Teamwork to reduce risk of delirium

April 2015 Vol 15 (2)

This edition’s critically appraised topic (CAT) looks at how best to reduce the risk of delirium in elderly patients without turning to drugs.


Non-pharmacological interventions – when the intervention is consistently implemented and has multiple components – reduce the incidence of delirium in elderly patients in acute hospitals by about 30 per cent. They may also substantially reduce the risk of accidental falls.


Elderly patients are at risk of developing delirium, particularly after surgery or severe illness. This is
distressing for the patient and their family, and increases the risk of poor outcomes. As a nurse leader you are also aware that caring for patients with delirium puts increased pressure on an already stretched nursing workforce. Non-pharmacological strategies are understood to prevent delirium but you are unsure of their effect and maybe you are doing enough already? To inform quality improvement you decide to search for evidence of their effectiveness.


In the hospitalised elderly, do non-pharmacological delirium prevention strategies prevent delirium when consistently implemented alongside standard care?


PubMed Clinical Queries (therapy, narrow): preventing delirium AND (nonpharmacological or non-pharmacological)


Martinez, F, Tobar, C, & Hill, N (2015). Preventing delirium: should non-pharmacological, multicomponent interventions be used? A systematic review and metaanalysis of the literature.Age Ageing, 44(2), 196-204. doi: 10.1093/ageing/afu173


A systematic review assessing the effect of multicomponent non-pharmacological interventions
in preventing incident delirium (developing during the hospital stay) in the elderly. Delirium was diagnosed using standardised National Institutes for Clinical Excellence (NICE) criteria.

Inclusion criteria for the review were:

  • Type of study: Randomised controlled trials involving elderly over 60 years
  • Studies excluded: Trials evaluating management of prevalence (existing) delirium, assessing pharmacological interventions or aimed at alcohol withdrawal delirium were excluded
  • Intervention: Multicomponent interventions were considered to be those with at least two components from the following domains: physical interventions including hydration (fluid therapy), electrolyte and nutrition, safe environment directives, drug reviews, cognitive stimulation programmes, daily reorientation activities, educational interventions for staff and family members, family involvement in patient care, and physical or occupational therapy during hospital stay
  • Comparison: Usual care defined as the standard care given to patients, including interventions
    specific to correction of underlying causes when already present 

  • Primary outcomes: Incident delirium at any point during hospitalisation

  • Secondary outcomes: Delirium duration, length of hospitalisation, accidental falls, institutionalisation rates and in-hospital mortality


Seven electronic databases were searched from inception to 31 December 2012 using a comprehensive iterative search strategy. No language restrictions applied. Unpublished studies were sought via grey literature repositories (OpenGrey, and by hand searching reference lists of delirium guidelines.

Two authors independently conducted the searches, quality assessment and data extraction. Data extraction included methods for delirium diagnosis and used a standardised form. Disagreements were solved by consensus, or third party arbitration.

The Cochrane Collaboration criteria was used to evaluate study quality and included assessment
of appropriateness of randomisation, allocation concealment, blinding, loss to follow-up, and intention to treat analysis. Publication bias was formally assessed using funnel plots and Egger’s test.

Overall it was a high-quality review of moderate to high-quality studies.


A total of 21,788 articles were screened, of which 602 were considered for the review. After abstract assessment, 28 full texts were assessed; seven studies involving 1,691 participants met the selection criteria and were included in the review. Studies involved elderly patients in an orthopaedic ward with hip fractures (three studies), in Summary of Results acute medical wards (two studies), in a coronary care unit (one study) and in intensive care (one study).

Studies were conducted in Australia, Sweden, United States, Spain, Chile and Italy and the majority were published from 2005 onwards.

Various interventions targeting known risk factors were involved. Common components were physiotherapy (70 per cent of trials), daily reorientation (60 per cent), family involvement in care (60 per cent), and avoidance of sensory deprivation (60 per cent).

Meta-analysis of seven studies found that multicomponent interventionssignificantly reduced incident delirium in hospitalised patients by nearly 30 per cent when compared with standard care (table). Two of these seven studies also measured accidental falls (these two studies were of moderate quality) and found that multicomponent interventions led to a 60 per cent reduction in accidental falls (table). Non-significant reductions in delirium duration, hospital stay and mortality were observed with delirium prevention interventions.


  • The significant reduction in delirium rates identified in this review supports the results from other published reviews that had included nonrandomised (lower quality) trials.
  • Delirium is a complex neuropsychiatric condition – prevention therefore requires complex interventions that target known risk factors.
  • Multidisciplinary teamwork, staff training in delirium prevention, and family involvement were
    key features of successful interventions. Many intervention components related to the consistent provision of high-quality nursing care.
  • Standardised protocols for reducing delirium incidence will require sound implementation strategies to ensure that practice change is adequately supported, resourced and sustained.


Cynthia Wensley RN MHSc is a PhD candidate at Deakin University. She also works at the School of Nursing, The University of Auckland as an honorary professional teaching fellow.

Summary of Results

delirium graph border

Post your comment


No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments