Use of Geographic Variations to Estimate Effectiveness with Non-Randomized Data

描述（由申请人提供）：在美国，不同地理区域的医疗支出和服务使用存在巨大差异。这种差异的很大一部分与使用的服务有关，而这些服务尚未在随机试验中得到研究。效益未知的各种服务差异很大，因此有人呼吁进行研究，利用观察数据来衡量医疗的相对有效性。估计非随机环境下的治疗效果需要仔细的数据收集和分析，因为患者组之间可能存在未观察到的差异。越来越多的研究人员正在利用服务使用的地理差异来推断其在个体患者中的有效性，假设患者特征在地理区域之间是平衡的。然而，在非随机研究中，人们对生态变异可用于获得有效治疗效果估计的环境知之甚少。利用监测流行病学和最终结果(SEER)-Medicare数据库中诊断为结直肠癌和前列腺癌的老年患者基于人群队列的数据，我们将开发、评估和应用先进的统计方法来评估癌症干预措施的有效性。癌症为这种分析提供了一个理想的环境，因为它是一种普遍而严重的健康状况，而新的、昂贵的、疗效未知的治疗方法正在不断地被开发和推广。这些目标将通过以下三个具体目标来实现：(1)开发和评估统计方法，利用地理居住地模拟治疗随机化来估计治疗效果；(2)应用这些方法评估干预措施对结直肠癌和前列腺癌患者的比较效果；(3)开发和传播软件，以促进在比较有效性研究中使用适当的统计方法。这些结果将提供有关最近医学研究所关于比较有效性研究优先事项的研究中强调的干预措施有效性的信息，并提供其他研究人员可以使用的方法来了解其他疾病治疗的有效性。

DESCRIPTION (provided by applicant): There is enormous variation in medical spending and use of services across geographic areas in the United States. A large portion of this variation has been linked to use of services that have not been studied in randomized trials. This wide variation in services of unknown benefit has led to calls for research using observational data to measure the comparative effectiveness of medical treatments. Estimating treatment effectiveness in non-randomized settings requires careful data collection and analysis because of possible unobserved differences between patient groups. Increasingly, researchers are exploiting geographic variation in use of services to infer their effectiveness in individual patients, under the assumption that patient characteristics are balanced across geographic areas. However, little is understood about settings in which ecological variation can be used to obtain valid estimates of treatment effects in non-randomized studies. Using data from population-based cohorts of elderly patients diagnosed with colorectal and prostate cancer in the Surveillance Epidemiology, and End Results (SEER)-Medicare database, we will develop, evaluate, and apply advanced statistical methods to estimate the effectiveness of cancer interventions. Cancer provides an ideal setting for such an analysis because it is a prevalent and serious health condition, and new and expensive treatments of unknown benefit are continuously being developed and diffused. These goals will be accomplished through the following three specific aims: (1) develop and evaluate statistical methods to estimate treatment effectiveness using geographic residence to mimic randomization to treatment; (2) apply these methods to estimate the comparative effectiveness of interventions for patients diagnosed with colorectal and prostate cancer; and (3) develop and disseminate software to promote the use of appropriate statistical methods in comparative effectiveness studies These results will provide information about the effectiveness of interventions highlighted in a recent Institute of Medicine study on priorities for comparative effectiveness research and also provide a methodology that other researchers can use to learn about the effectiveness of treatments in other diseases. PUBLIC HEALTH RELEVANCE: This project will provide valuable information on interventions in colorectal and prostate cancer highlighted in a recent Institute of Medicine study on priorities for comparative effectiveness. In addition, the methods developed will be applicable to a broad range of interventions for cancer and other diseases, providing a framework for future comparative effectiveness studies and ultimately changes in treatment guidelines.