Measurement in Health: Advancing Assessment of Delirium

Delirium, an acute clinical syndrome characterized by inattention and cognitive disturbances, is a common, preventable, and costly problem for older persons.

Over 3 million people develop delirium annually, accounting for over $160 billion in annual healthcare expenditures in the United States alone.

While prevention of delirium is possible in up to 30-40% of cases, under-recognition of the syndrome continues to occur in over two-thirds of cases. A problem plaguing delirium research is difficulty in identification.

Clinicians working in different disciplines have developed many tools to aid in delirium identification across varied clinical settings (for example, medical units, emergency room, intensive care units). The lack of consensus on key features and approaches has posed a major barrier to progress in the field of delirium.

To address this gap, the overarching goal of this proposal is to use advanced methods in measurement to improve the identification of delirium.

The specific aims are to: (1) determine the 4 most commonly used and well-validated instruments for delirium identification; (2) harmonize the 4 most commonly used and well-validated assessment instruments to generate an item bank, which is a collection of the individual instrument questions or ratings along with their parameter estimates from item response theory; (3) identify the threshold (cut-point), which will best identify delirium (vs. absence of delirium) rates for the harmonized item bank.

Accomplishing the first aim will occur through a rigorous systematic review of the published medical literature with methodologic quality rating of the articles.

The second aim will utilize modern psychometric methods and harmonization methods in a secondary analysis of the Better Assessment of Illness (BASIL) study and other studies which have simultaneously administered multiple delirium measures of interest.

The third aim will apply simulation methods to identify the optimal cut-points to determine whether a patient has delirium.

The significance of the proposed research is the creation of a harmonized item bank to provide a unified approach to identification of delirium that will help to advance both clinical care and research. The measure will allow for the comparison of studies regardless of their method of delirium identification.

Along with a single unified measure, crosswalks between delirium measures will be developed.

Ultimately, the measure will facilitate combining studies, paving the way for meta-analyses of existing studies, which are essential for guideline development and clinical decision-making.

Additionally, the harmonized item bank will provide the means to create big datasets for future largescale pathophysiologic studies (e.g., ?omics studies or population- based investigations) to advance our understanding of molecular, cellular, and physiological mechanisms underlying delirium.

Completion of these projects will help lay the foundation for my future career in aging research as a physician scientist.