Effect of an interactive E-learning tool for delirium on patient and nursing outcomes in a geriatric hospital setting: findings of a before-after study

Elke Detroyer, Fabienne Dobbels, Andrew Teodorczuk, Mieke Deschodt, Yves Depaifve, Etienne Joosten, Koen Milisen

Education of healthcare workers is a core element of multicomponent delirium strategies to improve delirium care and, consequently, patient outcomes. However, traditional educational strategies are notoriously difficult to implement. E-learning is hypothesised to be easier and more cost effective, but research evaluating effectiveness of delirium education through e-learning is scarce at present. Aim is to determine the effect of a nursing e-learning tool for delirium on: (1) in-hospital prevalence, duration and severity of delirium or mortality in hospitalized geriatric patients, and (2) geriatric nurses' knowledge and recognition regarding delirium. A before-after study in a sample of patients enrolled pre-intervention (non-intervention cohort (NIC); n = 81) and post-intervention (intervention cohort (IC); n = 79), and nurses (n = 17) of a geriatric ward (university hospital). The intervention included an information session about using the e-learning tool, which consisted of 11 e-modules incorporating development of knowledge and skills in the prevention, detection and management of delirium, and the completion of a delirium e-learning tool during a three-month period. Key patient outcomes included in-hospital prevalence and duration of delirium (Confusion Assessment Method), delirium severity (Delirium Index) and mortality (in-hospital; 12 months post-admission); key nurse outcomes included delirium knowledge (Delirium Knowledge Questionnaire) and recognition (Case vignettes). Logistic regression and linear mixed models were used to analyse patient data; Wilcoxon Signed Rank tests, McNemar's or paired t-tests for nursing data. No significant difference was found between the IC and NIC for in-hospital prevalence (21.5% versus 25.9%; p = 0.51) and duration of delirium (mean 4.2 ± SD 4.8 days versus 4.9 ± SD 4.8 days; p = 0.38). A trend towards a statistically significant lower delirium severity (IC versus NIC: difference estimate - 1.59; p = 0.08) was noted for delirious IC patients in a linear mixed model. No effect on patient mortality and on nurses' delirium knowledge (p = 0.43) and recognition (p = 1.0) was found. Our study, the first in its area to investigate effects of delirium e-learning on patient outcomes, demonstrated no benefits on both geriatric patients and nurses. Further research is needed to determine whether delirium e-learning nested within a larger educational approach inclusive of enabling and reinforcing strategies, would be effective. ISRCTN ( 82,293,702 , 27/06/2017).