Patients with Alzheimer’s disease are difficult to care for not only because of their deteriorating mental capacities, but also because of frequently accompanying behavioral problems like aggression.  It is quite common for nursing home clinicians to prescribe antipsychotic medications to these patients in order to make them easier to handle, despite the fact that a National Institute of Mental Health study found in 2006 that antipsychotics were only 9% better than placebo at reducing behavioral problems. It is possible that this type of prescription continues because doctors do not see a risk that outweighs this slight benefit. A study in this month’s Lancet Neurology, conducted by Dr. Clive Ballard, clearly demonstrates that antipsychotics do present a substantial risk.

Dr. Ballard gathered 128 patients with Alzheimer’s who were already being given antipsychotics (either thioridazine, chlorpromazine, haloperidol, trifluoperazine, or risperidone) in United Kingdom care facilities. Half of these patients continued their medication regimen, while the other half was secretly given placebos.

While the first year of the study showed only a small increase in mortality between antipsychotic and placebo groups (30% compared to 23%), this difference grew considerably over time. By the second year, mortality rates were 54% versus 29%, and by the third year the gap had grown to 70% versus 41%.

The building risk of antipsychotic medications argues for doctors to limit their use to short-term prescriptions. While it might be appropriate to use antipsychotics to make patients marginally more manageable during an emergency or a difficult transition, more than half of interminable prescriptions will end in death within 2 years, a cost that warrants rethinking.