SAN JUAN, P.R. — Atypical antipsychotics appear to have a modest effect on behavioral symptoms in elderly patients with dementia, but the effectiveness of nonpharmacologic treatments is less clear, according to a metaanalysis presented at the annual meeting of the American Association for Geriatric Psychiatry.
Dr. Mark B. Snowden of the department of psychiatry and behavioral sciences at the University of Washington in Seattle and his colleagues used metaanalysis techniques to compare the efficacy of nonpharmacologic treatments with that of pharmacologic therapies.
Articles from peer-reviewed, English language publications, including textbooks, from 1970 on were considered for the analysis. Nursing home residents had to make up at least half of the populations being studied. In addition to literature searches in several medical and nursing databases, the researchers submitted articles that they were aware of but that had not previously been identified. Articles were included only if they documented randomized, controlled trials.
The researchers identified five randomized, controlled trials of antipsychotic drugs and three randomized, controlled trials for nonpharmacologic interventions. The drug trials included four atypical drugs and one traditional antipsychotic drug.
The nonpharmacologic trials included 8 hours of nurses' aide training to communicate more effectively with patients with dementia, 8 hours of education/training with weekly follow-ups and hands-on activities of daily living care, 3 hours per day of psychosocial activities, and combined nonpharmacologic approaches.
The calculated effect size for nonpharmacologic interventions was − .088, which was not statistically significant. In comparison, the calculated effect size for pharmacologic interventions was − .23, which “would be considered small to modest at best,” Dr. Snowden said. “In this instance, the finding was consistent enough across studies that it is statistically significant.”
Only the pharmacologic studies provided data on the number of patients whose condition did or did not improve. giving a statistically significant mean odds ratio of 1.87; thus, patients had an 87% chance of improving with drug treatment.
The researchers also calculated the benefit-to-harm ratio for antipsychotic treatment. “For every 14 people who got a drug and improved, you would expect one excess death,” Dr. Snowden said. While Dr. Snowden pointed out that one excess death is not a trivial number, “when presented with this data, I have yet to have a nursing home family say that they don't want an antipsychotic drug given to their relative.”
In 2003, the American Geriatrics Society and the American Association for Geriatric Psychiatry released a consensus statement on the management of behavioral symptoms associated with dementia. In the statement, the two groups recommended the use of nonpharmacologic interventions as the initial treatment, as long as patients did not display psychotic symptoms and there was no immediate danger to the resident or to others. The statement iterated that antipsychotic drugs should only be considered for first-line treatment in cases with severe behavioral symptoms with psychotic features. Since then, the Food and Drug Administration has issued a public health advisory about increased mortality associated with off-label use of atypical antipsychotics in elderly patients.
“Given the modesty of the effect size, I think we probably need to remove the requirement for psychosis or danger. … Danger is a very high standard,” Dr. Snowden said. “If you say you can only use antipsychotics in someone who is dangerous, there are going to be a lot of people who are distressed that you're not going to treat.”