Delirium on stroke units: a prospective, multicentric quality-improvement project

Background Post-stroke delirium (POD) in patients on stroke units (SU) is associated with an increased risk for complications and poorer clinical outcome. The objective was to reduce the severity of POD by implementing an interprofessional delirium-management. Methods Multicentric quality-improvement project on five SU implementing a delirium-management with pre/post-comparison. Primary outcome was severity of POD, assessed with the Nursing Delirium Screening Scale (Nu-DESC). Secondary outcome parameters were POD incidence, duration, modified Rankin Scale (mRS), length of stay in SU and hospital, mortality, and others. Results Out of a total of 799 patients, 59.4% (n = 475) could be included with 9.5% (n = 45) being delirious. Implementation of a delirium-management led to reduced POD severity; Nu-DESC median: pre: 3.5 (interquartile range 2.6-4.7) vs. post 3.0 (2.2-4.0), albeit not significant (p = 0.154). Other outcome parameters were not meaningful different. In the post-period, delirium-management could be delivered to 75% (n= 18) of delirious patients, and only 24 (53.3%) of delirious patients required pharmacological treatments. Patients with a more severe stroke and POD remained on their disability levels, compared to similar affected, non-delirious patients who improved. Conclusions Implementation of delirium-management on SU is feasible and can be delivered to most patients, but with limited effects. Nursing interventions as first choice could be delivered to the majority of patients, and only the half required pharmacological treatments. Delirium-management may lead to reduced severity of POD but had only partial effects on duration of POD or length of stay. POD hampers rehabilitation, especially in patients with more severe stroke.