Healthy for Two / Healthy for You: Design and Methods for a Pragmatic Randomized Clinical Trial to Limit Gestational Weight Gain and Prevent Obesity in the Prenatal Care Setting

Abstract

Objective

To compare the effect of Healthy for Two/Healthy for You (H42/H4U), a health coaching program, in prenatal care clinics that serve a racially and economically diverse population, on total gestational weight gain (GWG) (vs. usual care). We hypothesize that compared to usual prental care, intervention participants will have lower GWG and lower rates of gestational diabetes mellitus (GDM).

Methods

We report the rationale and design of a pragmatic, parallel arm randomized clinical trial with 380 pregnant patients < 15 weeks gestation with overweight or obesity from one of 6 academic and community-based obstetrics practices, randomized to either H42/H4U or usual prenatal care in a 1:1 ratio. The study duration is early pregnancy to 6 months postpartum. The primary outcome is total GWG, calculated as the difference between first clinic-assessed pregnancy weight and the weight at 37 weeks gestation. Key maternal and infant secondary outcomes include GDM incidence, weight retention at 6 months postpartum, infant weight, maternal health behaviors and wellness.

Conclusions

This pragmatic clinical trial embeds a pregnancy health coaching program into prenatal care to allow parallel testing compared to usual prenatal care on the outcome of total GWG. The real-world design provides an approach to enhance its sustainability beyond the trial with the goal to ultimately improve maternal/child health outcomes and reduce future obesity.

Trial registration:

The study was first registered at clinicaltrials.gov on 1/26/21 (NCT04724330).

1. Introduction

Despite two decades of public health efforts to combat obesity, rates continue to rise and racial and socioeconomic disparities persist^1,2. There is an urgent need for obesity prevention efforts to target young adults, particularly reproductive age women^3–5. Twenty-three percent of women (vs. 13% of men) gain ≥ 20 kg in their childbeaing years, with the highest weight gain in African American women. Weight gain of ≥ 20 kg between 18 and 55 years is associated with the development of type 2 diabetes (DM) and greater mortality⁵.

Obesity prior to, during and after pregnancy has been implicated in worsening disparities in maternal mortality in the U.S., as obesity-related complications disproproprionately impact African American women.^6–8 Excess gestational weight gain (GWG) and retaining pregnancy weight in the postpartum period are common,^9,10 and associated with long-term obesity in women and their offspring, as well as short-term pregnancy complications, including gestational diabetes (GDM) and preeclampsia^11–14. Because most people seek prenatal care and are motivated to have a healthy baby, pregnancy and the postpartum period provide “teachable moments” to promote healthy behaviors (i.e., improved diet, exercise) to limit GWG, sustain long-term healthy behaviors for their families, and reduce future obesity and DM^15–17.

There is now strong evidence from the US Preventive Services Task Force (USPSTF) that limiting GWG is achievable and associated with improved exercise and dietary behaviors¹⁸. Importantly, the USPSTF recommends that clinicians “offer pregnant persons effective behavioral counseling interventions aimed at promoting healthy weight gain and preventing excess gestational weight gain in pregnancy (B recommendation).”¹⁹ Because previous studies were designed to test the efficacy of behavioral weight management on GWG, the interventions were resource intensive (i.e., in-person counseling and with limited technological components), without a focus on program implementation and sustainability in routine prenatal care settings and lacking intervention focus in the vulnerable postpartum period^18,20. To address this critical evidence gap and meet the USPSTF recommendations, there is a need to integrate and test a practical, evidence-based lifestyle intervention in pregnancy and continuing postpartum.

Our team previously developed and pilot-tested the intervention, Healthy for Two/Healthy for You (H42/H4U), an innovative, evidence-based pregnancy/postpartum health coach intervention that utilizes remotely-delivered phone health coaching, a web-based platform with learning activities and mobile phone behavioral tracking to support obstetric providers and clinics to optimize care for pregnant people at highest risk for obesity and future DM ²¹. The two aims of this paper are to: 1) Describe the design of a pragmatic randomized clinical trial to evaluate the effectiveness of H42/H4U when integrated into prenatal and postpartum care on important maternal and infant outcomes, compared to usual care in a diverse, clinical population; 2) Highlight the features of the trial that enable its implementation into ‘real world’ prenatal clinical practice and workflows, i.e, recruitment by nurses, training of existing clinical staff to deliver health coaching and providing an accessible user-friendly web-based platform for patients to enable sustainable behavior change. Pragmatic clinical trials provide an opportunity to improve the efficiency of clinical trials using data from usual clinical care enounters (e.g., electronic health record [EHR] data) and support ‘real world’ evidence generation that could rapidly lead to its wider implementation and sustainability into clinical practice^22,23. We applied the PRagmatic Explanatory Continuum Indicator Summary-2 (PRECIS-2) tool that has 9 domains to guide trialists, clinican and policymakers to make design decisions that facilitate integration of the trial into care and ultimately promote the healthcare relevance of its results²⁴.

2. Methods

Healthy for 2/Healthy for You (H42/H4U) is a parallel arm, pragmatic randomized controlled trial in early pregnancy through 6 months after delivery (N=380) that tests the effectiveness of the H42/H4U health coaching program integrated into prenatal care compared with a usual care comparison group (maintain Health in Pregnancy; mHIP) among pregnant people enrolled from academic and community-based prenatal clinics in a single health system. The primary outcome is total gestational weight gain (GWG) (calculated as 37-week gestation weight minus baseline pregnancy [≤ 15 week gestation] weight). Maternal secondary outcomes include proportion of participants with excessive GWG, gestational diabetes mellitus (GDM) incidence and postpartum weight retention at 6 months after delivery. Infant secondary outcomes include infant growth, measured by weight and length at birth, 4- and 6-months of age. Other outcomes include changes in maternal health behaviors (diet, physical activity, breastfeeding) and maternal wellness (depression, sleep and stress). The hypothesis is that women in the H42/H4U arm will have lower total GWG compared to usual care.

Approval was obtained fom the Johns Hopkins School of Medicine Institutional Review Board (11/9/2020).

2.2. Study eligibility and setting

We are recruiting and enrolling pregnant patients from 6 academic and community-based prenatal care clinics in and around the City of Baltimore that serve racially and economically diverse patient populations. In addition to including a multidisciplinary team throughout the trial design and implementation process, we conducted interviews and group meetings with clinical practice leaders, providers and staff at all 6 sites to ensure that the design of the trial was responsive to the needs of busy obstetrics practices²⁵.

As a pragmatic trial, we are applying broad eligibility criteria²⁴: 18 years of age or older, currently have a singleton pregnancy ≤ 15 weeks gestation with any parity status and BMI ≥ 25.0 kg/m². We will focus on patients with overweight or obesity because of the high prevalence in our clinical populations and because of prior evidence showing an impact of interventions in this population at high risk for obesity-related chronic disease¹⁸. Due to budgetary constraints to adapt intervention materials, participants must be able to engage in coaching calls in the English language. Exclusion criteria are multiple fetuses, pre-gestational diabetes, poorly controlled blood pressure, active substance use (excluding marijuana use), eating disorder, psychiatric hospitalization in last 12 months or diagnosis of severe mental illness that is not well controlled. Participants are not required to own a smartphone or iPad for the study, as the study is able to loan them to participants in need and the number of loaners are being tracked to guide future implementation efforts.

2.3. Participant recruitment

In close collaboration with clinical staff and providers, we designed the screening and recruitment procedures to emulate referral into a ‘real world’ clinical program, leveraging the “Nurse History Visit,” which most clinics have either through telemedicine or in-person. We designed a “Best Practice Advisory” (BPA) in the Epic electronic health record (EHR) to appear at the time of order entry to enable nurses to refer potentially eligible patients to study staff for further screening (see Figure 1). In addition, through conversations with the nurses at our involved clinics, we tailored referral methods to the workflows of the nurses using the Epic EHR, i.e, using features like Patient Lists and Secure Chat. In addition, we review the list of patients for whom the BPA was alerted but no action was taken to assess eligibility and then sent potentially eligible patients recruitment messages using the patient portal. Post-referral screening includes a brief screening process using the Epic EHR and then a phone call to the patient to confirm interest and eligibility. Following screening, eligible patients are consented electronically using a phone and online consent process.

Figure 1.

Screen shot of the Best Practice Advisory used to prompt clinic-based recruitment

2.4. Randomization and blinding

Because the included clinics’ populations are different and not comparable across sites, randomization is at the individual and not clinic (i.e., cluster) level. Because of the intervention intensity, risk of contamination to the usual care control group was deemed to be low. We designed a web-facilitated virtual randomization experience, with an option for in-person randomization as well. Randomization is stratified by practice and BMI (BMI ≥30.0 kg/m² vs. 25.0–29.9 kg/m²), and within each stratum using randomly varying block sizes. Due to the nature of the lifestyle intervention, participants and health care providers and intervention team cannot be effectively blinded to randomization assignment. Data collectors will be blinded to assignment. Until the end of the trial, all non-intervention team, including study co-investigators and the safety officer will also remain blinded.

2.5. Intervention: Healthy for Two / Healthy for You (H42/H4U)

In addition to our prior formative research to design the intervention²⁶, we are using a continuous quality improvement process. We are conducting mid-point interviews (aiming for 20–25) among racially and economically diverse intervention participants, selected because of their variation in level of platform and coach call engagement (inconsistent to “super” users). We use open ended questions about platform accessibility (including barriers to device or internet connection), ease of use, clarity of content, likeability, relevance and visual appeal (images, colors) and as well as their experience with coaching. We will also carry out 20 exit interviews following completion of the intervention to gain insight into their experiences.

Participants assigned to the H42/H4U intervention receive an approximately 10 month (early pregnancy to 3 months postpartum), remotely-delivered, behavioral lifestyle intervention including health coach phone contacts (with the option for videoconference) and an interactive web-based platform. H42/H4U was designed and pilot tested for feasibility, acceptability and signals of efficacy in clinical settings²¹. One of the most novel features of this trial is that the intervention is delivered “as in normal practice” by clinical staff and with the use of routinely available equipment (i.e., smart phones, work computers and phones)²⁴.

Table 1 describes the components of the H42/H4U intervention. The core intervention components are remote health coaching, a web-based interactive tool for educational activities and tracking of health behaviors using a home scale, provided by the study (EatSmart digital bathroom scale) and the Fitbit mobile application for physical activity and weight tracking.

Table 1.

H42/H4U Description and Approach

COACH Framework	Commitment to self, baby and program: ● Focus on 3 keys to a healthy weight gain goals (healthy eating, exercise, and tracking) ● Completion of learning activities ● Maintain contact with your Coach ● Attend visits with your provider and follow medical advice. Omit/Limit foods and drinks: added sugar, red/processed meats, alcoholic + caffeinated drinks, “unsafe” foods during pregnancy. Add ● Vegetables, fruits, whole grains, nuts, legumes and seafood ● 150 minutes exercise per week (30 min/5 days/week) Communicate with your Coach, OB providers, and support team Honor Your Wellness ● Manage your stress and mood ● Get enough sleep
Program Components
Health coaching	Starts as late as 15-weeks gestation to 3 months postpartum Phone calls using a patient-centered motivational interviewing approach (~20 mins) ● 7 weekly contacts in the 2nd trimester ● 1 contact in the 3rd trimester ● 1 contact postpartum ● Up to 2 optional additional contacts ● ~Optional quarterly topic groups (online) Coach case management meetings - twice monthly
Mobile phone and online behavioral tracking	Fitbit mobile app to track food and beverage intake and moderate-intensity exercise Weekly Weight Tracker with personalized feedback
Interactive online learning and goal setting platform	● Learning activities focusing on nutrition, exercise, behavior change strategies, wellness (e.g., sleep, mood), long-term obesity prevention ● SMART goal setting activity28
Prenatal care and EHR interface	Direct referral into program (Figure 1); EHR progress reports for providers
Obesity and Wellness Behavioral Goals
Pregnancy weight	Within recommended total weight gain based on IOM recommendations48
Postpartum 1	Return to pre-pregnancy weight and continue to lose an additional 0.5 to 2 lb./week until reaching 5% of pre-pregnancy weight lost49,50 Promote breastfeeding51
Physical activity	≥150 min/week moderate-intensity exercise 27
Diet	Diet patterns that are higher in vegetables, fruits, nuts, legumes, and fish and lower in added sugar, and red and processed meat 49,52

^1.Communication around postpartum weight loss is focused on healthy weight loss consistent with current guidelines and not based on a timeframe of reaching a weight loss goal by 6 or 12 month after delivery

We have partnered closely with participating clinic managers to assist with identifying and training of health coaches, existing clinical staff (e.g., nurses, medical assistants, dietitians, midwives), who then complete training and coaching during their usual work hours. Coaches do not receive additional salary support for their coaching. To acknowledge their important contribution to the program, we provide them with certificates (for completion of training and the program), intructions on represent their new role as “Health Coach” on their resume, letters from the PI commending their work for their Human Resources files, “credit” towards research (required by some units in the hospital), as well as additional “cafeteria gift cards” and holiday-time “thank you’s” whenever possible. Each coach receives 6.5 hours of training focused on delivering the behavioral intervention using a motivational interviewing approach, with topics focused on gestational weight gain and postpartum weight loss guidelines, pregnancy and postpartum nutrition, exercise and wellness. Study-supported health coach managers provide support, oversight, quality control and feedback to the coaches, including case management meetings every other week and as needed, individually. We determined the number of coaching contacts based on prior studies, including our own, showing that high (i.e., 12 or more contacts) and moderate (3 to 11 contacts) intensity interventions were more effective for GWG than low intensity interventions^18,21. The total number of planned contacts is 9 (“high-moderate”), with two additional optional “booster” contacts offered for participants who request or need more support. We also offer optional, supplemental, web-based groups led by Coach Managers and experts on topics (e.g., breastfeeding, gestational diabetes, managing holidays, mood/mindfulness). Health coaching contacts are ~ 20 minutes in length and occur weekly for the first 7 weeks in the second trimester, once in the third trimester, and once postpartum.

Figure 2 shows screen shots from the H42/H4U web-based platform, which was designed by our team to deliver educational learning activities timed with coach contacts and relevant to the pregnancy or postpartum phase of the intervention. Participants are offered a scale for home usage. Participants self-report weekly weights from their home scale using an email link called the Weekly Weight Tracker, which can be viewed by coaches (Figure 2). Participants receive guidance on downloading and using the mobile application and are advised, based on recommendations from the American College of Obstetricians and Gynecologists²⁷, to record moderate-intensity aerobic exercise in bouts of 10 minutes or more and track all food and beverages consumed daily. Prenatal care providers receive monthly progress updates via the Epic EHR to review patient progress, provide support and encourage participants to be actively engaged in the H42/H4U lifestyle intervention. The curriculum focuses on the core principles of the COACH model (Table 1) and behavioral strategies for lifestyle management, i.e., self-monitoring/tracking behaviors, controlling portion sizes, increasing exercise, communicating effectively, and problem solving²¹. Goal setting using SMART goals is also a key component of the intervention (Figure 2)²⁸.

Figure 2.

Screen shots of the H42/H4U platform with panels A-C

Participants’ adherence to the intervention is based on completion of coach contacts and web-based learning activities and use of the weekly weight tracker. Participants are periodically asked to allow some of their calls to be audio-recorded for review to ensure the quality and consistency of the health coaching.

2.6. Usual care comparison group: maintain Health in Pregnancy (mHIP)

Women assigned to the comparison group, Maintain Health in Pregnancy (mHIP), will receive typical, guideline-based care in their prenatal care clinic. The current recommended American College of Obstetrics and Gynecologists’ prenatal visit schedule for uncomplicated pregnancies consists of a visit every 4 weeks until 28 weeks, every 2 weeks until 36 weeks, and weekly until delivery.²⁹ “Usual Care” is a rational and important comparator in this study because it is the current standard of care; providing evidence of a meaningful improvement over standard care will demonstrate the effectiveness of the H42/H4U intervention. In the U.S., practice patterns for prenatal care visit frequency and intensity are similar, thus allowing for wide interpretability, applicability, and reproducibility.

2.6. Data Collection and Data Sources

Table 2 shows the measures, time-points and sources of data, described in more detail below. Data collection occurs during 3 time points in pregnancy (baseline, 24 and 37 weeks gestation) and 2 time points postpartum for mother and infant (4 and 6 months after delivery, based on timing of well-infant visits). Consistent with the PRECIS-2 guidance, we have designed data collection to be efficient and unobstrusive, i.e., obtained from EHR data and usual clinical care opportunities whenever possible with study staff validation and verification of key data (maternal height and weight)²⁴. We created a process of continuous data collection, integrating EHR data from prenatal and infant visits into a database to enable both careful tracking for when key data might be missing (e.g., missing weight because a patient missed a visit or changed clinics outside of the health system) and to optimize data cleaning and processing for end of study analyses. We are also verifying calibration of scales in clinics and providing in-clinic trainings for medical assistants on standard methods for obtaining height and weight, especially at the first prenatal visit, which represents our baseline data collection source. Because of concern for inaccurate (or missing) heights in the EHR, our team collects in-person height for verification of the height and weight at one prenatal visit (24 weeks gestation). When weights are missing from the EHR our study staff conduct home visits to obtain measures for all data collection time points (37 weeks gestation; 4 and 6 months postpartum)

Table 2.

Schedule of measures and outcome assessments

Procedures	Screen	BS/RZD	24 wks. gest.	37 wks. gest. (+ birth)	4 mo. pp	6 mo. pp
Maternal Anthropometric Measures
Height	EHR	EHR	EHR + staff assessed	EHR	EHR	EHR
Weight1	EHR	EHR	EHR + staff assessed	EHR	EHR	EHR
Maternal Data	X	X	X	X	X	X
Demographics	EHR	EHR	EHR	EHR	EHR	EHR
Encounter data including vital signs (weight, height, BP, Pulse), diagnoses, problem lists, past history	EHR	EHR	EHR	EHR	EHR	EHR
Laboratory tests	EHR	EHR	EHR	EHR	EHR	EHR
Medications	EHR	EHR	EHR	EHR	EHR	EHR
Labor and delivery course discharge summary	EHR	EHR	EHR	EHR	EHR	EHR
Infant Data
Demographics				EHR	EHR	EHR
Encounter data including vital signs (length, weight, head circumference), diagnoses, problem list)				EHR	EHR	EHR
Laboratory test results				EHR	EHR	EHR
Medications				EHR	EHR	EHR
Newborn, nursery or NICU discharge summary				EHR
Online Questionnaires
Demographics and Family Characteristics		X			X	X
Smoking, Marijuana and Alcohol (PRAMS)53		X
COVID19 pandemic wellness and vaccine54		X				X
Usual source of care (PRAMS)53		X				X
Pregnancy Intention (PRAMS)53		X
Dietary Behaviors (DSQ)55		X	X	X	X	X
Physical activity (IPAQ-SF)56		X	X	X	X	X
Anxiety and Depression (EPDS)57		X	X	X	X	X
Perceived Stress Scale (MSPSS)58		X		X	X
Sleep quality (PSQI)59		X		X	X	X
Breastfeeding practices (CDC-IFPS and PRAMS)53			X	X	X	X
Use of Community and Safety Net services (WIC) 53			X		X	X
Use of Community and Safety Net services (Home Visiting) 53			X		X
Life Events Inventory, modified (SRRS)60		X			X
Experiences of Discrimination (MDR)61						X
Postpartum Visit Attendance and Social Support (PRAMS)53					X
Postpartum contraception (PRAMS)53					X	X
Infant overall health question53					X	X
Safety surveillance survey			X	X	X	X
Study Satisfaction Survey – for Intervention Group						X

Abbreviations: BS=baseline; EHR=electronic health record; mo=month; NICU=neonatal intensive care unit; pp=postpartum; RZD=randomization

2.7. Retention strategies

Both intervention (H42/H4U integrated prenatal and postpartum care) and usual care groups have access to their own separate Facebook pages to promote communication, support and retention in the study. Participants are provided giftcards after each data collection visit: $10 at enrollment; $15 at 24-weeks, $20 at 37-weeks, $25 at 4 months postpartum and $40 at 6 months postpartum.

2.8. Safety Surveillance and monitoring

For active surveillance, we receive alerts in the Epic EHR to enable tracking for all maternal and infant hospitalizations, emergency room (ER) visits and labor and delivery triage evaluations. Participants complete a safety survey about recent hospitalizations or ER visits, problems with exercising or recent injuries. We also have a protocol to alert the team to scores indicating high levels of depressive symptoms based on survey timepoints. The study has an IRB and sponsor-approved Data Safety and Monitoring Plan and an independent Safety Officer to review study progress, intervention adherence and adverse events (mild, moderate, severe).

2.9. Analytic approach

2.9.1. Power and sample size

We based our sample size estimates on the following considerations: 1) 80% or greater follow-up rate for the main outcome of GWG; 2) Range of variance in the primary outcome of GWG; 3) Ability to detect effect sizes comparable to the published literature as well as our pilot data; 4) Two-sided type I α error=0.05; type II β error=0.20.

We expect to enroll at least 380 pregnant patients with overweight or obesity from prenatal care practices. Based on our past experience and review of prior studies, we anticipate 20% loss to follow-up for the 37-week visit, consequential to drop-out for various reasons. With this dropout rate and the assumption that the dropout is consistent with “missing at random”, we expect to retain an effective sample size of 304 participants for our primary outcome, by randomizing N=380 participants, with 190 participants in each arm. Standard deviations (SDs) for the minimum detectable differences (MDDs) evaluation were informed by previous studies of similar combined diet-physical activity behavioral interventions to limit weight gain in pregnancy^18,20,30. The MDDs with 80% power range from 1.45 to 2.32 kg with corresponding SDs for GWG between 4.5 and 7.2 kg, under the assumption of 20% random attrition of the proposed sample size of 380. Based on prior studies’ effect sizes, we believe our sample size calculation is conservative and our trial is sufficiently powered to detect clinically relevant effect on GWG by H42/H4U intervention compared to usual care. For the binary maternal categorical outcomes, we calculated 80% power to detect the proportion who have excess GWG (with minimum detectable odds ratios range from 0.48–0.55) and incident GDM (with minimum detectable odds ratios range from 0.10–0.40). All power calculations were performed using the Power and Sample Size (PASS) software.

2.9.2. Statistical analysis plan

The main analyses for the primary and secondary outcomes will follow the intent-to-treat principle. The primary outcome is GWG. The main analysis will utilize a mixed effects regression approach to model baseline and 37-week weight and assess between group difference in mean of total GWG through inclusion of group indicator, time indicator, and their cross-product interaction term in the mean model. This modeling approach has the flexibility to account for the stratification variables used for stratified randomization, to properly address and utilize the correlation between baseline and 37-week weight measures with participant, and include needed covariates to produce valid inferences in presence of missing 37-week weight outcome under the missing at random assumption. Maternal secondary outcomes include the proportion with excessive GWG, GDM incidence, and postpartum weight retention at 4 and 6 months postpartum, calculated as the difference between weight assessed at 4 and 6 months postpartum and baseline (early pregnancy) weight. The between group difference for the binary secondary outcomes, proportion with excessive GWG and GDM incidence, will be compared using generalized estimated equations (GEE) based logistic regression models. We will also assess the between group difference in mean postpartum weight retention over time after delivery using a similar mixed effects modeling approach as the primary outcome, with extended visit indicators in the mean model to represent the 4 and 6 month data in the postpartum period, and appropriate group by visit indicators cross-product interaction terms to capture the postpartum weight retention between groups at those specific time points, while accounting for needed covariates. We will use an unstructured variance-covariance model to allow full flexibility of outcome variability over time and different pairwise outcome correlations within participant between visits for valid statistical inferences.

Similarly, the between group difference in the infant secondary outcomes of birth weight, weight and length at 4 and 6 months of age will be assessed using seperate mixed effects regression models. To assess infants’ weight gain trajectory from birth to 6 months, for example, we will develop a mixed effects model that includes the H42/H4U intervention arm indicator, 2 visit indicators and 2 visit by intervention cross-product interaction terms, as well as appropriate covariates in the mean model. We will again use an unstructured variance-covariance model to allow full flexibility of outcome variability over time and different pairwise outcome correlations within infant between visits for valid statistical inferences. Missing data will be addressed through proper statistical modeling or multiple imputation (e.g., for the GEE analyses) under the missing at random assumptions as the main analyses, supplemented by sensitivity analyses using plausible missing not at random assumptions to examine the roubustness of the main analyses results.

3. Discussion

The goal of this trial is to apply principles of pragmatic trial design, consistent with the PRECIS-2 tool, to address several notable research gaps that have emerged from the evidence base applying behavioral strategies to limit GWG and reduce postpartum weight retention^20,24,31. Although prior studies show efficacy for impacting maternal and infant outcomes (e.g., gestational diabetes and infant macrosomia), few focused on program implementation and sustainability in prenatal care settings³¹. In addition, the various behavioral interventions were intentionally designed to be intensive, with in-person counseling sessions and even with meal replacements^20,32, but even before the pandemic there has been a growing need for lower touch (i.e. remote delivery) and technology-based (e.g. online and mobile) interventions, adapted to engage diverse populations³³. Finally, we designed H42/H4U to not only follow participants’ outcomes beyond delivery, but additionally to continue intervention into the vulnerable postpartum period when women need emotional support, as well as reinforcement of long-term preventive behaviors to reduce obesity risk³⁴.

Health care systems are utilizing a wide range of “population health” approaches to improve health outcomes and reduce health care costs, often through care management and wellness health coaching ³⁵. Although most health care systems do not yet have a workforce of health coaches, many insurance companies employ health coaches to deliver telephonic coaching programs at a low cost with measurable behavior change outcomes^36,37. In this trial, in addition to testing the effectiveness of H42/H4U, our goal is to demonstrate the value of coaching skills for clinical staff (i.e., nurses, dietitians, medical assistants) and their managers, including the value of training and serving as health coaches to high risk patients. In fact, the value potentially goes beyond the benefits to the patients receiving coaching, as coaches learn strategies for behavior change, healthy diet and exercise that they can apply in their own lives, and have shared personal gratification with getting to know and counsel their patients³⁸. Our team additionally plans to assess the implementation (i.e., costs, barriers, facilitators) of this prenatal care-based health coaching program in order to provide support for health coaching as a vital service we can offer to our patients. The cost analysis will be conducted from the perspective of an adopting organization or payer (e.g., Medicaid) to describe the monetary value for positive outcomes and promote insurance coverage of the program to enhance its future scalability.³⁹

Another unique aspect of this trial is the multiple methods for utilizing the EHR to recruit diverse patients from our clinics, obtain outcomes, monitor data collection completeness and communicate with participants, providers and all clinical staff. For recruitment, we designed alerts through the use of Best Practice Advisories that are triggered when a potentially eligible patient presents for early pregnancy care, and enables nurses to directly refer these patients to the study team for further screening. We have additional “back-up” recruitment strategies using the EHR, including using the patient portal to message patients we might have missed. For obtaining outcomes data, we designed a “real time” constant flow of EHR encounter and labor and delivery information into our secure database, to enable us to run reports, assess quality and determine when key data collection timepoints are missing (e.g., so study staff can call participants and arrange home visits). We designed EHR reports from the intervention team to keep providers updated about which patients are enrolled in the intervention arm, and their progress to provide encouragement, which has been effective in other weight management trials ⁴⁰. We also selected the “Fitbit” mobile tracking application because our health system has plans to integrate its data into the EHR, which will provide infrastructure for our intervention to view and report on these data to coaches.

Like many other studies funded early during the COVID-19 pandemic, we considered the limitations on in-person research in designing our screening procedures for recruitment and randomization, enabling both an in-person and remote process for each. We were fortunate that the H42/H4U intervention had already been intentionally designed to be delivered remotely, in response to prior input from patients about the need to reduce time burden and increase flexibility with coaching sessions. Because our study aims to recruit 50% African American and 40% participants with Medicaid, and also relies on referrals from clinic-based nurses, relationship-building with clinical staff has been a very important consideration when designing our recruitment methods. Because using patient-portal messaging alone could limit outreach to lower income patients ⁴¹ we have been working closely with clinican co-investigators, practice managers and clinical staff to integrate our recruitment processes directly into their workflows. The most challenging aspect of the pandemic for our project has been the shift in priorities, away from research and focused on day-to-day operations, like keeping our healthcare workers safe and healthy and keeping our clinics staffed. In fact, many of our clinics have experienced significant reductions in staff, due to budget cuts, attrition and staff illness, limiting our ability to fully engage with them, and also to recruit, train and retain health coaches, who need to spend time above and beyond their clinical duties coaching participants. In fact, a survey of healthcare workers from the Kaiser Family Foundation and The Washington Post reported that 30% considered no longer working in healthcare and more than 60% reported significant mental health concerns⁴². Ultimately, designing our protocol to have flexibility with virtual and in-person options, and to be responsive to the needs of our clinical partners, will strengthen the potential for the H42/H4U program to ultimately be sustained in our and other complicated health care systems, which are always changing⁴³.

The childbearing years provide a window into our current and future health. However, preventive care opportunities are often missed⁴⁴. To prevent obesity and reduce its intergenerational effects on children, there is a need to integrate and unify our systems of care, from prenatal care to the postpartum/early newborn transition into primary and pediatric care.^{15,34,45 46}H42/H4U provides an example of a needed strategy that engages women early in pregnancy when they are motivated toward a healthy outcome, provides person-centered motivational interviewing throughout pregnancy and into the often-challenging postpartum period and is inclusive and response to the needs of our clinic-based population, with African American, low-income pregnant people who are at highest risk for maternal morbidity. If shown to be effective, the public health implications will be to expand access to H42/H4U into other health systems, as well as public health safety net programs like early home visiting⁴⁷.

Conclusion

It is our goal that this study will advance a potentially powerful, prenatal care-based strategy to reduce obesity in young childbearing adults and thereby the intergenerational effects on their children. Our interdisciplinary team brings together engaged academic and community-based obstetricians and pediatricians, our health system’s population health program and researchers with experience in developing, testing and implementing behavioral interventions in both pregnant and non-pregnant adults in real world settings to enable wide dissemination.