Gianetta Rands, Martin Orrell

For patients with a diagnosis of vascular dementia there is evidence that aspirin is widely prescribed - in one study, completed by geriatricians and psychiatrists in the UK, 80% of patients with cognitive impairment (with vascular risk factors) were prescribed aspirin. However, a number of queries remain unanswered: Is there convincing evidence that aspirin benefits patients with vascular dementia? Does aspirin affect cognition or improve prognosis? In addition, does the risk of cerebral or gastric haemorrhage outweigh any benefit? To assess the evidence of effectiveness of the use of aspirin for vascular dementia. Computerized databases were searched independently by two reviewers. In addition, relevant websites were searched and some journals were handsearched. Specialists in the field were approached for unpublished material and any publications found were searched for additional references. The most recent search for trials was carried out in February 2000. All randomized controlled trials investigating the effect of aspirin for vascular dementia are included. Inclusion/exclusion of studies comprised systematic assessment of the quality of study design and the risk of bias. Data were extracted independently by both reviewers, using a previously tested data extraction form and, where required, authors were contacted for data not provided in the papers. The aim is to evaluate data recorded via tools assessing cognitive and behavioural changes along with mortality, morbidity and institutionalization data. No trials are eligible for inclusion in this review. There is no evidence that aspirin is effective in treating patients with a diagnosis of vascular dementia. Further research is needed to assess the effect of aspirin on cognition, and on other outcomes such as behaviour, and quality of life. At present there is no evidence relating to other queries about the use of aspirin for dementia (these are described in the Background section of this review). The most recent search for references to relevant research was carried out in February 2000, but no new evidence was found.