Characteristics of Speech-Language Pathology Evaluations Abstracted During An Autism Prevalence Study

Background:   The American Speech-Language-Hearing Association guidelines suggest speech-language pathologists (SLPs) can diagnosis an autism spectrum disorder (ASD) if trained in DSM-IV criteria, use ASD diagnostic tools, and are experienced in the diagnosis of developmental disorders.  The degree to which SLPs actively identify and assess ASD and associated behaviors is unknown.  In this study, we examine the ASD diagnostic patterns and characteristics of SLP evaluations abstracted during an ASD population-based prevalence study and compare to other provider types.

Objectives:   1) Among children with a previous ASD diagnosis, compare differences in evaluation patterns by provider type including average age at evaluation, average number of evaluations, and percent of total evaluations; and 2) Compare DSM-IV behaviors and autism discriminators as defined by the Centers for Disease Control Autism and Developmental Disabilities Monitoring (ADDM) Network on individual provider evaluations by provider type.

Methods: Children aged eight in 2008 were identified with an ASD (n=51) from a surveillance region in Northern Utah using the ADDM Network retrospective record review methodology.   Speech language pathologists were one of 37 provider types documented.   An ASD was identified based on information obtained from a cumulative set of evaluations scored by a qualified clinician reviewer using DSM-IV criteria for ASD.  Individual evaluations by provider type for children with a final case definition of ASD were used for the analyses. Descriptive statistics and goodness-of-fit tests were used to analyze patterns.

Results: Among ASD cases (N=51), 616 evaluations were abstracted.  A total of 103 evaluations (16.7%) were conducted by SLPs for an average of 2.2 evaluations per case. In comparison, 513 evaluations were conducted by non-SLP providers for an average of 10.1 evaluations per case.  Doctoral level psychologists conducted the second largest number of evaluations (13.02%) after SLPs.  Thirty-two unique SLPs were identified among all abstracted evaluations compared to 23 unique non-SLP providers.  The average age at which a SLP evaluated a child with ASD was 5.12 years compared to 4.61 for all other provider types.  Statistically significant differences were found in the presence of DSM-IV criteria and ASD discriminators as a function of provider type.  Speech language pathologists were significantly more likely to document a DSM-IV qualitative impairment in communication (item 2b) and significantly less likely to document the DSM-IV social behavior 1d and restricted, repetitive, and stereotyped patterns of behavior, interests or activities (3a-3d).  No differences were found on other social impairments as a function of provider type (1a-1c).  Speech language pathologists were significantly less likely to mention an autism discriminator in evaluations (19%) compared to other provider types (35%).  In cases, SLPs did not suggest or diagnose an ASD 79% of the time. 

Conclusions:  As was expected, Utah SLPs identify DSM-IV communication behavior 2b more frequently than other provider types in children with ASD.  Further training may be beneficial for SLPs to recognize and document other ASD related behaviors.