What’s Best When Prescribing Antipsychotics to Patients With Dementia?

On May 1, APA published its Practice Guideline on the Use of Antipsychotics to Treat Agitation or Psychosis in Patients With Dementia. The publication, which offers 15 evidence-based recommendations for how best to approach this critical pharmacological problem, is the result of hundreds of hours of review and discussion by a 10-member group of psychiatrists over the past two years.

UCSF

Victor Reus, M.D., a professor of psychiatry at the University of California, San Francisco, and chair of the writing group, recently spoke with Psychiatric News about the process of putting together the practice guideline and the conclusions reached by the group.

Why was the treatment of agitation or psychosis in patients with dementia selected as the next set of APA practice guidelines?

I was not in the original selection committee that prioritized these issues, but I know that the group tasked with deciding what issue to tackle looked at areas that are thought to be of clinically urgent concern and in which there is enough of an evidence base to provide meaningful recommendations for treatment.

Over the last 10 years, despite only modest evidence for short-term efficacy, and increasing evidence of morbidity and mortality associated with their use, antipsychotic drug use in the treatment of dementia symptomatology has remained remarkably prevalent, with the drugs continuing to be prescribed long after the patient has stabilized. Clearly this was a clinical scenario that called for a thoughtful and considerate analysis of what should be considered a best practice.

These are the first APA recommendations tailored toward a specific set of patients and symptoms. What sort of challenges presented themselves in developing these guidelines?

Our goal was to create a guideline that incorporated a critical review of the latest, best evidence on the topic but that was also informed by a consensus view of current clinical realities and exigencies. This can be difficult, especially in areas where the data are sparse or subject to differing interpretations and where the process requires almost unanimous agreement in the work group for a recommendation to go forward. The tension is between generating a guideline that addresses the clinical issue of concern but that may be more definitive in some statements than some would feel is warranted versus one that is more equivocating and generic in its recommendations but does little to change practice.

Was there any overarching theme among the numerous recommendations made?

Any antipsychotic might be useful for managing dementia-related psychosis or agitation in certain circumstances. The concern in the field at large has been that when people receive a prescription, it’s never changed and it continues on indefinitely, which over time carries increasing risks of morbidity and even mortality. The problem is that many patients are never reassessed. But if you look at the literature, a majority of patients who respond continue to do quite well once the drugs are discontinued.

What is the optimal length of time a patient should take antipsychotics for dementia-related psychosis or agitation? This was an issue of disagreement within the committee, with most data suggesting somewhere in the three- to six-month range. However, we were able to come to a consensus statement that captures the intent that patients given antipsychotics should be reassessed sooner rather than later.

Did you encounter any surprises as you went through the process of reviewing the large amount of literature?

The harms of antipsychotic use in people with dementia are backed by very consistent data, so there is no question that these drugs can have significant impairments in quality of life and cognition if used on a chronic basis and, in some individuals, carry a risk of sudden death, even in short-term usage.

I did find it unusual, given how widely antipsychotics are used, that there was not a better database of drug-to-drug comparisons that might have allowed for more specific recommendations.

The committee recommended against the use of haloperidol (a first-generation antipsychotic) as a first-line agent. But there were no second-generation antipsychotics that stood out as preferred first-line agents, correct?

This was a continuing point of discussion among the members of the committee. The problem is that most of the data we reviewed came from studies funded from pharmaceutical companies looking for drug approval, and often the comparator drug is one that they strongly think will do worse, like a first-generation antipsychotic. So we do not have a lot of head-to-head research comparing the atypical antipsychotics.

Take quetiapine as an example; what we found was that it is frequently used at a lower dose equivalent compared with other second-generation antipsychotics, so while the results seemed more favorable, they may not be directly comparable.

What do you hope clinicians will take away from the practice guidelines?

Given the very real harm statistics, I hope that we will see some significant alteration in practice within the next couple of years. Physicians have been reluctant to make changes in the past, and one of the reasons may be that many care settings don’t have the resources needed to offer the behavioral interventions we know can be often effective in treating these patients; antipsychotics are used as a quick and easy alternative, but that excess usage can create unacceptable risks.

Unfortunately, at this point there are not better pharmacologic alternatives than antipsychotics for some individuals, particularly in treating emergent agitation, which can be very harmful in its own right, so these medications still need to be a consideration. However, antipsychotics need to be handled in a more controlled and rigorous fashion than is currently the case. ■