Pilot Implementation Trial of Implementation Toolkits for Diabetes Coaching

Type 1 diabetes (T1D), one of the most prevalent chronic pediatric diseases in the United States, requires a complex treatment regimen. Youth with T1D, particularly adolescents, more frequently have higher glycemic levels. Less than 30% of youth achieve the recommended level of glycemic control. Compared to adults, youth are more at-risk for related complications (diabetic ketoacidosis) that can lead to heart disease, diabetic coma, and death. These complications often result from challenges completing T1D care tasks. Evidence-based behavioral interventions can increase T1D care task completion and lower glycemic levels (i.e., A1c). Integrating these interventions into standard T1D care may lead to better overall health for this population. Implementation science offers theory and rigorous methods to accomplish this goal, but few studies have employed them.

Behavioral Family Systems Therapy-Diabetes (BFST-D) is efficacious for increasing T1D care task completion and lowering A1c, particularly among adolescents who have higher glycemic levels (A1c≥9.0%). A brief 4-session, adapted version of BFST-D (Diabetes Coaching) delivered as part of standard care was associated with lowered glycemic levels. Despite these efficacy and initial effectiveness data, few receive this evidence-based care. Although Diabetes Coaching has been offered at the PI's institution, less than 10% of patients received it. Evaluation of strategies to improve reach of Diabetes Coaching in standard care is needed.

In a large national qualitative study of key informants (adolescents with T1D-caregiver dyads, medical and psychosocial healthcare providers, healthcare leaders) drawn from 5 children's hospitals, the PI identified determinants of implementing Diabetes Coaching (K23 Aim 1). Factors influencing implementation ranged across all domains of the Consolidated Framework for Implementation Research (CFIR), including those related to family needs and opportunity to engage in this care (evening, telehealth delivery, expert providers who are empathic, patient, encouraging, referrals from trusted provider when A1c rising, evidence of improved A1c), provider knowledge and self-efficacy in referring to and delivering Diabetes Coaching, and clinic resources (lack of streamlined referral processes; leadership buy-in), as well as external factors (insurance coverage).

Using these rich data and an evidence-based quality improvement (EBQI) approach with iterative feedback from Family and Healthcare Professional Diabetes Care Advisory Boards (F-DCAB; P-DCAB) drawn from Nemours Children's Hospital Delaware (NCHD), the PI developed a toolkit of implementation strategies (K23 Aim 2). Implementation strategies are methods for integrating evidence-based practices into standard care.20 Critical strategies needed to integrate Diabetes Coaching into standard T1D care at NCHD emerged and comprise a Base Implementation Toolkit (BIT). Pre-implementation BIT strategies include provider and staff education meetings (e.g., referring and delivering Diabetes Coaching), identification of and training of Champions for implementation, materials to increase demand for Diabetes Coaching among families (e.g., family-facing educational materials), and development of a workflow blueprint and of tools to monitor implementation. Active implementation BIT implementation strategies include feedback loop to clinical provider team regarding summary level data on uptake of referrals to and family engagement in Diabetes Coaching, as well as consultation to support problem-solving workflow and other implementation challenges with Champions. Two strategies (audit and feedback, adapting delivery format and timing of Diabetes Coaching) may be added to BIT (BIT-Enhanced) to improve adoption among medical providers with lower referral rates and/or to increase reach if few families attend Diabetes Coaching. BIT-Enhanced requires additional time and resources (e.g., Champion and providers review individual data, collaborate on ways to increase referrals). Thus, examining if this additional strategy meaningfully improves outcomes offers critical data for maintaining only the strategies that are necessary and sufficient. Key to maintaining implementation of any evidence-based practice, including Diabetes Coaching, is considering sustainability early on. The primary sustainability strategy involves training the Champions early and planful progression of their role in the implementation efforts, from first attending the PI-led (implementation and content expert) brief consultation meetings and data tracking to then independently leading these meetings (BIT-Sustainment)

Guided by the Exploration, Preparation, Implementation, Sustainment (EPIS) and the Reach, Efficacy, Adoption, Implementation, Maintenance (RE-AIM) frameworks, the goals of the current single site pilot implementation feasibility trial (K23 Aim 3) are to pilot study procedures and the implementation toolkits and track effectiveness of implementation toolkits in preparation for a fully powered trial. The overarching aim of this study is to gather preliminary data to support an R01 application to conduct a large, multi-site randomized hybrid implementation-effectiveness trial testing uptake of Diabetes Coaching and tracking patient health outcomes.

Aim 1 (Feasibility \& Acceptability): Conduct a pilot implementation feasibility study at the PIs institution (NCHD) to examine acceptability and feasibility of BIT for Diabetes Coaching and BIT-Enhanced for providers with lower adoption rates (<80%) and with low reach (<5 families per group), and BIT-Sustainment. In addition, study procedures will be piloted in preparation for a fully powered trial.

Aim 2 (Implementation Outcomes): Examine four exploratory implementation outcomes, (1) medical provider adoption of referring to Diabetes Coaching (i.e., number of providers vs. number of providers referring >80% eligible families), (2) psychologist fidelity in delivering Diabetes Coaching, and (3) reach of Diabetes Coaching to the target patient population (i.e., number families eligible vs. referred, number of sessions attended, completion rate). (4) Sustainability of adoption, reach, and fidelity will be examined following the removal of most research resources.