Delirium in stroke: systematic review and meta-analysis

Abstract

Due to the complexity of the symptoms, delirium detection poses achallenge in stroke patients. Alarge body of literature has established that neurospecific challenges can have aconsiderable impact on diagnosis and are underrepresented in screening. An analysis of current scientific literature on delirium screening tests and their applicability in stroke patients, acknowledging neurospecific challenges and evaluating diagnostic test accuracy. Asystematic literature search was conducted in PubMed, CINAHL, and Cochrane Library databases. Studies published between 2018 and 2021 were evaluated and the study quality was assessed according to the Institute for Clinical Systems Improvement. Furthermore, the specificity and sensitivity of delirium screening tests were pooled RESULTS: The systematic literature review found atotal of 2636articles, following areview of the inclusion and exclusion criteria. Thus, 18moderate-quality studies with atotal of 3320patients and 9distinct delirium screenings were identified. Within those 18studies, the prevalence of delirium was 34.2%. However, the delirium prevalence was significantly lower in 6studies that included patients with neurologic impairments (26.5 vs. 32.1%, p = 0.0004). Pooled sensitivity and specificity for the 4AT (Rapid assessment test for delirium ) were 82 and 77%, while these values were 72 and 93% for the CAM-ICU (Confusion Assessment Method for Intensive Care Units) and 79 and 72% for the ICDSC (Intensive Care Delirium Screening Checklist). Neurological impairments may influence the test quality of delirium screening in stroke patients. The CAM-ICU can be recommended for nonaphasic patients. The ICDSC can be used in all stroke patients on stoke units with an adjusted cut-off value of > 5points.