More Evidence Avatar Therapy Quiets Auditory Hallucinations

A novel digital treatment designed to reduce the frequency of auditory hallucinations and associated distress in patients with psychosis has been shown to be safe and effective, results from the largest study of avatar therapy to date show. 

The therapy allows patients to interact with a “digital embodiment” of the voice they hear, which is represented by a computer-generated face, also known as an avatar.

In the randomized, multisite, phase 2/3 AVATAR2 trial, patients who received AVATAR-Extended therapy, which included a personalized series of voiced dialogues based on their life history, plus treatment as usual (TAU) showed significantly greater improvement in distress and voice severity levels at 16-weeks vs those who received TAU only. They also had significant reductions in voice frequency at 16 and 28 weeks.

Patients in a third arm who were assigned to TAU plus AVATAR-Brief therapy, which included six sessions of a standardized version of the therapy, also showed improvements at 16 weeks compared with TAU alone — but the clinical impact was stronger with the extended version. 

“I was surprised at the extent to which the extended version seemed to be a more optimal version, and it should be the way forward with this therapy,” study investigator Philippa A. Garety, PhD, professor emerita of clinical psychology at the Institute of Psychiatry, Psychology, and Neuroscience at King’s College, London, told Medscape Medical News.

With more than 300 participants, AVATAR2 is the largest trial to access avatar therapy to date, Garety, noted.

“What’s unique about this work is that technology allows us to create safe face-to-face encounters with a representation of a person’s voice and allows them to relate to that voice in a new way,” she added. 

The findings were published online on October 28 in Nature Medicine. 

A Decade of Research

Auditory verbal hallucinations are common in patients with schizophrenia, but currently available therapies can be ineffective, investigators wrote. 

The therapy allows patients to customize how the avatar looks and sounds. Face-to-face dialogues are then conducted between the patients and avatars in order to build empowerment. A trained therapist provides support during these sessions. 

As reported at the time by Medscape Medical News, the creator of avatar therapy, Julian Leff, MD, presented promising results from a pilot study of 26 patients at the International Congress of the Royal College of Psychiatrists in 2014. 

“Opening up a dialogue between a patient and the voice they’ve been hearing is powerful,” Leff, who was emeritus professor at the Institute of Psychiatry in London at the time, told meeting attendees.

In 2018, a randomized single-site study (AVATAR1) of 150 participants showed that the intervention was associated with a greater decrease in voice severity at 12 weeks vs supportive therapy. Past research led to the idea of incorporating personalization to better optimize the experience. 

Garety noted that AVATAR2 is the largest trial to date of the therapy, as well as the first multisite trial to test the intervention, which was important in order to determine whether it could work outside of a research setting. 

The study included 345 participants (61.4% men; mean age, 39.6 years) from three sites in England and one in Scotland. All were randomly assigned to receive TAU alone (n = 115), TAU plus AVATAR-Brief (n = 116), or TAU plus AVATAR-Extended (n = 114). 

TAU typically included use of antipsychotics, as well as outpatient psychiatric visits and follow-up by case managers and care coordinators. 

“We didn’t interfere with treatment as usual. We wanted to test whether adding this therapy to [TAU] would enhance effects and provide better treatment for their voices,” Garety noted. 

AVATAR-Brief included a standardized process that focused on such things as self-esteem and assertiveness. AVATAR-Extended had two phases. In the first, participants received AVATAR-Brief therapy, whereas the second phase offered a more personalized intervention.

An ‘Unusual Finding’ 

The study’s primary outcome was voice-related distress at 16 and 28 weeks. Although the TAU plus AVATAR-Extended group did show a significant decrease in distress at 16 weeks vs TAU alone (-1.6 points; P = .029), the improvement was no longer significant at the 28-week follow-up (P = .175). The same was also true for the key secondary outcome of reduction in voice severity (-2.32 points; P = .009 at 16 weeks but P =.1 at 28 weeks).

The investigators note that this might be due to the number of dropouts in the AVATAR-Extended group by the 28-week timepoint. The completion rate for those patients was only 58%. The completion rate for the shorter, AVATAR-Brief group was 82%.

On the other hand, the other key secondary outcome of voice frequency was significantly reduced with AVATAR-Extended at both 16 weeks (-0.62 point; P = .01) and 28 weeks (-.89 point; P = .003).

“This is an unusual finding. We’re not aware of any other psychological therapy that shows a reduction in the occurrence of the voice,” Garety said. 

For TAU plus AVATAR-Brief, there were improvements at 16 weeks for distress (-1.05 points; P = .035) vs TAU alone. However, the researchers note that this version of the therapy was just below the prespecified threshold for a clinically significant change and was at the threshold for statistical significance. 

Although the shorter therapy was associated with a reduction in voice severity level at 16 weeks (-2.04 points; P = .017) vs TAU alone, there was no reduction in distress or voice severity at 28 weeks. There was no improvement in voice frequency at either timepoint. 

Both the brief and the extended versions of AVATAR therapy showed improved mood and anxiety levels at 16 weeks and sustained improvement in well-being and recovery, the researchers noted.

“The short version, as expected, did deliver benefits posttreatment, but clearly the extended, optimized version outperformed the brief version. It had stronger and more lasting effects across quite a wide range of outcomes that matter to people who hear voices,” Garety said. 

“In the extended version, people felt more empowered. And in just that version, the frequency of voices was reduced, which is a very important outcome,” she added. 

Safety Issues?

There were 58 serious adverse events (SAEs) in total, with 51% of those occurring in the AVATAR-Extended group. Two participants in that group died; however, independent reviews deemed these events as not related to the intervention. 

In addition, there were no “definitely related” SAEs and only a small number of “possibly related” SAEs, which typically included hospitalization with other contributory factors.

Garety noted during a press briefing that AVATAR therapy has now been demonstrated to be safe across two large trials. 

Study limitations cited included no direct comparison between AVATAR-Brief and AVATAR-Extended or between AVATAR therapy and a different type of psychological treatment. 

Overall, “we recommend that future development and provision of AVATAR therapy is primarily guided” by the AVATAR-Extended protocol, the investigators wrote.

Because the therapy was recommended by a National Institute for Health and Care Excellence Early Value Assessment, the investigators are now seeking to provide it in routine National Health Service settings to gather further real-world evidence of effectiveness over the next 3 years.

Next Steps

Although the intervention isn’t currently available to everybody who might be seeking it, “there’s a pipeline of movement from research into treatment and it’s moving towards the next stage of delivery,” Garety said. 

Investigators are also looking into cultural adaptations for the therapy so it can be used in different locales, including Ethiopia and India, she added. There isn’t a US version yet, but Garety noted that investigators in Canada are looking at similar research and suspects that will also occur in the United States soon. 

“We’re pioneers in this work, and it now needs to be going international and into services,” she said. “We have had many people who hear voices say what an amazing experience this has been. So, I feel very proud and excited to have been able to be part of this.” 

At the press briefing, Miranda Wolpert, director of mental health at Wellcome, which funded the study, noted that it is encouraging to see the development of a new intervention that could potentially change the lives of patients across the world. 

“We know that psychosis can start early in life, stopping people from having the jobs and relationships they want and from achieving the goals they want. This intervention was developed with those people to help them address an issue that really troubles them,” Wolpert told attendees. 

“For me, this represents part of a revolution we are starting to see in terms of mental health interventions and the potential impact on mental health science,” she added. 

Digital Placebo Effect?

Commenting on the findings for Medscape Medical News, John Torous, MD, a psychiatrist and director of digital psychiatry at Beth Israel Deaconess Medical Center, Boston, Massachusetts, said there is a need for new treatments for schizophrenia that work with different mechanisms.

“We have a lot of medication studies but not as many innovative therapy studies. I think it’s exciting that the results, at least in the shorter-term outcome, were positive. And I think that’s something that can give people hope in using these new technologies,” said Torous, who is also an assistant professor of psychiatry at Harvard Medical School and was not involved with the research. 

Still, he did note some study limitations, including whether there could have been some type of “digital placebo effect” from the therapy. 

“If you tell people they’re getting high-tech advanced digital care, that may have some effect,” he said, adding that “it’s always interesting” to tease out the benefit being delivered by the technology vs the delivery mechanism itself — or some combination of both.

Torous added, though, that it’s very difficult to have a rigorous digital control group. “It’s not necessarily a fault of their study, but it’s something to keep in mind when interpreting what the results are,” he said. 

He also noted that he would have liked to have seen a direct comparison between this new kind of psychological therapy vs standard psychological therapy, such as cognitive-behavioral therapy. 

In addition, he wondered about expenses and scalability of the intervention, and whether patients would need to go to a specialized center to undergo treatment. Torous mentioned that a version involving virtual reality could perhaps make this more scalable in the future. 

Overall, he said that what the investigators are currently doing is very innovative. “It’s exciting that we’re talking about the next steps. Giving people new options for psychological therapy that may be effective for their disorders is really wonderful to see,” Torous said. 

The study was funded by the National Institute for Health and Care Research (NIHR), the Wellcome Trust King’s Clinical Research Facility, the NIHR Maudsley Biomedical Research Centre and Maudsley NHS Foundation Trust, King’s College London, the Manchester Biomedical Research Centre, and NHS Research Scotland, as well as by a grant from Wellcome. Garety reports being an unpaid scientific adviser to Avatar Therapy Ltd. Financial disclosures for the other investigators are fully listed in the original article. Torous reports no relevant financial relationships.