Off-Label Prescribing of Second-Generation Antipsychotics to Elderly Veterans with Posttraumatic Stress Disorder and Dementia

Objectives: To determine whether elderly veterans with posttraumatic stress disorder (PTSD) and dementia are more likely to be prescribed second-generation antipsychotics (SGAs) than those with PTSD alone.

Design: National serial cross-sectional study.

Setting: Veterans Health Affairs inpatient and outpatient settings.

Participants: Veterans aged 65 and older with PTSD (excluding schizophrenia or bipolar disorder) with or without concomitant dementia who received care from the Veterans Health Administration between 2003 and 2010 were identified using International Classification of Diseases, Ninth Revision, codes (N = 93,068; 11.1% with dementia).

Measurements: Trends in SGA prescribing and odds of being prescribed an SGA were determined using a multivariable logistic regression model adjusted for clinical, sociodemographic, and geographic covariates.

Results: Between 2004 and 2009, SGA prescribing declined annually from 7.0% to 5.1% of elderly veterans with PTSD without dementia and 13.2% to 8.9% in those with dementia; findings over time consistently indicated that veterans with PTSD and dementia had at least twice the odds of being prescribed an SGA as those without PTSD (odds ratios 2.03 (95% confidence interval (CI) = 1.82-2.26) to 2.33 (95% CI = 2.10-2.58).

Conclusion: Although the prescribing of SGAs to elderly veterans with PTSD has decreased, prescribing an SGA to those with dementia remained consistently higher than for those with PTSD alone and is problematic given the high prevalence of medical comorbidities in this aging population coupled with the lack of compelling evidence for effectiveness of SGAs in individuals with dementia.