Design, Setting, and Population

The PROPS 2.0 programme is being offered at no cost as part of standard care to eligible patients from a variety of clinics affiliated with Brigham and Women’s Hospital (BWH), including Primary Care, Cardiology, Endocrinology, Orthopedic Surgery, Comprehensive Breast Health, Sleep Medicine, the Center for Weight Management and Wellness, and the Nutrition and Wellness Service. These clinics are in both urban and suburban locations throughout the greater Boston area and serve a racially and socioeconomically diverse population of patients.

PROPS 2.0 plans to enrol at least 5000 patients over 12–15 months. To be eligible, patients must be at least 20 years old, have a primary care provider, have received care from a provider at BWH in the last two years, and have a BMI of ≥30 kg/m² or a BMI of 25–29.9 kg/m² with at least one indicator of increased cardiovascular risk or other obesity-related condition (eg, type 2 diabetes, prediabetes, hypertension, hyperlipidaemia). Because the programme is being integrated as part of standard care, the goal is to offer it to any patients who may benefit from weight loss; therefore, there is no upper BMI limit, which is consistent with current clinical guidelines for recommending comprehensive lifestyle intervention.3 Patients also must speak English or Spanish and have a valid email address and regular Internet access (at least once a week) via a computer, smartphone or other device. Patients who are pregnant, have type 1 diabetes or are taking insulin for treatment of diabetes, or have active cancer or any serious medical condition for which weight loss is not recommended are not eligible. In addition, patients who have had weight loss surgery in the last 12 months are not eligible, due to the need to follow a strict diet for 12 months after surgery. However, patients who are participating in other weight loss programmes or taking anti-obesity medications are still eligible.

Although PROPS 2.0 is being conducted as a clinical implementation project, some components (eg, surveys, interviews) are considered to be research; therefore, we obtained approval from the Mass General Brigham Human Research Committee (HRC) and registered the programme with ClinicalTrials.gov (NCT05553925).

Patient Involvement

We conducted semi-structured interviews with 15 patients and 15 providers to obtain feedback about the programme prior to the beginning of recruitment; the methods and results from these interviews will be described separately. We also formed a Patient and Stakeholder Advisory Committee (PSAC) that includes patients, providers, and stakeholders from within and outside of BWH. The PSAC members provide input on the programme design, evaluation, and other issues related to the adaptation and integration of the intervention (see Aim 1).

Recruitment and Enrolment

Patients are being recruited for the programme through three main pathways, namely: (1) provider referral, (2) self-referral, and (3) identification via the electronic health record (EHR). An overview of the recruitment, enrolment and follow-up process is shown in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Overview of recruitment, enrolment and follow-up. CRM, customer relationship management; EHR, electronic health record; PCP, primary care provider.

For the provider referral pathway, we created an internal referral order within the EHR used at BWH (Epic). The referral order (figure 2) can be found with multiple search terms, including ‘PROPS’, ‘overweight’, ‘online,’ and ‘online weight’. We also developed two Best Practice Advisories (BPAs) within the EHR, one for patients with BMI ≥30 kg/m² and one for patients with BMI 25–29.9 kg/m², to alert providers about patients who may be eligible (figure 2). If one of the BPAs appears for a patient, providers can click on a button to go directly to the referral order. We visited participating clinics to educate providers (e.g., physicians, nurses, physician assistants) about the programme, and we distributed pocket-sized, laminated cards with information for providers about the programme, the eligibility criteria, and referral process. There also are clinician champions at some of the clinics who have helped with distributing materials, answering questions from other providers, and troubleshooting issues.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Internal referral order and Best Practice Advisories in Epic. BMI, body mass index; BWH, Brigham and Women’s Hospital.

After patients are referred by their provider, their information is imported from the EHR into a customer relationship management (CRM) platform (Microsoft Dynamics),15 which was configured for this project. The CRM is used to conduct outreach in bulk, automate and manage tasks, track data, and generate reports. After their data are imported into the CRM, patients are sent a welcome message explaining that their provider referred them, describing the programme, and asking them to complete a brief enrolment survey to verify their eligibility and interest. The welcome message is sent via the secure EHR patient portal (MyChart) to patients with active portal accounts. Patients without an active portal account are initially sent an email message with very limited information about the programme (and no personal health information); this type of message is deemed “low risk” by the healthcare system. Patients must then give consent to use their email address to receive more detailed information about the programme, including the welcome message and enrolment survey. If patients do not complete the enrolment survey after 14 days, the CRM automatically sends a reminder message.

For the self-referral pathway, we developed a website (https://props2.massgeneralbrigham.org) with information about the programme for patients, which is available in both English and Spanish. We distributed posters and business cards to all participating clinics with the URL and a QR code to the website and asked staff to display them in exam rooms and/or waiting rooms. From the PROPS 2.0 website, patients can complete the enrolment survey via REDCap (Research Electronic Data Capture) to verify eligibility and to confirm that they want to enrol, if eligible. After patients complete the survey, their information is reviewed and then imported into the CRM.

Finally, in the future, if enrolment targets are not reached from provider referrals and self-referrals, potentially eligible patients will be identified using local EHR data and other clinical systems. Information for these patients will be extracted from the Enterprise Data Warehouse (EDW) and imported into the CRM. An outreach message and link to the enrolment survey will then be sent via the secure EHR patient portal, or the “low-risk” message will be sent via email for patients who do not have an active portal account.

Independent of specific recruitment/entry mechanisms, if a patient is eligible based on the enrolment survey and decides to enrol, they receive a link to register for the online programme/app. The link is sent via the secure patient portal or email, depending on their communication preference indicated on the enrolment survey. Because the programme is being considered as part of standard care (i.e., offered to all eligible patients in their routine care), the Human Research Committee granted a waiver of the full informed consent form that typically is required for research studies. However, on the enrolment survey, patients still must give their consent to participate and to receive text and email messages from RestoreHealth, the online programme/app being used. They can also withdraw from the programme at any time.

Online Weight Management Programme/App (RestoreHealth)

Because the programme used for the original PROPS Study is no longer available, we chose RestoreHealth as the online programme for PROPS 2.0. RestoreHealth (HealthFleet Inc.) is a 12-month digital health programme that pairs an online programme/app with live one-on-one coaching.16 The programme is designed to support the behavior modification process via easy and simple interactions, highly personalised coaching, and a holistic curriculum that is based on cognitive behavioral therapy and focuses on overall physical and mental health (rather than focusing strictly on weight loss). It is available in English and Spanish and has interfaces for both patients and coaches.

After patients register for RestoreHealth, they can view brief lessons (a total of 118 lessons, available in written and video format) on specific topics related to weight management, including nutrition, exercise, sleep, and stress. None of the lessons are mandatory. They follow a general path in terms of the sequence, but the curriculum can be personalized for each patient; if a patient is interested in a particular topic, the coach can re-order the lessons to meet that request. Patients also can log their meals, get regular feedback from their coach, and interact with other users via discussion forums.

As part of the programme, RestoreHealth uses data from several connected devices, including scales and sleep/activity trackers, which can be connected to the platform, allowing patients’ data to be uploaded automatically and available to their coach. The coaches are behavioral change specialists who are certified through the National Board for Health and Wellness Coaching (NBHWC). After patients are paired with their coach, the coaches monitor patient engagement, weight, diet, physical activity, and stress and give personalized feedback. The coaches primarily correspond with patients via text messaging, but they also can schedule monthly calls with patients who prefer not to use text messaging. Patients can also participate in group coaching to provide additional peer support. The coaches work with patients for the entire 12 months of the programme. The coaching interactions are personalized, based on the needs of the individual patient, but coaches usually decrease the frequency of communication after the first 4–5 months of the progra, to help patients become more independent. Each patient determines how and how often they would like to interact with their coach and the programme. The patient is always able to engage with their coach via text message, and the coach will always respond. If any medical issues arise, coaches refer patients to their primary care provider.

Patient Navigator Support

Patient navigators, who are non-clinical staff members from the Mass General Brigham Accelerator for Clinical Transformation (ACT) team, are involved in enrolment of patients and provide some monitoring and support during the programme. The ACT team delivers remote care services through a combination of technology, patient navigators, and clinicians; ACT has run entirely remote disease management and drug optimization programmes for over 10,000 patients with hyperlipidaemia and/or hypertension.15 17–20 The navigators use the CRM platform to send outreach messages about enrolment to patients who have been referred. Once patients are fully enroled and registered in the programme, most of their communication is from the RestoreHealth coaches, but the navigators serve as the main connection with the healthcare system. For example, they enter documentation into the EHR at several time points, including when a patient enrols in the programme, after six months, and at programme completion.

In addition, the navigators monitor patients’ weight data throughout the programme and can help connect patients with additional weight management resources, such as through the Center for Weight Management and Wellness. There is a protocol in place for patients who are identified as having “rapid weight loss”, which is defined as weight loss >12% over the past 4 weeks or >10% over the past week. If a patient meets these criteria, a patient navigator conducts an outreach call to the patient to ask them a few standard questions about their health status. Depending on the patient’s responses, their information is sent to a clinical advisory group, which includes three physicians from the team. One of the physicians reviews the patient’s weight and other data and decides about next steps. The physician also notifies the patient’s primary care provider about any concerns.

Data Collection and Follow-Up

After patients enrol in PROPS 2.0 and register for RestoreHealth, they participate in the programme for 12 months and engage with the platform and their coach (figure 1). The programme has no in-person visits, other than visits that patients may have with their healthcare provider(s) as part of routine care. During the 12 months of the programme, patients continue to use RestoreHealth and to receive feedback from their coach. Data on patients’ engagement with the programme, as well as their data from the scales and trackers, are collected within the online programme/app. Data on demographic and clinical variables (e.g., weight, blood pressure, and other cardiometabolic measures) are extracted from the EHR. In addition, a sample of patients (50% of English-speaking, 100% of Spanish-speaking) are invited to complete an electronic survey (via REDCap) after registering for the programme and again at 12 months to assess changes in patient-reported outcomes such as diet,21 physical activity,22 health status and self-efficacy around weight loss,23 24 as well as satisfaction with the programme. Providers within all participating clinics also were invited to complete a brief electronic survey via REDCap before programme implementation, and they will be asked to complete a similar survey after implementation; the purpose of these surveys is to assess their knowledge, attitudes and behaviours about treatment of overweight and obesity and their satisfaction with the programme. In addition to these surveys, we also will conduct semi-structured interviews with patients and providers at the end of the programme to assess their overall experience, satisfaction, and feedback about the programme.

Outcome Measures

The evaluation of the programme will be mixed-methods, incorporating both quantitative and qualitative data, and is guided by the RE-AIM framework. RE-AIM is a well-established framework that was designed to address issues related to the implementation and external validity of health interventions, and it has been applied to questions related to health information technology and overweight and obesity.25 26

To assess reach, we will examine the number and proportion of patients who are referred to the programme by providers, refer themselves, or are identified via the EHR, as well as the number and proportion who are screened, eligible, enroled, and registered for the programme. To assess effectiveness, we will examine weight change over 12 months (primary outcome), calculated as the difference between each patient’s weight at enrolment and 12 months (±90 days) after enrolment, using weight data from RestoreHealth as well as from the EHR. We also will examine percent weight change, changes in cardiometabolic measures (e.g., blood pressure, haemoglobin A1c), and changes in patient-reported outcomes (e.g., diet, physical activity, health status) over 12 months, using data from the EHR and from patient surveys. To assess adoption at the provider level, we will examine the number and proportion of providers who refer patients to the programme. To assess implementation, we will examine patients’ engagement with the RestoreHealth programme (e.g., number of contacts with coaches, number of lessons viewed, number of meals logged). To assess maintenance (post-intervention outcomes), we will examine changes in patients’ weight and other outcomes at 18 months, six months after they have finished the programme. Other outcomes and data sources are included in table 1.

Analytical plan

The analyses correspond with the three specific aims, which are included in the Background section. The analyses for Aim 1 will be primarily qualitative, using data from interviews conducted prior to programme implementation; these will be described separately. In addition, descriptive statistics (e.g., means and standard deviations, percentages and 95% confidence intervals) will be used to summarize results from the provider surveys conducted before implementation.

For Aim 2, we will use descriptive statistics to summarize the reach, adoption, and implementation measures shown in table 1; we also will examine how these vary by patient characteristics (such as age, sex, race/ethnicity, and language), and we will compare characteristics of patients who enrol to characteristics of the entire BWH patient population. For effectiveness, we will examine changes in patients’ weight, cardiometabolic outcomes, and self-reported outcomes over 12 months. We will initially do this using paired t-tests for continuous outcomes (e.g., weight) and McNemar’s test for binary outcomes (e.g., excellent/very good health status), but we also will use linear and logistic regression models to examine the impact of patient characteristics on these outcomes. To account for missing data, we will use multiple imputation, in order to include all enroled patients in the primary analysis. We will use similar methods to examine changes in providers’ knowledge, attitudes, and behaviors before and after implementation of the programme.

Because this is a clinical implementation project, there is no true “control group” (i.e., all eligible patients can enrol in the programme). However, we will use a segmented regression model based on a discontinuity design27 to compare measured weight change among patients who enrol in the programme to their weight change projected from a “quasi-control” group of patients from the same practices/clinics. The “quasi-control” patients will be similar to participants in all possible respects, except that their BMIs do not quite reach programme eligibility. For enroled patients with BMI ≥30 kg/m², the quasi-control group will include patients with BMI 25–29.9 kg/m² with no indicators of increased cardiovascular risk or obesity-related conditions. For enroled patients with BMI 25–29.9 kg/m² with indicators of increased cardiovascular risk or obesity-related conditions, the quasi-control group will include patients with BMI 22–25 kg/m² who have at least one indicator of increased cardiovascular risk or an obesity-related condition. Linear regression models relating weight change to baseline BMI in these quasi-controls will be used to project the amount of weight change that would normally occur (without intervention) in people with higher BMIs. A segmented linear regression model will then be used to compare the actual pattern of weight change among intervention patients (i.e., those who enrol in the programme) to the projected pattern among quasi-control patients; separate models will be run for the two BMI groups, and only weight data from the EHR will be used in these analyses (as control patients will not have weight data from RestoreHealth). In the segmented regression models, the outcome for each patient will be their weight change at 12 months, and the primary predictors will be their initial BMI at enrolment, an indicator for intervention or control group, and an interaction between initial BMI and the indicator for intervention or control group. The models also will adjust for patient characteristics, such as age, sex, and race/ethnicity. Similar models will be run for some of the secondary outcomes (e.g., cardiometabolic measures). We also plan to conduct stratified analyses by age, sex, BMI, and race/ethnicity to examine whether the impact of the programme may differ within subgroups.

The analyses for Aim 3 will be similar to those for Aim 2, but they will use data collected six months after the end of the implementation period. For example, patients’ weight change over 18 months (six months after completion of the programme) will be examined using descriptive statistics, and segmented regression models will be used to compare their actual weight change over 18 months to the projected weight change from the same quasi-control patients described earlier.