Introduction

The COVID-19 pandemic has dramatically affected pediatric primary care. The intermittent suspension of in-person visits, the increase in use of telehealth, and the overall impact of COVID-19 has created new vulnerabilities for populations that are at risk of poor access to care [1]. For children with obesity, these changes have resulted in missed well-child visits and follow-up care to support weight management and related conditions, thereby exacerbating chronic medical conditions and increasing body mass index (BMI) [2,3,4]. To address the evolving needs and context changes in pediatric primary care, we adapted Connect for Health for telehealth and other impacts of the pandemic and related policies. Connect for Health is an evidence-based pediatric weight management program intended for delivery in primary care that leverages clinical and community resources to improve child body mass index (BMI) and family-centered outcomes [5, 6]. The program has family- and clinician-facing tools that required adaptations to ensure continued uptake.

Programs must be flexible and adaptable to respond to changing healthcare contexts to increase the likelihood of sustainability [7], but program developers also need to ensure that adaptations are equitable and reduce barriers to care [8, 9]. Telehealth and other program adaptations may offer many opportunities for patients to access necessary services, but they can also present with inequities and challenges when providing obesity-related care [10, 11]. For example, patient characteristics, such as limited English proficiency (LEP) and socio-economic status, have been identified as barriers to telehealth utilization [12, 13]. During the COVID-19 pandemic, childhood obesity interventions have rapidly transitioned to telehealth and made other necessary program changes. However, there have been limited reports of programs that systematically plan adaptations and engage multi-level stakeholders in this process to make sure modifications meet the needs of clinicians and families [14]. By using established adaptation frameworks from the implementation science literature to guide adaptations and document modifications [15, 16], we can evaluate changes to evidence-based programs to ensure we are providing children with obesity, effective, high-quality care.

The objective of this study is to describe the systematic process we undertook to make real-time adaptations in response to the COVID-19 pandemic to the Connect for Health pediatric weight management program to ensure continued and equitable program uptake. Our aims were to (1) systematically describe the planned adaptation process and results of that process and (2) examine the experiences and perceptions of parents and clinicians about using telehealth equitably for pediatric weight management to inform program adaptations. By understanding the experiences and perceptions of parents and clinicians and engaging stakeholders, we can make adaptations to the Connect for Health program to ensure it equitably meets the needs of clinicians and families who are at risk of poor access to care.

Materials and methods

Connect for Health weight management program

Connect for Health is a weight management intervention for children ages 2–12 years for delivery in the primary care setting. The program consists of clinician-facing tools including flagging of children with elevated BMIs, clinical decision support to guide management, and clinician education and training and family-facing tools including educational handouts about behavioral change, social and community-informed text messages, and community resource guides. Implementation strategies to support the uptake of the program are currently being tested; and the program components, implementation strategies, and study protocol have been described in detail previously [6]. The implementation of Connect for Health is currently being studied in four healthcare organizations that provide care in low-income communities and have high rates of obesity in Boston, MA, Greenville, SC, and Denver, CO. The study protocol was approved by the Mass General Brigham institutional review board.

Process for planned adaptations

In the pre-implementation phase, the healthcare organizations participating in the implementation of Connect for Health conducted extensive stakeholder engagement [17]. The program was then launched in late 2019, months prior to the nationwide lockdowns due to COVID-19 resulting in disruptions to pediatric primary care. During the early months of the pandemic, the healthcare organizations learned from clinicians and stakeholders that they wanted to continue to use the program due to the rapidly increasing rates of childhood obesity secondary to the pandemic [2, 3], but it was challenging to use the program due to the increased use of telehealth and other changes in pediatric primary care. The research study team made the decision to adapt the program during the implementation phase to align with telehealth usage and other pandemic-related policies and ensure continued equitable access to the program.

The iterative process we undertook to adapt the program for telehealth was guided by the phases of the Planned Adaptation Model that allowed us to balance the tension between retaining the program’s purpose and being responsive to the needs of stakeholders [16]. In the first phase, the program developers identified the core functions (i.e., purpose) and forms (i.e., activities) that would enable us to preserve the program’s purpose as we adapted it [18].

In the second phase, we identified the changing healthcare context that occurred during program implementation due to the pandemic. The purpose of this phase was to understand how pediatric primary care was impacted and the issues that needed to be addressed for continued program uptake and equitable delivery of care. We engaged stakeholders across the healthcare organizations (i.e., organizational level challenges) and conducted parent and clinician surveys (i.e., patient-level) challenges at Massachusetts General Hospital. We first sought feedback from clinical experts in pediatric weight management and telehealth in a series of three working groups from September to November 2020. The purpose of the working groups was to understand current challenges clinicians faced when delivering pediatric weight management care and when using telehealth for weight management and to provide input on the development of the clinician and parent surveys.

The purpose of the parent and clinician surveys was to understand their experiences, perceptions, needs, and preferences when using telehealth for pediatric weight management. In conducting surveys, we also wanted to ensure that the telehealth adaptations would not cause further healthcare inequities for patients who have historically been marginalized by the use of telehealth [10, 11]. We surveyed parents of children with a BMI ≥ 85th percentile who attended primary care or a specialty care weight management program either virtually or in-person from February to September 2021 which was prior to adaptations being made. Survey data was linked with electronic health record (EHR) data for socio-demographics. Surveys were conducted over the phone or via email and were available in English and Spanish. Surveys took approximately 20–30 min to complete and a $25 gift card was provided as compensation. The Research Electronic Data Capture (REDCap, Nashville, TN) was used to collect and manage data. We surveyed pediatric specialty and primary care clinicians who care for children with overweight and obesity from December 2020 to March 2021. Information about the survey was sent via email and participants completed the survey using REDCap. Surveys took approximately 10 min to complete, and $50 was provided as compensation.

Survey questions included constructs that have been previously reported in telehealth utilization studies [19]. Parent questions focused on technical quality, experience, and willingness to use telehealth in the future, and clinician questions focused on the effect on the interaction, usefulness, experience, and resources needed. Questions were developed using the Patient Assessment of Chronic Illness Care [20], the modified-Family Centered Care Assessment Tool [21], The Telehealth Usability Questionnaire [22], and questions used to assess patient experiences of care with telehealth [23]. Several questions were asked of both clinicians and parents to compare responses. We used the Checklist for Reporting of Survey Studies (CROSS) (see Additional file 1).

Following the working groups and conduct of surveys, we then engaged key stakeholders with expertise in telehealth and pediatric weight management and conducted planning sessions with sites implementing Connect for Health to review proposed adaptations, discuss feasibility, and generate new ideas. We initially presented them with preliminary findings of the clinician and parent surveys to understand if this was similar to their clinical experiences and a document with suggested adaptations. In meetings, we reviewed the information and sought additional input and ideas.

In the third phase, we adapted the intervention with input from stakeholders and survey results, using the Framework for Reporting Adaptations and Modifications-Enhanced (FRAME) to guide and document the adaptations [15]. Because we were making adaptations in real-time, we used FRAME to ensure adaptations did not impact core program mechanisms and that we made changes at the clinical- and family-level. For example, we assessed each adaptation as to whether it changed a core function of the program. We also initially had more clinician-level modifications and by using FRAME we could focus stakeholder discussions on family-level modifications. The final phase of the Planned Adaptation Model is to evaluate the adapted intervention the results of which will be reported a future publication when the study is complete.

Statistical analyses

We described clinician and parent characteristics who participated in the survey. Descriptive statistics, including frequency and percentage or mean and standard deviation, were then calculated for the survey questions. We described the participant characteristics and survey descriptive statistics overall and according to language of survey administration and used χ2 tests for categorical variables and t-tests for continuous variables to compare by language. To further examine the preferences of parents with LEP, we selected four questions that are key to delivering effective pediatric weight management care. The questions asked if a virtual or office visit is better for addressing health concerns, sharing private or personal information, setting goals and talking about healthy behaviors, and talking about mental health and well-being. We used multivariable logistic regression to determine the odds ratio and 95% confidence intervals and adjusted for parent education, visit mode, and child’s age. To test for statistical significance in all analyses, we used a 2-sided alpha level of 0.05. The R Studio Software (version 4.1.0) was used for the statical analyses.

Results

Identifying core functions

Following the Planned Adaptation Model phases, we initially identified the core functions and forms of Connect for Health (Table 1). This process allowed the research study team to clearly identify the current purpose of the program (core functions) and the activities (core forms) and to make decisions about what adaptations were needed.

Table 1 Connect for Health core functions, core forms, and adaptations
Full size table

Identifying changing healthcare contexts

Working groups

Results of the working group with pediatric weight management and telehealth experts revealed difficulty with families accessing telehealth platforms, ensuring families are not lost to follow-up which is essential in pediatric weight management care, and providing healthy lifestyle and community resources to families. The working group also provided direct input into the development of the parent and clinician survey questions and reviewed the final surveys.

Parent surveys

We surveyed 200 parents of children with a BMI ≥ 85th percentile and conducted 85 (43%) surveys in Spanish. More parents with LEP were a high school graduate or less (86%) and had an income < $20,000 (46%) than parents without LEP (Table 2). Table 3 describes the results of the parent survey. Fifty percent of participants had a virtual visit (defined as a video or phone visit) for their child and more parents with LEP had the visit conducted via phone (14%) as compared to parents without LEP (7%). Overall, parents did not report any concerns with joining the virtual visit or getting to an in-person visit. When asked about weight management visits, parents reported that they wanted care in the future to be a combination of in-person and virtual visits (80%). Of the parents who attended a virtual visit, a majority reported that the clinician listened to them (89%) and spent enough time with them (88%). Parents without LEP reported they “definitely agreed” that the clinician explained things in an understandable manner (94%) as compared to parents with LEP (74%). When asked if they prefer virtual visits or in-person visits for various reasons (Fig. 1), we found differences between parents with and without LEP. Parents with LEP preferred in-person visits to the virtual visits.

Table 2 Characteristics of parents by language (n = 200) and clinicians (n = 75) who completed the survey
Full size table
Table 3 Parent telehealth survey results by language (n = 200)
Full size table
Fig. 1
figure 1

Parent and clinician responses to virtual or in-person visit preferences

Full size image

In the multivariable logistic regression models (Table 4), we found parents with LEP had higher odds of affirming that office visits are better than virtual visits for ensuring their child’s health concern can be taken care of (OR: 2.91; 95% CI: 1.36, 6.21), feeling comfortable sharing private or personal information (OR: 3.91; 95% CI: 1.82, 8.43), talking to their child’s doctor about healthy behaviors and setting goals (OR: 3.09; 95% CI: 1.39, 6.90), and talking to their child’s doctor about their child’s mental health and overall well-being (OR: 4.02; 95% CI: 1.83, 8.87) than parents without LEP.

Table 4 Associations of telehealth preferences with limited English proficiency status in unadjusted and multivariable adjusted models
Full size table

Clinician surveys

We contacted 94 clinicians and had a response rate of 80%. Of the 75 clinicians surveyed, a majority were physicians (89%) and worked in the primary care setting (76%) (Table 2). More than half of clinicians had experience with virtual visits, had completed more than 50 visits, and had patients on their panels with LEP and limited digital literacy. When examining the effect on interactions (Table 5), clinicians generally felt that virtual visits were not a barrier for ordering labs (72%), providing educational materials (47%), and connecting families to resources (71%). Completing a physical exam (59%) and obtaining anthropometrics (43%) were strong barriers to caring for children with overweight and obesity. Clinicians were equivocal regarding identifying access to the internet, technological issues, privacy and security concerns, and access to medical interpreters as barriers.

Table 5 Clinician telehealth survey results (n = 75)
Full size table

Overall, clinicians found telehealth to be useful (Table 5) and strongly agreed or somewhat agreed that virtual visits provided enough time to provide necessary care (93%), enhanced pediatric weight management (60%), were an effective replacement for follow-up patients (82%), and improve patients’ access to healthcare (97%) and that they would miss it if it were no longer an option (82%). Clinicians affirmed the following resources as being useful to enhance virtual visits for pediatric weight management: guidance on best practices for virtual visits, virtual visit resources for families, improved methods for communicating with families, and resources to address patient telehealth accessibility.

Clinicians reported that virtual visits were better for patient adherence to treatment plans, amount of time required for visits, ease of ordering tests, workflow outside of the visit, and timely access to follow-up visits than in-person visits (Fig. 1). Fifty-five percent of clinicians felt that in-person visits were better for patient-clinician communication, and 61% felt the overall quality of in-person visits was better than virtual visits. When comparing parent and clinician responses, we found that parents with LEP felt in-person visits were better for sharing private or personal information (86%) as compared to clinicians (33%).

Stakeholder feedback

When we engaged stakeholders and shared preliminary findings of the clinician and parent surveys, we found their clinical experiences to be similar to the findings. We also learned that clinicians wanted multimodal educational resources, virtual resources, and guidance on how to conduct weight management visits virtually. Similar to the survey findings, they also wanted the flexibility to use the program in-person and virtually.

Adapting the program

Informed by the results of the stakeholder-engaged workgroups and parent surveys, we adapted the content and delivery of Connect for Health. We developed a list (using an Excel spreadsheet) that contained the ideas for modifying the program generated throughout the process. The program developers assessed the feasibility and relevance of the potential adaptations, consulted with experts as needed for additional input (for example, consulting with clinical informatics to assess if an electronic health record modification was possible), and made the final list of the planned adaptations (Table 1). During this process, we used FRAME to guide and document the adaptions (Table 6). As intended, the planned adaptations were all to the core forms (versus core functions) of the program, at both the clinician- and family-level to meet the needs of families who are at risk for poor access to care. The adaptations predominately involved refining and adding elements to the program and provided flexibility for the program to be used for in-person or telehealth visits.

Table 6 Program adaptations: Framework for Reporting Adaptations and Modifications-Enhanced (FRAME)
Full size table

The adaptations addressed telehealth as well as other pandemic-related challenges and allowed us to improve the reach and fit of the program given the changing context of the pandemic. Several of the adaptations were designed to ensure the program met the needs of families who are at risk for poor access to care. We culturally tailored materials regarding behavior change during stressful times, developed video-based educational materials in English and Spanish, provided additional clinician guidance and resources, and optimized the clinical decision support tools and flagging system to integrate into the workflow of virtual visits. The videos reviewed what to expect during a pediatric weight management visit to help familiarize and prepare families. Families could access the video on their mobile phone which based on our previous work was an acceptable method [17], and the brief video length under 2 min would keep data usage at a minimum. The video was created in response to the parent survey findings that in-person visits were better for communicating concerns and talking about healthy behaviors as we felt if families knew what to expect their visit experience would improve. The clinician guidance and resources were developed to address similar parent findings as well as the clinician survey finding that requested virtual weight management guidance to improve visit experience. The electronic health record modifications were in direct response to the clinician survey finding of additional tools to streamline workflow. When making program adaptations, we were mindful of organizational and policy changes that impacted telehealth usage, particularly initiatives to reduce inequities (for example, providing patients options of receiving in-person or virtual care, calling patients before visits to familiarize them with the technology platforms; registering them for patient portal platforms), and aligned clinician and parent guidance accordingly. Our findings from the surveys that families and clinicians wanted options for in-person and telehealth care reinforced organizational policies for the continued use of telehealth after the pandemic ended and made us carefully consider modifications that we made to ensure the flexible use of Connect for Health for all visit types.

Discussion

In this study, we described a systematic process for adapting the Connect for Health pediatric weight management program. The results of stakeholder engagement and parent and clinician surveys informed planned adaptations in response to changing healthcare contexts caused by COVID-19. We conducted working group meetings and engaged experts in pediatric weight management and telehealth to inform parent survey development, which in turn informed planned program adaptations. We found that parents wanted their care to be a combination of in-person and virtual visits and clinicians wanted virtual care to remain an option. When we further examined the preferences of parents with LEP, we found that in-person visits were better than virtual visits for ensuring their child’s health concerns can be taken care of, sharing private information, and talking about healthy behaviors and mental health as compared to parents without LEP. Clinicians felt virtual visits posed barriers to some aspects of care and they requested resources to improve virtual visits, particularly for families with limited digital health and English proficiency. Based on the information gathered through stakeholder engagement, our team adapted the program for telehealth and other contextual needs due to the COVID-19 pandemic. We made modifications to the core forms of the program that were tailored to both families and clinicians with special attention to ensure that the adaptations met the needs of families who are at risk of poor access to care.

During the COVID-19 pandemic as compared to a pre-pandemic period, childhood obesity rates have rapidly increased [2, 3], and racial, ethnic, and socio-economic disparities that were preexisting have widened [2]. The alarming rise underscores the importance of primary care management approaches and those approaches remaining adaptable to meet evolving needs. Telehealth is an approach to delivering care to children and their families that can promote equity by improving access and participation in weight management programs [24, 25]. Given the high prevalence of unmet social needs for children with obesity, providing care options that address transportation and time barriers can improve access and increase engagement. Studies prior to and during the pandemic have suggested that telehealth for pediatric weight management is acceptable to families, results in high satisfaction with care, improves show rates, and supports behavior changes [11, 26,27,28,29]. We similarly found that families perceived telehealth to be acceptable although we found differences in preferences for virtual or in-person care between parents with and without LEP. In a study conducted before the pandemic and the dramatic rise in telehealth, primary care and specialty care pediatricians were surveyed about their attitudes and experiences with telehealth [30]. Only 13% of clinicians had reported using telehealth compared to almost all of the clinicians in our survey. Sisk and colleagues [30] found that clinicians had identified reasons for not using telehealth including technological barriers, usefulness, and patient reluctance. We found many of these barriers to be of little concern to clinicians presently.

Despite the promise of telehealth improving access to care, telehealth is not without its own barriers to care that need to be addressed [10, 11, 31,32,33]. Although some barriers need to be addressed at policy levels (i.e., broadband infrastructure and reimbursement), other barriers should be addressed at the programmatic or organizational level. As programs are adapted for telehealth, finding the balance between expanding access to care while honoring the preferences of families will be a challenge. In this study, clinicians and parents affirmed that virtual visits improve access to care and wanted a combination of visits, but parents with LEP felt in-person visits were better for some aspects of care. We reconciled these differences by allowing for flexibility of the program to be delivered virtually or in-person, providing parent education about what to expect during a virtual pediatric weight management visit, and developing clinician virtual guidance to support families. Many of these tools could also be used asynchronously allowing families to engage with materials when it is convenient for them.

We found that parents with LEP reported less of a preference for virtual visits than parents without LEP. Similarly, studies have found that adults with LEP had lower rates of telehealth use and LEP has been identified as an important barrier to use [12, 13]. The barrier might be due to clinicians’ comfort and knowledge using telehealth for non-English speakers and integrating interpreters into visits [34, 35]. Additionally, adults with low incomes, less education, or who are racially and ethnically diverse have been shown to be less willing to use telehealth [36]. In our study, although we only examined differences between LEP status, we did find statistically significant differences between race/ethnicity, income, and education for parents with and without LEP as more parents with LEP were Hispanic/Latino, had an income less than $20,000, and had less education. Consistent with other studies, we also found that parents with LEP were more likely to have had a phone visit rather than a video visit [37,38,39] indicating that phone visits need to remain an option as solutions to helping families transition to video visits are designed. As we have done in our study, these factors are critical to systematically assess and address given the persistent racial-ethnic and socioeconomic disparities that exist in childhood obesity [11].

The evidence in the literature and our findings are convincing that primary care-based programs should be equitably adapted for virtual care to ensure all children are equally benefitting. Historically, fidelity to the original intervention and a tolerance for the intervention performing worse in other settings than the controlled setting for the trial was accepted [7]. However, the Dynamic Sustainability Framework posits that change should occur in interventions over time in response to the changing context of care and the setting and the broader system should dictate how the intervention is delivered [7]. Without allowing for change, interventions are less likely to produce improved health outcomes and be sustained. We began this process of adaptations during implementation to ensure the program met the changing care context due to the pandemic. We balanced the need for the adaptations while retaining the core functions that were essential to the effectiveness of the program. We surveyed and continually engaged stakeholders to understand their needs and ensure a good fit as well as considered the needs of populations who have been disproportionately affected by health inequities. Baumann and Cabassa [8] state that a critical element of addressing inequities in healthcare is by developing the science of adaptations and to view adaptations as an implementation strategy. This helps to normalize the adaptation process for researchers and clinicians and there is now an emerging literature of implementation science adaptations [40, 41], although few adaptation studies have focused on equity and childhood obesity [42].

Our study is not without limitations. Connect for Health is being implemented in four healthcare organizations, but the surveys were only conducted in one setting. To account for this, we iteratively sought feedback from the other organizations and conducted planning sessions with them. These other organizations serve both urban and rural areas, but the surveys were conducted in a region that is predominately urban; therefore, the unique needs of rural areas in regard to telehealth may not all have been addressed. In addition, the surveys helped us to understand the experiences, needs, and preferences and differences between parents with and without LEP. From the surveys, we do not understand the “why” of the differences, and conducting interviews in the future may be beneficial. Given the timeframe of the surveys, we also do not know how attitudes towards telehealth have changed over time.

Conclusions

Due to COVID-19-related disruptions to in-person care, we adapted Connect for Health and documented the process for making planned program adaptations. The results of the parent and clinician surveys and stakeholder engagement informed the adaptations. The adaptations were made to address the changing contexts due to the pandemic and related policies and focused on program flexibility for virtual or in-person visits as offering both modalities will enhance patient care for all. This key finding was pivotal for Connect for Health and other programs to ensure equitable access to childhood obesity interventions. The adaptations made were guided by implementation science frameworks and documented to assist with future program evaluation making the process replicable by others to support planned adaptations to meet the evolving and dynamic needs of healthcare settings for continued improvement of health outcomes.