The Causal Effect of Effective Depression Treatment on HIV Outcomes in CNICS

描述(由申请人提供)：抑郁症是艾滋病毒感染患者中非常常见的一种共病。抑郁症与一系列不良的健康行为和结果有关，包括抗逆转录病毒(ARV)药物依从性降低、HIV传播危险行为增加、病毒学失败增加和死亡率上升。抗抑郁药物和心理疗法在治疗艾滋病毒感染者的抑郁方面是有效的。因此，对于HIV临床医生、研究人员和政策制定者来说，一个重要的问题是抑郁症治疗在多大程度上可以改善HIV的行为、风险传播和临床结果，并帮助实现ARV治疗作为预防的全部潜力。有限的真实世界、强有力的证据表明，通过投资于艾滋病毒患者的高质量抑郁症治疗，可以在人口层面上对传播和健康产生影响。对抑郁症治疗对艾滋病毒结果的影响的准确和有效的因果估计，特别是从大的和可推广的样本中估计，对于确定这种投资的艾滋病毒相关回报是必要的。少数随机对照试验表明，在强化的基于心理治疗的抑郁症治疗干预后，ARV依从性有所改善，但这些试验是在小样本和不可推广的样本中进行的。一些观察性研究表明抗抑郁药物治疗与ARV依从性呈正相关，但这些研究通常没有使用适当的因果推断统计方法来解释抑郁症状和治疗之间的动态关系。到目前为止的研究也普遍将所有抗抑郁药治疗归类在一起，而不是区分适当的(最佳实践)治疗与临床惰性--非精神科提供者未能滴定抗抑郁剂的倾向 治疗持续症状的剂量。在这个项目中，我们将利用大量和多样化的CNICs观察队列中的丰富信息，包括来自全美8个CFAR临床站点的25,000名患者，以完成以下目标：(1)表征CFAR的频率 目的：(1)利用最先进的因果推理方法，评估抗抑郁治疗，特别是抗抑郁治疗对艾滋病毒相关行为和健康结局的影响程度；以及(3)评估适当的抑郁治疗对某些亚组的艾滋病毒结局的影响是否有所不同，例如，抑郁严重程度更高、同时存在焦虑或有问题的酒精或药物使用的人群。这项工作的范围很重要，因为我们将(1)定义目前艾滋病毒初级保健中的“抑郁症治疗差距”，(2)定义通过缩小这一差距可实现的人群水平对艾滋病毒结果的影响，以及(3)确定减少治疗差距可能对其最有益的亚群。我们的建议是创新的，它使用了先进的因果推理方法，并专注于区分抗抑郁药物治疗的适当性和缺乏性。CNICs所代表的庞大和多样化的患者群体将增强这项工作的发现在全美从事艾滋病毒初级保健的艾滋病毒感染患者群体中的普适性。

DESCRIPTION (provided by applicant): Depression is an exceedingly common comorbidity among HIV-infected patients. Depression has been associated with a range of adverse health behaviors and outcomes including reduced antiretroviral (ARV) medication adherence, increased HIV transmission risk behaviors, increased virologic failure, and higher mortality. Antidepressant medications and psychotherapy are effective in treating depression in HIV-infected patients. Consequently, an important question for HIV clinicians, researchers, and policy-makers is the extent to which depression treatment can improve HIV behavioral, risk transmission, and clinical outcomes and help realize the full potential of ARV treatment-as-prevention. Limited real-world, robust evidence exists about the population-level impact on transmission and health that could be achieved by investing in high-quality depression treatment for HIV patients. Precise and valid causal estimates of the effect of depression treatment on HIV outcomes, especially from large and generalizable samples, are necessary to define the HIV-related return on such investment. A small number of randomized controlled trials have demonstrated improvements in ARV adherence after intensive psychotherapy-based depression treatment interventions, but have been conducted in small and non-generalizable samples. Several observational studies have suggested a positive association of antidepressant treatment with ARV adherence, but these studies have generally not used appropriate causal inference statistical methods to account for the dynamic relationship between depressive symptoms and treatment. Studies to date have also generally grouped all antidepressant treatment together, rather than distinguish adequate (best-practices) treatment from clinical inertia - the tendency of non-psychiatric providers to fail to titrate antidepressant doses to address ongoing symptoms. In this project, we will draw on the rich information in the large and diverse CNICS observational cohort, encompassing 25,000 patients from 8 CFAR clinical sites across the US, to complete the following aims: (1) to characterize the frequency of adequate vs. inadequate antidepressant treatment provided within HIV clinical care; (2) to estimate the magnitude of the effect of antidepressant treatment, and specifically adequate antidepressant treatment, on HIV-related behavioral and health outcomes using state-of-the-art causal inference methods; and (3) to assess whether the effect of adequate depression treatment on HIV outcomes differs for certain subgroups, such as those with greater depressive severity, concurrent anxiety, or problematic alcohol or drug use. This scope of work is significant because we will (1) define the current "depression treatment gap" in HIV primary care, (2) define the population-level impact on HIV outcomes achievable by reducing that gap, and (3) define subpopulations for which the reduction of the treatment gap could be most beneficial. Our proposal is innovative in its use of advanced causal inference methods and its focus on distinguishing adequate from inadequate antidepressant treatment. The large and diverse patient population represented in CNICS will enhance the generalizability of this work's findings to the population of HIV-infected patients engaged in HIV primary care across the United States.