Economic burden of treatment-resistant depression in privately insured US patients with co-occurring anxiety disorder and/or substance use disorder.

Objective: To assess the burden of treatment-resistant depression (TRD) among privately insured patients with anxiety disorder and/or substance use disorders (SUD).

Methods: Adults <65 years old were identified in the Optum Health Care Solutions Inc. database (July 2009-March 2017). Among those with major depressive disorder (MDD) and antidepressant use, patients who initiated a third antidepressant (index date) after two regimens at adequate dose and duration were classified in the TRD cohort and patients without evidence of TRD were classified in the non-TRD MDD control cohort. The non-MDD control cohort comprised patients without MDD. In the non-TRD MDD and non-MDD cohorts, the index date was imputed to mimic the distribution of time in the TRD cohort from the first antidepressant to the index date or from the start of eligibility to the index date, respectively. Patients with <6 months of continuous insurance eligibility pre-/post-index, psychosis, schizophrenia, bipolar disorder and related conditions, dementia, and development disorders, and/or no baseline anxiety disorder and/or SUD were excluded. Patients with TRD were matched 1:1 to patients with non-TRD MDD and patients without MDD, based on exact matching factors (i.e. availability of work loss data) and propensity scores computed based on characteristics measured pre-index. Outcomes, including healthcare resource use (HRU) and costs, work productivity loss and related costs measured per patient per year ≤24 months post-index were compared between matched TRD, non-TRD MDD and non-MDD cohorts.

Results: A total of 3166 patients were identified in the TRD cohort and matched to non-TRD MDD and non-MDD cohorts. Among patients with TRD (mean age 39 years, 60.5% female), 87.3% had an anxiety disorder, 24.1% had SUD. The TRD cohort had higher HRU vs non-TRD MDD and non-MDD cohorts: 0.32 vs 0.20 and 0.14 inpatient admissions, 0.91 vs 0.73 and 0.58 emergency department visits, and 23.8 vs 16.8 and 11.6 outpatient visits, respectively (all  < .01). The TRD cohort had higher healthcare costs ($16,674) vs non-TRD MDD ($10,945) and non-MDD ($6493) cohorts (all  < .01). Among patients with work loss data ( = 310/cohort), patients with TRD had more work loss days (54) and higher work loss-related costs ($13,674) vs patients with non-TRD MDD (32 days; $7131) and without MDD (17 days; $4798; all  < .01).

Conclusions: In patients with an anxiety disorder and/or SUD, TRD was associated with higher HRU, healthcare costs, work loss days and work loss-related costs.