NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Goy E, Kansagara D, Freeman M. A Systematic Evidence Review of Interventions for Non-professional Caregivers of Individuals with Dementia [Internet]. Washington (DC): Department of Veterans Affairs (US); 2010 Oct.

Cover of A Systematic Evidence Review of Interventions for Non-professional Caregivers of Individuals with Dementia

A Systematic Evidence Review of Interventions for Non-professional Caregivers of Individuals with Dementia [Internet].

Show details



The review was requested by the VHA DSC, and the DSC served as the technical expert panel for guiding topic development and reviewing drafts of the report.

The objectives of this review are to address the following questions:

Key Question #1: Do CG interventions affect the CG’s knowledge and ability to manage problematic behavior, psychosocial burden, health and health behaviors, or outcomes in the individual with dementia?

Key Question #2: What are adverse effects of CG interventions?

Population: Non-professional CGs of individuals with dementia in all settings. CGs include spouses and other family members, as well as paid sitters or assistants hired by the family; professional staff is excluded.

Interventions: Categories of intervention include psychoeducational interventions, cognitive-behavioral interventions, counseling/case management, general support services, respite care, and multicomponent interventions. Specific interventions of interest targeted CGs and included telephone-based support groups and education, Home TeleHealth/Health Buddy home monitoring device; Internet-based resources and CG assistance programs, and physical activity.

Comparator: Usual care/no interventions directed at the CG.

CG outcomes:

  • Knowledge and ability to manage problematic behavior;
  • Psychosocial outcomes (burden/subjective well-being, depression, anxiety, perceived self-efficacy, positive experiences of caregiving, satisfaction with health care, quality of life);
  • Health behaviors (diet, exercise, sleep);
  • Health (reported health, symptoms, medication use/misuse, service use, mortality).

CR outcomes:

  • Use of psychotropic drugs;
  • Cognition, mood, behavioral disturbances, social function, or physical function;
  • Hospitalizations, institutionalizations, or health care visits including ER visits;
  • Accidents;
  • Health-related quality of life;
  • Satisfaction with health care.

Setting: Home, community living center.

The DSC served as the technical expert panel for guiding topic development and reviewing drafts of the report.


We conducted a search for systematic reviews of dementia CG interventions in MEDLINE (PubMed), using the following search terms: (“dementia”[MeSH Terms] OR “dementia”[All Fields]) AND systematic[sb]. We also searched in the Cochrane Database of Systematic Reviews and the Cochrane Database of Reviews of Effects (OVID) from database inception through July 2009, using the term In addition to the search for published systematic reviews, we contacted researchers within VA to identify important recent and ongoing studies of dementia CG interventions. We also examined recently published studies found in a compendium compiled by the Administration on Aging’s (AoA) Alzheimer’s Disease Supportive Services Program2 that were not captured in previous systematic reviews. All citations were imported into an electronic database (EndNote X2).


Three reviewers assessed the titles and abstracts of citations identified from literature searches. Full-text articles of potentially relevant abstracts were retrieved for further review. We selected systematic reviews of CG interventions, using the eligibility criteria shown in Appendix A. Eligible articles had English-language abstracts and provided data relevant to the key questions. Eligibility criteria varied depending on the question of interest, as described below.

The literature search identified four systematic reviews that focused on respite care, and six reviews on technology-based interventions. We selected one systematic review on respite care and three reviews on technology-based interventions that were the most comprehensive, recent, and relevant.

There were 10 systematic reviews that evaluated a variety of psychosocial interventions, including exercise, case management, behavioral management training, individual and group skills training, individual support or counseling, and multicomponent interventions. We examined the degree of overlap between articles included in systematic reviews and found that many of the primary studies were included in more than one review. We also found that the systematic reviews grouped psychosocial interventions in different ways, combining a variety of dissimilar therapies in some cases. This made it difficult to summarize the findings of previous systematic reviews on the effects of specific forms of treatment. We therefore retrieved the full-text articles for the primary studies included in these systematic reviews, and examined each study for quality (see Quality Assessment section below), design, and type of intervention. Out of concern that we might miss content from good quality primary studies that were reviewed elsewhere (in systematic reviews excluded from our sample due to poor quality), we retained one very comprehensive systematic review3 that had identified 127 CG intervention studies but did not meet our quality criteria for systematic reviews. This effort contributed an additional three RCTs to our review of primary data on psychosocial interventions.

Altogether there were 224 primary studies included among the 11 systematic reviews of psychosocial interventions. Of these, we selected RCTs rated good-quality by the respective systematic review and with sample size greater than 50, and analyzed the body of evidence for specific forms of treatment. This approach allowed us to identify the best evidence for specific psychosocial interventions, based on the literature searches and quality assessments previously conducted by existing systematic reviews.

Our expert review panel recommended additional individual studies published after the search dates of the respective systematic reviews, and were considered by panel members to demonstrate important advances in the field. We also examined studies found in the compendium of intervention studies compiled by the AoA Alzheimer’s Disease Supportive Services Program. These studies met our criteria for inclusion as good quality, RCTs, and were incorporated into our overall review.


For technology-based interventions and respite care, we summarized the findings of recent systematic reviews that had performed comprehensive, qualitative syntheses of the primary literature on these topics.

From RCTs on psychosocial interventions, we abstracted information about sample characteristics; the methods used for the intervention and control groups; the outcome measures used; and the results for CG and CR. We compiled evidence tables organized for the psychosocial interventions.


We rated the quality of systematic reviews using the criteria shown in Appendix B.4, 5 We selected good-quality systematic reviews based on the comprehensiveness and reproducibility of the search strategy, the use of standard methods to appraise the validity of included studies, and the absence of apparent bias in drawing conclusions.

As noted, we also included one systematic review3 that did not meet our quality criteria (the methods for quality rating were not reproducible as described), because it included the most comprehensive report of primary studies, and improved our confidence that we were not missing good quality primary studies. We examined 78 controlled trials of psychosocial interventions that had been previously rated high in quality by existing systematic reviews, by considering the following elements: the comparability of treatment groups; the adequacy of randomization; whether treatment allocation was concealed; whether eligibility criteria were specified; the use of blinding among patients, care providers, and outcome assessors; whether the analysis was intention-to-treat, or conducted with post-randomization exclusions, or with extensive or differential loss to follow-up; clearly defined interventions; and reliable outcome measurement (Appendix C).4 We applied criteria for randomization and adequate sample size (n≥50) to select uniformly the studies that would most likely represent the best evidence on a particular intervention, among the controlled trials that had been identified and screened by good quality systematic reviews.


We organized the literature into the following intervention categories:

  • Psychosocial interventions
    • Multicomponent interventions
    • Exercise training
    • Case management
    • Behavioral management training
    • Individual skills training
    • Group skills training
    • Individual, group, and combined individual/group supportive counseling
  • Technology-based interventions
  • Respite care

We selected systematic reviews on respite care and technology-based interventions that had performed recent, thorough assessments of the relevant evidence, and therefore represent the current knowledge available on these topics. We chose to present a summary of their findings directly in this report. For psychosocial interventions, which were far more heterogeneous, we critically analyzed primary trials selected to represent the best evidence on these topics according to the criteria listed above. We compiled a qualitative, descriptive synthesis of the evidence on specific forms of therapy: exercise, case management, behavioral management training, individual and group skills training, individual support or counseling, and multicomponent interventions.

A list of abbreviations is provided in Appendix D.


A draft version of this report was sent to the technical advisory panel and additional peer reviewers. Their comments and our responses resulted in updates to our review and are included in Appendix E.


Other titles in this collection

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...