Warning Letters Safely Lowered Antipsychotic Prescribing in Dementia

Overprescribing warning letters modestly cut the use of antipsychotic medications without apparent repercussion in people with dementia, a secondary analysis of a randomized trial indicated.

Compared with primary care physicians (PCPs) who received placebo letters, those who were sent three overprescribing warning letters significantly reduced quetiapine (Seroquel) use in their community-dwelling Medicare patients with dementia over a 90-day period (adjusted difference -1.5 days, 95% CI -1.8 to -1.1), Adam Sacarny, PhD, of the Mailman School of Public Health at Columbia University in New York City, and colleagues reported.

The series of warning letters also significantly reduced quetiapine use among nursing home patients with dementia as well, by an average of 0.7 days over that period (95% CI -1.3 to -0.1), the group wrote in JAMA Network Open.

But as quetiapine prescriptions were pared back, agitated or reactive behaviors -- often the main driver behind prescribing antipsychotics to people with dementia -- remained steady. On top of that, there was no change in the use of restraints. This outcome, however, was measured only in nursing home patients.

"About a million older adults with dementia are prescribed antipsychotics every year, but these drugs can cause a number of harms for these patients," Sacarny told MedPage Today. "There is a lot of interest in interventions that could safely reduce use of antipsychotics in dementia care."

Off-label use of antipsychotics in dementia is common, the researchers noted. A 2009 report estimated that 75% of quetiapine prescribing may lack a basis in clinical evidence as the drug is indicated only for schizophrenia, acute manic episodes, and adjunctive treatment for major depressive disorder. Its label carries a boxed warning about increased mortality in elderly patients with dementia-related psychosis.

An estimated one in seven nursing home residents receive an antipsychotic drug every quarter, Sacarny and co-authors pointed out. Moreover, a recent cohort study linked antipsychotic use in dementia with a wider range of adverse outcomes than previously thought, including pneumonia, acute kidney injury, venous thromboembolism, stroke, fracture, myocardial infarction, and heart failure.

The goal of the original trial was to test whether overprescribing warning letters from Medicare could reduce antipsychotic prescribing and improve quality of care in older and disabled adults. The secondary analysis focused on patients with dementia.

The researchers used Medicare fee-for-service claims, Part D prescribing, Minimum Data Set nursing home assessment, and Medicare enrollment data from 2014 to 2018. In the original trial, 5,055 prescribers were included: 2,528 randomized to the control group and 2,527 to the warning letter group. This included the highest-volume PCP prescribers of quetiapine in original Medicare, accounting for approximately 5% of all primary care prescribers of quetiapine.

Of the prescribers included, 1,885 had nursing home patients, 4,584 had community-dwelling patients, and some had patients in both settings. This equated to 84,881 patients living in nursing homes and 261,288 community-dwelling patients. Average baseline age was around 82 and the majority of dementia patients in both settings were female and white.

PCPs who received the series of three letters were told their quetiapine prescribing was under review by the Centers for Medicare & Medicaid Services (CMS) and was "extremely high" relative to within-state peers. The letter added that high quetiapine prescribing could be appropriate but was "concerning" for medically unjustified use. It encouraged physicians to review their prescribing patterns and said they could expect to receive future communications from CMS.

After the intervention period, the average days of quetiapine receipt for community-dwelling adults was 9.9 (from 10.4 days at baseline) in the control group and 8.2 (from 10.2 days) in the treatment group. This was higher among nursing home patients, with 10.3 (from 12 days) and 10.1 days (from 12.7 days) of quetiapine receipt for the control and treatment groups, respectively.

At baseline, nursing home patients were on a higher average dose of quetiapine (1,356 mg vs 1,225 mg for controls) than community-dwelling adults (1,161 mg vs 1,106 mg for controls).

At baseline, 1.5% of nursing home dementia patients and 1.9% of the community-dwelling sample were prescribed quetiapine. More than 3% of both groups were prescribed any antipsychotic, but the study didn't look at those prescribing patterns. Another limitation was that the study measured prescription drug fills only, not whether quetiapine was administered.

No adverse effects were detected during the 3-month period and the intervention was deemed safe. Patients of PCPs who were sent warning letters didn't have any change in metabolic indicators like weight, glucose, and lipid profiles. There was a very small but significant reduction in the risk of death for community-dwelling dementia patients (adjusted difference -0.1%, 95% CI -0.3 to 0.0, P=0.04).

Though there were "encouraging signs around improved mental health" marked by modest reductions, these outcomes didn't reach statistical significance, said Sacarny.

Comment