Optimizing Medical Decision Making for Older Patients with Type 2 Diabetes

项目摘要 目前，由于缺乏必要的证据，老年患者的2型糖尿病(T2D)的治疗受到阻碍 为有效的医疗决策提供信息。老年糖尿病患者具有高度的异质性 关于预期寿命、糖尿病病程、合并症、糖尿病并发症、残疾、功能 损伤和治疗偏好。目前的临床指南都承认这种异质性和 支持个性化糖尿病护理的概念，但就如何 个体化的血糖目标和药物治疗方案。这些冲突反映了缺乏证据表明 现实世界临床中治疗、健康状况和医疗决策之间的动态交互作用 练习人群。虽然患者的观点是共同决策的支柱，但人们对此知之甚少 老年患者的治疗偏好和当前糖尿病治疗经验的差异， 包括自我管理障碍和自我报告的药物不良事件。同样，缺乏基本的证据。 A1C和关键结果之间的关系如何因医学复杂性和不同类别而变化 降血糖药物。此外，忙碌的临床医生缺乏实用的、循证的工具来指导决策 关于个体化靶点和药物的制定。我们的首要目标是进行观测 研究提供必要的证据，为安全有效地护理患有T2D的老年人提供信息。这项研究 将基于具有良好特征的多种族人口145,894名≥65岁的T2D患者 凯撒长寿北加州。我们将使用NIA资助的调查结果(R56 AG051683)在 对约6,000名患者进行年龄分层随机抽样，以描述患者对治疗的看法 偏好、自我管理障碍和与患者相关的结果(例如，低血糖、跌倒、生活质量)。 在这项拟议的研究中，我们将把调查响应与基于EMR的暴露和结果联系起来(例如，实验室、 药物处方和依从性、并发症、死亡率)。这将使我们能够表征 老年患者对糖尿病治疗的经验和偏好，并考察它们之间的关系 在患者偏好和自我管理障碍与过去或未来的血糖控制、药物治疗之间 用途和结果(目标1)。为了为建立适当的血糖目标提供信息，我们将确定A1C 与关键不良糖尿病结局(微血管和大血管)风险最低相关的水平 并发症、死亡率和低血糖)，按健康状况、年龄、糖尿病病程和药物进行分层 类型(目标2)。最后，我们将创建一个决策支持工具，以鼓励糖尿病护理的个性化， 在患有T2D的老年人中最大限度地提高安全性和最大限度地减少过度治疗(目标3)。这一努力包括 开发和验证当代死亡率预测模型并将其与我们现有的 低血糖风险分层模型。拟议研究的完成将确保老年患者 他们的提供者拥有必要的临床证据和支持，可以为糖尿病做出明智的决定。

Project Abstract Management of type 2 diabetes (T2D) in older patients is currently hampered by the lack of evidence needed to inform effective medical decision making. Older patients with diabetes are highly heterogeneous with regards to life expectancy, duration of diabetes, comorbidities, diabetic complications, disabilities, functional impairments, and treatment preferences. Current clinical guidelines all acknowledge this heterogeneity and support the concept of individualized diabetes care, but provide conflicting recommendations regarding how to individualize glycemic targets and medication regimens. These conflicts reflect the lack of evidence regarding the dynamic interactions between treatments, health status, and medical decision making in real-world clinical practice populations. While the patient perspective is a pillar of shared decision making, little is known about the variation in older patients' treatment preferences and experiences with current diabetes treatments, including self-management barriers and self-reported adverse drug events. Similarly, basic evidence is lacking on how the relationship between A1C and key outcomes varies by medical complexity and by classes of glucose-lowering medications. Moreover, busy clinicians lack practical, evidence-based tools to guide decision making regarding individualized targets and medications. Our overarching goal is to conduct observational research to provide evidence needed to inform safe and effective care of older adults with T2D. This research will be based on a well-characterized, multi-ethnic population of 145,894 patients ≥65 years old with T2D from Kaiser Permanente Northern California. We will use the results of a NIA-funded survey (R56 AG051683) in an age-stratified, random sample of ~6,000 patients to characterize patient perspectives on treatment preferences, self-management barriers, and patient-related outcomes (e.g., hypoglycemia, falls, quality of life). In this proposed study, we will link survey responses to EMR-based exposures and outcomes (e.g., labs, medication prescribing and adherence, complications, mortality). This will allow us to characterize variation in older patient's experiences with and preferences for diabetes treatments and examine the relationships between patient preferences and self-management barriers with past or future glycemic control, medication use and outcomes (Aim 1). To inform efforts to establish appropriate glycemic targets, we will identify A1C levels that are associated with the lowest risk of key adverse diabetes outcomes (micro- and macrovascular complications, mortality, and hypoglycemia), stratified by health status, age, diabetes duration, and medication type (Aim 2). Finally, we will create a decision support tool to encourage the individualization of diabetes care, maximizing safety and minimizing overtreatment, in older adults with T2D (Aim 3). This effort includes developing and validating a contemporary mortality prediction model and integrating it with our existing hypoglycemia risk stratification model. The completion of the proposed studies will ensure that older patients and their providers have the clinical evidence and support necessary to make informed decisions for diabetes.