PRospective Observational POLIsh Study on post-stroke delirium (PROPOLIS): methodology of hospital-based cohort study on delirium prevalence, predictors and diagnostic tools

BACKGROUND Between 10 % to 48 % of patients develop delirium in acute phase of stroke. Delirium determinants and its association with other neuropsychiatric disturbances in stroke are poorly understood. The wildly accepted predictive model of post-stroke delirium is still lacking. METHODS/DESIGN This is a prospective, observational, single-center study in patients with acute phase of stroke. We aim to include 750 patients ≥18 years with acute stroke or transient ischemic attack admitted to the stroke unit within 48 hours after stroke onset. The goals of the study are: 1) to determine frequency of delirium and subsyndromal delirium in Polish stroke patients within 7 days after admission to the hospital; 2) to determine factors associated with incidence, severity and duration of delirium and subsyndromal delirium and to create a predictive model for post-stroke delirium; 3) to determine the association between delirium and its cognitive, psychiatric, behavioral and functional short and long-term consequences; 4) to validate scales used for delirium diagnosis in stroke population. Patients will be screened for delirium on daily basis. The diagnosis of delirium will be based on DSM-V criteria. Abbreviated version of Confusion Assessment Method and Confusion Assessment Method for the Intensive Care Unit will be used for delirium and sub-delirium screening. Severity of delirium symptoms will be assessed by Delirium Rating Scale Revised 98 and Cognitive Test for Delirium. Patients who survive will undergo extensive neuropsychological, neuropsychiatric and functional assessment 3 and 12 months after the stroke. DISCUSSION This study is designed to provide information on clinical manifestation, diagnostic methods and determinants of delirium spectrum disorders in acute stroke phase and their short and long-term consequences. Collected information allow us to create a predictive model for post-stroke delirium.