Development and Evaluation of a Depression Risk Calculator

描述(申请人提供)：数以百万计的老年人患有抑郁症。患者参与有效治疗的一个关键障碍是难以量化抑郁的严重程度和预测未来抑郁度过的时间。更精确的抑郁症风险模型将为患者提供关于他们未来进程的更好信息，并允许更有效地分享决策。我提出了一项为期四年、有指导的研究计划，以开发和评估“抑郁风险计算器”。这一工具将有助于识别有抑郁风险的患者，为他们提供更多关于他们抑郁轨迹的信息，识别与抑郁相关的其他症状，并加强患者和提供者之间的共同决策。为了实现这一目标，我将学习纵向数据分析和定性方法的课程，并获得在初级保健中应用共享决策的经验。使用三个关于老年人的大型纵向数据库(CHS、HRS和Impact)，我将为抑郁症风险计算器开发算法。我将征求五名之前接受过抑郁症治疗的患者的意见，了解来自计算器的信息将如何对他们产生最大好处，并将相应地调整输出。然后我将确定抑郁风险计算器的反馈对患者水平结果的影响，特别是患者的参与度和治疗方法的使用。我将从计算器向20名患者提供反馈，并将他们的经历与20名患者的对照组进行比较，跟踪观察一年。这项研究将为使用风险模型改善老年抑郁症治疗的更大规模的研究产生假设和产生初步结果。

DESCRIPTION (provided by applicant): Millions of older adults suffer from depression. A key barrier to patient engagement in application of effective treatments is the difficulty in quantifying depression severity and anticipating future time spent depressed. More refined models of depression risk would give patients better information about their future course, and allow more effective shared decision making. I propose a four-year, mentored research plan to develop and evaluate a "depression risk calculator". This tool will be useful for identifying patients at risk for depression, giving them more information about their depression trajectory, identifying other symptoms associated with depression, and enhancing shared decision making between patients and providers. In order to accomplish this end, I will take courses in longitudinal data analysis and qualitative methods, and gain experience in applying shared decision making in primary care. Using three large longitudinal databases of older adults (CHS, HRS, and IMPACT) I will develop algorithms for a depression risk calculator. I will seek input from five patients previously treated for depression about how information from the calculator would have the most benefit for them, and will tailor the outputs accordingly. I will then ascertain the effects of feedback from a depression risk calculator on patient-level outcomes, particularly patient engagement and use of treatments. I will provide feedback from the calculator to 20 patients, and compare their experiences with those of a control group of 20 patients, followed for one year. This research will generate hypotheses and produce preliminary results for larger studies of using risk models to improve treatment for geriatric depression.