REVIEW OF: Tiihonen J et al, JAMA Psychiatry 2019;76(5):499–507 and Stroup T et al, JAMA Psychiatry 2019;76(5):508–515

TYPE OF STUDY: Retrospective non-randomized controlled trials

Antipsychotic polypharmacy is discouraged in guidelines but common in practice. Up to 30% of patients with schizophrenia are prescribed multiple antipsychotics, and combinations of antipsychotics with other drug classes are even more common. Research on these practices is sparse. Two recent studies, both large retrospective non-randomized controlled trials, attempted to clarify whether polypharmacy brings greater benefits in schizophrenia, or just greater risks.

The first study collected data from a population-wide registry in Finland on 62,250 patients with schizophrenia who were hospitalized and followed between 1996 and 2015 (median age 46; male to female ratio equal).

Hazard ratios were calculated by comparing patients on one, multiple, or no antipsychotics. Within-individual analysis was used to eliminate selection bias (ie, patients were their own controls). Of the total cohort, 67% used antipsychotic polypharmacy at some point. To exclude switches between antipsychotics, data from the first 90 days of multiple antipsychotic use were censored. The primary outcome was psychiatric rehospitalization, and secondary outcomes were mortality and medical hospitalization.

The risk of psychiatric rehospitalization was 13% lower with polypharmacy than monotherapy (HR 0.87; CI 0.85–0.88). That risk was lowest with the combination of clozapine and aripiprazole: 58% lower than no antipsychotic use (HR 0.42; CI 0.39–0.46) and 14% lower than clozapine alone (HR 0.86; CI 0.79–0.94). Among the top 10 treatments with the lowest risk of rehospitalization, only one was monotherapy: clozapine. Remarkably, polypharmacy was also associated with a lower risk of hospitalization due to medical illness and mortality.

The second study evaluated the effects of adding different drug classes to standard treatment in schizophrenia. Using a Medicaid registry, 81,921 patients with schizophrenia on antipsychotic therapy were followed for one year after starting an additional psychotropic (mean age 41; 54% male). Patients who were already on multiple psychiatric medications or who filled their antipsychotic inconsistently were excluded from the sample (n = 241 and 579).

Hazard ratios were calculated by comparing patients based on whether they were prescribed antidepressants, benzodiazepines, or mood stabilizers vs additional antipsychotics. Patients in each of the treatment groups were demographically similar. Those who did not start a new psychotropic were not included in the comparisons, as it was thought they represented a group with fewer comorbidities and better prognosis. Dropouts were handled by analyzing data on an intent-to-treat basis. The primary outcome was psychiatric hospitalization, and secondary outcomes included medical hospitalization and mortality.

The risk of psychiatric hospitalization was 16% lower for patients who started an antidepressant (HR 0.84; CI 0.80–0.88). Patients started on benzodiazepines had a higher risk of psychiatric hospitalization (HR 108; CI 1.02–1.15), while those started on mood stabilizers had an equal risk (HR 0.98; CI 0.94–1.03). Antidepressants were associated with a lower risk of medical hospitalization (HR 0.87; CI 0.79–0.96), whereas no difference was found for benzodiazepines or mood stabilizers. Mood stabilizers were the only group associated with a statistically higher risk of mortality (HR 1.31; CI 1.04–1.66), and this risk was highest with gabapentin.

Both studies had similar weaknesses. With the lack of randomization, various confounding variables could have been overlooked. Factors not examined include reasons for changing medications, frequency of patient-provider contact, use of psychosocial interventions, and extent of medication adherence. Functioning and symptom severity were also not examined. On the other hand, patients prescribed multiple psychotropics are likely to have lower functioning and greater disease severity, so the fact these patients had favorable outcomes is impressive.

TCPR’S TAKE
Polypharmacy is often looked down on, but these results suggest it may be a viable strategy in schizophrenia. In combining antipsychotics, the best outcome was with clozapine and aripiprazole. This suggests prescribing antipsychotics with different receptor profiles may be a useful tactic. In terms of combining antipsychotics with other psychotropics, the results are even less definitive and more likely skewed. That limitation aside, antidepressants appear to have the greatest benefit and least risk. In contrast, mood stabilizers and benzodiazepines should be used with more caution.