ReCAPSE: Recurrence from Claims And PROs for SEER Enhancement

项目总结 国家癌症研究所的监测、流行病学和最终结果(SEER)计划 作为国家癌症监测的仪表盘，是癌症的权威资源 发病率和死亡率。然而，该登记缺乏关于癌症复发的信息，这是 一个影响疾病管理决策和政策的主要结果 个人和人口水平。拟议研究的目标是开发一种可扩展的方法 用于以人群为基础确定癌症复发，特别关注乳腺癌。这个 研究的第一阶段(UG3)将开发一种创新的算法，通过以下方式检测首次复发 将医疗索赔与患者报告的结果(PRO)相结合。此阶段将在#年进行 与西雅图-普吉特湾SEER登记处合作，并将利用正在进行的研究 与Puget Sound Payers的协议。将开发不同的算法，以允许在 使敏感度或特异度最大化。这些算法将针对黄金标准的重现进行验证 基于两家主要当地供应商的电子医疗记录提供的信息。研究参与者 2013-2016年间确诊的I-III期乳腺癌患者是否会以谁为“金标准”？ 关于复发的信息是可用的。将采用回溯性病例对照设计 将在每个验证站点征集相同数量的复发病例和非复发对照 关于他们的复发历史的专业人士。所有验证活动将显式调整是否缺少 参与者未对PRO征集做出回应所产生的数据。第二阶段(UH3)将是 是与肯塔基州癌症登记处合作进行的。在UG3中开发的算法 2015-2017年间，将在肯塔基州所有I-III期乳腺癌诊断病例上部署阶段 索赔数据可用。这一阶段将包括对照患者的医疗记录进行验证 在肯塔基大学马基癌症中心被诊断出在同一时期。在这 在这一阶段，研究小组将开发一个可视化界面，用于全面总结复发风险和 在特定的预后亚组内。该界面将允许用户检查其代表性 用于估计复发风险的患者队列，也将提供调整后的估计 对于参与者无响应。拟议的工作将是一项开创性的努力，将建立 医疗索赔数据与患者自我报告相结合是否构成一个有希望的途径 确定是否可以将癌症复发的关键终点添加到基于人群的研究中 癌症登记处。

PROJECT SUMMARY The Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute serves as a national dashboard for cancer surveillance and is an authoritative resource for cancer incidence and mortality rates. However, the registry lacks information on cancer recurrence, which is a major outcome impacting decisions and policies around management of disease at both the individual and the population levels. The goal of the proposed study is to develop a scalable approach for population-based ascertainment of cancer recurrence with a specific focus on breast cancer. The first (UG3) phase of the study will develop an innovative algorithm for detecting a first recurrence by combining medical claims with patient reported outcomes (PROs). This phase will be conducted in collaboration with the Seattle-Puget Sound SEER registry and will leverage ongoing research agreements with Puget Sound payers. Different algorithms will be developed to allow for flexibility in maximizing sensitivity or specificity. The algorithms will be validated against gold-standard recurrence information based on electronic medical records from two major local providers. Study participants will consistent of stage I-III breast cancer patients diagnosed in 2013–2016 on whom “gold standard” information on recurrence is available. A retrospective case-control design will be employed in which an equal number of recurrent cases and non-recurrent controls at each validation site will be solicited for PROs regarding their recurrence histories. All validation activities will explicitly adjust for missing data resulting from participant non-response to the PRO solicitation. The second (UH3) phase will be conducted in collaboration with the Kentucky Cancer Registry. The algorithms developed in the UG3 phase will be deployed on all Kentucky stage I-III breast cancer diagnoses in 2015–2017 on whom claims data are available. This phase will include a validation against medical records from patients diagnosed at the University of Kentucky Markey Cancer Center over a concurrent period. In this phase, the study team will develop a visual interface for summarizing recurrence risks overall and within specified prognostic subgroups. The interface will allow users to examine representativeness of the patient cohort used to estimate recurrence risks and will also provide estimates that are adjusted for participant non-response. The proposed work will represent a pioneering effort that establish whether medical claims data combined with patient self-report constitute a promising avenue towards determining whether the critical endpoint of cancer recurrence can be added to population-based cancer registries.