Defining the Phenotype of Complicated Delirium Associated with Long-Term Cognitive Decline

摘要 谵妄被认为是一种短暂的、可逆的综合征。然而，长期随访研究表明， 谵妄并不总是短暂和可逆的，并且与不良后果的风险增加相关，例如 持续和加速的认知能力下降、阿尔茨海默病或痴呆、死亡和过度服务利用- zation.我们已经证明，大约一半的人谁发展术后谵妄返回到术前， 手术后8周内的基线认知表现水平，但大约三分之一的人， 出现术后谵妄表现为术后36个月认知能力加速下降。我们 复杂性谵妄的工作定义，在我们之前的支持期间发展，是谵妄相关的 在长期随访中具有较高程度（或速度）的认知下降（即，≥ 2-3年）。在这个小组中， 这一速度与轻度认知障碍者的速度相似。定义复杂 就长期认知能力下降而言，谵妄是有问题的，因为多年来无法检测到结果。 因此，本研究将有助于识别复杂性谵妄的预测因子，以协助早期识别。我们 目的是：（1）使用专家小组确定早期识别复杂性谵妄的预测因子，（2） 使用经验数据识别早期识别复杂性谵妄的预测因子，以及（3）验证 独立样本中的预测模型（外部有效性）和针对临床结局的预测模型（预测性） 有效性）。我们将使用术前、术中和术后即刻可能获得的信息。 我们还将使用来自脑脊液、血清、神经生理学和免疫组织化学的生物标志物评估模型。 测量和手术前获得的神经成像。我们将实现我们的目标（a）的见解 在改良的德尔菲法中对谵妄的专家进行分析;（B）对已经收集的丰富数据进行二级数据分析 在SAGES I队列中，以及（c）在背景下收集的新观察和临床数据的验证 新的SAGES II队列研究。我们将开发多种模型，包括术前，围手术期， 术后预测变量集及其组合。我们的团队包括定性和 定量方法学和谵妄研究领域的领导者。这项工作的最终目标是提高 临床医生对谵妄的识别和治疗，并提高谵妄的预后重要性， 临床医生、患者和政策制定者。

ABSTRACT Delirium has been considered a transient, reversible syndrome. However, long-term follow-up studies suggest delirium is not always transient and reversible, and is correlated with increased risk of bad outcomes, such as persistent and accelerated cognitive decline, Alzheimer's disease or dementia, death, and excess service utili- zation. We have demonstrated that about half of people who develop postoperative delirium return to preoper- ative baseline cognitive performance levels within about 8 weeks of surgery, but about a third of those who develop postoperative delirium show accelerated cognitive decline out to 36 months following surgery. Our working definition of complicated delirium, developed in our prior period of support, is delirium associated with a higher degree (or pace) of cognitive decline in long-term follow-up (i.e., ≥ 2-3 years). In this sub-group, the pace is similar to that observed among persons with mild cognitive impairment. Defining complicated delirium in terms of long-term cognitive decline is problematic since the outcome cannot be detected for years. Therefore, this Project will help identify predictors of complicated delirium to assist with early identification. Our aims are: (1) to identify predictors for early identification of complicated delirium using an expert panel, (2) identify predictors for early identification of complicated delirium using empiric data, and (3) to validate the predictive models in an independent sample (external validity) and against clinical outcomes (predictive validity). We will use information that is potentially available before, during and immediately following surgery. We will also evaluate models using biomarkers derived from cerebrospinal fluid, serum, neurophysiologic measures, and neuroimaging obtained before surgery. We will accomplish our aims with (a) the insights of experts in delirium in a modified Delphi process; (b) secondary data analysis of the rich data already collected in the SAGES I cohort, and (c) validation with new observational and clinical data collected within the context of the new SAGES II cohort study. We will develop multiple models including preoperative, perioperative, and postoperative predictor variable sets, and their combination. Our team includes experts in qualitative and quantitative methodology and field leaders in the study of delirium. The ultimate goal of this work is to improve delirium recognition and treatment by clinicians, and heighten the prognostic importance of delirium among clinicians, their patients, and policy makers.