Mood stabilizer adherence patterns among adults with bipolar disorder: A single-center retrospective analysis

• Adults with bipolar disorder may exhibit poor adherence to lithium or valproate therapy following a psychiatric hospitalization. • Poor adherence emerged immediately after hospitalization and worsened with time. • Male sex and comorbid substance use disorder may contribute to poor mood stabilizer adherence. • Additional research needed to evaluate outpatient psychiatric interventions. Bipolar disorder (BD) is a severe psychiatric disorder frequently associated with psychotropic non-adherence. Few studies have quantified real-world adherence patterns among people prescribed lithium or valproate for BD. This single-center retrospective study examined adherence to lithium and valproate among adults hospitalized for BD between January 2017 and January 2022. The primary outcomes were the differences in the 6-month average proportion of days covered (PDC), for months 1 – 6 and 7 – 12 post-discharge, between the lithium and valproate groups. Secondary outcomes included the 12-month psychiatric readmission rate and demographic/comorbid associations with PDC. PDC was calculated by querying SureScripts fill records and confirming dispensing dates with pharmacies. Of 75 eligible adults, 56.0 % were discharged on lithium and 44.0 % on valproate. People discharged on lithium exhibited higher PDC during months 1 – 6 (0.36 versus 0.25 in the valproate group) and 7 – 12 (0.27 versus 0.16); these differences lacked statistical significance. Significant difference in ≥ 1 psychiatric re-admission during months 1 – 12 was observed between people discharged on lithium (9.5 %) and those discharged on valproate (30.3 %) (p = 0.035). Male sex and comorbid substance use disorder were associated with significantly lower PDC. People discharged on lithium or valproate are vulnerable to mood stabilizer non-adherence following psychiatric hospitalization. Future research should examine psychotropic adherence trajectories in larger subpopulations of people with BD. Data derived from medication fill records may overestimate adherence in people who miss doses after dispensing and underestimate adherence in people who switch mood stabilizer therapy under a clinician’s direction.