Protocol for the economic evaluation of the “Healthy for Two/Healthy for You” pragmatic lifestyle intervention in prenatal care to reduce gestational weight gain and gestational diabetes mellitus

Abstract

Background:

Despite policymakers’ demand for evidence of value and sustainability for scaling and implementing interventions, few studies have assessed the cost-efficiency and financial viability of lifestyle interventions for managing gestational weight gain and reducing gestational diabetes.

Objectives:

To describe our methodological approach for assessing the cost-effectiveness and return-on-investment (ROI) of the pragmatic (i.e., integrated into real-world prenatal care) Healthy for Two / Healthy for You (H42/H4U) lifestyle intervention versus the usual prenatal care comparison arm, using a trial-based health economic evaluation.

Methods:

Data from the 36-month randomized clinical trial and electronic health records will be used to estimate the intervention’s effectiveness (difference in total gestational weight gain [lbs]) and its associated incremental costs. ROI and cost-effectiveness analyses will be conducted from a U.S. payer’s perspective. Costs will include recruitment, intervention delivery via health coaches and an online platform, and downstream expenses. ROI and incremental cost-effectiveness ratio (ICER) will assess financial viability and cost-efficiency, respectively. Sensitivity analyses will assess the impact of uncertainties in effectiveness and costs on the ROI and ICER.

Discussion:

Our comprehensive methodological approach for evaluating the cost-effectiveness and financial viability of the pragmatic H42/H4U lifestyle intervention integrated into prenatal care aims to address gaps in health economic evidence and support future investment decisions. Ultimately the goal is to promote the successful implementation and uptake of lifestyle interventions for improving maternal health, preventing obesity and chronic disease.

Trial registration:

The clinical trial associated with this protocol paper is registered on ClinicalTrials.gov (NCT04724330). First posted on January 26, 2021.

BACKGROUND

The United States Preventive Services Task Force (USPSTF) recommends counseling for healthy weight gain in pregnancy, supported by evidence that lifestyle interventions improve diet and maternal health outcomes.¹ A review of 99 trials with 34,546 pregnant participants showed that antenatal lifestyle interventions reduced gestational weight gain (GWG), gestational diabetes, and preeclampsia.¹ Individual dietary interventions, especially at moderate intensity (6–20 sessions), were most effective in reducing GWG.² However, few evidence-based lifestyle interventions have been incorporated into prenatal care, in part due to limited data on cost-effectiveness, particularly in the U.S.

With rising financial pressures, U.S. policymakers and payers are increasingly expecting evidence of cost-efficiency and return on investment (ROI)³ to support scaling and translating research into practice.⁴ The Department of Health and Human Services encourages healthcare organizations to implement interventions targeting social determinants of health, including those that promote access to and eating nutritional foods, to enhance efficiency.⁵ Yet, few lifestyle interventions for controlling excessive GWG and reducing adverse maternal outcomes have undergone rigorous economic evaluation. Systematic reviews reveal inconsistent quality in economic evaluations for behavioral health interventions, with few studies assessing implementation costs.^6–8 For example, a 2007 review found only 6 of 30 studies measured implementation costs,⁶ and a 2019 review showed just 27% included cost data.⁸ A 2021 review found no studies addressed resource use for routine practice implementation in antenatal interventions,⁷ underscoring the need for thorough economic evaluations of implementation strategies.

HEALTHY FOR TWO / HEALTHY FOR YOU (H42/H4U) PRAGMATIC TRIAL: DESIGN, AND CHALLENGES IN ECONOMIC EVALUATION

Overview and design of the Healthy for H42/H4U intervention

The H42/H4U lifestyle intervention combines personalized health coaching, an online educational program, self-monitoring, and behavioral tracking to help pregnant individuals with pre-pregnancy overweight or obesity to meet Institute of Medicine gestational weight gain targets.^9,10 The intervention is being tested in a parallel-arm pragmatic trial with 380 participants, and aims to integrate evidence-based lifestyle interventions into routine prenatal care.¹⁰ The control group receives standard guideline-based prenatal care. Primary outcomes include total gestational weight gain, with secondary outcomes of gestational diabetes, postpartum weight retention at 6 months, and infant birth weight. Participants are enrolled from nine clinics in Maryland’s multi-specialty health system.

H42/H4U intervention components

Details on the design of the H42/H4U intervention have been published previously.¹⁰ Its design was similar to other theory-based remotely delivered weight loss interventions.¹¹ Key pragmatic features include training embedded clinical staff as health coaches, using the electronic health record (EHR) for program referrals and communication with providers, and integrating real-time EHR data into the study database. The behavioral goals focus on: 1) pregnancy weight gain within IOM-recommended limits;⁹ 2) postpartum weight reduction (return to pre-pregnancy weight and continue losing 0.5–2 lbs/week to achieve a 5% loss)^12,13 and breastfeeding promotion;¹⁴ 3) physical activity (≥150 min/week of moderate exercise);¹⁵ and 4) diet patterns linked to lower excessive weight gain, emphasizing vegetables, fruits, nuts, legumes, fish, and limiting added sugars, red, and processed meats.^12,16

Health coaching spans 10 months (early pregnancy to 3 months postpartum) and includes 20-minute, person-centered, telephone-based sessions led by trained clinic staff (nurses, dietitians, medical assistants). Coaches provide seven weekly calls in the second trimester, one in the third trimester, and one postpartum, with twice-monthly case management supervision by coach managers. Table 1 categorizes coaching activities for the economic evaluation into scheduled contacts, noncontacts, and unscheduled contacts, aligning with categories used in other trials employing health coaches as interventionist.¹⁷ The interactive online program covers nutrition, exercise, behavior change, wellness (e.g., sleep, mood), long-term obesity prevention, and Specific, Measureable, Acheiveable, Realistic, Time-sensitive (SMART) goal setting.¹⁸ An EHR interface with prenatal care practices enables program referral by nurses and providers and delivery of progress reports for providers. In the self-weighing and behavioral tracking component, participants track weekly weight using a web-based tracker and log food, beverages, and exercise through a Fitbit app.

Table 1.

Summary of Coaching Activities, By Contact Category and Activity Type.

Contact Category	Definition of Activity	H42/H4U Context
Scheduled contacts	• Direct coach-participant interaction, as aligned with the intervention procedures • Contacts occurred over phone or videoconferencing	• 9 remote contacts during pregnancy • Optional “boosters” • Optional Zoom group meetings
Scheduled noncontacts	• Without direct participant interaction • Involved health coach documentation and regular case management meetings	• Health coaches preparing for and documenting coach contacts afterwards • Two health coach managers and the intervention director met bi-weekly with health coaches for case management sessions to review participants’ weight changes, contact and web-based learning activity completion and coach identified concerns
Nonscheduled contacts	• Direct coach-participant interaction on an as-needed schedule	• “Booster contacts”: Up to two per participant for additional support or if missed regular contacts • Re-engagement phone calls/text messages for participants who miss contacts or without weight entry for more than two weeks
Nonscheduled noncontacts	• Any unscheduled activities not involving participant interaction	• Administrative duties by health coach managers: oversight of health coaches, quality control, recruitment of health coaches, troubleshooting technology issues, preparation of reports

The study was approved by the Johns Hopkins University School of Medicine Institutional Review Board (IRB00247895).

Challenges in economic evaluation

We designed an economic evaluation of the H42/H4U intervention to addresses key limitations of previous studies that mostly used pre-post designs, which are susceptible to regression to the mean.⁵ Although several prior RCTs assessed behavioral interventions’ effectiveness, few collected direct cost data.¹⁹ Instead, costs were estimated from pre-post models, overlooking the correlation between cost and effectiveness. This latter approach could lead to regression to the mean, where extreme initial results drift toward the average on re-measurement, especially in non-random samples, increasing error in cost estimates.^20,21

The pragmatic design features of the H42/H4U trial also introduce four key challenges for economic evaluation.²² First, real-world implementation of lifestyle interventions, such as those targeting outcomes of GWG, infant birth weight, or GDM incidence, presents measurement and timing challenges for height and weight data collection used in effectiveness analyses (challenge 1). Second, accurately valuing the time spent by health coaches embedded in care requires tracking and categorizing their activities across participants and sites, which can be inconsistent in pragmatic trials with variation across sites (challenge 2). Third, embedding the trial in routine care complicates attribution of costs and outcomes, as care delivery and resource use may also vary across settings, potentially displacing outcomes not evaluated in the study (challenge 3). Fourth, reliance on EHR data for cost measurement and outcome ascertainment introduces concerns about missing data, correlated costs and effects, baseline imbalances, skewed cost distributions, and variable adherence. EHR charges may not reflect actual costs, complicating accurate measurement of costs and outcomes (challenge 4).²³ These issues, while addressed in some hybrid-effectiveness studies, may not fully apply to pragmatic trial-based economic evaluations where costs and effects are jointly analyzed.

Contribution of the H42/H4U economic evaluation

This article describes a transparent approach to evaluate potential tradeoffs between the costs and health benefits of H42/H4U intervention. First, we outline our planned methodology for estimating the costs and value-for-money of the intervention, using cost-effectiveness analysis (CEA) and an ROI analysis to assess its financial viability. Next, we discuss how our approach mitigates the four aforementioned challenges (see Challenges in economic evaluation above) related to the value assessment of lifestyle interventions in the context of pragmatic trials. Our ultimate goal is to generate the needed evidence about the value of H42/H4U and similar pragmatic lifestyle interventions to improve decision-making by payers and health systems. With Medicaid covering 40% of U.S. pregnancies,²⁴ and extending postpartum coverage to 12 months in 41 states,^25,26 addressing literature gaps in resource use measurement and valuation is critical to scaling evidence-based lifestyle interventions during pregnancy and postpartum.

METHODS FOR HEALTH ECONOMIC EVALUATION

Overview of health economic evaluation of H42/H4U

We will assess the cost-effectiveness and ROI of H42/H4U versus usual prenatal care, evaluating its efficiency in reducing GWG and improving maternal health outcomes. ROI analysis will estimate financial returns to payers or adopting organizations to guide policy. Embedded within the trial, these analyses leverage its pragmatic, randomized design to control for confounding and account for the correlation between costs and outcomes, ²⁷ with costs assessed from payer and organizational perspectives.

All analyses will be conducted according to the intent-to-treat (ITT) principle, complemented by relevant subgroup analyses, when permissible.²⁸ Findings will be reported in accordance with existing cost and economic analysis reporting guidelines,^29–32 and the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) guidelines (Table 2).³²

Table 2.

Summary of activities, and time measurement units used in the cost calculations.

Activities	Cost Category	Measures and Units
Health coaches
Scheduled contacts • 9 remote contacts • Optional “boosters” • Optional Zoom group meetings	Intervention cost (direct nonmedical)	• Expected number of contacts per participant, n • Expected duration per contact, min (mean, SD)
Non-scheduled contacts • Reengagement (emails and text) • Technical assistance	Intervention cost (direct nonmedical)	• Average number of reengagement interactions (emails, texts) and technical assistance, n • Average time spent on these activities, min (mean, SD)
Scheduled noncontacts • Training  ○ Preparation  ○ Documentation	Implementation cost (direct nonmedical)	• Number of training sessions per health coach, n • Duration per training session, min (mean, SD)
• Case management  ○ Preparation  ○ Documentation	Intervention cost (direct nonmedical)	• Number of case managements per health coach, n • Duration per case management per health coach, min (mean, SD)
Non-scheduled noncontacts • Emails and text to coach managers outside of designated time	Implementation cost (direct nonmedical)	• Number of study participants, n • Average time spent on these activities per study participant, min (mean, SD)
Coach managers
Scheduled Contacts  Optional Zoom group meetings	Intervention cost (direct nonmedical)	• Expected number of contacts per participant, n • Expected duration per contact, min (mean, SD)
Scheduled noncontacts
Training of health coaches	Implementation cost (direct nonmedical)	• Number of sessions, n • Average duration of sessions, min (mean, SD)
Case management	Implementation cost (direct nonmedical)	• Number of sessions, n • Duration per session, min (mean, SD)
Epic provider reports	Implementation cost (direct nonmedical)	• Number of reports per participant, n (mean, SD) • Duration of preparation of each report, min (mean, SD)
Non-scheduled noncontacts Emails and text to health coaches outside of designated time	Implementation cost (direct nonmedical)	• Number of study participants, n • Average time spent on these activities per study participant, min (mean, SD)
Providers
• Additional visits	Intervention cost (direct medical)	EHR
Participants
• Healthcare utilization	Intervention and downstream costs (direct medical)	EHR
Web-based platform
• Integration into EHR • Maintenance	Implementation costs (direct medical)	Negligible

Abbreviations: n, number; SD, standard deviation; min, minutes; EHR, electronic health record.

Cost-effectiveness analysis (CEA) and rationale for the willingness-to-pay threshold

Cost-effectiveness will be assessed using the incremental cost-effectiveness ratio ($\text{ICER}$):

$$\mathrm{I}\mathrm{C}\mathrm{E}\mathrm{R}=\left(\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{D}\mathrm{i}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}\right)/\left(\mathrm{G}\mathrm{W}\mathrm{G}\mathrm{D}\mathrm{i}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}\right)$$ (1)

The $\text{ICER}$ will be compared to a willingness-to-pay (WTP) threshold to determine whether the intervention would be considered cost-effective, meaning, whether it generates sufficient incremental health benefits (measured in GWG difference) per additional dollar of healthcare and implementation resources used.

The WTP measures the value, i.e., “opportunity cost” of each unit of effectiveness outcome to the payer. As there is no universally agreed upon WTP threshold when the effectiveness outcome is measured in “natural units,” such as gestational weight difference, we will use ICER values from prior studies and existing interventions. For example, in a weight loss study that also used health coaches to deliver the intervention, the ICER values were estimated at $138/kg and $171/kg of weight loss per year (2014 US dollars) for in-person vs remote support, respectively.¹⁷ In other studies of behavioral interventions to promote weight loss, the range was $238–284/kg of weight loss per year,³³ and $150 to $262/kg of weight loss over 12 and 52 weeks, respectively.³⁴ Because no economic studies have produced WTP estimates for gestational weight gain, we used the evidence from these weight loss studies and assume a WTP of $250/kg of weight difference over a comparable period (~6 months gestation) in our primary analysis. An ICER below this threshold will be interpreted as indicative of “good value for money.”

The ICER metric can be challenging to interpret without the CEA plane, a graph that depicts the incremental costs and health outcomes of interventions.³⁵ For instance, a positive ICER could result from either negative costs and negative GWG differences, or from positive costs and positive GWG differences. In both cases, the WTP threshold guides decision-making, but in the latter, the intervention provides worse health benefits compared to standard care, despite being less costly. Similarly, a negative ICER could arise from positive costs and negative GWG differences, or vice versa. In the first case, the intervention is dominated by standard care, as it delivers worse health outcomes at higher costs, while in the second, it offers better outcomes at lower costs, indicating good value. There are also statistical challenges with ICER, such as constructing confidence intervals without methods like Monte Carlo sampling or bootstrapping, due to the unknown limiting distribution of the ratio of two random variables.³⁶ To address these concerns about the $\text{ICER}$, we will calculate the incremental net monetary benefit ($\text{INMB}$) and incremental net health benefit ($\text{INHB}$) of the active intervention relative to usual prenatal care, in addition to calculating the $\text{ICER}$.³⁶ The $\text{INMB}$ reflects the monetary value of the incremental health effects, adjusted for the incremental costs:

$$\mathrm{I}\mathrm{N}\mathrm{M}\mathrm{B}=\left[\right(\mathrm{G}\mathrm{W}\mathrm{G}\mathrm{D}\mathrm{i}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e})\times \mathrm{W}\mathrm{T}\mathrm{P}]–\left(\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{D}\mathrm{i}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}\right).$$ (2)

Similarly, the $\text{INHB}$ represents the delta between the incremental health effects and the incremental health opportunity costs³⁶:

$$\mathrm{I}\mathrm{N}\mathrm{H}\mathrm{B}=\left(\mathrm{G}\mathrm{W}\mathrm{G}\mathrm{D}\mathrm{i}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}\right)–\left[\right(\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{D}\mathrm{i}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e})/\mathrm{W}\mathrm{T}\mathrm{P}].$$ (3)

Both measures have desirable statistical properties lacking in the $\text{ICER}$ metric, as noted above.^36–38

Return-on-investment (ROI) analysis

We will calculate the ROI of H42/H4U, from the perspective of an adopting organization or payer that is in the position to implement the H42/H4U intervention and offer it to patients to demonstrate its direct financial return and inform decisions on future outcome improvement actions.¹⁰ The ROI is calculated as the ratio of the intervention’s net benefits to its initial investment costs:

$$\mathrm{R}\mathrm{O}\mathrm{I}=100\times \left(\mathrm{I}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{’}\mathrm{s}\mathrm{N}\mathrm{e}\mathrm{t}\mathrm{B}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{f}\mathrm{i}\mathrm{t}\mathrm{s}\right)/\left(\mathrm{I}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{s}\right),$$ (4)

where the net benefits are defined as the difference between the monetary values of positive outcomes from the intervention and the initial investment costs, expressed in net present value. We will also determine how long it will take, or how large the health benefits of H42/H4U will need to be, on average, for the intervention’s returns to offset its upfront and ongoing implementation costs.

Cost types and consideration of alternative perspectives

Costs and maternal and infant outcomes will be measured over 36 months (the trial duration) and reported in 2024 U.S. dollars. In primary analyses, both costs and outcomes will be discounted at 3% annually, reflecting the time cost of money and health benefits.^29,31 Annual discount rates will be varied between 0% and 5% in sensitivity analyses. Costs will be calculated from the perspective of a payer or adopting organization in the primary analysis, and from the U.S. healthcare sector in secondary analyses, using average unit costs of medical care. A societal perspective is not suitable for this trial-based economic evaluation due to the short follow-up period and challenges in accounting for societal spillover effects. We will calculate the incremental costs (positive or negative) of H42/H4U relative to standard care by subtracting the total costs for the control group from those for the intervention group, focusing on implementation, intervention, and downstream costs (Figure 1).³⁹ Costs beyond the trial period or related to research activities will be excluded. Total costs will be the sum of these categories.

Figure 1.

• Implementation costs.

These are costs related to implementing the intervention in clinical care, including administrative expenses (e.g., office space, utilities, staff salaries), training (e.g., training staff on protocols), and supplies and equipment.^29,40 Costs for program development, such as creating intervention materials and retention incentives, are excluded, as adopting organizations will adopt the program as a ready-to-use “off-the-shelf” intervention without needing to recreate these resources.

• Intervention costs and downstream costs.

This study will assess two types of intervention-related costs: intervention and downstream costs. Intervention costs are those incurred directly from implementing the intervention, including formal healthcare costs, such as direct medical expenses, as well as informal costs like transportation for appointments. Downstream costs are defined as direct medical expenses incurred by participants after the intervention period but during trial follow-up (from randomization to 6 months postpartum). Indirect, nonmedical, and intangible costs^41,42 willbe excluded from this analysis, given the chosen perspective.

Measurement of effectiveness and benefits

Intervention effectiveness and benefits will be measured in the natural units of the primary outcome measure of GWG (in pounds [lbs] as it is most interpretable unit to patients and providers). We will also measure effectiveness in terms of binary maternal categorical outcomes of excessive GWG prevalence, and GDM incidence as well as other secondary outcomes (postpartum weight retention at 4 and 6 months postpartum, infant birth weight and others). To incorporate these potential health impacts in the ROI analysis, we will estimate the average incremental costs associated with health care utilization across study participants who experienced these outcomes compared to matched participants who did not, as further described below in the costing approach.

Measurement and valuation of resource uses

Estimation of the implementation, intervention, and downstream costs involves calculating the direct medical and nonmedical costs (i.e., labor costs, and other costs) of the intervention and its implementation strategy.

Direct medical costs include medical expenses directly linked to the intervention, such as prenatal, delivery, postpartum care, and personnel costs.While delivery mode is not a prespecified secondary outcome it will still be reflected in cost estimates through billing data. Patients requiring closer monitoring, including those with overweight or obesity, may incur higher short-term medical costs due to more frequent follow-ups, while long-term benefits (e.g., reduced obesity and improved outcomes in future pregnancies) may take longer to manifest. To account for uncertain cost impacts, we will use EHR encounter charges, patient-level data, and fee schedules, applying econometric models to capture costs for physician visits, lab tests, medications, and hospitalizations during the trial period (randomization to 6 months postpartum).⁴³

Since this study is in the State of Maryland, which has a unique reimbursement model, we must consider state-specific factors in estimating healthcare costs and benefits. Maryland’s Health Services Cost Review Commission sets hospital rates and oversees reimbursement, using policies like Quality-Based Reimbursement, the Readmissions Reduction Incentive Program, and public payer reimbursement policies.⁴⁴ Under Maryland’s public payer policy, for instance, Medicare and Medicaid plans pay 92.3% of billed charges, less patient cost shares and a 2% sequestration discount, while Medicaid and Medicaid Managed Care Organizations apply a similar 7.7% differential. Costs will be calculated based on these policies, with sensitivity analyses using billed charges adjusted by charge-to-cost ratios for broader applicability.

Direct nonmedical costs are costs related to the intervention, but that are non-medical. To estimate these, we will follow the approach used in other studies that have health coaches who deliver the intervention and estimate labor costs and other costs like maintenance and infrastructure costs.^17,45–47 Labor costs reflect the time health coaches, coach managers (dietitians), and clinical staff spend on intervention activities (Table 2). These costs will be estimated by measuring time across four activity types: scheduled contact, scheduled noncontact, unscheduled contact, and unscheduled noncontact. For scheduled contacts, we will calculate the average time spent per contact from health coaches’ logs and multiply it by the average (or expected) number of contacts to estimate total time spent. Sensitivity analyses will explore the impact of time variability per contact and contact frequency. Estimation of clinical staff time will assume a single staff member led all sessions in the intervention arm (Table 2).

Time spent on nonscheduled contact, nonscheduled noncontact, and scheduled noncontact activities will be estimated using a multiplier approach,⁴⁸ due to variability across participants and limited tracking (Table 2). Multipliers, calculated as the ratio of recorded activity time to scheduled contact time, will be used to estimate nonscheduled and noncontact time by activity type and role. Phase-specific multipliers will not be calculated due to challenges in attributing time by phase.

We will calculate the total hours spent per participant by health coaches and managers for each intervention component by dividing the total intervention hours by the number of participants. Multiplying this time by the average salary for coaches and staff, adjusted for fringe benefits (e.g., a 0.464 benefit-to-wage ratio for 2023),⁴⁹ will provide the average monthly cost per participant and the total cost of each component. Salary estimates for coaches and managers will be based on wages for dietitians and nutritionists.⁵⁰ Other costs, including website development, hosting, and maintenance, will be adapted from a costing study by Smith et al.⁵¹ (Table 2).

Data sources

EHR data will serve as the source for all inputs necessary for the estimation of direct medical costs, including clinical outcomes and utilization data and charges. Utilization data and charges will include those associated with ambulatory visits (e.g., prenatal care), procedures (e.g., ultrasounds), emergency room visits, labor and delivery triage, and inpatient care for both mother and infant.

Statistical considerations

In primary analyses, we will estimate multivariate generalized linear mixed-effects models (GLMM),⁵² which extend the generalized linear model (GLM) approach,⁵³ to account for correlation of outcomes due to repeated measurements on the same individual and other clustering effects.⁵² In addition, the GLMM approach allows for joint modeling of longitudinal effectiveness and cost data as well as cost data from multiple sources (e.g. inpatient, outpatient, prescriptions).^54,55 This approach addresses two key limitations of other methods: (1) using aggregated total costs, which obscures covariate-specific impacts across cost categories; and (2) modeling cost categories separately, which ignores potential interdependence among the outcomes like cost and effectiveness.^52,54,56 Using this approach, we will assess four different correlation structures among the multiple cost outcomes using normal distribution based random effects: (1) shared random intercept, (2) shared random intercept and slope, (3) separate random intercepts from a joint multivariate distribution, and (4) separate random intercepts and slopes from a joint multivariate distribution. Model selection will be based on goodness of fit measures (AIC and BIC) and residual plots.

Sensitivity and scenario analyses

We will perform one-way and multivariate sensitivity analyses to identify key parameters affecting cost-effectiveness and ROI estimates and assess the combined impact of parameter uncertainties. Parameters will vary within uncertainty ranges (e.g., discount rate: 0%–5%, base 2%).⁵⁷ WTP thresholds will span a wider range of dollars per unit of weight loss per year. Probabilistic sensitivity analyses and bootstrapping will be used to quantify parameter uncertainty and generate 95% confidence intervals for costs, effectiveness, cost-effectiveness, and ROI estimates.⁵⁸ Uncertainty will also be represented in acceptability curves, showing the probability of adopting the intervention across different WTP and ROI thresholds.^59,60

DISCUSSION

This article outlined proposed methodologies for evaluating the H42/H4U health coaching program, integrated into prenatal care, using CEA and ROI analysis, to address a critical gap in the literature and for policymakers regarding the cost-efficiency and financial viability of H42/H4U and similar interventions targeting GWG and maternal health during pregnancy and postpartum. Our approach addresses the four challenges outlined in the introduction. It addresses challenge 1 by using mhealth applications⁶¹ and EHR integration to track outcomes like GWG, infant birth weight, and GDM incidence, and by leveraging the trial’s pragmatic design to refine measurement techniques as needed. Our approach accounts for the value of the health benefits generated by the intervention by monetizing the effectiveness measure of weight change (kg) using reference WTP thresholds, ensuring comparability with prior data. For challenge 2, we adapt the costing approach from another similar trial to categorize coaching activities (e.g., scheduled or unscheduled contacts) to estimate the monetary value of health coaches’ time. Challenge 3 is addressed by including a diverse patient population and conducting subgroup analyses to capture a broad range of effects, with a 12-month follow-up to track shifting outcomes. EHR integration further enables identification and measurement of displaced outcomes. Lastly, challenge 4 is handled using GLMM models to account for repeated measurements and clustering effects,^52,53 ensuring robust findings. By overcoming these challenges, we will produce rigorous evidence to support decision-making and facilitate the implementation of evidence-based healthy lifestyle interventions in U.S. prenatal clinical care settings.

Our planned economic evaluations aim to fill this evidence gap by providing insights into the cost-efficiency (via CEA) and financial viability (via ROI analysis) of the H42/H4U intervention. By adopting the perspective of a payer (e.g., Medicaid or other insurer) or adopting organization (e.g., health care system), we aim to inform decision-making at organizational levels, considering costs and health effects relevant to budgetary considerations. Our approach also addresses major methodological and practical challenges in conducting economic evaluations of behavioral health interventions. The multifaceted nature of the H42/H4U intervention, targeting various health behaviors and lifestyle aspects, complicates evaluation by requiring consideration of synergistic effects and participant adherence. Our study protocol employs rigorous data collection methods, integrating patient-level trial data and electronic health records to measure intervention costs and consequences effectively.

The societal perspective encompasses costs from various sectors, not just in the formal and informal healthcare sectors, and is arguably most appropriate for societal-level decisions. In contrast, the payer’s perspective focuses solely on intervention consequences (e.g., costs and health effects) affecting payers’ budgets, and is more useful for organizational-level decision-making. Hence, certain cost types, such as out-of-pocket expenses, informal healthcare sector costs, or productivity costs are often irrelevant under the payer’s perspective. While adopting this perspective may appear narrow, it is pragmatic, and allows us to avoid making unverifiable assumptions about the intervention’s effects on costs and health outcomes.

Additionally, this study protocol outlines a transparent approach to assess the cost-efficiency and financial viability of the H42/H4U intervention. By reporting our analytical approach ex ante, we aim to minimize the risk of bias in the analysis stage. The transparency enhances the reliability and comparability of our findings with other studies, promoting confidence in the results. ⁶² Economic evaluations are often secondary or tertiary aims and outcomes in many studies, exposing them to various biases. Our deliberate and transparent planning aims to mitigate these risks and encourage similar rigorous methodologies in future research.

Study limitations

This study has several limitations, typical of trial-based economic evaluations. First, relying on EHR charges and patient-level clinical data may overlook cost variations and indirect costs, affecting cost estimates. Additionally, higher medical resource use from follow-up visits may vary across participants. Second, long-term health benefits and cost implications, which may emerge over time, are not captured within the trial period.^63–67 Future research should extend follow-up or model outcomes beyond the trial. Third, nonmedical costs, especially labor, rely on assumptions due to inconsistent time tracking, risking inaccuracies despite planned sensitivity analyses. Fourth, assuming that clinical staff deliver sessions may limit generalizability. Lastly, excluding phase-specific cost multipliers simplifies costs, while adjusting salaries with standard fringe benefit ratios may not reflect regional variations, underscoring the need for setting-specific cost estimates.

Despite these limitations, this study offers valuable insights into the economic evaluation of the H42/H4U intervention and similar lifestyle interventions during pregnancy in the U.S. and identifies areas for future research to enhance economic evaluations in healthcare.

CONCLUSIONS

This study protocol outlines a transparent methodological approach for evaluating the cost-effectiveness and financial viability of the H42/H4U intervention. By addressing existing gaps in the literature, we aim to provide critical evidence to support investment decisions and promote the successful implementation of healthy lifestyle interventions that promote maternal health and reduce the risk of GDM.

Table 3.

CHEERS 2022 Checklist.

	Item	Guidance for Reporting	Reported in Section
TITLE
Title	1	Identify the study as an economic evaluation and specify the interventions being compared.	Title, pp. 1
ABSTRACT
Abstract	2	Provide a structured summary that highlights context, key methods, results and alternative analyses.	Structured Abstract, pp. 1–2
INTRODUCTION
Background and objectives	3	Give the context for the study, the study question and its practical relevance for decision making in policy or practice.	Introduction, Background, pp. 2
METHODS
Health economic analysis plan	4	Indicate whether a health economic analysis plan was developed and where available.	Methods, pp. 3–7
Study population	5	Describe characteristics of the study population (such as age range, demographics, socioeconomic, or clinical characteristics).	Introduction, Overview and design of the Healthy for Two / Healthy for You (H42/H4U) intervention, pp. 2 Methods, pp. 3–7
Setting and location	6	Provide relevant contextual information that may influence findings.	Introduction, Overview and design of the Healthy for Two / Healthy for You (H42/H4U) intervention, pp. 2 Methods, pp. 3–7
Comparators	7	Describe the interventions or strategies being compared and why chosen.	Introduction, Overview and design of the Healthy for Two / Healthy for You (H42/H4U) intervention, pp. 2 Methods, pp. 3–7
Perspective	8	State the perspective(s) adopted by the study and why chosen.	Methods, pp. 3–4
Time horizon	9	State the time horizon for the study and why appropriate.	Methods, pp. 3–4
Discount rate	10	Report the discount rate(s) and reason chosen.	Methods, pp. 4
Selection of outcomes	11	Describe what outcomes were used as the measure(s) of benefit(s) and harm(s).	Methods, pp. 3–7
Measurement of outcomes	12	Describe how outcomes used to capture benefit(s) and harm(s) were measured.	Methods, pp. 6–7
Valuation of outcomes	13	Describe the population and methods used to measure and value outcomes.	Methods, pp. 6–7
Measurement and valuation of resources and costs	14	Describe how costs were valued.	Methods, pp. 6–7
Currency, price date, and conversion	15	Report the dates of the estimated resource quantities and unit costs, plus the currency and year of conversion.	Methods, pp. 4
Rationale and description of model	16	If modeling is used, describe in detail and why used. Report if the model is publicly available and where it can be accessed.	N/A
Analytics and assumptions	17	Describe any methods for analysing or statistically transforming data, any extrapolation methods, and approaches for validating any model used.	Methods, pp. 3–7
Characterizing heterogeneity	18	Describe any methods used for estimating how the results of the study vary for sub-groups.	Methods, pp. 3
Characterizing distributional effects	19	Describe how impacts are distributed across different individuals or adjustments made to reflect priority populations.	Methods, pp. 3
Characterizing uncertainty	20	Describe methods to characterize any sources of uncertainty in the analysis.	Methods, pp. 7
Approach to engagement with patients and others affected by the study	21	Describe any approaches to engage patients or service recipients, the general public, communities, or stakeholders (e.g., clinicians or payers) in the design of the study.	N/A
RESULTS
Study parameters	22	Report all analytic inputs (e.g., values, ranges, references) including uncertainty or distributional assumptions.	Not applicable, Study protocol
Summary of main results	23	Report the mean values for the main categories of costs and outcomes of interest and summarize them in the most appropriate overall measure.	Not applicable, Study protocol
Effect of uncertainty	24	Describe how uncertainty about analytic judgments, inputs, or projections affect findings. Report the effect of choice of discount rate and time horizon, if applicable.	Not applicable, Study protocol
Effect of engagement with patients and others affected by the study	25	Report on any difference patient/service recipient, general public, community, or stakeholder involvement made to the approach or findings of the study	Not applicable, Study protocol
DISCUSSION
Study findings, limitations, generalizability, and current knowledge	26	Report key findings, limitations, ethical or equity considerations not captured, and how these could impact patients, policy, or practice.	Discussion, pp. 7–8
OTHER RELEVANT INFORMATION
Source of funding	27	Describe how the study was funded and any role of the funder in the identification, design, conduct, and reporting of the analysis	Funding, pp. 8
Conflicts of interest	28	Report authors conflicts of interest according to journal or International Committee of Medical Journal Editors requirements.	Competing interest, pp. 8

LIST OF ABBREVIATIONS

AIC

Akaike information criterion

BIC

Bayesian information criterion

CEA

Cost-effectiveness Analysis

CHEERS

Consolidated Health Economic Evaluation Reporting Standards

EHR

Electronic health record

EQ-5D-3L

3-Level EuroQol 5-Dimensional

GLM

Generalized linear model

GLMM

Generalized linear mixed-effects model

GDM

Gestational diabetes mellitus

GWG

Gestational weight gain

H42/H4U

Healthy for Two / Healthy for You

ICER

Incremental cost-effectiveness ratio

INHB

Incremental net health benefit

INMB

Incremental net monetary benefit

ITT

Intention-to-treat

ROI

Return-on-investment

USPSTF

United States Preventive Services Task Force

WTP

Willingness-to-pay

Availability of data and materials

The full protocol will be available from the corresponding author on reasonable request. This manuscript does not contain participant-level data.