BACKGROUND: Research has highlighted the problems for carers of people with dementia. These include the effective loss of companionship and support of a life partner, social isolation and complex financial, legal and social decision making. The burden of caring is financially, emotionally and physically significant. Studies of the effects of ways of supporting carers and reducing the burden of caring often examine a range of outcomes. Probably because of these characteristics, no formal overview of this area of service provision has been undertaken. Moreover, the term 'intervention' in relation to caregivers of people with Alzheimer's disease is open to wide variations in interpretation at the level of service provision. Interventions range from the application of 'hi-tech' computer technology for socially isolated carers, to the formation of specialist support groups and respite services. OBJECTIVES: The objective of this review is to provide an assessment of the effectiveness of health and/or social interventions designed to help support the carers of people with Alzheimer's-type dementias. SEARCH STRATEGY: The Cochrane Controlled Trials Register was searched using the terms 'carer*' and 'caregiv*'. See the Cochrane Dementia & Cognitive Impairment group's details for the full search strategy. SELECTION CRITERIA: All randomised trials in which primary carers of people with Alzheimer's disease are allocated to either intervention or non-intervention/control groups and where the intervention was provided by healthcare and/or social services. DATA COLLECTION AND ANALYSIS: Data was extracted by both reviewers and was found to be unsuitable for quantitative synthesis. A qualitative overview is presented and structured according to the comparisons used in the review. Weighted Mean Differences (weighted by sample size) for each outcome and its subcategories are also presented. MAIN RESULTS: The results of the review are inconclusive. No evidence was found for the following comparison interventions:1) individualised service assessment and planning versus conventional support2) technology-based carer networking (via computers or telephones) versus conventional support3) carer-education/training versus conventional support4) multi-faceted/dimensional strategies (such as specialised carer assessment and training) versus conventional supportWhilst the overview suggests little or no evidence that interventions to support caregivers of people with Alzheimer's disease are of quantifiable benefit, the poor quality and small sample sizes of the studies, as well as the diversity of interventions examined and outcomes reported, mean that this conclusion cannot be put forward without the need for caution, particularly as some of the studies put forward qualitative evidence which contradicts this conclusion. AUTHORS' CONCLUSIONS: With the limited nature of the research evidence in mind, it is not possible to recommend either wholesale investment in caregiver support programmes or withdrawal of the same. With the addition of further studies in future updates of this review, expected in early 1999, this presently inconclusive picture may become clearer.A number of conclusions relating to future areas of research can be put forward with more conviction. Specifically,1) Future trials need to examine interventions included in the existing knowledge base on Alzheimer's carer-supportive interventions.2) Outcome measures used should mirror those in similar studies.3) Outcome measures need to be clinically and 'lay' relevant.4) Trials need to be of longer duration given the 7-10 years median life expectancy of people with Alzheimer's Disease.5) Sample sizes need to be increased and calculated properly given the likelihood of a moderate intervention effect (if any).7) Blinding at the outcome assessment stage needs to be a part of future basic trial designs if bias is to be avoided.8) If reviews are to have their power increased then quality of reporting results needs to be improved.