Feasibility of an observational procedure to enhance early identification of autism spectrum disorder in paediatric settings: A mixed-methods study on an Ecuadorian sample.

The children tested did not have ASD, so the study cannot show whether the tasks detect autism or distinguish autistic children from non-autistic children. The samples were small: 125 children in Study 1 and 33 children in Study 2, recruited from Quito-area preschools and clinical settings. The authors say the findings cannot be generalised to Andean plateau or Amazonian areas, which cover 48% of Ecuador's territory and about 5% of its population. Social referencing did not produce expected responses and was dropped; pointing responses varied by socio-economic level, residency area and age, so that item needs caution. In the clinical setting, eye contact responses fell from 97% to 62%, and fidelity was lowest when professionals did not first get the child's attention, so delivery quality matters. Professionals were not blinded to the study aims and completed ratings in the researcher's presence, so their positive attitudes may reflect a desire to please. The study did not measure sensitivity, specificity, diagnosis, referral or intervention outcomes, so it does not establish clinical effectiveness.

The authors reported no financial support for the research, authorship or publication, and declared no potential conflicts of interest.

Do not conclude that this procedure detects autism or improves early identification. It was tested only in children without ASD, did not measure sensitivity or specificity, and the authors say these accuracy measures still need to be assessed in children who receive an ASD diagnosis.