Evaluating the Implementation of the Diabetes Prevention Program in an Integrated Health System

The prevalence of prediabetes and obesity among U.S. adults age 40 and older is significant, with over 30% having prediabetes and over 40% having obesity.\[1,2\] Prediabetes and obesity increase the risk for diabetes, cardiovascular disease, and poor quality of life, and are responsible for substantial healthcare costs.\[3\] In response to the multi-level burden of prediabetes and obesity, there have been several efforts to prevent diabetes at the population level and reduce healthcare costs,\[4,5\] including national implementation and reimbursement of the successful Diabetes Prevention Program (DPP). Beginning in April 2018, the Centers for Medicare \& Medicaid Services (CMS) made a landmark decision to reimburse clinical and non-clinical settings for providing DPP to Medicare beneficiaries (i.e., Medicare DPP); this coverage is currently for in-person DPP only and not digital DPP.\[6,7\] CMS's decision to cover DPP among older adults with prediabetes further catapulted efforts within healthcare organizations to address the increasing number of individuals with diabetes receiving care in their facilities. However, few studies have examined the sustainability of providing DPP based on maintenance of the effect (i.e., long-term change in weight and HbA1c), healthcare costs, participant experience, and organizational support. In addition, attracting individuals to DPP and similar lifestyle change interventions remains a significant challenge and identifying useful approaches is important.\[8-11\] Lastly, whereas the effectiveness of in-person DPP is well-established, prior studies evaluating the effect of digital DPP identified positive outcomes but had significant methodological limitations, such as a single arm pre- / post-test design and participant-reported outcomes.\[12,13-19\] In 2017 Kaiser Permanente Northwest (KPNW), a large, integrated health system serving Oregon and southwest Washington, began piloting both digital and in-person versions of DPP for its adult health plan members with prediabetes and obesity. The purpose of this mixed-methods, natural experiment is to evaluate this large health system initiative by assessing the effects of both digital and in-person DPP on change in weight and HbA1c, health behaviors, and psychosocial factors. Also, sustainability based on cost-effectiveness and patient and healthcare stakeholder perspectives will be examined. KPNW patients eligible to participate in DPP (digital or in-person) will be identified and recruited using the electronic health record (EHR). Both the digital and in-person (group-based) DPP programs will be delivered over 12 months. Demographic and clinical data to be included in analyses will also be extracted from the EHR. Behavioral and psychosocial questionnaires will be administered to DPP enrollees and non-enrollees online using REDCap. Semi-structured qualitative interviews will be conducted with a subset of DPP enrollees and non-enrollees to understand reasons for enrollment and likes/dislikes about the program. Healthcare system providers and stakeholders will also be interviewed to capture factors related to sustainability of offering DPP within the health system. For the primary analysis, investigators plan to model 12- and 24-month weight and HbA1c trajectories using a linear mixed effects model using time since baseline as the time axis. Because randomization is not feasible in this real-world implementation of DPP, propensity score adjustment will be used to control for potential confounding. Furthermore, investigators will conduct an economic evaluation over the 12-month follow-up period for both the digital and in-person DPP cohorts as well as over the 24-month period for the digital DPP cohort from the perspective of the health plan, following best practices,\[20\] and guided by previous economic analyses of DPP interventions.\[21-26\] The mixed-methods, natural experiment design investigators will use to evaluate KPNW's implementation of digital and in-person DPP will build on existing evidence related to DPP effectiveness across the two delivery modes on change in weight and HbA1c over time. In addition, the cost-effectiveness analysis will determine the impact of digital and in-person DPP on return on investment for healthcare systems and sustainability of the program. Findings from our evaluation will therefore inform best practices for implementing and sustaining DPP within large healthcare systems.