Feasibility and Efficacy of a Virtual Reality Social Prediction Training in Children and Young Adults with Congenital Cerebellar Malformations.

The pandemic prevented the planned 42 participants from being reached, so the study was underpowered for some primary analyses. The assessor was not blind to group allocation. The sample was small, single-centre, and included varied cerebellar malformations and cognitive levels, so results may not apply to all patients with cerebellar malformations. The action-prediction beta index was not significant after training, and the follow-up result was marginal or exploratory. Parent-reported emotional-behavioural problems did not differ, and the parent measure was not designed to detect change in neurodevelopmental disorders. The VR-Spirit group also received motor games, so the trial tested a combined social-prediction and motor-game programme, not the social-prediction game alone.

The paper reports funding from Ministero della Salute and Ministero dell’Istruzione, dell’Università e della Ricerca. It does not report other conflicts of interest.

Do not read the significant VR prediction score as proof that the training improves every social outcome or changes behaviour. The study did not reach its planned 42 participants, the action-prediction score was not significant after training, and parent-reported emotional-behavioural problems did not differ.