Schizophrenia is a serious psychiatric illness which occurs throughout the world and affects one in 100 people, often appearing in adolescence or early adult life. Medication that targets the symptoms of schizophrenia improves the lives of many patients, but all drugs have side effects, such as causing problems with mobility. One in four patients also fails to respond to any of the existing drugs.
These people often experience very distressing “voices” which abuse them, command them to hurt themselves and sometimes to hurt other people. The voices can continue throughout the day making it very difficult to concentrate, to work and to sustain social relationships. The continual harassment leads one in 10 patients with schizophrenia to commit suicide.
Family carers of people with schizophrenia also have to deal with the daily stress and financial burden of this illness. The absence of effective medication has led to attempts to develop forms of talking therapy to help people affected by persecutory voices.
American psychologist Aaron Beck developed what he called cognitive behaviour therapy for people suffering from depression, which has proved very successful in relieving depressive symptoms without the use of medication. During the past 15 years psychologists in the UK have worked on adapting Beck’s approach for people with persecutory voices - known as cognitive behaviour therapy for psychosis (CBTp).
Scientific trials of CBTp have shown that it reduces the patients’ distress, and equips them with coping strategies to deal with the voices, but does not reduce the frequency or volume of the voices. This therapy is given in sessions of an hour over the course of six months to one year, and requires an extensive training for the therapists.
Developing a dialogue with the voices
In thinking about possible solutions for this condition, I found from published papers that most patients when asked about the worst aspect of the voices answered, “the helplessness”. However those who managed to develop a dialogue with the voices felt much more in control. This set me thinking about ways of stimulating a dialogue between the patient and the voice, and it struck me that I could use an avatar to represent the voice - a speaking human image.
Through the avatars - and with help from colleagues with software expertise - I can speak to the patient. The National Institute of Health Research in UK then provided a grant and we worked for seven months to develop the computerised system. The system enables a patient to choose a face that resembles the person they think is talking to them and a voice to match the voice they hear. The voice is chosen from a range of variations on the therapist’s voice, so that the therapist can talk through the avatar in real time.
The effectiveness of the therapy was tested by a randomised controlled trial comparing patients who received the therapy with those who continued with their usual treatment of medication and visits to their psychiatrist. The therapy was planned to be six, 30 minute sessions followed by a repeat assessment of their symptoms one week after the last session. The control group was also reassessed after seven weeks and then offered the therapy.
The avatar process
The patient begins by creating their avatar with assistance from Geoff Williams, an IT specialist. In the first session, the avatar speaks with the usual critical and harsh phrases that the patient normally hears.
In addition to voicing the avatar, the therapist can speak to the patient in their normal voice, urging the patient to oppose the avatar and to tell it to go away and leave the patient alone. Over the next five sessions the avatar progressively becomes less threatening, stops the verbal abuse, and agrees to leave the patient in peace.
It also begins to be friendly to the patient and offer advice. All sessions are recorded and the recordings are transferred to an MP3 which is given to the patients to keep and to use to help them counteract the voices.
To my surprise three of the patients stopped hearing the voices, two of them after the second session and the third after the fifth session, despite hearing voices for 16, 13, and four years. Most of the other patients experienced a reduction in the frequency and volume of the voices, and many were able to live more normal and productive lives. The man who had heard the devil for 16 years thanked us for giving him his life back.
It is now necessary to run a much larger trial of 140 patients with newly trained therapists to determine whether avatar therapy can be integrated into standard clinical practice.