Long-Term Diagnostic Stability of Autism Before the Concept of Spectrum: An Italian Hub-Center Experience

The diagnosis of autism is considered more longitudinally stable compared to other diagnoses for children and adolescents, although significant heterogeneity in longitudinal outcomes emerges between studies based on temporal intervals and diagnostic instruments. The introduction of the concept of Autism Spectrum Disorder (ASD) by DSM-5 may have impacted diagnostic stability. The current study aimed to investigate the longitudinal diagnostic stability of different diagnoses of autism according to the ICD-10 diagnostic system before the institutional move towards the utilization of a unique ASD diagnosis according to DSM-5.

A retrospective study based on medical records and the ELEA electronic database was performed on children attending the Autism Center of the Child and Adolescent Mental Health Service of the Azienda USL-IRCCS, Reggio Emilia, Italy, from January 1, 2005, to December 31, 2015. Diagnostic stability was evaluated according to the diagnosis at follow-up, which was set at 10 years from baseline or December 31, 2020.

A total of 380 children were included at baseline, and 341 were included at follow-up (with a mean temporal interval of 8.74 ± 2.7 years). Among them, 251 individuals maintained a diagnosis within the F84 spectrum, resulting in a diagnostic stability of 73.6%. Specifically, diagnostic stability was 79.8% for the F84.0 diagnosis and 68.2% for the F84.9 diagnosis.

Before the adoption of DSM-5 diagnostic criteria for ASD, ICD-10 diagnoses related to autism were relatively stable in the long run, with an 8-year follow-up. The next step is to assess the longitudinal diagnostic stability of ASD after the adoption of DSM-5 criteria.