Rexulti (brexpiprazole) was recently also studied for agitation associated with Alzheimer’s dementia, which is different from its use in depression or schizophrenia. It is not approved for memory or cognition — its role is specifically for managing agitation. Because of its increased risk of death, guidelines recommend Rexulti only when agitation is severe, distressing, and non-drug approaches have not worked. Here’s a clear overview of the non-drug (behavioral and environmental) strategies that are recommended before starting medications like Rexulti:
Step 1: Identify Triggers (Always Start Here)
Step 2: Try Non-Drug Strategies First
Step 3: Evaluate Response
Step 4: Consider Medication (Last Resort)
Step 5: Monitor & Reassess
Rexulti was approved in 2023 for the treatment of agitation in Alzheimer’s dementia. Clinical trials showed that patients receiving Rexulti had significantly greater reductions in agitation scores (measured by the Cohen-Mansfield Agitation Inventory).
About 1 in 7–8 patients taking Rexulti show meaningful improvement in agitation compared with placebo.
This risk is thought to be similar across all atypical antipsychotics.
When considering atypical antipsychotics (like risperidone, olanzapine, quetiapine, aripiprazole) for psychosis or behavioral symptoms in dementia, the key concern is increased risk of harm.
Here’s what the data show on NNH (Number Needed to Harm):
Large meta-analyses (e.g., Schneider et al., NEJM 2005; FDA pooled data) found about a 1.6–1.7-fold increased risk of death in elderly dementia patients taking atypical antipsychotics compared with placebo.
Absolute risk increase ≈ 3–4% over 10–12 weeks.
That translates to a NNH ≈ 25–30 → meaning for every 25–30 dementia patients treated with an atypical antipsychotic for ~10–12 weeks, 1 extra death occurs compared to placebo.
Risk of cerebrovascular adverse events (stroke, TIA) is also increased.
Absolute risk increase ≈ 1–2%.
NNH ≈ 50–100 over ~10 weeks.
Effect size is modest: about 18% improve on drug vs 10% on placebo (Cochrane, 2021).
So NNT ≈ 12 for symptomatic benefit.
NNT for benefit: ~12 (some reduction in aggression/psychosis).
NNH for death: ~25–30.
NNH for stroke: ~50–100.
This is why guidelines recommend only using antipsychotics in dementia when symptoms are severe, dangerous, or distressing, and always for the shortest possible time after non-drug approaches fail.