Antipsychotic polypharmacy across psychiatric disorders: a nationwide cohort study

Background

Antipsychotic polypharmacy (APP) is widely used despite concerns regarding adverse effects and limited evidence for additional efficacy. While most evidence derives from schizophrenia, antipsychotics are prescribed across a broad range of psychiatric conditions. A transdiagnostic characterization of APP use is needed to identify populations with high cumulative exposure and to differentiate evidence-based augmentation from potentially avoidable prescribing patterns.

Methods

We conducted a nationwide register-based cohort study including all individuals in Denmark redeeming ≥1 prescription for an antipsychotic (ATC N05A) between 1997 and 2024. Treatment episodes were constructed using the PRE2DUP method. APP was defined as concurrent use of ≥2 antipsychotics for ≥60 consecutive days. Individuals were assigned to time-varying, hierarchical diagnostic groups (schizophrenia-spectrum disorders, bipolar disorder, intellectual disability, dementia, other psychiatric disorders, and individuals without records of psychiatric diagnoses). Outcomes included (i) one-year APP prevalence by diagnostic group and over time, (ii) proportion of antipsychotic treatment time spent on APP, (iii) duration of APP episodes, and (iv) APP composition, including APP involving sedative antipsychotics, clozapine, and partial dopamine-receptor agonists.

Results

The cohort comprised 747,044 individuals contributing 2,319,768 person-years of antipsychotic exposure. Overall, 176,778 APP episodes were identified, and 11.4% of total antipsychotic treatment time occurred under polypharmacy, ranging from 4.1% in individuals without a recorded psychiatric diagnosis to 20.1% in schizophrenia-spectrum disorders. In 2024, the one-year prevalence of APP was highest in schizophrenia-spectrum disorders (14.1%), followed by intellectual disability (8.2%), bipolar disorder (7.1%), dementia (3.2%), other psychiatric disorders (2.4%), and individuals without a recorded psychiatric diagnosis (1.0%). Schizophrenia-spectrum disorders accounted for 67% of total APP exposure, yet approximately one-third of all APP exposure occurred outside schizophrenia-spectrum disorders.

Overall, median duration of APP episodes was 7.4 months (interquartile range: 3.4–18.1), with 56.2% of episodes lasting ≥6 months, 35.3% ≥12 months, and 19.2% ≥24 months. Long-term APP (≥24 months) was most common in intellectual disability (26.0%) and schizophrenia-spectrum disorders (20.9%), and least common in dementia (10.1%).

APP combinations were dominated by quetiapine-containing regimens across all diagnostic groups. The three most frequent combinations during 2020–2024 were aripiprazole–quetiapine (14.3% of all APP episodes), risperidone–quetiapine (10.0%), and olanzapine–quetiapine (9.5%), together accounting for approximately one-third of all APP episodes. Sedative antipsychotics were involved in 53.6–73.8% of APP episodes across diagnostic groups. Clozapine was involved in 15.5% of APP episodes in schizophrenia-spectrum disorders, compared with 1.6% in bipolar disorder and fewer than 3% in all other diagnostic groups. Partial dopamine-receptor agonists were involved in 24.7% and 25.6% of APP episodes in schizophrenia-spectrum disorders and bipolar disorder, respectively, compared with 7.1–17.8% in the remaining groups.

Conclusions

APP was common, frequently long-term, and observed across all psychiatric diagnoses, with approximately one-third of all exposure occurring outside schizophrenia-spectrum disorders. Combinations involving sedative antipsychotics were predominant across all diagnostic groups, suggesting that APP often reflects pragmatic symptom management rather than evidence-based augmentation. These findings underscore the need for coordinated, transdiagnostic deprescribing initiatives targeting sedative antipsychotic add-ons across psychiatric settings.

References

No references