factors associated with postpartum clinical care engagement

Compared with White women, Black and American Indian/Alaska Native women had lower odds of postpartum attendance. These results extend prior evidence documenting lower postpartum care utilization among Black women and add to limited data on AI/AN women, who face higher maternal mortality, particularly in rural areas, as compared to non-Hispanic White women.^22,31

Interestingly, being a recent immigrant was associated with a higher likelihood of postpartum clinic attendance. While prior work describes language, cultural, and logistical barriers to accessing clinical care among immigrant populations, community-based interventions may partially mitigate these barriers.³² These findings align with the “immigrant paradox,” whereby non–US-born immigrants often demonstrate better health outcomes than their US-born counterparts.³³ Together, this suggests that barriers such as language, cultural tailoring, and navigation support may be effectively addressed through home visiting, motivating immigrant mothers to engage in postpartum follow-up.

4.3. Timing of enrollment and longitudinal engagement

A key strength of this study is its evaluation of how timing of home visiting enrollment relates to clinical care engagement. Prenatal enrollment in PAT was associated with a two-fold greater likelihood of postpartum clinic attendance, emphasizing the value of early outreach and sustained relationship-building. This association was reinforced by a temporal trend showing that increases in prenatal enrollment coincided with higher postpartum attendance. Although prenatal enrollment was associated with higher attendance, this relationship should not be interpreted as evidence that early home visiting increases postpartum clinic attendance. It is as likely that individuals who engage earlier in pregnancy are more proactive, better resourced, or more motivated to seek care across the perinatal and postpartum continuum. With this consideration, home visiting may still provide a feasible, broad-reaching mechanism to enhance postpartum and interpregnancy surveillance and may help support linkage to clinical care, particularly for those at increased risk of adverse pregnancy and cardiovascular outcomes.

Longitudinal participation across multiple pregnancies was also associated with progressively higher attendance, which may reflect sustained engagement among participants who remain connected to both home visiting and clinical care systems over time.³⁴ While most healthcare-oriented home visiting programs target first pregnancies, examining multiple pregnancies is important, as uncontrolled risk factors across pregnancies contribute to morbidity, mortality, and long-term cardiometabolic disease.³⁵

Temporal patterns were also notable. The decline in 2021 likely reflects effects of the COVID pandemic, possibly because telehealth clinical visits were not consistently captured. Rates of both prenatal enrollment and postpartum clinic visit attendance rose sharply in 2022, coinciding with PAT’s new prenatal and postpartum education materials as well as initiation of remote home visiting.¹⁶ By 2024, postpartum clinic attendance rose to approximately 50%, representing a substantial increase from prior years. Sustained home visiting has been shown to reinforce postpartum education and adherence,³⁴ suggesting that structured program-level initiatives can have population-level impact. Overall, substantial shifts during the COVID-19 period likely reflect a combination of care delivery changes (including telehealth expansion), documentation workflows, and programmatic modifications. Secular trends should therefore be interpreted cautiously.

Previous work has noted that the timing of postpartum care delivery is rarely studied, despite its centrality to ACOG and AHA guidelines on postpartum care delivery. Despite guidelines recommending contact within 3 weeks and a comprehensive clinic visit by 12 weeks, most participants reported visits between 6–8 weeks, even in later years of the study, reflective of prior standards.⁶ These findings suggest opportunities for system-level improvement, including encouraging earlier clinical engagement for those at highest risk and ensuring insurance coverage up through 12 weeks postpartum.

4.4. Clinical context: Usual source of care

Utilization of emergency room care or urgent care as a participant’s primary source of healthcare when not pregnant, as reported at PAT enrollment, was associated with reduced likelihood of postpartum attendance. Prior studies have linked higher pre-pregnancy care utilization to greater postpartum and preventative engagement, but few distinguish between types of pre-pregnancy care.^36,37 Understanding health behaviors across the preconception-to-postpartum continuum is critical for linking women to longitudinal clinical care.

4.5. Policy implications

These findings have several implications for policy and practice. Ensuring adequate reimbursement for evidence-based home visiting services during pregnancy and through the first year postpartum may be a critical step to support sustained engagement during a high-risk period. Prioritizing maternal-child health and postpartum care as performance measures can strengthen education during home visits, improve patient knowledge, and promote positive health behaviors; extending this priority across home visiting funding sources may be beneficial. In parallel, maintaining continuous health insurance coverage for mothers for at least 12 months after delivery is essential to facilitate access to postpartum clinical care. Additional strategies such as culturally tailored patient navigation, telehealth expansion, and structured referral pathways from maternity care to primary care warrant prospective evaluation to determine their impact on postpartum care continuity.

4.6. Limitations

Several limitations in this study warrant careful consideration. First, postpartum clinic attendance and high-risk pregnancy classification were based on participant report and standardized documentation by trained parent educators rather than direct medical record abstraction. However, PAT educators receive structured training, utilize standardized data entry protocols, and participate in routine quality assurance review conducted by the PAT National Center, which supports data reliability. Nonetheless, variation in data collection methods across sites cannot be excluded. Furthermore, the substantial missing data on substance use and postpartum depression highlight areas for higher data quality and further investigation.

Second, the primary analytic cohort required documentation through at least 12 weeks postpartum, which may preferentially select participants retained in home visiting. Because engagement behaviors may correlate with healthcare utilization, selection bias is possible. However, sensitivity analyses including pregnancies that exited prior to 12 weeks and with less than 3 documented visits demonstrated non-significant differences in demographic distributions and comparable postpartum attendance rates, suggesting that exclusion criteria did not materially alter overall attendance estimates. Third, the study did not capture other forms of postpartum clinical contact (e.g., telehealth visits, specialty care, emergency care), limiting conclusions about overall healthcare utilization.

Because participation in home visiting is voluntary, individuals who enroll earlier or remain engaged may differ systematically in motivation, stability, health literacy, or access to care. These engagement characteristics plausibly influence both program participation and postpartum clinic attendance. Therefore, observed associations should be interpreted as reflecting correlated clinical engagement behaviors rather than direct program effects. Furthermore, enrollment duration was used as a proxy for program engagement; analysis of visit-level data revealed substantial individual-level variability in program dose (median 3 visits, IQR 1–5, range 1–9), precluding its use as a standardized exposure measure and warranting examination of visit intensity as a primary outcome in future studies.

4.7. Future directions

Despite these limitations, home visiting data capture a segment of the postpartum population that is often underrepresented in clinic- and claims-based datasets, including individuals with limited or fragmented access to healthcare. Although postpartum visit attendance was documented within a community-based system rather than verified through direct medical record linkage, this infrastructure allows observation of care engagement among families who may not consistently interface with traditional clinical healthcare systems.

Large-scale clinical verification of postpartum care engagement within home visiting populations has not yet been systematically implemented. As such, the descriptive and multivariable analyses presented here reflect the methodologic limits of the current data infrastructure. Nevertheless, home visiting data represent one of the few scalable approaches to examining postpartum care patterns in populations that are often underrepresented in clinic- and claims-based datasets, and provide substantially larger community-based samples than most single-system clinical studies. Future integration of home visiting records with electronic medical records - or development of enhanced clinical verification processes within home visiting systems - could strengthen data validity and enable more robust evaluation of postpartum care delivery and longitudinal health outcomes.

Future work should also explore care delivery models addressing persistent postpartum barriers, including the demands of caring for a newborn, and evaluate not only strategies that increase postpartum follow-up but also those that improve the quality, continuity, and effectiveness of the care delivered. Differentiated service delivery models, such as hypertension interventions in Black barbershops³⁸ in the United States and international models of home-based postpartum health surveillance³⁹ may provide useful conceptual frameworks for future adaptation and evaluation.

4.8. Conclusion

This study demonstrates that postpartum clinical care engagement remains low - even among high-risk pregnancies - in a large, national, community-based sample. Key sociodemographic and clinical factors identified highlight priority populations for intervention, while the association of prenatal enrollment and longitudinal participation with greater clinical follow-up highlights the overlap between program engagement and healthcare engagement. These findings point to the value of home visiting as a potential complementary care model to extend early care beyond clinical settings and promote sustained engagement in clinical follow-up.

Supplemental material

Supplemental material - The Role of home visiting in addressing postpartum care gaps: Evidence from a large community sample

Supplemental material for The Role of home visiting in addressing postpartum care gaps: Evidence from a large community sample by Aishwarya Vijay, Allison Kemner, Karen Hoerchler, Abigail Edwards, Jeannie C. Kelly, Rachel Tabak, Mark D. Huffman and Zainab Mahmoud in Women’s Health.

Footnotes

Acknowledgments

The authors acknowledge the Parents as Teachers National Center and participating local affiliates for their partnership and support of this work. We also thank the home visiting staff and program leaders who contribute to data collection and quality improvement efforts that make this research possible.

ORCID iD

Aishwarya Vijay

Ethical considerations

All records were de-identified. This retrospective study was reviewed by the Washington University Institutional Review Board and was deemed exempt from full review.

Consent to participate

The requirement for informed consent was waived due to the use of existing, de-identified data.

Author contributions

AV conceptualized and designed the study, conducted data extraction and analysis, and drafted the manuscript. KH contributed to preliminary data analysis. AE, JK and RT provided critical revision of the manuscript for important intellectual content. AK, ZM and MDH contributed to manuscript writing, supervision, and overall study oversight. All authors reviewed and approved the final manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: AV is supported by T32HL007081. MDH is supported by K24HL175228. ZM is supported by K23HL173684. JCK is supported by R01HD113199, R21DA057493, and R61DA062321. This work was made possible with support from Washington University in St. Louis WU-CDTR (Grant Number P30DK092950 from the NIDDK). The content is solely the responsibility of the authors and does not necessarily represent the official views of the WU-CDTR or NIDDK.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article

Data Availability Statement

The data that support the findings of this study are not publicly available due to data use agreements and protection of participant confidentiality but may be available from the corresponding author upon reasonable request and with permission of Parents as Teachers National Center and Washington University.*

Supplemental material

Supplemental material for this article is available online.

Appendix

References

Hossain

Hussain

Patel

, et al. Trends and Disparities in Cardiovascular-Related Maternal Mortality in the United States: 1999-2022. Journal of the American College of Cardiology 2025; 85: 2455. https://doi.org/10.1016/s0735-1097(25)02939-0

Attanasio

Ranchoff

Cooper

, et al. Postpartum visit attendance in the United States: a systematic review. Women's Health Issues 2022; 32: 369–375. https://doi.org/10.1016/j.whi.2022.02.002

Sukmanee

Liabsuetrakul

. Risk of future cardiovascular diseases in different years postpartum after hypertensive disorders of pregnancy: a systematic review and meta-analysis. Medicine 2022; 101: e29646. https://doi.org/10.1097/MD.0000000000029646

Lui

Jeyaram

Henry

. Postpartum interventions to reduce long-term cardiovascular disease risk in women after hypertensive disorders of pregnancy: a systematic review. Frontiers in cardiovascular medicine 2019; 6: 160. https://doi.org/10.3389/fcvm.2019.00160

Countouris

Mahmoud

Cohen

, et al. Hypertension in pregnancy and postpartum: current standards and opportunities to improve care. Circulation 2025; 151: 490–507. https://doi.org/10.1161/CIRCULATIONAHA.124.073302

Care

. ACOG committee opinion. Obstetrics Gynecol 2019; 134: e84–e89.

Mehta

Sharma

Creanga

, et al. Call to action: maternal health and saving mothers: a policy statement from the American Heart Association. Circulation 2021; 144: e251–e269. https://doi.org/10.1161/CIR.0000000000001000

Cohen

Daw

. Postpartum cliffs—missed opportunities to promote maternal health in the United States. In: JAMA Health Forum. American Medical Association, 2021, pp. e214164.

Rokicki

Oviedo

Perreault

, et al. Home visits and the use of routine and emergency postpartum care among low-income people: a secondary analysis of a randomized clinical trial. JAMA Network Open 2024; 7: e2451605. https://doi.org/10.1001/jamanetworkopen.2024.51605

10.

Lewey

Beckie

Brown

, et al. Opportunities in the postpartum period to reduce cardiovascular disease risk after adverse pregnancy outcomes: a scientific statement from the American Heart Association. Circulation 2024; 149: e330–e346. https://doi.org/10.1161/CIR.0000000000001212

11.

Levine

Nkonde-Price

Limaye

, et al. Factors associated with postpartum follow-up and persistent hypertension among women with severe preeclampsia. Journal of Perinatology 2016; 36: 1079–1082. https://doi.org/10.1038/jp.2016.137

12.

Arsenault

Wright

Taddele

, et al. The maternal and newborn health eCohort to track longitudinal care quality: study protocol and survey development. Global health action 2024; 17: 2392352. https://doi.org/10.1080/16549716.2024.2392352

13.

Tabak

Kandula

Angell

, et al. Implementation of Evidence-Based Behavioral Interventions for Cardiovascular Disease Prevention in Community Settings: A Scientific Statement From the American Heart Association. Circulation 2025; 152: e75–e87. https://doi.org/10.1161/CIR.0000000000001349

14.

Parents as Teachers . Parents as teachers evidence-based home visiting model, 2012.

15.

Von

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. The Lancet 2007; 370: 1453–1457. https://doi.org/10.1016/s0140-6736(07)61602-x

16.

Center PaTN . Supporting prenatal and postpartum families: Release of new resources and content, 2022.

17.

Campos

Kapp

Simoes

. The evidence base for the maternal, infant, and early childhood home visiting program constructs. Public Health Reports 2018; 133: 257–265. https://doi.org/10.1177/0033354918764383

18.

American College of Obstetrics and Gynecology . Gestational hypertension and preeclampsia: ACOG practice bulletin, number 222. Obstet Gynecol 2020; 135: e237–e260.

19.

Hardcastle

Ford

Bellis

. Maternal adverse childhood experiences and their association with preterm birth: secondary analysis of data from universal health visiting. BMC Pregnancy and Childbirth 2022; 22: 129. https://doi.org/10.1186/s12884-022-04454-z

20.

Fabi

Mena

Desir

, et al. Public Coverage of Postpartum Services for Immigrants. In: JAMA Health Forum. American Medical Association, 2025, p. e250702.

21.

Kebede

Wana

Tirore

, et al. Determinants of dropout from the maternal continuum of care in Ethiopia, multilevel analysis of the 2016 demographic and health survey. PLOS Global Public Health 2024; 4: e0003641. https://doi.org/10.1371/journal.pgph.0003641

22.

de Bocanegra

Braughton

Bradsberry

, et al. Racial and ethnic disparities in postpartum care and contraception in California’s Medicaid program. American journal of obstetrics and gynecology 2017; 217: e41–.

23.

Williams

Franco-Rowe

Lopez

, et al. Structural and relational factors for successful cross-sector collaboration in home visiting: a multiple case study. BMC Health Services Research 2024; 24: 316. https://doi.org/10.1186/s12913-024-10719-4

24.

Lewey

Levine

Yang

, et al. Patterns of postpartum ambulatory care follow‐up care among women with hypertensive disorders of pregnancy. Journal of the American Heart Association 2020; 9: e016357. https://doi.org/10.1161/JAHA.120.016357

25.

Danilack

Brousseau

Paulo

, et al. Characteristics of women without a postpartum checkup among PRAMS participants, 2009–2011. Maternal and child health journal 2019; 23: 903–909. https://doi.org/10.1007/s10995-018-02716-x

26.

Missouri Department of Health and Senior Services . A Multi-Year Look At Maternal Mortality: 2018- 2022 Missouri Pregnancy-Associated Mortality Review, 2025.

27.

Quay

Harbert

Schramm

, et al. Maternal Mortality in Missouri: A Comparison of Definitions and Data Sources. Journal of Public Health Management and Practice 2025; 31: 368–371. https://doi.org/10.1097/phh.0000000000002091

28.

Miller

. Nearly 60% of New Mothers Miss Essential Postpartum Follow-Up Care. Global News Wire: Cedar Gate, 2025.

29.

Ahrens

Palmsten

Grantham

, et al. Acute health care utilization in the first 24 months postpartum by rurality and pregnancy complications: A prospective cohort study. Health Services Research 2024; 59: e14247. https://doi.org/10.1111/1475-6773.14247

30.

Sonenberg

Mason

. Maternity care deserts in the US. In: JAMA Health Forum. American Medical Association, 2023, p. e225541.

31.

Heck

Jones

Bohn

, et al. Maternal mortality among American Indian/Alaska Native women: A scoping review. Journal of Women's Health 2021; 30: 220–229.

32.

Rogers

Hogan

Coates

, et al. Responding to the health needs of women from migrant and refugee backgrounds—Models of maternity and postpartum care in high‐income countries: a systematic scoping review. Health & social care in the community 2020; 28: 1343–1365. https://doi.org/10.1111/hsc.12950

33.

Lines

Weech-Maldonado

. The Immigrant Paradox: Health Advantages and Health Barriers Among Foreign-Born Americans. LWW, 2024, pp. 703–705.

34.

Finello

Terteryan

Riewerts

. Home visiting programs: what the primary care clinician should know. Current problems in pediatric and adolescent health care 2016; 46: 101–125. https://doi.org/10.1016/j.cppeds.2015.12.011

35.

Erondu

Dunlop

. Interpregnancy care: an opportunity to improve women's health and reduce the risk of maternal morbidity and mortality. Journal of Public Health Management and Practice 2021; 27: S155–S158. https://doi.org/10.1097/PHH.0000000000001319

36.

Wouk

Morgan

Johnson

, et al. A systematic review of patient-provider-and health system-level predictors of postpartum health care use by people of color and low-income and/or uninsured populations in the United States. Journal of Women's Health 2021; 30: 1127–1159.

37.

Kruk

Sabwa

Lewis

, et al. Population assessment of health system performance in 16 countries. Bulletin of the World Health Organization 2024; 102: 486–497B. https://doi.org/10.2471/BLT.23.291184

38.

Victor

Blyler

, et al. Sustainability of blood pressure reduction in black barbershops. Circulation 2019; 139: 10–19. https://doi.org/10.1161/CIRCULATIONAHA.118.038165

39.

Deshmukh

John

Arora

. Utilization of postnatal healthcare services delivered through home visitation and health facilities for mothers and newborns: an integrative review from developing countries. The Indian Journal of Pediatrics 2020; 87: 207–216. https://doi.org/10.1007/s12098-019-03101-4

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.50 MB