Antipsychotic selection in schizophrenia remains an important, high-stakes decision, yet clinical guidelines often offer limited guidance on relative efficacy. A recent large, head-to-head randomized trial directly compared seven antipsychotics in over 3,000 hospitalized adults aged 18–45 experiencing acute exacerbation. For six weeks, patients were treated with one of seven medications: olanzapine, risperidone, quetiapine, aripiprazole, ziprasidone, perphenazine, and haloperidol.

After six weeks of monotherapy, olanzapine and risperidone showed the greatest improvement in total Positive and Negative Symptom Scale (PANSS) scores, with response rates of 68% each. Both medications were significantly more effective than quetiapine, aripiprazole, and ziprasidone, but not significantly better than haloperidol or perphenazine. Olanzapine had the lowest all-cause discontinuation rate (20%) and was superior to haloperidol and ziprasidone in retention. However, it carried the highest risk of weight gain (30%), while risperidone had the highest rate of hyperprolactinemia (95%). Aripiprazole stood out for its low rates of sedation (4%) and metabolic adverse effects. Aripiprazole was alsol the most effective at normalizing prolactin levels in patients with baseline elevations.

FGAs performed as expected: haloperidol had the highest risk of extrapyramidal symptoms. Perphenazine had better tolerability, with lower rates of movement disorders and side effect–related discontinuation. All SGAs were more tolerable than haloperidol, with aripiprazole and ziprasidone having the most favorable side effect profiles overall. Across all medications, higher discontinuation due to inefficacy was observed with aripiprazole and ziprasidone.

Overall, this trial offers high-quality, direct comparative data to inform antipsychotic selection in the acute phase. While olanzapine and risperidone may offer the best chance of symptom reduction, their side effect burdens remain a trade-off. Aripiprazole and ziprasidone may be better suited for patients with metabolic risk or sensitivity to sedation, albeit with potentially lower efficacy. These findings support more nuanced, individualized prescribing; prior guidelines have been unable to endorse one med over the other due to limited evidence.

Reference:

Reference: Zhao G et al. Am J Psychiatry. 2025:appiajp20250111. Abstract