Les performances du dépistage des troubles du spectre autistique (TSA) restent insuffisantes malgré l’identification des facteurs impactant son efficacité. Une des pistes d’amélioration est celle de l’inclusion des enseignants dans ce processus de repérage. Les recherches internationales montrent en effet que les enseignants sont des informateurs potentiellement fiables, de par la richesse des observations pouvant s’effectuer en milieu scolaire. Notre étude vise à examiner cette hypothèse, en comparant la qualité des observations réalisées par des enseignants d’école maternelle à celle de parents et de professionnels de l’autisme à partir d’un outil de dépistage commun.

La cohorte principale est composée de 90 enfants avec TSA vus en Centre Ressources Autisme pour une évaluation diagnostique. L’Autism Discriminative Tool(ADT) a été rempli en aveugle par les professeurs et les parents en début de bilan. Leurs observations respectives ont été comparées à celles de l’équipe diagnostique au terme du bilan, soit au référentiel complet du tableau clinique de l’enfant.

Les enseignants s’avèrent être d’excellents témoins des signes de TSA, avec un pouvoir de prédiction de 83,3 %, proche de celui des parents. La nature et le nombre de signes autistiques qu’ils observent (x_moy=16,64) rejoignent celles des experts (x_moy=16,78). Leurs observations sont particulièrement pertinentes en ce qui concerne les déficits sociaux en groupe, là où les parents se concentrent sur des aspects plus individuels et plus subtils.

La tripartite enseignants–parents–professionnels est le modèle à privilégier pour améliorer le dépistage des enfants présentant un TSA.

Autism screening remains largely inefficient despite the identification of multiple biases and shortages. One way to improve this process may be to include teachers as informants. Indeed, international studies suggest that teachers are good observers of developmental particularities, including in the field of autism spectrum disorders. Furthermore, school setting seems like the right environment to notice socio-communicative difficulties. Our study examines the relevance of including preschool teachers in the ASD screening process, by comparing the quality of their observations to those of parents and experts in autism with an identical detection tool.

The examined cohort is composed of 90 preschoolers aged two to six years of age, seen in a specialised autism clinic for diagnostic purposes. The Autism Discriminative tool, a new screening test, was blindly and independently rated by teachers and parents at the very beginning of the assessment. Their observations were compared to those gathered by diagnostic team members at the end of the assessment procedure and during the consensual decision process which finalizes the diagnostic status of the child. In this configuration, ADT profile depicted by professionals was considered as the true clinical picture of the child and therefore the referential against which parents and teachers’ observations were compared.

Data show that teachers are excellent informants when it comes to detect symptomatology of autism spectrum disorders in preschoolers. Their screening predictions are 83.3% accurate for children who end up receiving the diagnosis of ASD at the end of the assessment procedure. In addition, teachers are as efficient observers as experts in the field, with a high concordance in symptoms observation (ADT X_mean=16.64 versus X_mean=16.78). Teachers are particularly accurate when it comes to observe skills that are paramount for group and social interaction. For example, they are better placed to pinpoint difficulties with peers and reactions to changes. Parents’ observations within the home setting and during family outings are different from those gathered from school. Nevertheless, their observations are complementary to those of teachers, with a better focus on the one-to one relationship. As such, they are more able to notice subtle difficulties such as deficit of eye-contact, lack of social smile or fingers mannerisms.

In the domain of ASD, observations of the child's difficulties are very much context-dependent. Therefore, the isolated questioning of a parent, a school intervener or the sole observation of the clinician only partially describes the child's psychopathology. As the combination teachers–parents–professionals ensures a better detection of ASD, we strongly suggest that this model be promoted in official screening policies.