ROUTE-T1D: A Behavioral Intervention to Promote Optimal Continuous Glucose Monitor Use Among Racially Minoritized Youth with Type 1 Diabetes: Design and Development

Abstract

Background:

Type 1 diabetes management is often challenging during adolescence, and many youth with type 1 diabetes struggle with sustained and optimal continuous glucose monitor (CGM) use. Due to racial oppression and racially discriminatory policies leading to inequitable access to quality healthcare and life necessities, racially minoritized youth are significantly less likely to use CGM.

Methods:

ROUTE-T1D: Research on Optimizing the Use of Technology with Education is a pilot behavioral intervention designed to promote optimal CGM use among racially minoritized youth with type 1 diabetes. Intervention strategies include problem solving CGM challenges and promoting positive caregiver-youth communication related to CGM data.

Results:

This randomized waitlist intervention provides participants with access to three telemedicine sessions with a Certified Diabetes Care and Education Specialist. Caregiver participants are also connected with a peer-parent coach.

Conclusion:

Hypothesized findings and anticipated challenges are discussed. Future directions regarding sustaining and optimizing the use of diabetes technology among racially minoritized pediatric populations are reviewed.

Background and Rationale

Type 1 diabetes (T1D)¹ affects nearly 1/400 youth and is one of the most common childhood chronic illnesses [1–3]. Youth with T1D have complex daily medical regimens that include careful monitoring of insulin delivery, blood glucose, nutrition, and physical activity [4]. Less than 17% of adolescents reach the current American Diabetes Association hemoglobin A1c (A1c) recommended target of ≤7% [5]. The steepest rate of deterioration in A1c occurs between ages 10 to 16 [5–7], making adolescence a period of increased risk for T1D-related complications [8].

As a result of centuries of racial oppression and discriminatory policies leading to inequitable access to quality healthcare and resources (e.g., housing, education, employment, clean air/water [9]), it is unsurprising that racialized health inequities exist across several medical conditions, including T1D. Compared to non-Hispanic, white (NHW) youth, Black/African American and Latina/o/x/Hispanic (racially minoritized) youth with T1D are disproportionately more likely to experience elevated A1c, diabetic ketoacidosis, and severe hypoglycemic events [10]. Further, racially minoritized youth are less likely to use advanced T1D technology (e.g., insulin pumps/continuous glucose monitors (CGMs)) to manage diabetes [11,12], regardless of socioeconomic status, insurance type, psychosocial variables, diabetes-specific variables, and parental education [13].

Institutionalized racism in healthcare contributes to, exacerbates, and maintains racial health inequities in medical outcomes for minoritized youth with T1D [14]. The SEARCH study demonstrated that the prevalence of T1D among racially minoritized youth is increasing at a steeper rate than NHW youth [15]. Diabetes technology use improves health outcomes, yet most T1D technology studies have been conducted with samples of >80% NHW [16]. Thus, there is a critical need to reduce the exclusion of racially minoritized youth from T1D studies.

Although diagnosed with T1D at similar rates as English-speaking, NHW youth, racialized health inequities also exist for Latina/o/x/Hispanic families², who are significantly less likely to use diabetes technology compared to English-speaking, NHW families [17]. According to the 2020 Census, 18.7% of the US population identifies as Latina/o/x/Hispanic, and of those, 28% identify as an English language learner [18]. Yet, a recent study found that 91% of pediatric research conducted between 2012–2021 excluded non-English speaking participants, highlighting the need to reduce the exclusion of Latina/o/x/Hispanic families from T1D research [19].

These systemic barriers impact T1D management behaviors, such as inconsistent glucose monitoring, which predicts a high-risk trajectory of negative glycemic outcomes throughout adolescence [20]. CGMs allow for regular glucose monitoring by providing continuous glucose readings every one to five minutes through a sensor-inserted device paired with a receiver that can provide alerts for impending or actual hypo-/hyperglycemia [21]. Consistent CGM use, regardless of insulin regimen, has been shown to improve glycemic and health outcomes [22–25]. However, 10–20% of youth who start a CGM discontinue its use, and 30% of adolescents wear the device less than the recommended minimum 5 days per week [22,23]. Further, healthcare systemic barriers (e.g., insurance status and type, healthcare practice limitations, etc.) and health care professional bias results in fewer CGM and T1D technology recommendations for minoritized youth [13,26]. Compared to NHW youth, minoritized youth are 2.2 times less likely to initiate CGM, and once initiated, are 3.9 times more likely to discontinue CGM, even when controlling for insurance type [27] and other medical and psychosocial variables [13,28].

Regarding family-level characteristics [29], less developmentally appropriate, diabetes-specific caregiver involvement is associated with decreased adolescent T1D management adherence and worse glycemic outcomes [30–36], with racially minoritized families being more impacted [37,38]. Higher adolescent perceived CGM burden and caregiver-adolescent conflict about glucose data have been associated with CGM discontinuation [39–42]. Constant access to shared glucose data also can contribute to diabetes-specific family distress [39,43,44]. Individual level factors can also impact CGM initiation and maintenance, such as annoyances with alerts, interruptions from daily life, body image insecurities, sensor failures, and chronic illness stigma are further associated with inconsistent CGM use among adolescents [28,45].

Tailored education programs delivered by Certified Diabetes Care and Education Specialists have shown improvements in glycemic outcomes [46]. Additionally, peers in the T1D community often serve as role models, which may enhance emotional support for caregivers [47]. Connections with peer parent coaches with lived experiences with the unique cultural nuances to T1D management may be particularly beneficial for racially minoritized youth and their caregivers and may promote sustained CGM use [47].

Thus, there is a significant need to evaluate developmentally- and culturally-tailored behavioral interventions to optimize CGM use and improve health outcomes for racially minoritized youth with T1D. Behavioral strategies that promote peer support and caregiver-youth communication, specifically related to CGM data, may enhance caregiver-youth involvement in diabetes management and support optimal CGM use.

Theoretical Frameworks: Systems-levels, family- and individual-level factors

The foundation for the current study was influenced by several schools of thought, including multilevel frameworks, social determinants of health, and the different generations of racial health equity research [48]. Our team previously published a model on the impact of systemic, healthcare, and individual factors that contribute to long-term T1D health outcomes [49], which was adapted for the current study based on the work of Agarwal and colleagues [13,26,50] (see Figure 1).

Figure 1.

The impact of systemic, healthcare, and individual factors that influence CGM specific behaviors and long-term health outcomes. This figure provides an overview of systemic factors that influence health care professional and youth-caregiver behavior. Youth and caregiver behaviors and communication are reciprocal and influenced by modifiable factors, which in turn, contribute to short-term diabetes self-management behaviors and long-term health outcomes.

The intervention components of Research on Optimizing the Use of Technology with Education (ROUTE-T1D) were developed based on the Self- and Family- Management conceptual framework, which uses a systems and developmental focus for disease management of chronic illnesses (see Figure 1) [51]. This model highlights key contributors of the health care system that may influence T1D outcomes and incorporates multiple ecological systems, such as peer support, cultural considerations, and relationship with health care team members, to improve health outcomes. At the family-level, this framework further highlights the interplay between youth and caregiver efforts for T1D care and aims to improve youth and family interactions around CGM decisions and support problem solving with glucose data.

At the individual level, the ROUTE-T1D intervention is further informed by the Information-Motivation-Behavioral Skills model, which attends to three main factors: an individual’s knowledge about a behavior, performance of the behavior, and skills needed to exhibit the behavior across a variety of settings and situations. The Information-Motivation-Behavioral Skills model, which has previously been used to increase diabetes management behaviors and has particular relevance to the initiation of diabetes technology [52,53], is specifically applied in this study to CGM initiation by providing families with personalized CGM education and support [50].

Study Aims

ROUTE-T1D is a pilot behavioral telemedicine intervention and is one of the first studies to focus on increased recruitment of racially minoritized youth to sustain and optimize CGM use through promoting positive caregiver-youth CGM communication and problem solving. The study aims are:

Aim 1.

To evaluate the feasibility and acceptability of a behavioral CGM support intervention. We hypothesize that the CGM support intervention will be feasible and acceptable (e.g., >50% minoritized youth, high recruitment (>60% of eligible participants) and high retention (>85% at 12 months), and satisfaction (e.g., >80% would recommend the study to others)). We will explore whether feasibility and acceptability rates differ between NHW and minoritized youth, and youth with private vs. public insurance.

Aim 2.

To evaluate the preliminary impact of the intervention on T1D health outcomes. Primary outcomes include A1c and CGM data (i.e., % time in range (70–180 mg/dL), below range (<70 mg/dL), and above range (>180 mg/dL), average glucose, and % time in use or ‘wear time’) at 6-months post-randomization [54]. Secondary outcomes include the perceived CGM benefits and burdens, diabetes management, diabetes distress, and family conflict.

We hypothesize that at 6-months post-randomization, youth who received the immediate behavioral intervention will evidence improved glycemic outcomes (lower A1c, higher CGM-derived time in range, and higher CGM wear time) compared to those in the delayed intervention group (Hypothesis 2.1). We also hypothesize that at 6-months post-randomization, youth in the immediate intervention will evidence reduced CGM burdens, diabetes distress, and family conflict with increased CGM benefits and diabetes management compared to the delayed intervention group (Hypothesis 2.2).

Key demographics (e.g., age, child race/ethnicity, insurance), medical (e.g., insulin regimen type), and behavioral (e.g., T1D adherence) factors may moderate intervention impact (Hypothesis 2.3).

Aim 3.

To examine participant experiences with the intervention and CGM use through qualitative interviews. Youth and caregiver reports will be integrated to provide a rich, detailed evaluation of intervention components and impact. To further evaluate associations of race/ethnicity, insurance, and intervention experience, we also investigate associations among key demographic characteristics with intervention engagement (e.g., session attendance; parent coach contact).

Research Design and Methods

Study Population

ROUTE-T1D is registered on ClinicalTrials.gov (NCT05564481) and is Institutional Review Board-approved. The study has a Safety Officer who will review trial data twice per year. A power analysis indicated the target sample as 60 youth with T1D and their caregivers, to be recruited from an academic medical center in the mid-Atlantic, with >50% of recruited youth identifying as Black/African American or Latina/o/x/Hispanic. The T1D population from which this sample is drawn identifies as 39% white, 29% Black/African American, and 32% identify as either another race, mixed race, or have an unknown race. Families who identify as Latina/o/x/Hispanic make up 14% of the T1D clinic population, and 6% of the clinic population speaks Spanish as the primary language. Over 40% of patients have public insurance. At the time of publication of this paper, data collection is underway.

Youth are eligible if they are (1) between 10–15 years old, (2) have been diagnosed with T1D for ≥6 months, (3) speak and read English or Spanish, (4) are new to CGM within the last 3 months, restarted CGM after >12 months of discontinuation, or wear CGM <75% of the last 3 months determined by Dexcom Clarity Data Sufficiency or caregiver reports, and (5) have access to an internet-enabled device. Youth are not eligible if they have any other major chronic illness, psychiatric disorder, or cognitive limitation. The age range represents a key developmental period that is associated with a decline in T1D management behaviors and glycemic outcomes [5]. Youth diagnosed for ≥6 months were selected to avoid the honeymoon period [55].

Caregivers are eligible if they are (1) the youth’s primary caregiver for diabetes management, (2) are currently living with the youth, and (3) read and speak English or Spanish. In addition to the ROUTE-T1D intervention, participants who are new to CGM proceed with concurrent routine clinical care (standard CGM education).

Recruitment of Participants

To identify potentially eligible participants, a research team member checks the hospital electronic health record for (1) diabetes follow-up appointments 6 weeks in advance of clinic visits and (2) weekly CGM education classes. A trained Research Assistant mails potentially eligible families recruitment letters, and then follows up by email, phone, text, and/or in-clinic to verify eligibility and family interest. Families may be contacted up to 15 times per month of an upcoming appointment (1 letter, 5 phone calls, 4 emails, 4 texts, and 1 in-clinic approach). Phone calls are made at various times to reach caregivers outside of common work hours. We do not systematically exclude families who have missed multiple previous diabetes follow-up appointments (as has previously been done [56]). If the youth and caregiver meet eligibility criteria, the Research Assistant explains the study in detail and offers ample opportunities to ask questions. Once interest is confirmed, an orientation visit is scheduled.

Given that language barriers are a barrier to participation in research for Hispanic/Latina/o/x populations [57], for Spanish-speaking families, recruitment materials were translated by a certified medical interpreter (hospital employee), and longer documents (e.g., questionnaires, consent forms, flyers, and letters) were translated by an outside company, who provided a Certificate of Translation. English study materials were first reviewed by the hospital translator to ensure cultural relevance prior to translation. Additionally, all documents involving patient/family communication met readability and health literacy standards of a 6^th grade reading level [58–60]. In addition to the recruitment methods used for English-speaking families, a Spanish language interpreter, whose presence has been conceptualized to decrease linguistic barriers to research participation and reduce health inequities [61], is present for all recruitment calls/in person approaches. The Spanish language interpreter is a hospital employee who has participated in at least 40–60 hours of trainings, including those relevant to cross-cultural communication and healthcare interpretation. Lastly, we consulted with one physician in the diabetes clinic, who provides medical care in Spanish, on culturally-affirming recruitment methods for Spanish-speaking people, as recruitment by one’s own doctor is a facilitator of research participation for racially minoritized populations [57].

Parent Coach

Prior to youth/caregiver enrollment, we will recruit parent coaches, who are lay persons and parents of youth with T1D who have been using a CGM for ≥1 year. Similar to recruitment and training for parent coaches in other studies [47], parent coaches will be recruited by either nomination by a diabetes care team member or prior participation in a research study. Parent coaches are selected to match the anticipated demographics of study participants. Once recruited, a licensed psychologist with background in T1D psychology delivers a one-hour virtual training session to parent coaches that includes using active listening and basic helping skills, embodying cultural humility, learning about diabetes, mental health and crisis resources, mandated reporting requirements, and delineating parent coach vs. medical team responsibilities. Parent coaches meet with the licensed psychologist for ongoing supervision. After a participant completes the first intervention call, parent coaches make up to 3 attempts per month to contact their assigned caregivers monthly via phone call, text, and email for three months. Coaches complete a brief questionnaire each month tracking the number of times they contacted their assigned participant and a survey reporting their contacts at the end of three months intervention period. Coaches receive $50 in compensation for their training and $70 for each assigned participant.

Orientation Visit

Given that mistrust is a barrier to research participation for racially minoritized groups, a Research Assistant leads a specific, one-hour visit with caregiver/youth participants via Zoom video conferencing, a secure telemedicine platform for the purposes of obtaining informed consent. The Research Assistant obtains informed consent regarding potential risks and benefits of the study. Written consent is obtained via REDCap, a secure, web-based data management questionnaire tool [62]. Once consented, the caregiver and youth each separately complete baseline questionnaires via REDCap. After the questionnaires are complete, the Research Assistant has an open discussion with the dyad to elicit anticipated CGM benefits and burdens, expectations for CGM use, current diabetes management practices, and psychosocial challenges. The purpose of this discussion is to introduce the three intervention sessions and to personalize the parent-coach pairing.

Randomization

Youth are randomly assigned after study enrollment to either Immediate Intervention (n=30) or Delayed Intervention (n=30). Randomization is stratified by youth race/ethnicity (NHW youth or minoritized youth) to ensure balance across conditions. A biostatistician generated the randomization scheme.

Immediate vs. Delayed Intervention Groups

The active phase of the intervention is three-months and study participants are followed for 12 months (see Figure 2).

Figure 2.

Timeline of data collection and study activities for the immediate and delayed intervention group.

Immediate Intervention Group

Caregiver and Youth.

Caregivers and youth complete three individualized sessions with a Certified Diabetes Care and Education Specialists. Given that prior studies show time commitment as a barrier to research participation for racially minoritized people [47], the sessions take place approximately once per month to allow for both practice of skills taught, and for ease of scheduling at the family’s convenience. The goal is completion of all three sessions within the 3-month intervention window. Each session takes approximately 30–45 minutes, and sessions are recorded with permission. At the beginning of each session, youth and caregivers complete a brief questionnaire (created for the intervention) on CGM benefits and burdens for that month so that session content can be personalized. During the session, any adjustment to insulin dosing is documented in the medical chart. For Spanish-speaking caregivers, an interpreter is present during each session. See Table 1 for a breakdown of topics discussed during the three ROUTE-T1D intervention sessions.

Table 1.

Interventions used during the three ROUTE-T1D intervention sessions.

Intervention Session	1	2	3
Topics Discussed	In the first telemedicine intervention session, the Certified Diabetes Care and Education Specialists highlights the importance of developmentally appropriate caregiver involvement in T1D care, introduces diabetes problem-solving and reviews steps to conduct a weekly joint caregiver-youth review of CGM data (“Tech Tune Up”).	The second telemedicine intervention session reviews and problem solves any CGM challenges and elicits feedback from weekly Tech Tune Ups. The session is focused on communication skills training (e.g., active listening) and conflict resolution (e.g., practicing effective communication, setting realistic expectations, and positive reinforcement).	The third and final telemedicine intervention session continues to apply a problem-solving model to address remaining CGM challenges, identify support persons for diabetes management, and create a plan for CGM management beyond the intervention.

Parent Coach.

Caregiver participants are matched with a parent coach at this time.

Delayed Intervention Group

The Delayed Intervention group receives no intervention beyond standard clinical care for the first six months post-randomization. They receive the same study intervention, including the three program sessions and access to a parent coach, starting at 6-months post randomization.

Measures.

Youth and their caregivers complete all questionnaires at baseline, 3-, 6- and 12- months post randomization. Participants have 1 month to complete all questionnaires. For Spanish-speaking participants, all questionnaires are provided in Spanish. See Table 2 for a description of each measure and questionnaire.

Table 2.

Data collected and questionnaires used in the ROUTE-T1D study.

Data Collected	Measure
Demographics	Caregivers report on family sociodemographic information, as well as child diabetes related variables. Youth report their gender identity and youth date of birth/age is obtained from the electronic health record.
Glycemic Outcomes	Hemoglobin A1c, a measure of average blood glucose levels over the past 6–12 weeks, is obtained during each diabetes clinic visit. If a clinic A1c is unavailable, participants are sent a remote A1c kit from CoreMedica Laboratories to complete at each study time point [63,64]. A1c may also be approximated using average CGM glucose readings over the last 90 days (Glucose Management Indicator) when sensor usage is >70%. Both at-home A1c kits and Glucose Management Indicators have shown strong concordance to point-of-care A1c values [65].
CGM Uploads	Participants’ 5CGM devices are uploaded during each clinic visit or participants are invited to share CGM data on the Dexcom Clarity cloud-based platform. RAs extract this information and enter it into the REDCap database. CGM data are collected at baseline (for the proceeding 30 days) and 3-, 6-, and 12-month post-randomization (for the proceeding 14 days). 30-days are collected at baseline to account for those who are new to CGM and adjusting to CGM use.
Health Care Utilization	T1D related complications (e.g., hospitalizations, school absences, severe hypoglycemia, diabetic ketoacidosis) are reported by caregivers, and a Research Assistant also reviews medical records.
Perceived CGM Benefits and Burdens	The Benefits of CGM (BenCGM) and Burdens of CGM (BurCGM) is a validated questionnaire that includes 8 Likert-type items per scale (Benefits and Burdens) to assess the benefits/burdens of CGM, with positively/negatively worded statements starting “I think…” and then a listing a perception related to wearing a CGM that may be good or bad [66]. Higher scores indicate greater perceived benefits or burdens.
Family Conflict	The Diabetes Family Conflict Scale-Revised (DFCS-R) is a caregiver and youth report form including 19-items that assesses caregiver and youth perceptions of disagreements related to T1D care [67]. Both caregivers and youth individually rate each item on a 3-point Likert-scale. Higher scores indicate a greater frequency of conflict.
Diabetes Distress	The Problem Areas in Diabetes Teen Version (PAID-T; 14-items) and Child version (PAID-C; 11-items) assess diabetes-specific emotional distress using teen report (13+ years) and child-report (8–12 years), respectively [68,69]. Using a 6-point Likert scale, teens and children self-report how much each problem associated with diabetes applied to them over the past month. The Problem Areas in Diabetes-Parents scale (PAID-PR) is an 18-item caregiver-report questionnaire, where caregivers report their own distress about their child’s diabetes [44]. Higher scores indicate more diabetes-specific distress.
T1D Management	The Diabetes Management Questionnaire (DMQ) is a 20-item youth- and caregiver-reported questionnaire that assesses diabetes care behaviors (e.g., blood glucose monitoring, adjusting care behaviors to specific situations) over the past month [70]. Higher scores indicate greater adherence to diabetes management.
Qualitative Exit Interviews	The purpose of the exit interview is to give “voice” to our participants as a means to include participant feedback into the next iteration of our study [71]. The interview is audio-recorded and thematically coded to gain insight into experiences with the intervention, including satisfaction, perceived impact, ease of adhering to study components, and challenges associated with caregiver-youth interactions around CGM use and diabetes management.
Parent Coach Questionnaires	Parent Coaches self-report on family sociodemographic information, child diabetes related variables, as well as current diabetes management prior to their training. While they have participants who are actively enrolled in the study, they complete a monthly text questionnaire to report on any contact between themselves and their participant. After completion of the 3 months being connected with a participant, coaches complete another questionnaire about contact and topics discussed. Once a parent coach ends their study participation, they complete a final questionnaire about study feedback. They also have the option to complete an exit interview to share overall experiences with assigned participants, including satisfaction and perceived impact.

Strategies to promote engagement and retention throughout the follow-up period include maintaining the same research staff that participants interact with, which helps build trust, safety, and rapport, and flexibility when scheduling recruitment calls and intervention sessions to reduce time burdens. Text messages, phone calls, and email are used to communicate with participants depending on participant preference to increase convenience. Families are financially compensated (via ClinCard) at each data collection point and can earn up to $200 for participating in the study, Additionally, before questionnaires are sent at each follow-up period, retention items with the ROUTE-T1D logo (e.g., water bottle, pen, phone holder) are sent to both the caregiver and youth.

Data Analysis

Power

Sample size was determined by feasibility considerations and prior experiences with similar studies. The proposed sample of 60 dyads will be sufficient to assess the intervention’s effect at 6-months post-randomization. A delayed intervention group, with no intervention in the initial 6-months post-randomization, will be used as the comparison sample. Based on clinical sample and prior studies, our expected Mean±SD of A1c at baseline is 8.8±1.7%. Thus, a 1:1 study with 60 dyads and 85% retention rate will result in a precise estimate of intervention’s efficacy compared to standard care, measured by difference in Mean A1c between arms, with standard error of 0.3% before any adjustments. Estimating effect size and sampling variability in this study is a critical step in sample size and power determination of a future randomized controlled trial to inform feasibility and health impacts.

Planned Statistical Analyses

Missing Data

We will make several attempts to minimize missing values on all outcomes and document possible reasons for missingness. Our analyses also will handle missing outcome values using multiple imputation methods and generating 15 imputed datasets given all measured baseline characteristics before carrying out analyses using library miceadds in R [72]. The resulting effect size estimates and standard errors will account for sampling variability and variability.

Aim 1.

Descriptive statistics will be used to assess study feasibility/acceptability. Feasibility and acceptability include enrollment of >50% minoritized youth, and overall increased uptake of CGM among racially minoritized youth and youth with public insurance compared to NHW youth or youth with private insurance, respectively. Feasibility and acceptability benchmarks include: recruitment (>60% of eligible participants), high retention (>85%), sessions attended (>80%), as well as parent coach contacts (>80% with ≥1 contact).

Satisfaction benchmarks include post-intervention satisfaction questionnaire at 3- and 12-months post-randomization for the immediate and delayed groups, respectively (80% reporting satisfaction and perceived utility and benefit).

We will also examine whether rates of feasibility, acceptability, and retention at 3- and 12-months post-randomization differs by key demographic characteristics.

Aim 2.

Primary clinical outcomes at 6-months post-randomization are A1c and CGM statistics. Secondary outcomes at 6-months post-randomization include youth and caregiver total scores on the BurCGM, BenCGM, DMQ, PAID, and DFCS.

For Hypothesis 2.1 and Hypothesis 2.2 examining clinical and psychosocial outcomes, we will compare the Immediate Intervention arm to the Delayed Intervention arm at 6-months post-randomization using unadjusted and adjusted analyses. We expect unadjusted analyses to estimate effect size accurately and adjusted analyses to increase precision of these estimates. Unadjusted univariate analyses for A1c, CGM, and psychosocial (continuous) measures will compare the two arms at 6-months post-randomization using 95% confidence interval for mean difference. If there are departures from normality (e.g., variance increases with mean) for CGM statistics, we will report the 95% confidence interval of the rate ratio using exact methods.

In univariate analyses for group comparisons, we will compare measures at baseline with those at 6-months post-randomization using a paired t-test. In multivariate analyses, we will quantify the impact of the intervention on clinical outcomes with mixed effect models and report model-based 95% CIs for baseline/post intervention effect for the immediate intervention group as compared to delayed intervention at 6-months post-randomization. We can conduct similar univariate and multivariate models for within subject comparisons. The dependent variable will be the primary outcome post-intervention, the independent variables will include an indicator of whether the measurements were taken from baseline or post-intervention, the first measurement in the study at baseline, a linear trend for time post-intervention, and a random effect for participant to adjust for correlation of measures within the same participant. For CGM outcomes, which are counts or rates, we will report the 95% confidence interval of the rate ratio using exact methods and fit a multivariable regression after log transformation or use a Poisson mixed effect regression accounting for overdispersion.

For Hypothesis 2.3, we will examine the effect of modification of the interventions on each outcome by different baseline characteristics including demographics, medical, and changes in youth- and caregiver-reported T1D adherence. We will summarize the mean outcome for different subgroups defined by each characteristic. We will compare association between change in outcomes from baseline to post-intervention with age, child race/ethnicity, insurance, regimen, and T1D adherence using independent t-test. A p-value of >0.10 will indicate possible effect modification. We will add those characteristics that show (univariate) association with intervention on a particular outcome (p-value >0.10) to multivariable models and assess interaction of characteristic with intervention.

Aim 3.

We will use qualitative coding to evaluate youth and caregiver themes related to intervention and CGM experiences. Interviews will be transcribed and coded for content by a senior research staff member. Thematic prevalence will be determined by the proportion of interviews to which a code is applied [73]. Youth and caregiver report will be integrated to provide a rich, detailed evaluation of intervention components and impact.

Discussion

ROUTE-T1D is one of the first studies to examine behavioral support for sustained and optimal CGM use while centering racially minoritized youth with T1D. This novel approach may improve diabetes outcomes and prevent CGM discontinuation or resistance later in adolescence as youth experience more independence with T1D care.

The current study places a strong emphasis on evaluating the intervention with racially minoritized youth who are historically excluded from T1D technology research and may experience unique, barriers to CGM uptake and sustained use [27,74]. Further, the demands of using a CGM may be a source of conflict for both youth and caregivers [40,41]. This intervention provides an opportunity for more frequent support, which may help equip youth and caregivers with tools to solve the problems of managing T1D with a CGM.

This study is limited by using a single center. Also, the study is limited by the setting in which it is conducted, including CGM education and procedures specific to this academic medical center, all limiting generalizability. Further, our study focuses on reducing a specific systemic barrier, which is the historical exclusion of minoritized youth from T1D technology research and health care professional bias; however, many other systemic barriers exist when it comes to equitable medical care for minoritized youth, and this study did not specifically address these other barriers [13]. Additionally, our study did not specifically recruit individuals identifying as Indigenous, Asian/Asian American, Native Hawaiian, Alaskan Native and Pacific Islanders, Middle Eastern/North African, mixed race or another racial/ethnic identity, who have all been historically subject to systemic racism and experience racial health inequities as a result; future studies should also aim to examine the unique experiences of these racial/ethnic groups.

Conclusion

The current study aims to evaluate the acceptability, feasibility, and effect of a behavioral intervention designed to promote optimal CGM use among racially minoritized youth with T1D. This study provides the opportunity for clinically meaningful change in access and continued use of CGM technologies among youth with T1D, particularly those who have been historically excluded from T1D technology research. The Certified Diabetes Care and Education Specialists-delivered intervention combined with accessible, low-cost delivery methods (i.e., telemedicine) offers a strong potential for impact and translatability. The results of this pilot study will aid in addressing the needs of racially minoritized youth related to diabetes technology and have the potential to improve health outcomes and clinical diabetes education and care. If successful, the next step would be a larger multi-site randomized trial to better evaluate the intervention’s efficacy and longer-term impact on glycemic outcomes among racially minoritized youth.