PATIENT OUTCOME IN RA WITH EARLY ANTIRHEUMATIC THERAPY

类风湿性关节炎（RA）的医疗管理取决于许多药物，单独或联合应用顺序。因此，未来的管理改进将不仅来自于使用最有效的药物，而且来自于湿润有效序列。该项目将确定RA患者的最佳治疗策略，通过将疾病过程分解为多个离散的，特定的治疗过程，以最佳顺序使用最佳药物的策略，称为治疗段。治疗段从药物或药物组合治疗的开始日期开始，并在药物治疗发生变化的日期结束，无论是由于停药、删除药物或添加药物。典型的类风湿性关节炎患者在病程中可能有5到15个治疗阶段。将根据曲线下面积（AUC）残疾、疼痛、整体健康、累积毒性和成本来评估最佳治疗。我们将完成五个结果中的每一个的治疗区段结果，每个主要DMARD和DMARD组合，并将确定影响这些结果价值的患者和组变量。我们将首先比较采用所有dmard的替代治疗策略，不考虑序列，使用多变量回归模型对长期结果进行比较。然后，我们将使用回归分析和分类树来评估RA患者特征对结果的影响，并考虑治疗阶段和这些阶段的顺序的影响。需要评估的特征包括：社会人口学变量、疾病持续时间、既往治疗反应、既往DMARD治疗和既往毒性。我们的下一个任务将是评估序列的敏感性和个体患者特征变化的结果。这项工作将产生个体结果（AUC残疾、疼痛、全球、健康毒性和成本）的统计模型。最后，我们将为每个片段和每个序列构建成本效益和有效性/毒性比。结果将鼓励临床决策在有效性，成本/效果，效果/毒性的基础上作出，这可能是个别患者的个体化。该项目建立在大型、长期、高质量的数据集和该团队开发的有效性、毒性和成本评估技术的基础上，以提供独特而重要的数据来指导RA患者的治疗选择，以改善其预后。这项研究将提供累积（AUC）结果的重要性和大小，最适合改善结果的药物序列，考虑个体患者可变性的必要性，并将为临床医生提供更好的信息，以帮助他们对RA患者的治疗选择做出决策。

The medical management of rheumatoid arthritis (RA) depends on a number of medications, applied sequentially either alone or in combination. As a result, future improvement in management will come not only by use of the most effective drugs but also the moist effective sequences. This project will identify optimal therapeutic strategies for patients with RA, strategies which use the best drugs in the best order, by breaking the disease course into multiple discrete, specific treatment courses called therapeutic segments. A therapeutic segment starts with the date of onset of treatment with a drug or drug combination and ends on the date of a change in the drug treatment whether due to discontinuation, deletion of an agent or addition of an agent. A typical RA patient may have 5 to 15 therapeutic segments over the disease course. Optimal treatment will be evaluated in terms of area under the curve (AUC) disability, pain, global health, cumulative toxicity, and costs. We will complete therapeutic segment outcomes for each of the five outcomes, for each major DMARD and DMARD combination, and will identify patient and group variables which affect value of these outcomes. We will first compare alternative therapeutic strategies employing all DMARDs, without regard to sequence, for long-term outcomes using multivariate regression models. We will then evaluate the impact of RA patient characteristics on outcomes using regression analyses and classification trees and consider the effects of therapeutic segments and of the sequence of these segments. Characteristics to be evaluated include: sociodemographic variables, disease duration, prior response to treatment, prior DMARD therapy, and prior toxicity. Our next task will be to evaluate the sensitivity of the sequences and the outcomes to variation in individual patient characteristics. This work will result in statistical models for the individual outcomes (AUC disability, pain, global, health toxicity, and costs). Finally, we will construct cost- effectiveness and effectiveness/toxicity ratios for each segment and for each sequence. Results will encourage clinical decision to be made on the basis of effectiveness, cost/effectiveness, and effectiveness/toxicity which may be individualized for the particular patient. This project builds upon large, long-term, high quality data sets, and techniques for estimation of effectiveness, toxicity, and costs developed by this team, in order to provide unique and important data to guide therapeutic choices toward improved outcomes for RA patients. This research will provide knowledge of the importance and magnitude of cumulative (AUC) outcomes, the drug sequences best employed for outcome improvement, the necessity of considering individual patient variability, and will provide clinicians with better information to assist them in decision-making regarding treatment choices for their RA patients.