Community-based non-pharmacological interventions delivered by family caregivers reduce behavioural and psychological symptoms of dementia
- Correspondence to: Dr Ann M Kolanowski
School of Nursing, Pennslylvania State University, 106 Health & Human Development East, University Park, PA 16802, USA;
Commentary on: Brodaty H, Arasaratnam C. Meta-analysis of nonpharmacological interventions for neuropsychiatric symptoms of dementia. Am J Psychiatry 2012;169:946–53.
Implications for practice and research
Non-pharmacological interventions (NPIs) should be the first line of treatment for the behavioural and psychological symptoms of dementia (BPSD).
Multicomponent NPIs tailored to individual and caregiver needs can reduce BPSD in community settings.
Current instruments that measure BPSD lack precision, diluting the observed effect of NPIs; more precise outcome measures are needed.
Quality-of-life indicators may be more sensitive measures of NPI benefits than reduction in negative behaviours.
Behavioural and psychological symptoms of dementia (BPSD) are prevalent, impose a high burden on caregivers and greatly increase the cost of care. The use of pharmacological treatments is a common practice, but these drugs have not demonstrated efficacy and they carry a substantial risk for increased mortality in frail older adults.1 In group residential settings, non-pharmacological interventions (NPIs) show effectiveness as a first line of treatment for BPSD. Less is known about their utility in community settings where most people with dementia reside. The purpose of this meta-analysis was to determine the effectiveness of family caregiver interventions for reducing BPSD in people with dementia who live in the community and to assess their effect on caregiver reactions to the symptoms.
Searches were conducted in MEDLINE, Embase, PubMed, PsycINFO and Scopus for studies published between 1985 and July 2010 that met the authors’ inclusion criteria: peer-reviewed; English language; more than five participants; tested non-pharmacological intervention involving family caregiver; outcomes relevant to BPSD; and, ranked as level II (randomised controlled trial) or III-1 (pseudo-randomised controlled trial) using National Health and Medical Research Council criteria. Data from these studies were included in the meta-analysis which was performed using Review Manager. Effect sizes (Cohen's d) were calculated as the standardised mean difference between treatment and control groups or between preintervention and postintervention assessments.
Twenty-three studies met the review criteria and were included in the meta-analysis. For the 17 studies examining symptom outcomes, the medium pooled-estimate effect size was 0.34, indicating a significant and positive effect. For the 13 studies examining caregiver outcomes, the medium pooled-estimate effect size was 0.15, indicating a significant but small positive effect.
Possibly no other area of dementia care is more fraught with scientific, ethical and policy concerns than how to best treat the BPSD. Consensus groups have long argued for the use of NPI as the first line of treatment,2 and advocacy groups are quite vocal about the dangers of chemical restraints. Complicating matters, data from a recent study indicate that the discontinuation of antipsychotics is associated with return of BPSD.3 Thus, this study by Brodaty and Arasaratnam is timely and important. The investigators have conducted a well-designed meta-analysis of NPI for BPSD. With a few exceptions, they followed the PRISMA statement guidelines for reporting systematic reviews and meta-analyses.4 The thoroughness of the review and analysis helps ensure confidence in the validity of their findings: multicomponent NPIs (skill-training for caregivers, activity planning and environmental redesign, enhanced caregiver support or caregiver self-care techniques) delivered by family caregivers in the community have the potential to reduce BPSD and caregiver distress without any adverse events. Equally important, the effect sizes reported were comparable with, or better than, those found in studies of antipsychotic medications. There are a few caveats that deserve mentioning. The restrictive parameters for the selection of studies included in the review (ie, only randomised clinical trials (RCTs), and those in which the family member lived with the participant) may limit the generalisability of the findings. Reliance on RCTs as the sole source of evidence for the effectiveness of NPIs excludes much of the available data. Despite the limitations imposed by the review criteria, the investigators conducted commendable work and provided research recommendations that will advance the science of NPIs. These include the need for better tailoring of NPIs and more precise measurement of BPSD, and use of methodological approaches that can help determine the active component of treatments and the optimal dose for effectiveness.