Personalized Medicine in Comparative Effectiveness

描述（由申请人提供）：传统上，CER主要关注广泛人群的平均效应。然而，干预措施在风险和/或益处方面的有效性往往因患者亚组而异。最近科学技术的进步导致许多与疾病结局和治疗反应相关的生物和遗传标记的发现。这些新的标志物与传统的临床评估相结合，在确定最有可能从特定治疗中受益或具有高风险毒性的患者亚组方面具有很大的潜力，从而可能导致个性化或量身定制的药物。该项目将开发用于CER个性化医疗的统计方法。这些方法可用于指导和定制治疗或疾病筛查策略的个别患者。这些方法将为确定每个患者最有效的临床选择提供基础，从而提高未来的CER和公共卫生质量。该提案的具体目标是：1。利用随机临床试验的数据，开发在个体水平上评估治疗效果的方法，这些随机临床试验有：(a)单一结果，(b)多维结果，量化风险和收益。我们将开发系统的统计程序，以确定未来的患者将受益于新疗法，而不是标准治疗。2. 开发和应用控制结肠癌和前列腺癌疾病早期检测的随机模型。我们将根据年龄和风险状况制定最佳的筛查检查策略。筛查策略将包括基于风险的建议，而不是固定时间的建议。我们还将调查终止筛查的年龄上限。3. 开发和评估新的诊断和预后模式的患者水平的增量价值。我们将开发定量方法来评估新预测模式的增量价值在不同亚群中的变化，并利用临床试验或观察性研究的数据确定从新模式中获益最多或最少的亚群。4. 制定方法来比较不同研究中实施的治疗方法的有效性。在这种情况下，我们还将制定针对患者的治疗选择策略。该提案是由哈佛大学公共卫生学院（HSPH）生物统计学系的主要研究人员提交的，他们在CER研究方面具有互补但综合的专业知识，为CER的方法发展提供了良好的基础设施和环境。研究人员还与著名的临床试验网络（例如艾滋病临床试验小组和东部肿瘤合作小组）和其他可用于应用所开发方法的数据源建立了牢固的联系，使HSPH处于确保提案成功的独特地位。

DESCRIPTION (provided by applicant): Traditionally CER has focused primarily on the average effects across broad populations. However, the effectiveness of interventions with respect to risk and/or benefit often varies by patient subgroups. Recent advancement of science and technology has led to the discovery of many biological and genetic markers associated with disease outcomes and treatment responses. These new markers combined with traditional clinical assessments hold great potential for identifying subgroups of patients who are most likely to benefit or are at high risk for toxicity from a particular therapy and thus may lead to personalized or tailored medicine. This project will develop statistical approaches to personalized medicine in CER. The methods can be used to guide and tailor the treatment or disease screening strategies for individual patients. These methods will enhance future CER and improve the quality of public health by providing the foundation for identifying the most effective clinical options for each individual patient. The specific aims of the proposal are: 1. To develop methods for assessing treatment effects at an individual level using data from randomized clinical trials with: (a) a single outcome, and (b) multi-dimensional outcomes that quantify both risks and benefits. We will develop systematic statistical procedures to identify future patients that would benefit from a new therapy vs. for example, the standard care. 2. To develop and apply stochastic models governing the early detection of disease to colon and prostate cancers. We will develop optimal screening examination strategies as a function of age and risk status. The screening strategies will involve risk-based recommendations rather than fixed-time recommendations. We will also investigate upper age limits for ending screening. 3. To develop and evaluate the patient-level incremental value of new diagnostic and prognostic modalities. We will develop quantitative methods for assessing how the incremental value of new predictive modalities may vary across sub-populations and for identifying sub-populations that benefit the most or the least from the new modalities using data from clinical trials or observational studies. 4. To develop methods to compare the effectiveness of treatments implemented in different studies. We will also develop patient-specific treatment selection strategies in this setting. The proposal is submitted by leading researchers with complimentary but integrated expertise in CER research from the Department of Biostatistics at the Harvard School of Public Health (HSPH) which provides a well- established infrastructure and environment for methodological development in CER. The researchers also have strong ties to prominent clinical trial networks (e.g., AIDS Clinical Trials Group and the Eastern Cooperative Oncology Group) and other data sources that can be utilized to apply the developed methods, putting HSPH in a unique position to ensure the success of the proposal. PUBLIC HEALTH RELEVANCE: We will develop novel statistical methods for personalized medicine in comparative effectiveness. These methods will allow more robust evidence-based decisions in clinical practice that are tailored to individual patients based on their personal characteristics, so that the best clinical decisions are made for individual patients and the efficiency in public health practice is optimized.