BACKGROUND:
Although increasing numbers of war veterans are seeking treatment for posttraumatic stress disorder (PTSD) at the U.S. Department of Veterans Affairs (VA), information on the role of psychotropic pharmacotherapy in their treatment has not been available.
METHOD:
Records of psychotropic prescriptions for all VA patients diagnosed with ICD-9 PTSD (N = 274,297) in fiscal year 2004 (October 1, 2003, to September 30, 2004) were examined. Descriptive statistics and multivariable logistic regression were used to identify veteran characteristics and measures of service use that were associated with receipt of any psychotropic medication and, among users of such medications, with use of each of 3 medication classes: antidepressants, anxiolytics/sedative-hypnotics, and antipsychotics.
RESULTS:
Most veterans diagnosed with PTSD received psychotropic medication (80%), and among these, 89% were prescribed antidepressants, 61% anxiolytics/sedative-hypnotics, and 34% antipsychotics. Greater likelihood of medication use was associated with greater mental health service use and comorbid psychiatric disorders. Among comorbidities, medication-appropriate comorbid diagnoses were the most robust predictors of use of each of the 3 medication subclasses, i.e., depressive disorders were associated with antidepressant use, anxiety disorders with anxiolytic/sedative-hypnotic use, and psychotic disorders with antipsychotic use. Use of anxiolytics/sedative-hypnotics and antipsychotics in the absence of a clearly indicated diagnosis was substantial.
CONCLUSIONS:
Diverse psychotropic medication classes are extensively used in the treatment of PTSD in the VA. While disease-specific use for both PTSD and comorbid disorders is common, substantial use seems to be unrelated to diagnosis and thus is likely to be targeted at specific symptoms (e.g., insomnia, anxiety, nightmares, and flashbacks) rather than diagnosed illnesses. A new type of efficacy research may be needed to determine symptom responses to psychotropic medications as well as disorder responses, perhaps across diagnoses.