Nursing homes administer largely pointless and potentially harmful drugs to a majority of residents with advanced dementia, according to findings in JAMA Internal Medicine.

Out of more than 5,400 residents under consideration, about 54% received a “medication with questionable benefit,” the investigators determined. Alzheimer’s disease drugs such as donepezil (Aricept) and memantine (Namenda) were the most commonly administered. There is little evidence that they improve cognitive functioning for people with advanced stages of dementia, and potentially put residents at risk for falls or urinary tract infections.

About 20% of the residents were on a lipid-lowering agent. These also are associated with a host of troublesome potential side effects, including confusion and muscle fatigue, and may do little to extend the life of these residents.

The costs associated with these medications is substantial, the study authors found. The mean 90-day expenditure for these drugs was $816. This represents more than a third of the medication expenditures for residents with advanced dementia.

“While it can be difficult for family decision-makers to discontinue medications that treat the chronic diseases of their loved ones as they transition toward comfort care, minimizing questionably beneficial interventions is an important therapeutic option consistent with recommendations by the Institute of Medicine about care quality at the end of life,” the authors wrote.

Hospice patients with a do-not-resuscitate order were less likely to be on a questionable medication, they determined. They did not find any “facility-level association” in terms of having a dementia special care unit, more beds, or a higher percentage of residents with a DNR. The findings should prompt physicians to reconsider their prescribing practices for late-stage dementia patients, the Regenstrief Institute’s Greg A. Sachs, M.D., wrote in an accompanying JAMA editorial.

The findings are based on an analysis of a nationwide long-term care pharmacy database linked to the Minimum Data Set. The numbers came from 2009 and 2010. The authors are affiliated with a variety of institutions, including the University of Massachusetts Medical School in Worcester and the Wistar Institute at the University of Pennsylvania. Findings were posted online Monday ahead of print.