EEG-based Neurofeedback for Auditory Verbal Hallucinations (HALFEED)

This study's primary objective is to perform a randomized controlled pilot study to assess the feasibility of using EEG-based neurofeedback to reduce the severity of treatment-resistant auditory verbal hallucinations ('hearing voices') in patients diagnosed with schizophrenia. Patients will be randomized to receive either EEG-based neurofeedback or treatment-as-usual.

Auditory verbal hallucinations (AVH) are experienced by up to 80% of patients diagnosed with schizophrenia, where they can cause significant occupational and social impairment. Current treatments are incompletely effective. Around 25-30% of AVH are refractory to antipsychotic drugs, and cognitive behavioural therapy only shows a small-medium effect size. Initially promising studies of neurostimulation have shown smaller effect sizes as better controlled trials have been conducted. There is hence the need for innovative new treatments. One potential option is neurofeedback training. The primary objective of study is to perform a randomized, controlled, rater-blinded pilot trial (n=40) of EEG neurofeedback for AVH in patients with treatment-resistant schizophrenia, to assess trial process, which will then inform a future definitive trial. The secondary objective is to calculate a 95% confidence interval that will allow interpretation of statistical difference between neurofeedback and treatment-as-usual groups to assess neurofeedback for reducing auditory verbal hallucinations. Participants will be randomly allocated to either a neurofeedback (plus treatment-as-usual) or treatment-as-usual alone condition. Neurofeedback will employ Z-score based LORETA (Low Resolution Brain Electromagnetic Tomography). After a baseline assessment, twenty sessions of personalized neurofeedback training will be delivered over a period of approximately four months. This is the first registered trial of EEG neurofeedback for hallucinations. The primary focus of the pilot trial is on feasibility. However, a 95% confidence interval will be determined for the difference on PSYRATS-AH and AHRS scores between neurofeedback and treatment-as-usual to help inform a future definitive trial.