Population Health Impact of a Self-Insured Employer's Policy Change to Cover Weight Reduction and Diabetes Prevention Interventions for Employees, Dependents, and Retirees with Prediabetes

摘要： 2015年9月，密歇根大学（U of M）改变了其医疗保健福利，以涵盖干预措施 用于减轻体重和预防糖尿病，无需为约20，000名超重或肥胖者支付自付费用 在其约85，000名员工、家属和退休人员中， 退休人员我们将首先比较这一大规模政策变化对人口健康的影响， 四种策略的收益率被用来识别非糖尿病员工，家属和退休人员 糖尿病前期这些策略包括使用索赔，HbA 1c和BMI水平在M的自我U， 资助的健康保险数据库，以确定糖尿病前期个体;应用经验证的筛查算法 在同一个数据库中，以确定其他年龄≥40岁的非糖尿病个体的最高风险 糖尿病前期的HbA 1c检测;通过要求个人使用CDC Prediabetes自我筛查糖尿病前期 筛查试验（如果阳性，则进行HbA 1c检测）;以及进行糖尿病前期筛查和HbA 1c检测 作为工作场所健康计划的一部分进行测试。我们还将评估量身定制的 反馈，让初级保健医生参与病例发现，以及对会员参与的经济激励 在筛选中。其次，我们将描述干预的吸收和达到，并评估个人的偏好， 个人和互联网为基础的生活方式干预和二甲双胍治疗。我们将试图解释 基于健康信念模型的干预参与，并进一步描述和比较吸收， 依从性、保留率和与每种治疗相关的结局（体重、BMI、HbA 1c和生活质量）变化 干预我们还将评估干预措施对人口健康的影响。最后，我们将对 在一年和两年期间，干预措施相对于不干预措施的有效性、成本和成本效用， 在模拟的5年和10年的时间范围内，使用经过验证的计算机模拟模型。严谨的 对肥胖政策的这一大规模变化的评估将涉及研究小组之间的合作， 患者、提供者、提供者团体、干预提供者、密歇根大学及其健康保险公司。的结果 评估将促进科学知识，并将对福利设计，卫生政策， 人口健康。

ABSTRACT: In September 2015, the University of Michigan (U of M) changed its healthcare benefits to cover interventions for weight reduction and diabetes prevention at no out-of-pocket cost for the ~20,000 overweight or obese employees, dependents, and retirees with prediabetes among its ~85,000 employees, dependents, and retirees. We will evaluate the impact of this large scale policy change on population health by first comparing the yield of four strategies being used to identify nondiabetic employees, dependents, and retirees with prediabetes. These strategies include using claims, HbA1c, and BMI levels available in the U of M's self- funded health insurance database to identify prediabetic individuals; applying a validated screening algorithm to the same database to identify additional nondiabetic individuals ≥40 years of age at highest risk for prediabetes for HbA1c testing; by asking individuals to self-screen for prediabetes using the CDC Prediabetes Screening Test (and to have HbA1c testing if positive); and by performing prediabetes screening and HbA1c testing as part of a worksite wellness program. We will also evaluate the incremental benefits of tailored feedback, engaging primary care physicians in case finding, and financial incentives for member participation in screening. Second, we will describe intervention uptake and reach and assess individual preferences for in- person and internet-based lifestyle interventions and for metformin therapy. We will attempt to explain intervention participation based on the Health Belief Model, and further describe and compare the uptake, adherence, retention, and change in outcomes (weight, BMI, HbA1c, and quality-of-life) associated with each intervention. We will also assess the impact of the interventions on population health. Finally, we will model the effectiveness, costs, and cost-utility of the interventions relative to no intervention over one and two years and over simulated 5- and 10-year time horizons using a validated computer simulation model. The rigorous evaluation of this large scale change in obesity policy will involve collaboration among the research group, patients, providers, provider groups, intervention providers, the U of M, and its health insurer. The results of the evaluation will advance scientific knowledge and will have major implications for benefit design, health policy, and population health.