A trial of voice assistant speech therapy for young people with dysarthria

The therapy approach being trialed is a four-week period of semi-structured, guided use of a voice assistant (specifically Amazon’s ‘Alexa’) for speech therapy. We will recruit participants with the speech disorder dysarthria, from two subgroups: (i) Perisylvian PMG; (ii) FOXP1 syndrome. This is a cross over trial - all participants will complete 4 weeks of treatment or 4 weeks of control, then swap to the opposite condition. There will be 4 weeks of washout between each condition. Each subgroup will include 6 participants, who will be randomised equally to each of the two sequences (Sequence A treatment then control (n=3 PMG and 3 FOXP1); Sequence B control then treatment (n=3 PMG and 3 FOXP1)). This therapy will be completed by the participant in their home. Participants assigned to the intervention arm will be sent the following items: 1. A voice assistant device by courier. 2. An introductory email containing: a. A prerecorded video showing how to set up the device in the family’s home (e.g. how to connect it to the internet, where to place the device to reduce background noise). b. A therapy plan outlining the family’s timeline in the study, important dates (e.g. assessment dates) and the questions for the participant to ask Alexa during each speech practice session. An accompanying video of an example session will also be produced and sent to families. During the four-week period that the participant is assigned to the voice assistant treatment arm, they will be asked to complete at least 1 session a day. We anticipate that each session will take 10-15 minutes. Parents will facilitate each session as per the therapy plan provided (detailed above), according to the following general structure: 1. Repeated sentences (5): Parent tells (verbally) participant five sentences to ask Alexa. These sentences will be different each day, and prescribed by the research team in the PDF therapy plan. It will be up to the participant to include the wake word (“Alexa”). For example, the parent would ask “Tell me what the weather is today”; ? child: “Alexa, tell me what the weather is today.” 2. Participant generated sentences (5): Parent will prompt child to ask Alexa five additional sentences, which will be generated by the participant. For participants who cannot generate a novel question, parents will be given prompts in the PDF therapy plan to ask participant. For example: Parent “Can you ask Alexa about the football score?” ? child: “Alexa, what is the footy score?” Participants will be encouraged to use the device outside of the facilitated session per day, however additional use will not be prescribed. In this sense, the facilitated session is acting as the minimum dosage. Any additional usage will be additional dosage. Dosage will be tracked using written transcripts which are automatically recorded by Alexa. We will ask parents to send us these transcripts and audio recordings at the close of the four-week intervention period. Parents will be guided to do this during the telehealth assessment session at the close of the intervention period. These recordings will be saved by the study team. Parents will be asked not to interact with Alexa (on the device used for the trial), to ensure recordings and transcripts relate to the trial. Should parents or other members of the house use Alexa on the trial device accidentally, parents will be given an opportunity to delete these incidental recordings, prior to sending these to the study team. Parents will be guided to do this over the phone or Zoom by a member of the research team.