Pharmacological treatment patterns for bipolar disorder have changed during recent years, but for better or worse?

To investigate the comparative real-world effectiveness of antipsychotics and mood stabilisers in bipolar disorder.

Register-based cohort study including all Finnish residents aged 16–65 with a diagnosis of bipolar disorder from in-patient care, specialised out-patient care, sickness absence and disability pensions registers between 1996 and 2018, with a mean follow-up of 9.3 years (s.d. = 6.4). Antipsychotic and mood stabiliser use was modelled using the PRE2DUP method and risk for hospital admission for psychiatric and non-psychiatric reasons when using versus not using medications was estimated using within-individual Cox models.

Among 60 045 individuals (56.4% female; mean age 41.7 years, s.d. = 15.8), the five medications associated with lowest risk of psychiatric admissions were olanzapine long-acting injection (LAI) (aHR = 0.54, 95% CI 0.37–0.80), haloperidol LAI (aHR = 0.62, 0.47–0.81), zuclopenthixol LAI (aHR = 0.66, 95% CI 0.52–0.85), lithium (aHR = 0.74, 95% CI 0.71–0.76) and clozapine (aHR = 0.75, 95% CI 0.64–0.87). Only ziprasidone (aHR = 1.26, 95% CI 1.07–1.49) was associated with a statistically higher risk. For non-psychiatric (somatic) admissions, only lithium (aHR = 0.77, 95% CI 0.74–0.81) and carbamazepine (aHR = 0.91, 95% CI 0.85–0.97) were associated with significantly reduced risk, whereas pregabalin, gabapentin and several oral antipsychotics, including quetiapine, were associated with an increased risk. Results for a subcohort of first-episode patients (26 395 individuals, 54.9% female; mean age 38.2 years, s.d. = 13.0) were in line with those of the total cohort.

Lithium and certain LAI antipsychotics were associated with lowest risks of psychiatric admission. Lithium was the only treatment associated with decreased risk of both psychiatric and somatic admissions.