Although clinical trials evaluate pharmacotherapeutic interventions under highly controlled conditions, there remains a need to evaluate medication use in actual practice.
Patients prescribed atypical antipsychotic medications in the VA system during a 4-month period in 1999 (n = 73,981) were classified into 32 groups on the basis of clinical diagnosis and recent level of inpatient use. Variation was e-amined across groups in drug costs, agents, dosages, and duration of use. The potential impact of these medications on VA costs was estimated by calculating medication costs and subtracting estimated inpatient savings.
A majority of patients were diagnosed with schizophrenia (57.2%), but substantial off-label use of these medications to treat other psychiatric illnesses was also evident (42.8%). Compared with published trials reporting average annual costs from $3,000 to $7,000, average annualized pharmacy costs were only $1,395 per patient because of a 58.5% VA price discount; relatively low dosing, especially for people with diagnoses other than schizophrenia; and medication prescription coverage for only 75% of the days in the study period. The sample averaged only 6.96 inpatient days; as a result, potential inpatient savings were limited. Assuming 0% to 18% inpatient savings, annual net drug costs are estimated to range from $500 to $1,152 per patient.
Medication costs in actual practice can be substantially lower than in clinical trials. Atypical antipsychotic medications in actual VA practice incur net costs estimated at $500 to $1,152 per patient per year with substantial variation across clinical subgroups.