Randomized observation study of biologic therapy for rheumatoid arthritis

描述（由申请人提供）：最近医疗保健研究和质量机构（AHRQ）的执行摘要表明，需要更好的系统来确定现有的与新的和昂贵的类风湿性关节炎（RA）生物药物治疗的相对优点。目前还没有明确的范例说明这些不同的生物疗法应该如何用于临床。考虑到治疗类风湿性关节炎的多种治疗方法，需要CCE研究和生物标志物预测来为生物治疗选择提供基本的算法。目前有8种昂贵的生物疗法被FDA批准用于治疗类风湿性关节炎。针对药物治疗的比较和成本效益（CCE）的研究受到几个因素的阻碍。一方面，依赖随机药物试验数据的CCE研究不能提供真实世界医疗成本的最佳估计，并且在接受生物治疗的RA患者中，随机研究和观察性研究的CCE结果存在显著差异。然而，观察性研究中缺乏随机化对CCE数据的分析施加了限制。例如，现实世界中患者的经验和非随机治疗选择因混杂或选择偏差而变得复杂，这些混杂或选择偏差显著影响任何CCE分析的结果，包括组间疾病严重程度、药物依从性和受试者分布到不同治疗类别的差异。最后，类风湿关节炎的CCE研究由于缺乏预测对一种治疗相对于另一种治疗的反应性的生物标志物而受阻。CCE研究的理想系统将允许在随机分配到类似疗法的患者中捕捉真实世界的成本。我们提出了一种新的方法来解决当前的障碍，不仅进行CCE研究，而且还简化了个性化医疗的努力。这一概念的本质是，在随机化之后，所有其他方面的护理都将由患者和他们的医生管理，包括药物剂量、监测和停药的决定。需要随机化来均匀分布患者并消除偏差，需要真实世界的观察来获取有关实际成本和治疗效果的准确信息。我们将本研究设计称为随机观察。我们计划利用匹兹堡大学医学中心（UPMC） RA比较有效性研究（RACER）系统进行随机观察性研究，以比较不同治疗策略在现实世界中的有效性，从而克服CCE研究在RA中的障碍。最初，我们将从哈佛大学的合作者和UPMC RACER那里获得CCE数据。UPMC RACER系统将利用UPMC风湿病学家的大型网络，这些网络已经通过电子病历（EMR）系统连接起来；电子病历将用于识别类风湿性关节炎患者，并获取有关治疗、医疗费用和临床实验室数据的信息。结合对哈佛大学BRASS注册的1110多名RA患者的分析，我们将展示UPMC RACER系统在治疗RA的生物疗法分析中的实用性，我们将利用这一分析的结果来设计未来RA生物疗法的随机观察研究。该项目涉及与哈佛大学和匹兹堡大学的研究人员合作，目的是建立系统，以便有效地对RA患者进行现实世界的成本效益研究。我们还将收集生物标本用于机制研究和潜在生物标志物的分析，因此我们也可以提供可能适用于其他研究的数据，这些研究将重点放在为RA患者提供个性化药物的研究上。

DESCRIPTION (provided by applicant): A recent Agency for Healthcare Research and Quality (AHRQ) executive summary indicated that better systems are needed to determine the relative merits of existing versus new and expensive biologic drug therapies for rheumatoid arthritis (RA). There is currently no clear paradigm for how these different biologic therapies should be used in the clinic. CCE studies and biomarker predictions are needed to provide rationale algorithms for biological therapy selection given the extensive array of therapies for treating RA. There are now 8 (a ninth in the pipeline) expensive biological therapies approved by the FDA to treat RA. Research that addresses the comparative and cost effectiveness (CCE) of drug therapies is hindered by several factors. On the one hand, CCE studies that rely on randomized drug trial data do not provide the best estimates of real world health care costs and there is significant disparity between CCE results from randomized versus observational studies in RA patients treated with biologic therapies. However, the absence of randomization in observational studies imposes limitations on the analysis of CCE data. For example, the empiric and non-random selection of therapies for patients in the real world is complicated by confounding or selection bias that significantly impact the outcome of any CCE analysis including differences between groups in disease severity, medication compliance, and subject distribution into different treatment categories. Finally, CCE research in RA is hindered by a lack of biomarkers that predict responsiveness to one therapy versus another. The ideal system for CCE research would allow real world costs to be captured in patients that were randomly assigned to comparable therapies. We propose a Novel approach to address these current obstacles for performing not only CCE research but also streamlining efforts for personalized medicine. Essential to this concept is that following randomization, all other aspects of care would be governed by the patient and their physician, including decisions about drug dosing, monitoring and discontinuation. Randomization is needed to distribute patients evenly and remove biases, and real world observation is needed to capture accurate information on actual costs and therapeutic effectiveness. We refer to this study design as randomized observation. We plan to overcome the barriers to CCE research in RA by utilizing the University of Pittsburgh Medical Center (UPMC) RA Comparative Effectiveness Research (RACER) system to perform randomized observational studies to compare real world effectiveness of different treatment strategies. Initially, we will obtain CCE data from collaborators at Harvard and also from the UPMC RACER. The UPMC RACER system will utilize a large network of UPMC rheumatologists that are already linked by an electronic medical record (EMR) system; the EMR will be used to identify RA patients and to capture information on treatment, medical costs and clinical laboratory data. In conjunction with the analysis of over 1,110 RA patients followed at Harvard in the BRASS registry, we will demonstrate the UPMC RACER system's utility in an analysis of biologic therapies for treating RA and we will use the results of this analysis to design a future randomized observation study of biologic therapy for RA. This project involves a collaboration with researchers at Harvard University and the University of Pittsburgh with the goal of establishing the systems in order to effectively perform real-world cost-effectiveness research in patients with RA. We will also collect biological specimens for analyses of mechanistic studies and potential biomarkers so we can also provide data that potentially will be applicable for additional studies that will focus on research that will lead to personalized medicine for patients with RA.