DiaBetter Together: Clinical trial protocol for a strengths-based Peer Mentor intervention for young adults with type 1 diabetes transitioning to adult care

Abstract

Background:

Type 1 diabetes (T1D) management is challenging for young adults, who are expected to transfer from the pediatric to adult T1D healthcare system while also managing typical developmental demands (e.g., social, financial, work/school, residential). Many young adults have extended gaps in care before following up in adult care, increasing risk for poor health outcomes. There are few evidence-based programs to support young adults with T1D to promote a timelier transition during this period. This paper reports on the design of DiaBetter Together, a randomized controlled trial to evaluate a 12-month Peer Mentor-delivered intervention compared to usual care among young adults with T1D during the transfer from pediatric to adult care.

Methods:

One-hundred young adults (age 17–25) with T1D and 29 Peer Mentors enrolled in this randomized clinical trial. Peer Mentors are experienced, older young adults with T1D, trained by the study team to share transition experiences and strategies to successfully navigate the adult healthcare system, help young adults prepare for the first adult care visit, and use strengths-based support strategies to teach and model skills for managing T1D-related challenges.

Results:

The primary outcome of the trial is HbA1c, and secondary outcomes include time to adult care, engagement in diabetes self-management behaviors, and psychosocial well-being.

Conclusion:

The goal of this research is to evaluate a developmentally appropriate, supportive intervention that can improve T1D self-management and successful transfer of care during the difficult young adult years and promote optimal T1D health outcomes.

1. Introduction

About 244,000 youth age < 20 and 1.6 M adults in the U.S. are diagnosed with type 1 diabetes (T1D), and the prevalence is increasing [1,2]. T1D is a chronic condition requiring intensive management, including insulin dose calculations and administration, glucose monitoring, attending to nutrition, regular healthcare appointments, and managing medical supplies [3]. Few adolescents (17%) and young adults (14%) with T1D have HbA1c values (representing overall glycemic control) below the American Diabetes Association (ADA) target of <7.0% [4].

Young adults experience social/interpersonal, financial, educational, occupational, and residential changes [5], which co-occur with T1D-related changes, such as assuming greater responsibility for self-management and transferring from pediatric to adult healthcare systems [6,7]. For young adults with T1D, diabetes distress [8,9], higher HbA1c [4], and increased hospitalization risk [10] are common, and nearly one-third experience early diabetes-related complications [11]. Extended gaps (>6 months) between pediatric and adult diabetes care are a significant barrier to achieving optimal outcomes [10,12–14]. Many young adults are not adequately prepared to leave pediatric healthcare [15,16], yet transition interventions for young adults with T1D are not readily available or easily accessible [7,10]. The few existing interventions primarily comprise logistical resources (e.g., appointment reminders) [17,18] or an intermediate transition to a young adult clinic [19] and have modest outcomes. Novel methods are needed to facilitate a timely transfer to adult care and promote optimal glycemic outcomes [7,10].

Peer support has been identified as a helpful resource for young adults with T1D [20–22], and they express interest in receiving peer support to help overcome transition barriers [23]. Potential benefits of diabetes-related social support include reduced dropout from medical care, less diabetes burden and distress, and improved health outcomes [17,20,22–24]. However, young adults with T1D report receiving less support than peers without T1D [25,26]. Among adults with type 2 diabetes, peer-led behavioral interventions have demonstrated efficacy in reducing isolation, increasing perceived support, enhancing self-management, and improving adherence and glycemic outcomes [27–31]. Peer-led interventions have also demonstrated promise in improving HbA1c and clinic attendance in adolescents with T1D [32]. A systematic review of peer support interventions concluded that youth with T1D reported feeling emotionally supported, encouraged to seek additional resources, and more confident in managing T1D and working through diabetes experiences [33]. By sharing experiences, providing emotional support, facilitating access to resources, and promoting self-efficacy, peer mentorship can reduce barriers to T1D self-management [20,29,33,36]. Greater engagement in T1D self-management behaviors leads to improvement in diabetes-related outcomes, including glycemic outcomes [28,37]. Peer-based interventions also improved well-being and retention in first year college students [34], highlighting the developmental appropriateness of this approach. Additionally, peer-led interventions benefit peer mentors’ own health [32,35]. Despite these findings, peer support approaches have not been applied to healthcare transition interventions for young adults with T1D.

1.1. Study aims

The aim of the “DiaBetter Together” trial (NCT04247620) is to evaluate the impact of a novel, strengths-based, social support-focused transition intervention for young adults with T1D during the transfer between pediatric to adult care, compared to usual care. DiaBetter Together is delivered by Peer Mentors, or “near peers” [38], older young adults who have T1D and share experiences around T1D management and transitioning to adult diabetes care.

Hypothesis 1. Intervention participants will have lower HbA1c (primary), shorter time to first adult care visit, higher engagement in diabetes self-management behaviors, and better quality of life (secondary) at one year.

Hypothesis 2. Intervention participants will report more transition readiness skills, perceived support, diabetes strengths, and lower diabetes distress and depressive symptoms at 6 months, and these factors will mediate intervention impact on primary and secondary outcomes at 12 months.

We will also explore the impact of delivering the intervention on Peer Mentors’ own self-management behaviors, diabetes distress, perceived social support, and quality of life.

2. Study design and methods

2.1. Overview

This paper describes the study design, intervention protocol, and baseline characteristics of young adult participants and Peer Mentors. Recruitment and baseline data collection took place from February 2021 through June 2023, and the intervention is ongoing through June 2024. We use past tense for activities through baseline and future tense for other ongoing study activities.

Participants were enrolled and randomized to the intervention or usual care for the 12-month intervention period. Data are collected at baseline, 6- and 12-months post-randomization. Participants receive financial compensation via a reloadable debit card for completing data collection throughout the study (maximum $235). The Institutional Review Board at Baylor College of Medicine approved this study protocol. We convened a DSMB semi-annually throughout the duration of the study.

2.2. Conceptual framework

DiaBetter Together has theoretical foundations in Social Cognitive Theory [39] and the Diabetes Resilience Model [40]. The intervention targets health behaviors including self-efficacy, outcome expectations, goal-setting, and social and environmental factors, as Peer Mentors provide social support, peer modeling, accountability, reinforcement, and relevant information. The Diabetes Resilience Model posits that people with diabetes achieve resilient outcomes (i.e., low HbA1c, high engagement in self-management behaviors) by using protective processes to reduce the impact of risk factors [40]. This intervention uses a strengths-based approach to reinforce young adults’ strengths and successes during the transition process to promote positive outcomes [41].

2.3. Recruitment and enrollment

The target sample for this study was n = 150 young adults (ages 17–25) with T1D, beginning the transfer from pediatric diabetes to adult care. Based on the clinical importance of a 0.5 percentage point difference in HbA1c [42,43], the power analysis indicated a sample size of 128 participants (64/group) would provide 80% power to detect this difference using an independent two-sample t-test assuming equal variances, common SD = 1, and two-sided α = 0.05. We planned to enroll and randomize 150 young adult participants (75/group) to allow for up to 15% attrition based on retention rates ranging between 86 and 96% in our previous intervention studies. The target sample was adjusted to n = 100 due to delays and adaptations made related to the COVID-19 pandemic [44,45]. This study will have about 70% power to detect a 0.5-point difference in mean HbA1c between treatment arms if the observed SD = 1. This will decrease to about 24% power if SD = 2. The detectable effect sizes (80% power) will be about 0.57 and 1.13, respectively.

Eligibility criteria includedbeing diagnosed with T1D for ≥1 year, currently receiving T1D care at Texas Children’s Hospital (TCH), fluent in reading and speaking English, and must have completed their expected final pediatric diabetes care appointment at TCH. Participants were randomized if consent and baseline data completion occurred within 2 months following this appointment. Recognizing that young adults may return to pediatric care in some circumstances (e.g., hospitalization for DKA, insulin prescription refill), enrolled participants who subsequently receive care in the pediatric setting do not have a change in eligibility. Exclusion criteria included having serious medical, cognitive, or mental health comorbidities that would preclude ability to provide informed consent or participate in study activities.

To identify potentially eligible young adults, study staff reviewed upcoming diabetes care appointments and systematically pre-screened electronic medical records (EMR) for inclusion and exclusion criteria. Study staff requested confirmation from pediatric providers about transition plans to adult care for upcoming appointments. If an upcoming TCH diabetes appointment was confirmed to be the last before transferring to adult care, study staff sent recruitment letters (including intervention summary, timeline, contact information) via mail and EMR patient portal. After their final TCH diabetes visit, study staff met with young adults by phone or in-person to confirm eligibility, explain the study in detail, and obtain informed consent. For 17-year-olds, a parent/legal guardian provided consent, and the participant provided assent and signed consent after turning 18. Study staff emailed a link to an electronic consent form on a secure, HIPAA-compliant web portal operated by Baylor College of Medicine’s Clinical Trials Management Systems (CTMS). If they did not complete the final appointment or indicated a plan to return for another appointment, study staff monitored the schedule and re-approached the young adult after the subsequent diabetes care appointment.

Peer Mentors were also enrolled as participants to collect process and outcomes data. Eligibility criteria for Peer Mentors included age 20–35, diagnosed with T1D before age 18, diabetes duration of ≥1 year, previously treated in pediatric diabetes care and currently treated in adult care, and fluency in reading and speaking English. Exclusions included serious medical, cognitive, or mental health comorbidities that would preclude ability to deliver the intervention or complete questionnaires. Study staff recruited potentially eligible Peer Mentors through referrals from pediatric providers at TCH, advertisements in local adult diabetes centers, online diabetes groups, and social media. After potential Peer Mentors expressed interest, study staff met with them by phone or video to confirm eligibility and determine their ability to effectively deliver the intervention. After a detailed description of the study, including the responsibilities of being a Peer Mentor and data completion, those who were eligible and interested were emailed a link to the electronic consent form on the CTMS portal.

2.4. Baseline data collection

Following consent, participants were emailed a new link to access baseline questionnaires via the CTMS portal. HbA1c was obtained from EMR or home HbA1c kit.

2.5. Randomization

Young adults were randomized to intervention or usual care in a 1:1 ratio using a permuted-block design. The 12-month study period began on the date of randomization. Those randomized to intervention were assigned to a trained Peer Mentor following randomization.

3. Intervention arms

3.1. Usual care

Participants in both arms receive usual diabetes care, which consists of routine diabetes treatment and transition resources from the pediatric care team, including a list of local adult diabetes care providers. The pediatric practice is available to patients as needed during the transition from TCH to an adult care provider, including access to medical providers, diabetes educators, and social workers. In both arms, study participation does not impact participants’ ability to contact the pediatric diabetes team or other medical services to receive care. Participants randomized to the usual care comparator arm participate in all study activities related to data collection and only receive usual medical care, without the peer mentor intervention.

3.2. Peer mentor intervention

Before enrolling Peer Mentors, an advisory board of 6 young adults with T1D who had previously transitioned to adult care reviewed topics related to transitioning to adult healthcare and provided feedback about length of surveys and amount of contact between Peer Mentors and participants. Advisory board members shared their experiences and thoughts about barriers to the transition process, including what they felt would have helped them as they transitioned from pediatric to adult care. Their input helped us select topics to include in the intervention manual for Peer Mentors. Advisory board members also reviewed early drafts of the study manual and provided feedback about the pertinence and importance of the topics and resources included.

Peer Mentors follow a DiaBetter Together intervention manual with specified topics, skills, and resources (Table 3), and connect with their assigned young adult over the 12-month intervention period to encourage successful transition to adult diabetes care. Within 1–2 weeks of being paired, each Peer Mentor meets with their assigned young adult virtually using the participant’s preferred medium (e.g., video, telephone) to introduce the intervention, share personal experiences with transition and living with T1D, learn about the participant’s experiences with diabetes, and discuss goals for the transition process. Peer Mentors then contact each participant approximately weekly during the first 1–3 months, every 2–3 weeks during months 4–6, then approximately monthly for the remainder of the year. Participants can contact their Peer Mentors more often with questions or for additional support. Connections take place via the communication method(s) preferred by the young adult. Due to COVID-19 pandemic restrictions, all Peer Mentor-participant contacts are remote.

Table 3.

Peer Mentor Intervention Topics and Skills.

Theme	Topics	Skills	Resources	Approx Timing
Strengths-based Support	Positive, encouraging discussion and resources for transition to adult care Encouragement and accountability for goals: praise for successes, assistance with problem-solving barriers	Recognizing successes	Links to relevant resources from Got Transition, The Diabetes Link (previously College Diabetes Network), JDRF, American Diabetes Association [ADA], and other resources	Across 12 months
Establishing Peer Mentoring Relationship	Rapport-building Personal experiences with diabetes Transition goals, concerns, plans, questions	Establish communication plan Sharing personal experiences	Any from manual, as relevant to follow-up on conversation	First few weeks
Establishing Adult Care: Before First Medical Appointment	Goals for the transition process Selecting adult healthcare provider Making appointment Navigating adult healthcare system Preparing for first appointment	SMART goals Differentiating adult provider types Identifying personal considerations and preferences in choosing a healthcare provider Insurance options and cards Planning ahead, time management Safety plan for T1D emergencies	Reminder apps Adult endocrinologist lists from pediatric providers Links to relevant resources from The Diabetes Link, ADA Information regarding the use of telehealth Study created checklist including what to prepare and bring to first adult visit (lists of prescriptions, insurance card, etc.)	First 4 months Emphasize in subsequent months if no appointment made by 4th month.
Establishing Adult Care: After First Medical Appointment	Debrief after first appointment Handling missed appointment	Self-Awareness, Fit with adult provider Build relationship with provider Plan for second visit Preventing missed appointments	Links to relevant resources about establishing a relationship with your doctor and tips for successful medical visits	After completion of adult health care appointment
Healthcare Resources	Health Insurance Planning for emergencies	Seeking support Learning about insurance and accessing insurance Preparing for emergencies Making an emergency kit Understanding symptoms of illness- (DKA, COVID-19, etc.) Organizing supplies	Links to relevant resources about types of insurance from JDRF and Beyond Type 1 CMS marketplace Patient assistance programs- list created by study team Healthcare.gov Links to resources about managing sick days from the Diabetes Link Study-created sick day management guidelines and supplies	Across 12 months
Taking the Lead in Self-Management	Leading diabetes self-management Teach Problem Solving Maximizing Technology Sharing diabetes data; setting boundaries Prioritizing diabetes	Stress management Problem solving Seeking support Celebrate successes Finding apps to help Finding trusted online resources and information How to set boundaries with others Setting goals Time management	Links to relevant resources from Beyond Type 1, The Diabetes Link, DiaTribe Medical ID on phone	Across 12 months
Managing Stress and Psychosocial Concerns	Strengths based stress management Mood and diabetes burnout Drinking with diabetes	Self-awareness Identify and use personal strengths and resources Know signs of struggling and how to ask for help Seek Social Support Practice Gratitude Relaxation strategies: sleep, exercise, deep breathing	Links to relevant resources from The Diabetes Link, Beyond Type 1, Diabetes Strong	Across 12 months
Social Support and Communication	Mobilizing support network Finding new support Developing a working relationship with medical providers Self-advocacy	Identify and use support network, growing network How to ask for help Connecting with online community Telling others about diabetes Building a relationship with medical provider Speaking up for self	Links to relevant resources from Psychology Today, The Diabetes Link, JDRF, Everyday health, ADA	Across 12 months

The Peer Mentor-young adult relationship aims to provide peer support around the transition to adult care, including navigating the transfer of medical care between a pediatric and adult healthcare setting, prioritizing diabetes self-management among competing demands, shifting relationships, and integrating life with T1D into everyday experiences of young adulthood. With a focus on supporting participants to establish adult diabetes care, soon after leaving pediatric care, Peer Mentors guide participants in identifying an adult diabetes care provider, scheduling and preparing for their first appointment, and debriefing after. In addition to discussing skills such as problem-solving, goal setting, time management, and stress management, Peer Mentors share their transition experiences, provide positive feedback, share relevant resources, and encourage participants to access support from their social networks (Table 3).

3.3. Peer mentor training

Peer Mentor Liaisons (clinical social worker and psychologist) oversee all aspects of Peer Mentors’ experiences in the study, including recruitment, screening, training, assigning participants, collecting data, and providing supervision and feedback. Peer Mentors were enrolled and trained between December 2020 and August 2022. Due to the COVID-19 pandemic, all Peer Mentor activities were completed remotely. The training protocol included self-paced readings of intervention materials and two approximately 2-h online training sessions. Following best practices for training peer leaders [46–48], training included guidance about working with young adults with T1D, an overview of the study protocol, orientation to the intervention manual and resource lists, interactive training in active listening skills, instruction about the limitations and scope of the peer mentor role (e.g., not to offer medical or mental health advice), and role-plays of skills. Trainers emphasized research ethics and provided instruction in protecting participant confidentiality and following protocols related to mental health or safety concerns. Following the training, Peer Mentors have at least monthly phone calls with the Peer Mentor Liaison to review contacts with their assigned young adult(s), discuss concerns, and explore relevant topics for discussion with participants.

4. Data collection

4.1. Assessment schedule

Baseline assessments were completed within two months following participants’ final pediatric diabetes care appointment, prior to randomization. Follow-up assessments take place 6- and 12-months post-randomization, each with a 2-month window. Participants complete self-report questionnaires via CTMS, a HIPAA-compliant survey portal. Study staff extract clinical data from the participants’ pediatric EMR during the study. At 6- and 12-months, participants self-report the names and specialties (e.g., endocrinologist, primary care) of adult healthcare professionals (including urgent care or emergency department visits) they have seen for diabetes care during the study and provide a signed release of information form. With this release, study staff collect information about completed appointments and clinical data from adult care.

4.2. Measures

4.2.1. Diabetes data

4.2.1.1. Glycemic outcome.

HbA1c, a measure of average blood glucose over the previous 3–4 months, is the primary indicator of overall diabetes-related health recommended by the ADA [49]. Study staff extracted HbA1c values from the pediatric EMR (performed using point-of-care DCA 2000 HbA1c Analyzers [Siemens-Bayer] or outside professional labs) at baseline. We mailed a home HbA1c kit to participants without a documented HbA1c within 2 months of their final pediatric appointment. At the 12-month time point, all participants receive a home HbA1c kit and pediatric and adult medical records are reviewed for HbA1c values documented since baseline. Home HbA1c kits include filter paper to provide a dried blood spot, a single-use lancet, alcohol wipe, and instructions. Participants return the completed card (packed with a desiccant pack to minimize risks to sample integrity during shipping) to the study team using postage-paid return mailing materials. The TCH Clinical Chemistry laboratory processes the samples on the VITROS 5600 HbA1c Assay, which is correlated semi-annually with the point-of-care DCA 2000 HbA1c Analyzers at TCH. Systematic review and meta-analysis results demonstrate high correlations between HbA1c assays from dried blood spot samples and standard venous samples [50,51]. Intra assay precision of the dried blood spot methodology was demonstrated to be 0.9–1.5% [52]. Dried blood spot samples using the same or similar methods have been used in other clinical trials with adolescents and young adults with T1D [53–55].

For participants using a continuous glucose monitor (CGM), glucose time in range (TIR) is also collected as an exploratory, secondary measure of a glycemic outcome: the percent of time glucose is within the target range of 70–180 mg/dL. Study staff download CGM data for 14 days at baseline and 12 months. Percent glucose TIR is based on a target range of 70–180 mg/dL.

4.2.1.2. Follow-up in adult care.

At 12 months, participants self-report dates of all adult healthcare appointments completed since baseline. Study staff confirm completed appointments from adult medical records to calculate time to first follow-up appointment in adult care.

4.2.2. Participant reported measures

See Table 4 for a detailed list of psychological, behavioral, and fidelity measures used.

Table 4.

Psychological, behavioral, and fidelity measures.

Measure	Construct(s) Measured	Time Points	Number of Items	Notes	References
QUALITY OF LIFE
Type 1 Diabetes and Life (TIDAL)	TID-specific health-related quality of life	Baseline, 12 months	Adolescent version (age 17): 23 items; Young Adult version (18-−25): 27 items	Total scores calculated on a 100-point scale; higher scores indicate better quality of life	[64,65]
Satisfaction with Life Scale	General quality of life	Baseline, 12 months	5 items	Each item rated on a 7-point scale from “Strongly Disagree” to “Strongly Agree”	[66]
DIABETES SELF-MANAGEMENT
Self-Care Inventory-Short Form (SCI-SF)	Engagement in self-management behaviors	Baseline, 6 months, 12 months	9 items	Higher scores indicate higher adherence to diabetes self-management guidelines	[67]
TRANSITION READINESS
The Readiness for Emerging Adults with Diabetes (READDY)	Transition readiness	Baseline, 6 months (Navigation subscale only), 12 months	Knowledge: 13 items; Navigation: 13 items, Health Behaviors: 9 items	Each item is rated using a 5-point scale ranging from "“I Haven’t Thought About This"” to "“Yes, I Can Do This,"” higher scores indicate greater transition readiness	[68]
SOCIAL SUPPORT
Patient Reported Outcomes Measurement Information System (PROMIS) Short Forms	Emotional support, informational support, social isolation	Baseline, 6 months, 12 months	Emotional Support: 4 items; Informational Support: 4 items; Social Isolation: 1 item	Total raw score is calculated by summing the items; raw score is converted to a scaled score; higher scores reflect greater levels of feelings of support or isolation	[69]
Brief 2-Way Social Support Scale	Perceptions of giving and receiving instrumental and emotional social support	Baseline, 6 months (Receiving Support subscales only), 12 months	12 items	Each item is rated on a 6-point scale ranging from "“Not at All"” to "“Always"”	[70]
DIABETES STRENGTHS
Diabetes Strengths and Resilience (DSTAR)	Adaptive attitudes and behaviors related to living successfully with T1D	Baseline, 6 months, 12 months	Teen version adapted for Young Adults: 16 items	Each item is rated on a 5-point scale from "“Never"” to "“Almost Always"”	[71]
MOOD CONCERNS
PROMIS Depression Short Form 4a	Symptoms of depression	Baseline, 6 months, 12 months	4 items	Each item rated on a 5-point scale from "“Never"” to "“Always;"” a total raw score is calculated by summing items; raw score is converted to a scaled score; higher scores reflect greater levels depressive symptoms	[69,72,73]
Diabetes Distress Scale for Adults with T1D (DDS-T1)	Diabetes distress, emotional burdens related to living with T1D	Baseline, 6 months, 12 months	28 items	Each item is rated on a 6-point scale ranging from "“Not a Problem"” to "“Very Serious Problem;"” higher scores indicate greater diabetes distress	[74]
STRESSFUL LIFE EVENTS
Stressful Event Questionnaire	Number and types of stressful life events experienced during the past year	Baseline, 12 months	10 items	Participants select events from a list and make note of any not listed; total number of selected events are summed	Study created
SLEEP
Pittsburgh Sleep Quality Index (PSQI) - Revised	Dimensions of sleep health	Baseline, 12 months	4 items about sleep time to calculate sleep latency, efficiency, and duration; 4 items about sleep disturbances (including 2 items created for this study about diabetes-related sleep disturbances)	Abbreviated version. Each item is rated on a 4-point scale ranging from "“Not During Past Month"” to "“Three or More Times a Week,"” and 1 item on subjective sleep quality on a 4-point ranging from "“Very good"” to "“Very bad"”	[75]
PANDEMIC EXPERIENCES
COVID-19 Experiences Questionnaire for Young Adults with T1D (CEQ-YAD)	Impact of the COVID-19 pandemic on various life circumstances, health behaviors, psychosocial concerns, T1D management, and transition to adult healthcare for young adults with T1D	Baseline, 12 months	33 items	Uses scaled, yes/no, and open-ended items	Study created
PREVENTION BEHAVIORS
Oelsner MESA COVID-19 - Revised	Engagement in social distancing and hygiene behaviors during the COVID-19 pandemic	Baseline, 12 months	11 items	Rates how often participants engaged in specific precautionary behaviors at the peak of the pandemic (Spring/Summer 2020) and over the past month; item responses ranged from "“Never"” to "“Always"”
VACCINATION
COVID-19 Vaccination	COVID-19 vaccine status	Baseline, 12 months		Report of the month/year and manufacturer of any COVID-19 vaccinations received; if no vaccinations reported, participants respond to an open-text question about the reason(s)	N/A
PEER MENTOR MEASURES
PEER MENTOR OUTCOMES
Type 1 Diabetes and Life (TIDAL)	TID-specific health-related quality of life	Baseline, Study Completion	Young Adult version (age 18–25); Adult-1 version (age 26–45): 27 items	Each item is rated using a 5-point scale ranging from “No, Not At All True” to “Yes, Very True.” Total scores are calculated on a 100-point scale; higher scores indicate better quality of life	[64,65]
Self-Care Inventory-Short Form (SCI-SF)	Engagement in self-management behaviors	Baseline, Study Completion	9 items	Each item is rated using a 5-point scale from “Never” to “Always;” higher scores indicate higher adherence to diabetes self-management guidelines	[67]
Brief 2-Way Social Support Scale	Perceptions of giving and receiving instrumental and emotional social support	Baseline, Study Completion	12 items	Each item is rated on a 6-point scale ranging from “Not at All” to “Always”	[70]
Diabetes Distress Scale for Adults with T1D (DDS-T1)	Diabetes distress, emotional burdens related to living with T1D	Baseline, Study Completion	29 items	Each item is rated on a 6-point scale ranging from “Not a Problem” to “Very Serious Problem;” higher scores indicate greater diabetes distress	[74]
FIDELITY
Audio-recording of Peer Mentors' first session with each participant	Peer Mentor's fidelity to intervention	After Peer Mentor's first session with each participant	9 items	Checklist of intervention components and delivery of skills as outlined in the intervention manual, filled out by Peer Mentor Liaisons based on audio-recording of Peer Mentors' first session with each participant. If first session not recorded, Peer Mentor Liaison will discuss details of the call with Peer Mentor and complete the fidelity form accordingly.	Study created
Peer Mentor Brief Fidelity Form	Brief fidelity form regarding contacts with participants over the previous 6 months	Intervention midpoint (6 months), post-intervention (12 months)	9 items	Peer Mentor fidelity form tracks frequency of contacts, who initiated contacts, participant responsiveness to Peer Mentor outreach attempts, challenges Peer Mentors encountered in connecting with young adults, and ratings of perception of closeness of the mentoring relationship	Study created

4.2.3. Participant characteristics

4.2.3.1. Demographics.

Participants completed a demographic questionnaire at baseline and provide relevant updates at 12 months. Demographic information includes gender, race, ethnicity, and marital/family information. Indices of socio-economic status include school financial aid qualification, work/volunteer status, education level, and health insurance coverage/source. Additionally, we calculated the CDC/ATSDR Social Vulnerability Index (SVI) [56] at baseline, based on each participant’s primary address. We used ArcGIS Pro Desktop software and the Street Map Premium package to batch geocode and transform addresses into geographical coordinates, which were intersected with census tract polygons and linked to data tables from the US Census Bureau. SVI includes 15 variables including average household composition, proportion of residents with a disability, racial and ethnic groups, language, housing status, and transportation status.

4.2.3.2. Clinical characteristics.

At baseline, participants reported their date of diagnosis, method of insulin administration, use of CGM, diabetes-related healthcare utilization over the past 12 months, experiences with severe hypo- and hyperglycemia, diabetes complications and comorbidities, and engagement with mental/behavioral healthcare or peer social support programs for diabetes. At 12-months, they update any clinical data that have changed. Study staff verify clinical characteristics via EMR review.

4.2.4. Intervention satisfaction and feedback

Post-intervention, all participants complete a satisfaction survey rating their experiences in the study (e.g., frequency and time demands of questionnaires, interactions with study staff). At 6- and 12-months, intervention arm participants also complete a survey reporting which topics their Peer Mentor discussed, contact frequency, and mode of communication. At 12-months, they rate their satisfaction with the intervention overall, on a 5-point Likert scale from ‘not at all satisfied’ to ‘very satisfied,’ in terms of how useful they found it and how satisfied they are with it. Semi-structured interviews are conducted with a subset of participants in both study arms and Peer Mentors to contextualize satisfaction ratings, discuss experiences with the intervention, identify any barriers to engagement, and obtain suggestions for improvement. Interview recordings are transcribed verbatim and coded to identify themes in experiences with and suggestions to improve the intervention.

4.3. Peer mentor assessment schedule

Peer Mentors complete questionnaires about themselves and their experiences with T1D at baseline (pre-training) and when they finish as a Peer Mentor. Peer Mentors also complete monthly surveys documenting contacts with participants. All questionnaires are completed via CTMS web portal.

4.3.1. Peer mentor measures

See Table 4 for description of Peer Mentor measures.

5. Planned statistical analyses

5.1. Participant characteristics

Baseline participant demographics and clinical characteristics will be summarized by means with standard deviation, medians with minimum and maximum values, or frequencies with percentages as appropriate. Summary statistics will be stratified by treatment arm and compared using independent, two-sample t-tests, Wilcoxon rank sum, chi-square or Fisher exact tests.

5.2. Aim 1

The primary outcome is 12-month HbA1c. An independent, two-sample t-test will compare mean 12-month HbA1c between treatment arms assuming homogeneity of variance and normally distributed responses. Model assumptions will be evaluated using Levene’s test, quantile-quantile plots, and Shapiro-Wilks test. Departures from model assumptions will be addressed by data transformation or non-parametric methods (e.g., Wilcoxon rank sum).

Secondary analysis will estimate the mean (95% CI) HbA1c at 12-months by treatment arm adjusting for baseline HbA1c using a multiple linear regression. The model will include fixed effects for treatment arm, baseline HbA1c, and the arm-baseline HbA1c interaction term. If the interaction term is not significant at the 0.05 level, then it will be excluded from the model.

Secondary outcomes include time to first adult visit from date of randomization, self-management behaviors, and health-related quality of life. The time to first adult care visit will be compared between treatment arms using the log-rank test statistic for comparing Kaplan-Meier curves. Participants who do not complete an adult diabetes care appointment will be censored at the 12-month time point. A Cox Proportional Hazards regression model will estimate the Hazards Ratio (95% CI) for adult care follow-up in the intervention versus usual care arms adjusting for clinically important characteristics. Self-management behaviors and health-related quality of life will be analyzed using general linear mixed models similar to HbA1c.

5.3. Aim 2

A regression analysis will test mediation between study arm and primary and secondary outcomes at 12-months. We will use a 4-step mediation testing approach [57]. Indirect effects will be assessed by computing the difference between two regression coefficients, and standard errors will be estimated using Bootstrapping.

5.4. Aim 3

General linear mixed models will test associations between changes in measures over time with demographic and clinical characteristics. Models will be similar to that described earlier. Covariates will be independently entered into the multiple regression model. Covariates significant at the 0.05 level will be included in the final multiple regression model.

5.5. Statistical considerations

Statistical analysis will be conducted in SAS 9.4. The single, primary endpoint for this study is 12-month HbA1c. Secondary and exploratory endpoints will be assessed without adjusting for multiple hypothesis testing across endpoints. However, Holm’s step-down Bonferroni correction will be used for p-value adjustment within endpoints if appropriate. All statistical significance will be assessed at the two-sided alpha = 0.05 level. All self-report measures will be scored according to published instructions, including rules for handling missing item responses. Analysis will proceed using Intention to Treat, and all available observations will be used for analyses. No imputation will be used for this analysis.

6. Baseline data

6.1. Recruitment and enrollment

As detailed in the CONSORT diagram (Fig. 1), n = 100 young adults were randomized: 50 to the Peer Mentor intervention and 50 to usual care. As detailed in Fig. 2, n = 29 Peer Mentors enrolled. Two did not complete training and were not paired with a participant.

Fig. 1.

Young Adult Participant Recruitment and Enrollment CONSORT Diagram.

Fig. 2.

Peer Mentor Recruitment and Enrollment CONSORT Diagram.

6.2. Baseline participant characteristics

Table 1 presents young adult participants’ medical and demographic details. Briefly, the sample included 46% young adults who identified as being from a racially or ethnically marginalized group, 34% had public, county, or no insurance, and 13% were unsure about their insurance status. The mean HbA1c was 8.8 ± 2.0% at baseline and mean diabetes diagnosis duration was 9.8 ± 4.7 years (n = 99). Table 2 presents Peer Mentors characteristics.

Table 1.

Young Adult Participant Demographic and Clinical Characteristics, n = 100 (50 per group).

Variable	Overall M ± SD or N (%)	Peer Mentor Intervention	Usual Care Group
Age at consent (years)	19.9 ± 1.3	20.0 ± 1.3	19.9 ± 1.3
Gender
Female	58 (58%)	33 (66%)	25 (50%)
Male	42 (42%)	17 (34%)	25 (50%)
Race/Ethnicity
African American/Black,
non-Hispanic	12 (12%)	8 (16%)	4 (8%)
American Indian/Alaskan
Native, non-Hispanic	1 (1%)	1 (2%)	0 (0%)
Asian, non-Hispanic	3 (3%)	1 (2%)	2 (4%)
Hispanic	25 (25%)	8 (16%)	17 (34%)
White, non-Hispanic	54 (54%)	29 (58%)	25 (50%)
Another or multiple	5 (5%)	3 (6%)	2 (4%)
Insurance
Private/Commercial	53 (53%)	27 (54%)	26 (52%)
Public/County/Other/No
Insurance	34 (34%)	18 (36%)	16 (32%)
Not sure	13 (13%)	5 (10%)	8 (16%)
Qualified for Financial Aid (N, %)	46/80 58%)	24/44 (55%)	22/36 (61%)
Diabetes Duration (years)	9.8 ± 4.7	9.7 ± 4.8	9.9 ± 4.6
Method of insulin delivery (N, %)
Pump	56 (56%)	27 (54%)	29 (58%)
Injections	44 (44%)	23 (46%)	21 (42%)
Type of pump
Regular pump	22/56 (39%)	10/27 (37%)	12/29 (41%)
Automated pump	34/56 (61%)	17/27 (63%)	17/29 (59%)
CGM Use	76 (76%)	42 (84%)	34 (68%)
HbA1c %	8.8 ± 2.0	9.1 ± 2.0	8.5 ± 2.0

Note. CGM = continuous glucose monitor; HbA1c = hemoglobin A1c; if HbA1c recorded as >14, then value is replaced with 14.1 (n = 3).

Table 2.

Peer Mentor Characteristics.

Variable	Overall (N = 29)
Age at consent (years)	25.1 ± 3.3
Gender (N,%)
Female	21 (72%)
Male	8 (28%)
Race/Ethnicity (N,%)
African American/Black, non-Hispanic	1 (3%)
American Indian/Alaskan Native, non-Hispanic	0 (0%)
Asian, non-Hispanic	2 (7%)
Hispanic	6 (21%)
White, non-Hispanic	20 (69%)
Another or multiple	0 (0%)
Insurance (N,%)
Private	27 (93%)
Public/County/Other/No Insurance	2 (7%)
Diabetes Duration (years)	14.8 ± 4.5
Method of insulin delivery (N,%)
Pump	25 (86%)
Injections	4 (14%)
Type of pump
Regular pump	7/25 (28%)
Automated pump	18/25 (72%)
CGM Use (N, %)	27 (93%)
HbA1c (%)	6.5 ± 0.8

Note. CGM = continuous glucose monitor; HbA1c = hemoglobin A1c.

7. Discussion

Given the developmental challenges of young adulthood and health risks young adults with T1D face, they may benefit from tailored resources and support to help them transition to adult health care and achieve resilient diabetes outcomes. To fill the gap in T1D intervention research, our use of Peer Mentors to deliver a strengths-based intervention represents an innovative approach to transition support. Intervention delivery by “near peers” has potential to be developmentally appropriate, well-received, and useful, based on peer-based intervention research in other populations [20,38,58] and interest expressed by young adults with T1D [23].

Strengths of the DiaBetter Together study design include a targeted focus on intervening during the transition process, a strong foundation in behavioral theories [39,40], and input from young adults with T1D. The 12-month intervention duration allows ample opportunity for young adults to receive peer support across a wide range of transition-related experiences. Moreover, conducting the study in Houston, Texas, USA, generated a racially, ethnically, and socioeconomically diverse sample, which will increase the representativeness and generalizability of the study’s results.

Potential limitations also warrant discussion. Recruitment for young adult participants was geographically limited to Texas, which has unique social, political, and healthcare policy considerations [59], and results may not apply directly to other states or countries with distinct healthcare systems and resources. The study excluded non-English speakers, which limits generalizability and does not account for the heightened challenges people who do not speak English face navigating the healthcare system in the United States [60,61]. Future work should include participants who speak and read Spanish and other languages to advance more inclusive and generalizable research [62]. The COVID-19 pandemic introduced challenges to recruitment and intervention delivery which required adaptations that reduced sample size and statistical power [44,45]. Finally, it was difficult to ascertain young adults’ last pediatric visits, and appointment no-shows or cancelations were common, as has been documented among young adults [63].

In summary, despite the challenges of the COVID-19 pandemic and inherent structural challenges of reaching young adults during the period of transfer between pediatric and adult care, the DiaBetter Together trial is evaluating a novel intervention in a diverse sample of 100 young adults with T1D. Prior to this study, a Peer Mentor-delivered strengths-based approach has not been applied to healthcare transition interventions for young adults with T1D. Results will provide valuable information about novel strategies to improve the transition process from pediatric to adult diabetes healthcare. If successful, this intervention approach could be implemented by healthcare systems and/or community outreach groups to support young adults during this challenging and vulnerable transition period.

Data availability

Data will be made available on request.