This study examined monitoring for metabolic side effects among older outpatients with dementia starting a new antipsychotic.
In a retrospective cohort analysis of U.S. Department of Veterans Affairs data, monitoring, as recommended by the American Diabetes Association and the American Psychiatric Association, was examined between October 1, 2005, and September 30, 2011. The sample included outpatients aged ≥60 years with dementia but without a psychotic disorder (N=3,903) and outpatients with a psychotic disorder but without dementia (N=5,779) who were prescribed a new antipsychotic. Because dementia patients differed from psychosis patients in all observed patient characteristics, especially age, metabolic monitoring of dementia patients was compared with a propensity score–matched sample of outpatients with psychosis (1,576 matched pairs).
At baseline (±30 days from the index prescription), 68% of the matched dementia patients were weighed, compared with 63.7% of the matched psychosis patients (odds ratio [OR]=1.28, 95% confidence interval [CI]=1.03–1.48). Monitoring for glucose or glycosylated hemoglobin (HBA1c) and low-density lipoprotein (LDL) was not significantly different between the groups: glucose or HBA1c, 41% versus 44%; LDL, 24% versus 27%. At three months (±30 days), metabolic monitoring for all three parameters was significantly lower for the dementia group: weight, OR=.86, CI=.75–.99; glucose or HBA1c, OR=.83, CI=.71–.97; and LDL, OR=.69, CI=.57–.85.
Monitoring rates for metabolic side effects were low for both dementia and psychosis groups, with lower rates for dementia patients at follow-up compared with matched psychosis patients. Quality improvement efforts are needed to improve monitoring, especially for patients with dementia.