Deriving high-quality evidence from national healthcare databases to improve suicidality detection and treatment outcomes in PTSD and TBI

项目总结 创伤后应激障碍(PTSD)具有复杂的共病症状(共病)。 并与高自杀风险有关，特别是在退伍军人中，这是导致死亡的主要原因。 创伤后应激障碍药物治疗严重缺乏进展，这从非标签药物的使用增加可见一斑。 药物和多药联用(多种药物同时使用)。随之而来的关于 治疗的相对风险和好处造成了创伤后应激障碍管理的危机。此外，创伤后应激障碍及其主要 在电子健康中，共病[创伤性脑损伤(TBI)和自杀]通常仍未记录在案 记录(EHR)。疾病结果的可预测性也很差，因为 药物治疗和多种修饰合并症。我们的长期目标是提高诊断水平， 通过增强EHR对创伤后应激障碍及其合并症的二级/三级预防和治疗结果 利用率。为了实现我们的目标，我们将分析退伍军人的电子健康记录和行政索赔数据 管理(VA)和非VA数据库，总共涵盖&gt；200万创伤后应激障碍患者和&gt；200万创伤性脑损伤患者。 具体地说，我们的目标是：(1)识别EHR中未检测到的创伤后应激障碍、创伤后应激障碍和自我伤害(使用机器学习 使用和不使用自然语言处理)，以指导卫生服务的改进。(2)预测 通过对疾病轨迹的新建模研究退伍军人中的创伤后应激障碍临床病程 时变治疗和偏差(3)比较创伤后应激障碍精神药物单一疗法的有效性， 多药联用和心理治疗，以指导改善患者预后的治疗选择。通过 增强和验证我们团队开发的机器学习方法，我们将归因于未记录的 来自两个数据集的创伤后应激障碍、创伤后应激障碍和自残，并描述与文档相关的因素 差距。我们将用增强的潜在类别分析对疾病轨迹进行建模，重点放在自我伤害上， 药物滥用和创伤后应激障碍的精神科住院治疗。随着本地控制方法的创新，我们 将比较患有和不伴有脑损伤的退伍军人患创伤后应激障碍的风险。最后，我们将表演最大的 (到目前为止)创伤后应激障碍治疗在&gt；100单一疗法和多药联用中的疗效比较研究 养生法加上心理治疗干预。这些研究将提供高质量的证据来证明 住院、滥用药物和自杀行为/自残。成功完成这些调查 将提高提供者和患者的决策质量，并引导改进的服务提供 患有创伤后应激障碍/创伤性脑损伤和/或高自杀风险的退伍军人和非退伍军人。

PROJECT SUMMARY Post-traumatic stress disorder (PTSD) has complex profiles of co-occurring medical conditions (comorbidities) and is associated with high risk of suicide, particularly among Veterans, in which it is a leading cause of death. There is a critical lack of advancement in PTSD pharmacotherapy, as illustrated by increased use of off-label medications and polypharmacy (multiple drugs used simultaneously). The consequent limited evidence on the relative risks and benefits of treatments creates a crisis in PTSD management. Moreover, PTSD and its major comorbidities [traumatic brain injury (TBI) and suicidality] often remain undocumented in electronic health records (EHR). There is also poor predictability of disease outcomes since there are frequent changes in pharmacological treatment and multiple modifying comorbidities. Our long-term goal is to improve diagnostics, secondary/tertiary prevention, and treatment outcomes of PTSD and its comorbidities via enhanced EHR utilization. To achieve our objectives, we will analyze EHR and administrative claims data from Veterans Administration (VA) and non-VA databases, collectively covering >2M PTSD and >2M TBI patients. Specifically, we aim to: (1) Identify undetected PTSD, TBI, and self-harm from EHRs (using machine learning with and without natural language language processing) to guide health service improvements. (2) Predict PTSD clinical course in the VA population through novel modeling of disease trajectories that account for time-varying treatments and biases (3) Compare the effectiveness of PTSD psychotropic monotherapies, polypharmacy, and psychotherapy to guide the choice of treatment for improved patient outcomes. By enhancing and validating a machine learning approach developed by our team, we will impute unrecorded PTSD, TBI, and self-harm from both datasets, and characterize factors associated with documentation disparities. We will model diseases trajectories with enhanced latent class analysis, focusing on self-harm, substance misuse, and psychiatric hospitalization in PTSD. With Local Control methodology innovations, we will compare the risk of PTSD in veterans with and without comorbid TBI. Finally, we will perform the largest comparative effectiveness studies (to date) of PTSD treatments on >100 monotherapy and polypharmacy regimens plus psychotherapy interventions. These studies will provide high-quality evidence on the risk of hospitalizations, substance misuse, and suicidal acts/self-harm. Successful completion of these investigations will improve the quality of decision making for providers and patients, and guide improved service delivery to the population of veterans and non-veterans with PTSD/TBI, and/or high risk of suicide.