Purpose: To evaluate the effect of non-pharmacological interventions versus standard care on incidence and duration of delirium in critically ill patients.
Methods: We searched electronic and grey literature for randomised clinical trials up to March 2018. Two reviewers independently screened, selected and extracted data. Meta-analysis was undertaken using random effects modelling.
Results: We identified 15 trials (2812 participants). Eleven trials reported incidence of delirium. Pooled data from four trials of bright light therapy showed no significant effect between groups (n = 829 participants, RR 0.45, 99% CI 0.10-2.13, P = 0.19, very low quality evidence). Seven trials of various individual interventions also failed to report any significant effects. A total of eight trials reported duration of delirium. Pooled data from two trials of multicomponent physical therapy showed no significant effect [n = 404 participants, MD (days) - 0.65, 99% CI - 2.73 to 1.44, P = 0.42, low quality of evidence]. Four trials of various individual interventions also reported no significant effects. A trial of family voice reorientation showed a beneficial effect [n = 30, MD (days) - 1.30, 99% CI - 2.41 to - 0.19, P = 0.003, very low quality evidence].
Conclusions: Current evidence does not support the use of non-pharmacological interventions in reducing incidence and duration of delirium in critically ill patients. Future research should consider well-designed and well-described multicomponent interventions and include adequately defined outcome measures.
Keywords: Critical care; Delirium; Meta-analysis; Non-pharmacological interventions; Systematic review.