Health Outcome Effects of Common Medications in Elders with Multiple Conditions

描述（由申请人提供）：为了告知患有多种慢性疾病（MCC）的人的临床决策，我们必须比较各种疾病的治疗效果。在本研究中，我们超越了单一疾病治疗对单一疾病特定结局的比较获益问题，开始探索多种疾病治疗对多种普遍健康结局（例如功能、症状和生存期）的比较有效性，这些结局对患者有意义，并受所有健康状况的影响。多个普遍的结果是重要的探索，因为治疗可能会影响结果不同，老年人与MCC在他们的健康结果的优先事项不同。虽然复杂，但这一调查路线是对RFA-AG-13-003的响应，并且与患有MCC的老年人最相关。该项目建立在最近的工作基础上，我们已经：1）确定了一组普遍的健康结果; 2）确定了几种常见条件对这些结果的影响; 3）开发了一种方法来确定并发条件对死亡的贡献百分比; 4）确定了最常见的竞争条件对;和5）确定一种病症的治疗对潜在竞争性病症的影响。在进行这项工作时，我们已经认识到方法上的挑战，这有助于解释尽管这一领域的临床重要性，但缺乏研究。建议的方法建立在我们最近的工作，在适应联合建模的纵向结果和发展的纵向延伸的平均归因分数。使用创新的分析方法，具体目标是：1）估计9种最常用的指南推荐药物类别对一组常见和病态慢性疾病的5种普遍健康结局的影响; 2）估计每种药物对每种普遍结局的贡献百分比。参与者将是大型的、特征良好的、具有全国代表性的医疗保险当前受益人调查（MCBS）队列的成员，他们患有9种研究条件中的2种。从临床的角度来看，这条调查线最终将允许一个基于证据的方法，通过确定具有最佳整体效益的药物，为MCC患者提供多药治疗。结果将通过引入新方法，确定有希望的治疗方法来比较MCC患者，并确定效应大小和功效估计，为未来的研究提供信息。由于MCBS代表了MCC可变组合的全国老年人群，因此结果将提供研究药物相对效应的人群水平估计值。

DESCRIPTION (provided by applicant): To inform clinical decision-making for persons with multiple chronic conditions (MCC), we must compare the effects of treatments across conditions. In this study, we go beyond the question of comparative benefit of treatments for a single condition on single condition-specific outcomes to begin exploring the comparative effectiveness of treatments across multiple conditions on multiple universal health outcomes (e.g. function, symptoms, and survival) that are meaningful to patients and that are affected by all health conditions. Multiple universal outcomes are important to explore because treatments may affect outcomes differently and older adults with MCC vary in their health outcome priorities. While complex, this line of investigation is responsive to the RFA-AG-13-003 and of utmost relevance for older individuals with MCC. The project builds on recent work in which we have: 1) identified a set of universal health outcomes; 2) determined the effect of several common conditions on these outcomes; 3) developed a method for determining the % contribution to death of co-occurring conditions; 4) identified the most common pairs of competing conditions; and 5) determined the effects of treatment for one condition on a potentially competing condition. In conducting this work, we have come to appreciate the methodological challenges that help explain the dearth of research in this area despite its clinical importance. The proposed methodology builds on our recent work in adapting joint modeling of longitudinal outcomes and developing the longitudinal extension of the average attributable fraction. Using innovative analytical methods, the Specific Aims are to: 1) estimate the effect of the 9 most commonly used guideline recommended medication classes for a set of common and morbid chronic conditions on 5 universal health outcomes; and 2) estimate the percent contribution of each medication to each universal outcome. Participants will be members of the large, well-characterized, nationally representative Medicare Current Beneficiary Survey (MCBS) cohort who have e 2 of the 9 study conditions. From a clinical perspective, this line of investigation eventually will allow an evidence-based approach to polypharmacy for patients with MCC by identifying medications with the best overall benefits. Results will inform future studies by introducing new methods, identifying promising treatments to compare in persons with MCC, and determining effect size and power estimates. Because MCBS represents the national older population with variable combinations of MCC, results will provide a population-level estimate of the relative effects of the study medications.