Patient Navigation and Postpartum Health Care Outcomes

Key Points

Question

Does postpartum patient navigation improve postpartum health care quality among people with low incomes?

Findings

In this randomized clinical trial of 405 pregnant people, patient navigation did not result in a higher frequency of achieving all 6 components of postpartum care, but it did improve multiple elements of postpartum care, including retention in postpartum care, the transition to primary care, and the proportion of care components received.

Meaning

The trial found that patient navigation for individuals with low incomes resulted in improvements in multiple aspects of postpartum health care quality in the short and long term.

Abstract

Importance

Postpartum care in the US is fragmented and inadequate, with major disparities in quality. Patient navigation is a promising health services intervention to improve health care provision and reduce disparities.

Objective

To evaluate whether postpartum patient navigation improved postpartum health care quality among people with low incomes.

Design, Setting, and Participants

This randomized clinical trial was conducted at a single urban academic medical center from January 2020 to July 2024 and included English- or Spanish-speaking pregnant people 16 years or older with Medicaid insurance. Data analysis started in October 2024.

Intervention

Participants were randomly assigned to 1 year of patient navigation vs usual care. Those randomized to navigation received individualized services that were designed to overcome postpartum-specific barriers to care from a trained lay navigator.

Main Outcomes and Measures

The primary outcome, as ascertained from medical records, was receipt of 6 care elements considered to be essential to optimal postpartum care by 12 weeks postpartum. Secondary outcomes included the proportion of components received by 12 weeks and receipt of health services at 11 to 13 months.

Results

A total of 405 people (mean [SD] age, 28.2 [5.7] years; 10 [2%] were Asian, multiracial, or another race; 166 [41%] were Hispanic; 202 [50%] were non-Hispanic Black; and 27 [7%] were non-Hispanic White) were randomized (203 [50%] in patient navigation; 202 [50%] received usual care). The frequency of the primary outcome (ie, receipt of all 6 essential components of postpartum care) was similar between the groups (19 [9.4%] in patient navigation vs 16 [7.9%] in usual care; P = .60). However, the mean (SD) proportion of components received was significantly higher among navigation recipients (71% [17%] vs 64% [23%]; P = .002). Three components drove this difference: postpartum visit completion (195 [96%] vs 161 [80%]; P < .001), receipt of all indicated anticipatory guidance (131 [65%] vs 104 [51%]; P = .01), and postpartum depression screening and care (174 [86%] vs 145 [72%]; P < .001). At 11 to 13 months, participants randomized to navigation were more likely to have attended a primary care visit, be using their desired family planning method, and have received depression screening and linkage, cardiometabolic screening, and recommended vaccines.

Conclusions

The results of this randomized clinical trial suggest that although patient navigation did not result in a higher frequency of achieving all 6 components of postpartum care, the proportion of care components received was significantly higher. At 1 year postpartum, those randomized to patient navigation demonstrated multiple improved health services outcomes, including more frequent transition to primary care.

Trial Registration

ClinicalTrials.gov Identifier: NCT03922334

This randomized clinical trial examines whether postpartum patient navigation improved postpartum health care quality among people with low incomes.

Introduction

Engaging with the health care system during the postpartum period, a time of rapid and intense life change, enables individuals to receive important medical, emotional, and social support. Despite the 2018 call from the American College of Obstetricians and Gynecologists for changes in postpartum care provision, postpartum care remains fragmented and inadequate.^1,2,3,4,5 As few as 60% of individuals with Medicaid return for postpartum care,^6,7 and this risk is even greater for individuals with a minoritized racial or ethnic identity due to complex social and structural determinants of health.^4,8,9,10,11 Furthermore, a minority of individuals successfully transition to primary care during the first postpartum year.^{1,9,12,13,14,15,16} Inadequate postpartum health care translates to increased and disparate maternal morbidity and mortality during the early postpartum period and over the life course.^{17,18,19,20,21}

One mechanism to provide comprehensive support for patients with complex medical needs is including individuals who are not traditional health care professionals but who have been trained to help patients overcome barriers to care.^22,23 Specifically, patient navigation is a barrier-focused, longitudinal, and patient-centered intervention that offers support for a defined set of health services.^23,24,25 Navigation, which is designed to focus on the interaction between social needs and coordination of care, has improved health outcomes and reduced disparities in various nonpregnant populations.^{26,27,28,29,30,31,32,33,34} Navigation models have been increasingly studied in obstetrics and benign gynecologic care.^{35,36,37,38,39,40} Although observational studies, including those by our group, have suggested that it is a promising approach to improve postpartum care provision, to our knowledge it has not yet been rigorously tested in a randomized fashion.^22,41,42,43

To address this gap, we developed Navigating New Motherhood 2 (NNM2), a multifaceted program designed to provide individuals with low incomes with navigation services during their first postpartum year. The objective of this trial was to evaluate whether postpartum patient navigation, compared with usual care, improved postpartum health care quality among people with low incomes. Secondary objectives included evaluating whether postpartum patient navigation improved health services outcomes 1 year after delivery. We hypothesized that individuals randomized to patient navigation would have improved postpartum care and more frequent transition to primary care.

Methods

Trial Oversight

NNM2 was a randomized clinical trial that was conducted at a single large, academic, tertiary care center in Chicago, Illinois, between January 2020 and July 2024. The protocol (Supplement 1) was approved by the Northwestern University institutional review board before initiation (the statistical analysis plan can be found in Supplement 2). All participants provided written informed consent before randomization. The institutional review board and sponsor deemed an independent data and safety monitoring committee unnecessary for this low-risk intervention. The authors vouch for the accuracy and completeness of the data and the fidelity of the trial to the protocol. Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines were followed.⁴⁴

Eligibility, Randomization, and Study Visits

Eligibility criteria included being 16 years or older, English- or Spanish-speaking, pregnant (>20 weeks of gestation) or postpartum before hospital discharge, and having Medicaid insurance as the primary payer. Participants were recruited from a practice that provides obstetric and gynecologic care for underserved individuals. Individuals with planned delivery outside of the study site, who were receiving enhanced perinatal HIV case management, who were enrolled in other navigation or postpartum health behavior intervention studies, or who had been enrolled in NNM2 during a prior pregnancy were excluded. After trial initiation, minor protocol adjustments were made to accommodate the SARS-CoV-2 pandemic (eMethods 1 in Supplement 3).

Participants were recruited as outpatients after 30 weeks of gestation or during the postpartum hospitalization. On enrollment, participants completed an in-person study visit, including surveys and anthropometric measurements. Participants were then randomized 1:1 (computer-generated block randomization with random block sizes) to either usual care or patient navigation. Randomization was stratified by antenatal vs postpartum enrollment. Randomization assignment was not masked, as it was an essential part of the intervention to document the availability of navigation in the electronic medical record (EMR). For individuals in patient navigation, clinicians received navigator contact information, and navigators began their intervention by introducing themselves to the participant and care team.

Study visits occurred at 4 to 12 weeks and 11 to 13 months postpartum. The 4- to 12-week study visit was designed to occur concurrently with clinical care when possible. On completing each study visit, participants received gift cards for compensation for their time.

Intervention

The NNM2 conceptual framework incorporated 4 pillars: addressing social determinants of health, enhancing access and entry to care, promoting self-efficacy and health literacy, and sustaining long-term engagement (eMethods in Supplement 3).^22,45 NNM2 used 2 full-time, bilingual lay patient navigators who underwent extensive training in conducting obstetric patient navigation.^41,45 On the same day as randomization, those assigned to the intervention met with the navigator, who provided individualized, intensive navigation services that were designed around postpartum-specific barrier ascertainment and reduction. The intervention continued through 1 year (11-13 months) post partum. Navigators identified social needs and related support to mitigate those needs, coordinated health care appointments, assisted with insurance, provided health education, and served as patient advocates, including support for person-centered decision-making. Navigators maintained detailed logs of their work as well as standardized transition and communication documents that were cocompleted with participants and clinicians.^46,47,48,49

Navigators were embedded in the clinical care team and had full access to the EMR, including scheduling capacity, the ability to communicate via EMR messaging with clinical and administrative staff, and the ability to document in the patient’s record. Navigators directly scheduled postpartum obstetric appointments; for appointments in other departments or facilities, navigators provided scheduling support and coordination but did not have direct scheduling capacity. Navigators prepared participants for transitions from obstetric to primary care, helped to find accessible sources of primary care, provided appointment and social needs support, and provided information about pregnancy events to primary care clinicians to facilitate hand-offs between teams. Navigators attended appointments in person with participants to support educational, advocacy, and coordination needs. Finally, navigators worked closely with participants to prepare them to exit NNM2 by developing a self-efficacy plan that addressed long-term health care needs and engagement.

Outcomes and Sample Size

The primary outcome was the frequency of achieving 6 components of postpartum care that were considered essential to receiving optimal care by 12 weeks post partum, which were based on recommendations from the American College of Obstetricians and Gynecologists.¹ Achieving the primary outcome required achievement of all components, which included (1) completing a comprehensive postpartum visit, (2) receiving all indicated anticipatory guidance, (3) receiving a desired family planning method, (4) undergoing screening for postpartum depression and linkage to care if indicated, (5) initiating and maintaining breastfeeding, and (6) receiving all indicated vaccinations. Each component was also evaluated individually (eMethods 3 in Supplement 3). Attendance at the comprehensive postpartum visit was not required for the achievement of other elements of the outcome, which could have been achieved at other care interactions. As a secondary outcome, we examined the mean proportion of the postpartum care components achieved by participants in each group. We also analyzed group differences in each element of anticipatory guidance.

Another secondary outcome included health services received by 11 to 13 months post partum. The primary outcome during this period was attendance at a primary care visit for preventive health care between 12 weeks and 13 months post partum. Additional outcomes included receipt or continued use of a desired family planning method, maintenance of breastfeeding, and receipt of diabetes testing (among those who had gestational diabetes), cardiometabolic screening, and indicated vaccinations.

Outcomes data were collected via EMR abstraction. For the breastfeeding outcome, if a participant’s reported breastfeeding status at the study visit conflicted with EMR data, participant self-report was used. Abstraction of data from the antenatal, intrapartum, and postpartum periods was completed by trained research staff members who were not involved in care provision or navigation. Records were obtained for all outside medical care received during the study period. Abstractors were unable to be masked to the participant’s assignment due to EMR documentation. However, to reduce potential ascertainment bias, ascertainment of the primary outcome was independently abstracted and confirmed by a second reviewer.

To determine sample size, we estimated that 50% of participants receiving usual care would achieve the primary outcome⁴¹ and hypothesized that patient navigation would result in at least a 20% absolute increase in the primary outcome. Allowing for 30% attrition (α = .05 and 90% power), at least 173 participants per group were needed. Therefore, we planned to enroll at least 400 participants.

Statistical Analysis

Analyses were performed according to the intention-to-treat principle. Baseline characteristics and obstetric outcomes (eTable 1 in Supplement 3) were compared by randomized assignment using bivariable tests. We compared continuous outcomes using the Wilcoxon signed-rank test and categorical variables using the χ² or Fisher exact tests, as appropriate. No interim analysis was performed. For each binary outcome, a log-binomial regression model was constructed to estimate the relative risk and 95% CI. For the continuous outcome (proportion of postpartum care components achieved), we used t tests and a linear regression model to examine its relationship with the intervention; the mean difference and 95% CI are reported. We planned a priori to perform adjustment in the secondary analyses for timing of randomization (antenatal vs postpartum) and, if detected, characteristics with significant randomization imbalances. All analyses were performed with R, version 4.2.1 (R Foundation), and SAS (version 9.4; SAS Institute).

Results

From January 2020 to June 2023, the study team approached 598 potentially eligible participants, of whom 405 (mean [SD] age, 28.2 [5.7] years; 10 [2%] were Asian, multiracial, or another race; 166 [41%] were Hispanic; 202 [50%] were non-Hispanic Black; and 27 [7%] were non-Hispanic White) consented to randomization (203 patient navigation; 202 usual care); participants were followed up through July 2024 (Figure). All participants had records available for outcome abstraction; 385 (95%) completed the 4- to 12-week study visit and 362 (89%) completed the 11- to 13-month study visit.

Figure. CONSORT Diagram.

^aIncludes individuals with multiple exclusion criteria.

A total of 365 participants (90%) were enrolled antenatally and spoke English as a first language (Table 1). Sixty-two of 405 (15%) had a college degree, 195 (48%) had an annual household income less than $25 000, and 151 (37%) were uninsured prepregnancy. The population had multiple comorbidities, with a high prevalence of obesity and chronic hypertension. Baseline characteristics between groups were similar except for parity; 67 individuals (33%) assigned to navigation were nulliparous compared with 93 (46%) of those assigned to usual care.

Table 1. Demographic and Clinical Characteristics^a.

Characteristic	No. (%)
	Usual care (n = 202)	Navigation (n = 203)
Age, y
<19	7 (4)	4 (2)
19-24	61 (30)	55 (27)
25-29	61 (30)	69 (34)
30-34	43 (21)	56 (28)
≥35	30 (15)	19 (9)
Self-reported race and ethnicity
Asian, multiracial, or other	2 (1)	8 (4)
Hispanic, any race	84 (42)	82 (40)
Non-Hispanic Black	99 (49)	103 (51)
Non-Hispanic White	17 (8)	10 (5)
Partnered (married or living with partner)	98 (49)	106 (52)
Working full or part time for pay	73 (36)	80 (39)
English as primary language	191 (95)	194 (96)
Highest level of educational attainment
Less than a high school diploma	16 (8)	18 (9.1)
High school diploma	61 (31)	71 (36)
Some college or associate degree	93 (47)	77 (39)
College or graduate degree	30 (15)	32 (16)
Born in the US	167 (83)	165 (81)
Household income, $
<25 000	94 (47)	101 (50)
25 000-50 000	57 (28)	48 (24)
>50 000	18 (8.9)	23 (11)
Did not know	33 (16)	30 (15)
Uninsured prepregnancy	77 (38)	74 (36)
Nulliparous	93 (46)	67 (33)
Body mass indexb at enrollment
<25	27 (13)	19 (9.4)
25-29.9	46 (23)	43 (21)
30-34.9	63 (31)	58 (29)
35-39.9	32 (16)	35 (17)
≥40	34 (17)	47 (23)
Chronic hypertension	25 (12)	29 (14)
Pregestational diabetes	11 (5.4)	7 (3.4)
Multifetal gestation	4 (2)	8 (3.9)
Initiation of prenatal care after 20 wk of gestation	27 (13)	24 (12)
Enrolled postpartum	20 (9.9)	20 (9.9)
Gestational age at enrollment (of those who enrolled while pregnant), median (IQR), y	35.3 (33.6-37.0)	35.4 (33.3-37.0)

^a All comparisons were nonsignificant (P > .05) except for nulliparous status, which was P = .01.

^b Calculated as weight in kilograms divided by height in meters squared.

Achievement of the primary outcome (Table 2) was substantially lower than anticipated (35 [8.6%] overall) and was similar between groups (19 [9.4%] in patient navigation vs 16 [7.9%] in usual care; P = .60). However, the mean (SD) proportion of postpartum care components achieved was significantly higher among navigation recipients (71% [17%] vs 64% [23%]; P < .001; mean difference, 7.6; 95% CI, 3.7-11.5). Three components drove this difference: postpartum visit completion (195 [96%] vs 161 [80%]; P < .001), receipt of anticipatory guidance (131 [65%] vs 104 [51%]; P = .01), and receipt of postpartum depression screening and linkage to care (174 [86%] vs 145 [72%]; P < .001). Participants randomized to navigation were significantly more likely to have received anticipatory guidance for mental health, social needs, resumption of sexual activity, family planning, sleep management, physical recovery, and health maintenance (Table 3).

Table 2. Postpartum Care Outcomes at 4 to 12 Weeks in Participants Receiving Patient Navigation vs Usual Care.

Outcome	No. (%)		P value	Relative risk or mean difference (95% CI)
	Usual care (n = 202)	Navigation (n = 203)
Primary outcome: optimal postpartum carea	16 (7.9)	19 (9.4)	.60	1.2 (0.6-2.2)
Components of primary outcome
Completed comprehensive postpartum visitb	161 (80)	195 (96)	<.001	1.2 (1.1-1.3)
Received all indicated anticipatory guidancec	104 (51)	131 (65)	.01	1.3 (1.1-1.5)
Received desired family planning methodd	180 (89)	191 (94)	.08	1.1 (1.0-1.1)
Screening for postpartum depression and linkage to care, if indicatede	145 (72)	174 (86)	<.001	1.2 (1.1-1.3)
Initiated and maintained breastfeedingf	119 (60)	125 (62)	.50	1.0 (0.9-1.2)
Received all indicated vaccinations (influenza, Tdap, measles-mumps-rubella, human papillomavirus)g	64 (32)	54 (27)	.30	0.8 (0.6-1.1)
Proportion of postpartum care components achieved, mean (SD), %	64 (23)	71 (17)	<.001	7.6 (3.7-11.5)

Abbreviation, Tdap, tetanus, diphtheria, and acellular pertussis.

^a Optimal postpartum care was defined as all 6 components of the primary outcome achieved.

^b Defined as present when a visit with an obstetric clinician occurred that addressed physical health, mental health, and other postpartum needs within 4 to 12 weeks post partum. Visits solely for wound care, blood pressure checks, or other single-issue topics were not considered comprehensive postpartum visits.

^c Domains included mental health; infant feeding; social needs; resumption of sexual activity; reproductive life planning; prevention of future pregnancy complications; family planning; sleep management; physical recovery; chronic disease management; and health maintenance. See Table 3 for details.

^d Participants were considered to have received their desired family planning method if they had been given a prescription for a method, had a device inserted/implanted, had received male or female permanent contraception, chose a barrier method, or declined a method by 12 weeks post partum. Participants were ineligible for this outcome if they had undergone a hysterectomy or had a same-sex partnership (n = 2).

^e Participants were considered to have been screened if they completed a Patient Health Questionnaire–9 (by 12 weeks post partum; among those with a score of 9 or greater, linkage was considered to have occurred if they received further mental health care).

^f Participants who were exclusively or partially feeding infants with their own milk within the 4- to 12-week postpartum period were considered to be breastfeeding. Participants who did not have custody of their infant for the entirety of the study period or whose infant was deceased were ineligible for this outcome (n = 2).

^g All participants were considered eligible for Tdap or influenza vaccination. Participants were considered eligible for measles, mumps, and rubella vaccination if they were nonimmune to measles or rubella. Participants were considered eligible for human papillomavirus vaccination if they had never initiated or completed the series. Vaccinations were considered received if they occurred during pregnancy or until 12 weeks post partum.

Table 3. Receipt of Indicated Postpartum Anticipatory Guidance Topics by 12 Weeks Post Partum.

Topic	No. (%)		P value
	Usual care (n = 202)	Navigation (n = 203)
Mental healtha	194 (96)	203 (100)	.004
Infant feedingb	195 (99)	202 (100)	.20
Social needsc	147 (73)	183 (90)	<.001
Resumption of sexual activity	171 (85)	198 (98)	<.001
Reproductive life planningd	145 (85)	159 (86)	.50
Prevention of future pregnancy complicationse	38 (69)	44 (68)	.70
Family planningf	195 (97)	203 (100)	.01
Sleep managementg	177 (88)	198 (98)	<.001
Physical recoveryh	174 (86)	196 (97)	<.001
Chronic disease managementi	49 (94)	47 (92)	.90
Health maintenancej	194 (96)	203 (100)	.004

^a Includes mood, smoking cessation, substance use, and mental health conditions.

^b Excludes individuals without custody of an infant or with neonatal death (usual care: n = 197; navigation: n = 202).

^c Includes family/child material needs and social services resources.

^d Includes pregnancy spacing and reproductive life planning. Excludes individuals who received permanent contraception or underwent a hysterectomy (usual care: n = 180; navigation: n = 185).

^e Among individuals who had gestational diabetes, hypertensive disorders of pregnancy, or preterm birth (usual care: n = 62; navigation: n = 65).

^f Includes family planning preferences and options.

^g Includes sleep, fatigue, and social support regarding sleep.

^h Includes physical activity and recovery from birth.

ⁱ Includes management of chronic diseases and prevention of morbidity among those with chronic conditions (usual care: n = 54; navigation: n = 51).

^j Includes health maintenance and vaccinations.

At 11 to 13 months, participants randomized to navigation had a greater likelihood of receiving multiple health services. Specifically, they were more likely to have attended a primary care appointment (115 [57%] vs 60 [30%]; P < .001; relative risk, 1.9; 95% CI, 1.5-2.4), be using their desired family planning method (118 [65%] vs 77 [43%]; P < .001), and have had depression screening/linkage (76 [37%] vs 43 [21%]; P < .001). Those who had gestational diabetes were more likely to have had any diabetes screening (15 [94%] vs 6 [40%]; P = .002). Individuals in navigation were also more likely to have had cardiovascular screening, completed HPV vaccination, and received a COVID-19 vaccination (Table 4). Adjusted models accounting for the timing of randomization and nulliparity demonstrated similar findings (eTable 2 in Supplement 3).

Table 4. Postpartum Outcomes at 11 to 13 Months in Participants Receiving Navigation vs Usual Care.

Outcome	No. (%)		P value	Relative risk (95% CI)
	Usual care (n = 202)	Navigation (n = 203)
Attended primary care visit	60 (30)	115 (57)	<.001	1.9 (1.5-2.4)
Received or continued use of chosen family planning method	77 (43)	118 (65)	<.001	1.5 (1.2-1.8)
Depression screening and linkage	43 (21)	76 (37)	<.001	1.8 (1.3-2.4)
Maintained breastfeedinga	27 (26)	29 (26)	>.90	1.0 (0.6-1.5)
Any diabetes screening (OGTT or other) among patients with GDb	6 (40)	15 (94)	.002	2.3 (1.2-4.4)
Cardiometabolic screening
Blood pressure evaluation	125 (62)	172 (85)	<.001	1.4 (1.2-1.5)
Lipids	26 (13)	57 (28)	<.001	2.2 (1.4-3.3)
Hemoglobin A1c or OGTT	33 (16)	63 (31)	<.001	1.9 (1.3-2.8)
Receipt of indicated vaccines
Influenza	9 (5)	15 (7)	.20	1.7 (0.7-3.7)
Tdap	5 (3)	6 (3)	.80	1.2 (0.4-3.9)
Human papillomavirus series	4 (3)	21 (25)	<.001	4.8 (1.7-13.4)
SARS-CoV-2 initial or booster	15 (7)	37 (18)	.001	2.5 (1.4-4.3)
Other (pneumococcal, MMR, varicella, other)	39 (83)	41 (82)	.90	1.0 (0.8-1.2)

Abbreviations: GD, gestational diabetes; MMR, measles, mumps, and rubella; OGTT, oral glucose tolerance test; Tdap, tetanus, diphtheria, and acellular pertussis.

^a Included participants who self-reported or had a medical record review that indicated sustained breastfeeding during the 11- to 13-month postpartum period.

^b GD: n = 15 for usual care, n = 16 for navigation.

Discussion

This randomized clinical trial of postpartum navigation vs routine care for individuals with low incomes found that the frequency of achieving all 6 essential components of optimal postpartum care was uncommon for all participants. Accordingly, while the point estimate for receiving all components was higher in the navigation group compared with the usual care group, this study was underpowered to detect our hypothesized difference in the primary outcome. In contrast, patient navigation improved multiple individual components of postpartum care, including attendance at a comprehensive postpartum care visit, receipt of all key elements of anticipatory guidance, and screening for mood disorders. Navigation improved the transition to primary care, with a greater likelihood of not only making but keeping a primary care appointment, as well as multiple health service outcomes at 1 year post partum.

Results in Context

The maternal morbidity crisis in the US, which disproportionately affects patients with low incomes with medical or social complexity, demands novel strategies to reverse extant trends and improve outcomes. Patient navigation is one strategy that may reduce gaps in care availability, accessibility, acceptability, and quality.^22,23 A major component of NNM2 was individualized attention to social needs. Recent studies of patient navigation in other clinical settings have demonstrated similar results from comprehensive wraparound interventions, although programs were typically short term, were rarely studied in a randomized fashion, and had services that varied widely. For example, Brown et al⁵⁰ retrospectively evaluated an urban 1-month postpartum navigation program and identified a reduction in postpartum hospitalizations. A trial of 150 Black women similarly found that postpartum visit attendance was improved among those randomized to navigation and a behavioral incentive intervention through 12 weeks postpartum.⁵¹ Formative work in other populations has identified the potential benefit of navigator services among postpartum refugees, postpartum individuals with hepatitis B, and individuals in need of perinatal mental health care.^{42,52,53,54,55,56} In contrast, in an observational study by Hernandez et al,⁵⁷ individuals offered a navigator were not more likely to attend a postpartum visit or engage in remote hypertension monitoring than a historic cohort. Collectively, our findings and those of others suggest that the optimal features of interventions include longitudinal relationships with patients, tailored service delivery, and attention to social needs alongside health system navigation. Future efforts are required to understand which elements of a patient navigation bundle may be most impactful, scalable, cost-effective, and acceptable to patients and clinicians.

Pregnancy and postpartum care are often thought to be a window of opportunity through which to engage and connect people with clinicians who can serve their long-term health needs, yet many social and structural barriers exist to these transitions of care.¹⁴ Interventions such as patient navigation, virtual curricula, text messaging, personalized care plans, and behavioral nudges all have been proposed as potential strategies to bridge these gaps in care.⁵⁸ A notable finding in NNM2 was that individuals randomized to navigation were significantly more likely to receive primary care after pregnancy. Our sample included many patients who were uninsured before pregnancy and had multiple comorbidities; therefore, the 27% higher primary care follow-up among participants in navigation represents a potentially major benefit to long-term health. In the navigation group, most engaged in primary care within the first postpartum year, nearly all with gestational diabetes completed screening for diabetes, and rates of cardiometabolic screening increased. These findings suggest that navigation is an effective strategy to engage people in preventive health care during the postpartum period.

Program Implementation, Limitations, and Future Directions

The high recruitment rate in NNM2 suggests that the proposed intervention was highly feasible and acceptable to potential participants; this high degree of acceptability is a key finding for future dissemination and implementation of the intervention. Another key aspect of this intervention was integration of navigators into the obstetric team.^46,59 By training with the clinical team and working side by side with clinical and administrative staff, navigators were welcomed as part of the patient’s clinical team.^45,60 Key to the successful implementation of this study was the navigators’ access to the EMR, which allowed them not only to remain up to date with clinical events but also to directly schedule patients and communicate with the care team. This feature distinguished them from non–health system–embedded ancillary staff, such as community health workers or home visitors, who often do not have such systems awareness or access.

Although many outcomes were improved with navigation, several outcomes did not differ based on receipt of navigation. Receipt of one’s desired family planning method was frequent and similar between groups. However, in health care settings with less comprehensive family planning services, navigation may serve as a beneficial intervention, as has been suggested in family planning and adolescent health studies.^61,62 While navigators were trained to support breastfeeding education and link participants to experts as needed, there was no difference in breastfeeding frequency. Interventions specifically designed to address breastfeeding may have a greater effect on infant feeding choices, particularly in this population with substantial social barriers to lactation.^63,64

Navigation may be necessary during the antenatal period to improve outcomes when decisions about health behaviors are made earlier. A combined perinatal navigation approach to assisting complex patients has the potential to affect maternal and neonatal morbidity, as well as the patient experience. In an ongoing trial (NCT06941974), whose design was informed by the implementation science findings of NNM2,^{46,47,48,49,60} our team is evaluating whether antenatal navigation improves health outcomes. This study, together with the current results, has the potential to define how low-cost, high-touch health-services interventions, such as patient navigation, may be used to optimize maternal health. Future work should continue to explore the best approaches to implementation, including the cost-effectiveness of different models,⁶⁵ tailoring of navigation to additional settings, and adaptation of navigation to other medically or socially complex populations.^39,52,66,67 Future work must also consider how to scale navigation for different levels of social needs and delivery types.

Strengths and Limitations

This study is one of few that has evaluated rigorously designed postpartum patient navigation in a randomized fashion for a diverse population.^46,47,59 The study implemented a primary outcome measure designed to assess the quality of postpartum care, and the findings revealed opportunities for improvement in assessing postpartum care in future work. The study team adjusted protocols and navigation practices to accommodate the COVID-19 pandemic pragmatically and worked closely with an implementation advisory board.⁶⁰ However, there were limitations. The study was underpowered to assess differences in the primary outcome due to the very low frequency of achieving all 6 components; although many components of postpartum care improved with navigation, ultimately it was uncommon for individuals to achieve all components of the composite, which may reflect an ideal but rarely obtained outcome. This unexpected finding speaks to the challenges of providing comprehensive postpartum care even in well-resourced settings; it additionally has implications for the need to disaggregate future quality metrics when evaluating the quality of postpartum care. Although a strength of the study was the use of the EMR to abstract outcomes, it is possible that not all elements of anticipatory guidance were documented, which would bias findings toward the null. The study was conducted in a single academic medical center in an urban Midwest location; the findings may not generalize to other settings. Inherent in any patient navigation study is the limitation that the intervention is highly context dependent; therefore, the intervention may be challenging to replicate broadly. However, our study setting reflected the realities of many large, urban academic medical centers that serve patients with low incomes. Further, this study launched shortly before the onset of the COVID-19 pandemic, and pandemic-related changes in health care may have contributed to the null primary outcome. Finally, the study was designed to evaluate the effect of a scalable, practice-based program but not other changes, such as home visits,⁶⁸ integrated maternal-newborn care,⁶⁹ or changes in clinical care protocols, nor was it designed to measure differences in rare outcomes, such as severe maternal morbidity and mortality.

Conclusions

In this randomized clinical trial, the NNM2 intervention was designed to integrate a person-centered mindset, attention to shared decision-making, health systems–level expertise, and awareness of community resources. Patient navigation for individuals with low incomes resulted in improvements in multiple aspects of postpartum health care quality in the short and long term. This study demonstrates the potential of patient navigation to affect change in postpartum care and be an effective, sustainable, and scalable strategy to address the maternal health crisis.

Supplement 1.

Trial protocol

Supplement 2.

Statistical analysis plan

Supplement 3.

eMethods 1. Trial modifications for SARS-CoV-2 pandemic

eMethods 2. Details of Navigating New Motherhood 2 intervention

eMethods 3. Primary outcome definitions

eTable 1. Obstetric outcomes of Navigating New Motherhood 2 participants

eTable 2. Adjusted relative risk of primary outcome and components

eReferences

Supplement 4.

Data sharing statement