Abstract
Service providers and researchers in autism spectrum disorders (ASD) are challenged to categorize clinical variation in function. Classification systems for children with cerebral palsy have enabled clinicians and families to describe levels of function. A web-based survey engaged international ASD stakeholders to advise on considerations of function and constructs needed to classify functional subgroups of preschool children with ASD. All respondents (n = 56) believed that evaluating function was important; 95% reported that function was one way to subclassify preschool children with ASD. Two domains judged most important were communication language (55%) and social function (22%). Respondent comments indicated that these domains were difficult to disentangle.
Children with autism frequently are subcategorized based on symptom profiles and developmental history (e.g., Asperger disorder) on the basis of early language milestones. Researchers typically subclassify autism based on cognitive function (Allan et al., 2001; Nicholas et al., 2008; Tsatsanis et al., 2003). However, using cognitive skills as a proxy for clinical severity may not adequately characterize variations in ability and functioning among children with autism spectrum disorders (ASD). Diagnostic categories have not been reliably differentiated as separate groups in ways that are independent of the defining criteria, nor are they informative in terms of prognosis or treatment needs (Boucher, 1998; Schopler, 1996).
One method to categorize function is to examine characteristics of “functioning” as defined within the International Classification of Functioning, Disability and Health (ICF) framework (World Health Organization, 2001). Classification systems based on the ICF concept have been successfully developed for children with cerebral palsy (CP; e.g., the Gross Motor Function Classification System [GMFCS], Palisano, Rosenbaum, Bartlett, & Livingston, 2008). The GMFCS provides clinicians and families with a consistent communication tool to describe a child’s functional ability (in the CP hallmark area of gross motor function), as well as a stratification system to enable the child to receive the most appropriate intervention for gross motor function. Stratification also has been used in research to minimize confounding that can affect outcomes in intervention studies. Longitudinal research using the GMFCS has led to the creation of motor growth curves for children with CP, illustrating that classification systems can predict function over time (Rosenbaum et al., 2002). These findings have been extremely useful for clinicians and parents to set realistic intervention goals, have an evidence-informed understanding of prognosis of gross motor abilities, and identify the appropriate intervention(s) for the child based on one critical construct in CP, gross motor function. Such a system in ASD, based on the characteristic(s) that are the hallmark of the disorder, could provide the consistency needed in both clinical practice and research.
Method
The survey focused on the function of children with ASD who were 3 to 5 years old. This age range was selected based on the common age of ASD diagnosis (>3 years; Volkmar, State, & Klin, 2009), while maintaining sufficient homogeneity to allow a meaningful initial exploration of the construct of “function” in ASD. Potential informants were recruited to complete the online survey using a variety of processes, such as identifying authors from a literature review related to diagnosing and classifying ASD, identifying experienced clinicians through established autism clinics, and identifying nonclinical experts (i.e., parents of children with ASD and child educators) through administrative staff at local and provincial autism associations. These individuals were considered stakeholders because in addition to being knowledgeable about ASD and factors that affect day-to-day functioning in this population, they each had a strong personal and or professional interest (or “stake”) in issues related to ASD diagnosis and care. Among the 54 stakeholders contacted, 50 expressed interest in participating and suggested another 6 informants. Stakeholders who did not participate (n = 4) either declined due to other commitments or had inaccessible email addresses. A total of 56 respondents participated in the survey, with 45 surveys (more than 80%) fully completed.
An initial draft of the survey involved an iterative process of drafting items after reviewing the literature and the categories in the Diagnostic and statistical manual of mental disorders (4th ed.; DSM-IV; American Psychiatric Association [APA], 1994). The key areas addressed by the survey included the perceived importance of evaluating functional ability, characteristics of subgroups based on severity, examination of existing labels of functional severity, and the impact of environment on functional ability. The survey was pilot tested with six local clinical stakeholders (one speech/language pathologist, three occupational therapists, and two developmental pediatricians) experienced in ASD and/or childhood disability. Several changes were made to ensure readability and clarity of the survey items. The final survey had 11 items and involved various item formats (e.g., forced choice, rating scales). Research Ethics Board approval was received from McMaster University. Stakeholders were sent the link to the survey, which was available on www.surveymonkey.com.
Survey respondents varied by years of experience (1–5 years [12%], 6–10 years [11%], and 11+ years [57%]); country of origin (Canada [36%], the United States [18%], Australia [16%], and the United Kingdom [9%]); and professional background (pediatrician, psychiatrist, developmental pediatrician, child neurologist [32%], psychologists [21%], therapist, nurse, social worker [27%], and parent, educator, administrator [14%]). Some respondents had multiple roles as a professional and administrator and/or parent. Respondents interacted with children with ASD in the following settings (included more than one setting for some): ASD clinic (48%), ASD center/organization (23%), community assessment or treatment clinic (16%), child development center or children’s treatment center (32%), consultant (41%), researcher (16%), involvement in community/school/nurseries (4%), and involvement in the home (2%).
Results
Responses on a discrete item showed that 100% of respondents (n = 56) believed that evaluating functional level of ability provides important clinical information. Furthermore, on a 5-point Likert-type scale (1 = not very important to 5 = extremely important), respondents (n = 55) rated highly the importance for evaluating functional ability as 49% and 38% showed scores of 5 and 4, respectively. Similar ratings were provided for quality of functional ability, where 100% of respondents (n = 53) indicated that it provided important clinical information and rated level of importance to evaluate quality of function as extremely important (52%) and important (38%).
Stakeholders (74% of 53 respondents) reported that children were subcategorized (formally or informally) to determine the level of severity. Among this 74%, the respondents selected the following method(s) they believed were necessary to categorize subtypes of ASD: classification systems (e.g., DSM-IV [APA, 1994], International Classification of Diseases–10th Revision) = 56%; diagnostic tools (e.g., Autism Diagnostic Observation Schedule, Childhood Autism Rating Scale) = 72%; developmental comparisons with other children with disabilities = 28%; developmental comparisons with other children with ASD = 46%; and functional measures (e.g., cognitive, language, adaptive behavior, and/or motor scales) = 95%.
Respondents ranked the most important constructs or domain areas to classify subtypes of children with ASD as (a) language/communication (55%), (b) social function (22%), (c) cognitive function (21%), and (d) repetitive behaviors (0%).
For “mild ASD” and “high functioning” labels, respondents consistently identified the social functioning characteristic as the most impairing functional domain, 67% and 58% of respondents, respectively. The top-ranked characteristic of impairment identified by respondents in both “severe ASD” and “low-functioning” labels was cognitive function, 47% and 57%, respectively.
In all, 98% (n = 42) of respondents believed that features of the environment can impair functional ability. Each environment (i.e., home, community, and school) was rated as to its impact on functional characteristics of the child. Respondents (n = 37) judged that the impact of environment varied depending on the functional domain being considered, as follows: home environment (language function = 59%); community environment (social function = 67%); and school environment (social function = 50%).
Discussion
The majority of stakeholders in our sample believed that it is important to consider all attributes of function (level and quality) in children with ASD in clinical practice. The majority of stakeholders also endorsed that functional measures are used and/or ought to be used (independently or in conjunction with symptom-based measures) to help classify levels of ASD severity. The majority (55%) reported that the language/communication domain was the most important construct for classifying ASD subtypes. The second most highly ranked construct was social function (22%). These results can be used to suggest combining these domains (together identifying 77% of respondents) and can be characterized by the single construct of “social communication.” This aligns with the trend of the proposed DSM-V that collapses these two domains into one—“social communication” (APA, 2009).
Our work here not only describes the attributes of function that influence severity of ASD in preschool children but also highlights “social communication” as a major construct of function on which to subclassify children with ASD. This early work has generated new knowledge on how stakeholders in ASD perceive attributes of function and has fuelled more questions about characteristics of severity and function, within the construct of social communication. This work marks the beginning of the exploration of the construct of social communication as the basis of an Autism Classification System of Functioning (ACSF), analogous to the GMFCS in CP. The ACSF will aim to provide new knowledge to the understanding of the natural history of ASD and brings a fresh perspective in categorizing the phenotypic diversity in ASD.
Footnotes
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The authors received no financial support for the research, authorship, and/or publication of this article.