Authors' reply

Gupta & Kamboj correctly note that akathisia is a clinically relevant adverse effect with oral aripiprazole because it causes distress and is associated with an increased risk of agitation and suicide in patients with schizophrenia. We did not want to discount a higher rate of akathisia with aripiprazole 400 mg once monthly as being of ‘unknown clinical relevance’, but rather questioned the clinical relevance of the absolute 0.11-point group difference on the 5-point Barnes Akathisia Global Scale. We appreciate that this could have been stated more clearly. In our study, ¹ 10.6% of patients treated with aripiprazole 400 mg once monthly reported akathisia as a treatment-emergent adverse event (TEAE), as did 6.8% of patients treated with oral aripiprazole and 8.4% of patients treated with a sub-therapeutic dose of aripiprazole once monthly; no patients discontinued because of akathisia. Rates of agitation, reported as a TEAE, were low among all treatment groups (aripiprazole 400 mg: 2.6%; oral aripiprazole: 0.8%; aripiprazole 50 mg: 0%). As noted in our manuscript, Clinical Global Impression Severity of Suicide (CGI-SS) scores and Columbia Suicide Severity Rating Scale (C-SSRS) suicidal ideation intensity total scores remained stable across treatment groups (see Table 4 in the published article ¹ ).

Gupta & Kamboj note that the rate of TEAEs with aripiprazole 400 mg once monthly may not be similar to the rate with a sub-therapeutic dose of aripiprazole once monthly if psychotic disorder and schizophrenia are removed from the list of TEAEs. They also suggested that psychotic disorder and schizophrenia are not TEAEs and are efficacy outcomes. In this context, we note that the regulatory authorities in Europe and the USA require accurate recording of all patient-reported TEAEs, and psychotic disorder, psychotic symptoms, and schizophrenia are Medical Dictionary for Regulatory Activities (MedDRA)-defined adverse events and typical patient-reported TEAEs in antipsychotic trials.

Gupta & Kamboj also state that reporting efficacy outcomes using the LOCF method for an ITT analysis could be misleading because discontinuation rates differed between the two active treatment groups. Gupta & Kamboj recommend a comparison of ITT and per protocol results for a more informative assessment of efficacy outcomes in our 38-week trial. LOCF was not used for the primary efficacy outcome or for the additional efficacy outcomes. As reported in our paper, the Kaplan-Meier estimated impending relapse rate at week 26 (primary study endpoint, ITT population: aripiprazole 400 mg: 7.12%; oral aripiprazole: 7.76%; aripiprazole 50 mg: 21.80%) was similar to week 26 in the per protocol results (i.e. for observed impending relapse rates, aripiprazole 400 mg: 6.79%; oral aripiprazole: 7.14%; aripiprazole 50 mg: 18.32%).

Lastly, we trust that readers of the BJPsych do not base their treatment decisions on ‘superficial reading’.