The treatment of unipolar major depression with psychotic features using combination therapy versus monotherapy : a study of adherence, persistence, health care utilization and expenditures, and medication-related adverse events

The purpose of the study was to assess medication adherence, medication persistence, suicide ideation/attempts, health care utilization and costs, and medication adverse events in Texas Medicaid patients with psychotic depression who utilized antidepressant monotherapy (AD cohort) or antidepressant plus second-generation antipsychotic therapy (AD/SGA cohort). Using prescription and medical records from September 2007 to December 2012, beneficiaries were included if they were aged 18-63 years, had no confounding psychiatric disorders, had a 6-month pre-index period with no antidepressants plus a 12-month post-index period, and had a diagnosis for unipolar major depressive disorder with psychotic features (ICD-9-CM 296.24 or 296.34). The index date was the first claim date for an antidepressant. All participants had at least two antidepressant claims, and those in the AD/SGA cohort also had at least two SGA claims. Study covariates included: age, race/ethnicity, gender, residence, Charlson Comorbidity Index score, tobacco use and/or dependence, and antidepressant persistence. A total of 926 participants met study criteria (AD cohort n=510; AD/SGA n=416). Overall, the mean age (±SD), Charlson Comorbidity Index score, and antidepressant persistence rate was 40.5 (±13.2) years, 0.6 (±1.3), and 172.3 (±130.4) days, respectively. The final sample included 66.8% females, 25.2% Caucasians, 34.9% African Americans, 36.7% Hispanics, 79.5% urban dwellers, and 19.7% with known tobacco use/dependence. The AD/SGA cohort had a 53% significantly higher likelihood of being adherent to antidepressant therapy, compared to the AD cohort (p=0.006). Similarly, the AD/SGA cohort had a 23% significantly lower hazard of antidepressant nonpersistence (based on persistence with a 45-day gap) (p=0.001). Alternatively, the AD cohort had a significantly lower rate of psychotic depression-related outpatient/emergency department visits (p<0.001), as well as significantly lower psychotic depression-related costs (medication, medical, and total) and all-cause medication costs (p<0.001). There were no differences in suicide ideation/attempts or rates of incident dyslipidemia or diabetes mellitus between cohorts. Evidence of incident extrapyramidal symptoms were rare (n=12). In conclusion, the AD/SGA cohort had better outcomes associated with antidepressant adherence and persistence, and the AD cohort had lower rates of health care utilization and costs. These real-world estimates should help increase the understanding of appropriate treatment for psychotic depression.