Long-term effectiveness of BPH/LUTS pharmacological therapies and using machine learning based predictive analytics to tailor treatment

项目摘要/摘要 该职业发展奖将支持马文·兰斯顿博士向独立过渡 翻译研究人员。他的长期研究目标包括了解该病的病因和结果 以前列腺癌(前列腺癌、前列腺炎和下尿路)为重点的泌尿系统疾病 继发于良性前列腺增生症(BPH/LUTS)；开发和验证临床工具以 改进和个性化治疗方案；并在“真实世界”的临床环境中实施这些工具。 BPH/LUTS在老年男性中是一种非常常见的疾病，可以被认为是一种 进行性疾病，主要特征是LUTS随着时间的推移而恶化，造成相当大的麻烦 极少数病例的生活质量下降、尿潴留和死亡。传统药理 在约30%的接受治疗的患者中，治疗策略未能防止长期临床结果。尽管有一个 大量的药理学研究，在治疗BPH/LUTS方面的进展非常有限 最近几年。因此，需要创新和精心设计的研究来理解长期 现代BPH/LUTS药物治疗的疗效比较 可以从电子健康记录等“大数据”来源中获取的协变量。 为了解决这些限制，在目标1中，兰斯顿博士将评估长期比较 来自Kaiser的一项观察性研究中BPH/LUTS药物治疗的有效性 永久会员人口，一个完全整合的医疗网络。如果设计良好，研究 证明现代药理学方法并不能更有效地预防长期的 BPH/LUTS的进展比单纯的传统方法更好，那么个性化的治疗选择可能 最终为减少BPH/LUTS的长期进展提供最佳策略。因此，在《目标2》中，Dr。 朗斯顿将利用现有的人口统计学、临床和实验室数据开发和验证预测模型 在诊断BPH/LUTS时确定长期BPH/LUTS临床风险增加的患者 使用基于机器学习的预测分析的结果(急性尿潴留和BPH手术)。 随着电子数据的数量、可获得性和访问权限的增加，科学家对 正确使用这些数据并将其转化为改进的泌尿外科护理的内容和方法专业知识。 这款K01为兰斯顿博士提供了作为翻译研究人员迎接这一挑战的培训 成功争取未来的R01资金。

PROJECT SUMMARY/ABSTRACT This Career Development Award will support Dr. Marvin Langston in his transition to independence as a translational researcher. His long-term research goals include understanding the etiology and outcomes of urologic conditions with an emphasis on prostate diseases (prostate cancer, prostatitis, and lower urinary tract symptoms secondary to benign prostatic hyperplasia (BPH/LUTS)); developing and validating clinical tools to improve and personalize treatment options; and implementing these tools in “real world” clinical settings. BPH/LUTS is an extremely common set of conditions among older men and can be considered a progressive disease, mainly characterized by deteriorating LUTS over time which causes considerable bother and decreased quality of life, urinary retention, and death in rare cases. Conventional pharmacological treatment strategies fail in preventing long-term clinical outcomes in about 30% of treated patients. Despite a high volume of pharmacological research, there have been very limited advances in therapy for BPH/LUTS in recent years. Therefore, innovative and well-designed studies are needed to understand the long-term comparative effectiveness of contemporary BPH/LUTS pharmacological therapies while leveraging the clinical covariates that can be captured in “big data” sources such as the electronic health record. To address these limitations, In Aim 1, Dr. Langston will evaluate the long-term comparative effectiveness of BPH/LUTS pharmacological therapies in an observational study drawn from the Kaiser Permanente membership population, a fully integrated healthcare network. If well designed studies demonstrate that contemporary pharmacological approaches are not more effective at preventing long-term BPH/LUTS progression than conventional approaches alone, then personalized treatment options may ultimately provide the best strategy for reducing long-term BPH/LUTS progression. Therefore, in Aim 2 Dr. Langston will develop and validate predictive models using demographic, clinical, and laboratory data available at the time of BPH/LUTS diagnosis to identify patients at increased risk of long-term BPH/LUTS clinical outcomes (acute urinary retention and BPH surgery) using machine learning based predictive analytics. As the volume, availability, and access to electronic data increases, so does the need for scientists with the content and methodologic expertise to correctly use and translate these data into improved urology care. This K01 provides the training for Dr. Langston to meet this challenge as a translational researcher and compete successfully for future R01 funding.