Network for Investigation of Delirium: Unifying Scientists (NIDUS) Junior Investigator Pilot Award

Title: Advancing understanding of delirium severity: the contribution of domains of the delirium syndrome to outcomes ($49,871.45).

Abstract
Background: Delirium severity scores predict outcomes such as mortality, and thus add valuable additional prognostic information beyond that provided by a binary diagnosis. Published severity scales provide a single score reflecting summed ordinal measures of various domains of the delirium syndrome. Yet, delirium severity is a complex construct, and relationships between individual domains and prognosis are under-researched, potentially limiting our understanding of severity and prognosis. Our main aims are (a) to understand the state of the science through a systematic review and (b) in a substantial combined dataset to test the primary hypothesis that reduced level of arousal is associated with mortality and length of stay.

Study methods: Project 1: the systematic review will examine literature linking individual domains of delirium with outcomes. We will perform a meta-analysis if there are sufficient studies, seeking data from authors to allow an individual participant data analysis. Project 2 will involve analysis of two datasets (combined N=1200) which comprise detailed assessment of delirium components including the Delirium Rating Scale–Revised–98, cognitive assessments (attention tests including digit span, Vigilance 'A', DelApp), assessment of arousal (Richmond Agitation-Sedation Scale and Observational Scale of Level of Arousal), orientation, psychotic features, short- and long-term memory, and also follow-up data (12-week mortality, length of stay and institutionalisation). Data will be pooled using individual patient data methods. Logistic and Poisson regression models will be fitted to determine the associations between delirium domains and outcomes.

Outputs and future work: These projects will help to advance understanding of delirium severity in relation to prognosis through examining the specific contribution of domains of the delirium syndrome. Future work in larger and more diverse patient samples will lead to more precise prognostic knowledge in relation to domains of delirium, and will facilitate future studies investigating the relative contributions to prognosis of other factors such as delirium duration and etiologies.