Effectiveness of a Robot-Assisted Psychological Intervention for Children with Autism Spectrum Disorder.

Analyses mainly used the 44 who completed the full protocol, not all 51 allocated, so children who dropped out or received a partial dose may have responded differently. The study was small, with 51 allocated and 44 analysed, and was conducted in two centres in Greece. Many participants also received speech therapy, occupational therapy, physiotherapy or special education, so the contribution of the robot-assisted component cannot be separated from those supports. The therapist was central in both arms, could change the flow and used cues to control the robot, so the robot alone cannot be credited with the change. Follow-up data were very limited, with only 22 completed follow-ups, so durability of effects is unclear. Teacher reports were limited because lessons were online, and teachers could not observe social skills as well as in a classroom. Robot recordings had technical problems and could not be validly compared with therapist recordings. The CARS-2 change was statistically significant but did not reach the authors' stated clinical meaningfulness threshold. Baseline group differences existed for processing speed and inhibition, although overall characteristics were similar.

Funding came from the Action Research-Develop-Innovate programme under the Operational Programme Competitiveness, Entrepreneurship and Innovation and the National Strategic Reference Framework of Greece 2014-2020. The supplied text does not state author conflicts of interest.

Do not conclude that the robot adds benefit over therapist-only care. On most primary outcome comparisons the groups did not differ significantly, the CARS-2 change was too small to matter clinically, and one child showed stress at first robot contact. Many children also received speech therapy, occupational therapy, physiotherapy or special education, and the therapist guided every session, so the robot's contribution cannot be isolated.