WASHINGTON – Monitoring for metabolic side effects proved to be infrequent for outpatients with dementia or delirium who had been prescribed antipsychotic medications at 32 Veterans Affairs medical centers.

Antipsychotic medications are approved by the Food and Drug Administration for use in patients with psychotic disorders, not for those with dementia and delirium. However, they are widely used off label to control behavior disturbances among dementia/delirium patients, despite the lack of long-term data. In 2004, the American Psychiatric Association (APA), the American Diabetes Association (ADA), the American Association of Clinical Endocrinologists, and the North American Association for the Study of Obesity issued consensus recommendations for monitoring of glycemic control, lipidemia, and weight gain among all patients taking antipsychotics (Diabetes Care 2004;27:596-601).

Subsequently, in 2005 and 2008, the FDA issued black box warnings about the increased risk for mortality in dementia patients with the use of typical and atypical antipsychotics. Other risks include cerebrovascular events, parkinsonism, pulmonary infections, and metabolic dysfunction

"To my way of thinking, these patients are getting a medication that is not FDA approved for this indication and it has significant risks, so if I’m a prudent clinician, if I prescribe something off label, I have to be more careful," Dr. Dinesh Mittal said in an interview.

But his findings in a large, retrospective cohort analysis suggest that this is not happening. Of patients identified as having received a new antipsychotic prescription and used it for at least 60 days, 916 had dementia or delirium but no psychosis, whereas 3,446 had psychosis but no dementia/delirium. The dementia patients were a mean age of 69 years, nearly all were male (95%), and most were married (56%). Two-thirds of the dementia/delirium group (64%) were aged 65 and older, compared with just 9% of the psychosis patients, said Dr. Mittal, a research scientist at the VA’s Center for Mental Health Outcomes Research and staff psychiatrist at Central Arkansas Veterans Healthcare System, Little Rock.

At baseline (defined as within 30 days of the index prescription), monitoring of either glucose or hemoglobin A_1c had been done for only 45% of those with dementia/delirium who were without psychosis. The proportion was not significantly different (47.5%) for the patients with psychosis but no dementia/delirium. Measurement of LDL cholesterol was even less common in both groups and was significantly lower for the dementia/delirium patients (27% vs. 34% of those with psychosis). Also, significantly more dementia patients than psychosis patients had their weight checked at baseline (71% vs. 66%).

The difference in weight measurement might simply relate to the fact that dementia patients were being seen more often for office visits and were having weight measured routinely, rather than being monitored specifically for metabolic side effects of antipsychotics, noted Dr. Mittal, who is also with the University of Arkansas, Little Rock.

The proportions being monitored for all side effects had dropped by 90 days, with glucose/HbA_1c measurements in just 27% of both the dementia/delirium group and the psychosis group; LDL cholesterol in 12% and 17%, respectively; and weight in 48% and 50%. Of those, the only significant difference between groups was the lower LDL measurement for the dementia/delirium patients, Dr. Mittal and his associates reported in a poster at the annual meeting of the American Association for Geriatric Psychiatry.

In the interview, Dr. Mittal commented that these results are particularly concerning given the absence of data to support the use of antipsychotics beyond a few days in dementia/delirium patients. The study sample was selected based on these patients’ having been prescribed antipsychotics for at least 60 days, which places them under the metabolic monitoring guidelines of the ADA and APA. "There are studies, but the only data [are] for short-term use for agitation of only a few days. That’s why we wanted to select a sample [of] prescribed antipsychotics long enough to get away from the data showing benefit," he explained.

The dilemma, he said, is that stopping an antipsychotic in a patient with dementia for whom they helped curb the agitation requires clinical judgment and is far more difficult in the outpatient setting, which was the focus of this study.

Although neither the psychotic nor the delirium/dementia patients were being adequately monitored for metabolic adverse effects, the delirium patients are older and might be at greater risk for morbidity and mortality while taking antipsychotics, Dr. Mittal said.

"Despite the risks of prescribing antipsychotics for patients with dementia, adherence to recommendations for monitoring metabolic side effects remains low. ... They’re being monitored as frequently as those with psychosis, but it’s not enough in either group."

Such monitoring must be accomplished via collaboration between primary care/endocrinology and mental health, he advised.

The study was funded by a health services grant from the VA’s Health Services Research & Development Service. Dr. Mittal stated that he had no other disclosures.

Monitoring Antipsychotic Side Effects Infrequent Among Dementia Patients