3D CAM: Deriving and Validating a 3 minute Diagnostic Assessment for Delirium

描述（申请人摘要）：谵妄（急性精神错乱）是一种病态和昂贵的综合征，影响30-40%的住院老年人。在NIH的支持下，我们在定义谵妄的流行病学和制定预防和治疗策略方面取得了实质性进展。然而，在大多数临床环境中，谵妄仍未得到充分认识。混淆评估法（CAM）算法已成为谵妄诊断的“金标准”。然而，CAM在完成之前需要进行精神状态检查（MSE），而推荐的MSE，即迷你精神状态检查加注意力测试，在临床实践中被广泛采用的时间太长。开发更短的MSE，使CAM能够准确诊断谵妄，将对临床实践和研究大有裨益。利用PI最近完成的nih资助研究中谵妄评估的两个大型数据库，我们建议开发，完善和验证3D-CAM：使用CAM算法对谵妄进行3分钟诊断评估。我们提出了两个发展和两个验证的具体目标：1)使用在急性后护理中获得的4744个谵妄评估数据集，我们将使用因子分析将MSE项目映射到关键

DESCRIPTION (applicant's abstract): Delirium (acute confusion) is a morbid and costly syndrome that affects 30-40% of hospitalized elders. With NIH support, we have made substantial progress defining the epidemiology of delirium and developing strategies for its prevention and treatment. However, in most clinical settings, delirium remains distressingly under-recognized. The Confusion Assessment Method (CAM) algorithm has become the "gold standard" for diagnosis of delirium. However, the CAM requires a mental status examination (MSE) prior to its completion, and the recommended MSE, the Mini-Mental State Examination plus attentional testing, is too long for widespread adoption in clinical practice. Development of a shorter MSE that allows accurate CAM diagnosis of delirium would be of great benefit for clinical practice and research. Leveraging two large databases of delirium assessments from the PI's recently completed NIH-funded studies, we propose to develop, refine, and validate the 3D-CAM: a 3-minute diagnostic assessment for delirium using the CAM algorithm. We propose 2 development and 2 validation Specific Aims: 1) Using a dataset of 4744 delirium assessments obtained in post-acute care, we will use factor analysis to map MSE items to key cognitive domains of delirium, and item response theory to identify a subset of items that maximize the screening efficiency for each domain. 2) Using the items identified in Aim 1 and multivariable model selection methods, we will develop the 3D-CAM. We will refine the 3D-CAM using simulations in an independent dataset of 752 delirium assessments conducted after cardiac surgery. 3) We will prospectively validate the 3D-CAM and test its inter-rater reliability in a new cohort of 600 elderly hospitalized patients. We will compare the performance of the 3D-CAM with two gold standards: the full CAM assessment and a DSM-IV-based clinical diagnosis of delirium made by an experienced geriatric clinician after a detailed assessment. 4) We will compare the performance of the 3D-CAM with the CAM-ICU, another brief screening protocol for CAM-defined delirium that does not use verbal responses. Our proposed research has numerous strengths, including our ability to leverage 2 large databases of rigorously performed delirium assessments, use of state-of-the-art measurement methodology, and the expertise of our investigative team. Most importantly, the 3D-CAM will be a critical tool for recognition of delirium, thereby improving its clinical management among hospitalized elders. The 3D-CAM will also facilitate new quality improvement initiatives to improve patient safety in hospitals, as well as education and research. PUBLIC HEALTH RELEVANCE: Delirium (acute confusion) affects 30-40% of hospitalized elders, and leads to poor clinical outcomes and higher costs; yet, only 20% of cases are recognized by the treating physicians and nurses. The goal of our research is to derive, refine and validate the 3D-CAM, a 3 minute diagnostic assessment for delirium. The 3D-CAM will provide a short, valid, and reliable assessment that can be readily integrated into clinical care, thereby facilitating accurate diagnosis, and appropriate evaluation and management of this common, morbid, and costly problem. Our research has the potential to improve the quality of clinical care and outcomes of millions of elders who are hospitalized each year.