OBJECTIVES: To evaluate and compare healthcare resource utilization (HCRU) and costs between bipolar I disorder (BD-I) patients diagnosed with and without comorbid substance abuse (SA).

METHODS: Using the Truven Health Analytics MarketScan® Medicaid, Commercial, and Medicare Supplemental databases (7/1/15-6/30/16-Medicaid, 7/1/15-3/31/16-Commercial and Medicare Supplemental), adult patients ≥18 years with BD-I were identified if they initiated an atypical oral antipsychotic treatment and had ≥6-month continuous enrollment before (baseline) and after (follow-up) the first day of treatment (the index date). Presence of comorbid SA was determined based upon having ≥1 claim with a relevant ICD-9/10 or procedure code in baseline. Multivariable regression models were conducted to estimate adjusted utilization and costs, controlling for demographic and clinical characteristics, insurance type, baseline medication, and baseline hospitalization.

RESULTS: Of 18,388 identified patients with BD-I who initiated atypical antipsychotics, 4,307 (23.4%) had comorbid SA; the remaining 14,081 (76.6%) were without SA. At baseline, patients with SA were younger [mean (SD) 38.0 (13.3) years vs. 41.0 (14.8) years], had a higher general disease burden measured by mean (SD) Charlson Comorbidity Index [0.9 (1.6) vs. 0.7 (1.4)], and higher hospitalization rate (55.1% vs. 19.4%) (p<0.001) than those without SA. Controlling for baseline differences, patients with SA had statistically significantly higher adjusted all-cause and psychiatric-specific hospitalization rates [17.8% vs. 13.0% (all-cause); 10.6% vs. 7.0% (psychiatric)] and medical costs [$11,142 vs. $9,332 (all-cause); $5,369 vs. $3,271 (psychiatric)] during the follow-up period (p<0.001).

CONCLUSIONS: BD-I patients with comorbid SA had higher all-cause and psychiatric-specific hospitalization rates and costs. Efforts to address comorbid SA in BD-I patients may help reduce HCRU and costs.