Feasibility of a Smartphone Application to Identify Young Children at Risk for Autism Spectrum Disorder in a Low-Income, Community Setting in South Africa

Background:

More than 90% of children with autism spectrum disorder (ASD) live in low- and middle-income countries (LMIC) where there is a great need for culturally appropriate, scalable and effective early identification and intervention tools. Smart phone technology and application (‘apps’) may potentially play an important role in this regard. Here we investigated the feasibility of a mobile application that elicits and quantifies social referencing and positive emotional behaviours in young children to detect risk for ASD.

Objectives:

Our aim was to determine the technical feasibility and cultural appropriateness of the Autism&Beyond iPhone application to determine risk for ASD in children aged 12-72 months in a naturalistic, low-income South African community setting.

Methods:

Parents of typically-developing African children were recruited from community centres in Khayelitsha Township, Cape Town, South Africa. With appropriate ethics and consent, parents completed a short technology questionnaire, followed by an iPhone-based demographic and ASD-related questionnaires. Next, children were shown 4 videos of 30s each on a study smartphone while sitting beside or on a parent’s lap. The smartphone front facing (“selfie”) camera recorded video of the child’s facial expressions and head movement. After home-based administration of the app, a subset of families were invited to participate in focus group discussions. For the purpose of this presentation, we examined 1) parental familiarity with and access to smartphones, internet and apps, and 2) parental feedback about the cultural acceptability and accessibility of smartphone technology using apps in the local community.

Results:

A total of 45 families participated in this part of the feasibility study. Parent participants had a mean family income less than $30 per week, representing the lowest socio-economic group in South Africa. The majority of parents (38/45, 84%) owned a cellphone, of which 79% were smartphones. Interestingly, no participant had an iPhone. 13/45 (29%) had email, 24/45 (53%) had a social media account, and 20/45 (44%) used WhatsApp. 10/45 (22%) had internet access at home and 6/45 (13%) and 6/45 (13%) access to WiFi. Interestingly, 27/45 (60%) used mobile data. 25/45 (55%) rated their proficiency using a cellphone as good to very good. Thematic analysis of focus group discussions indicated that families found the app relatively easy to use, and would recommend it to others in their community. Parents did not have any concerns about the fact that videos were ‘American’, but made a number of recommendations to localise the app. A key theme from families was the importance that the app must be free, and that data transfer should not be at a cost to them.

Conclusions:

The technical feasibility and cultural appropriateness evaluation in this very low-resource community in South Africa suggested that smartphone technology and apps may be a useful tool to reach families for early identification of ASD risk, provided the need for an android platform and free access to app and data to make this feasible. Further evaluation of the children’s performance on the app, accuracy of the automated coding, and comparison with US data will extend these early observations.