Profile and speech-language diagnosis in children of the autism spectrum : checklist use

Purpose: The use of a different protocol to assess the same aspects of the Functional Communication Profile (FCP) may contribute to a faster and less expensive determination of individual profiles of abilities and inabilities. The purpose of this study was to verify the applicability of a checklist to replace the aforementioned complete protocol as a way to facilitate clinical and therapeutic follow-up processes. Methods: The participants in this study were 50 children aged from 3 to 12 years, with diagnoses within the autism spectrum who were receiving specialized speech-language therapy for at least six months. The participants were filmed while interacting with the speech-language pathologist, and the data were transcribed to the FCP protocol. After the recording and prior to the transcription, the speech-language pathologists were asked to answer the checklist of Communicative Functions. Results: All answers on the checklist and on the FCP were compared. The results indicated that there were statistical differences in nine of the 20 communicative functions, and in nine of the 50 children. These results suggest that the checklist is efficient to describe a group of children but not to characterize them individually. Therefore, it is possible to identify differences in the communicative profile but not to specify the frequency with which each function occurs. Conclusion: The checklist can be used as a tool in the therapeutic follow-up processes of children with autism spectrum disorders, but it does not replace the complete FCP protocol.


INTRODUCTION
The importance of language in disorders within the autism spectrum is unique, because, contrary to other extensive development alterations in which language disorders are either symptoms or consequences of other deficits, language alterations in the autism spectrum correspond to one out of the three fundamental criteria that determine this diagnosis.
For this reason, they have been investigated (1) and revised by several authors for decades (2,3) .These studies have evolved toward the notion that the central point of language alterations within the autistic spectrum is related to language's functional use, and have thus counted on contributions from pragmatic theories, which enable the articulation of language and aspects of interaction and cognition as an individual's development unfolds; this scenario comprises exactly the triad of fundamental dominions that is used to diagnose autism spectrum disorders (4) .
In regard to aspects of conceptualization and diagnosis, the concept of autism spectrum (5) includes the axes that compose the diagnostic systems proposed by the American Psychiatric Association in the Diagnostic and Statistical Manual of Mental Disorders (6) , and by the World Health Organization in the International Classification of Diseases (7) .The diagnosis criteria in question always entail behavior observation and identification, since a biological marker has not yet been identified for autism (8) .
The concept of autism spectrum allows for the complex interrelation among several clinical situations and not only their juxtaposition (9) .But a discussion still exists concerning exactly which disorders must be included in the autism spectrum, although there is little disagreement about the existence of a spectrum.
The implications of diagnosis criteria based mainly on clinical observation have been extensively discussed (10) , as well as differential diagnoses among the clinical scenarios that compose the autism spectrum (11) .
The notion that the most significant contribution that speech-language pathology and audiology can offer, both to research studies on the etiology of autism and to diagnostic processes, is the determination of a clearer language phenotype, has guided research concerning the best criteria to describe this population's language, the best way to obtain them, alternatives to elicit better performances, and thorough data analysis.
The aim of this study was to verify the applicability of a checklist to replace the complete protocol as a facilitating element in clinical and therapeutic follow-up processes.The participants were 50 children between 3 and 12 years of age who were assisted at a specialized service center on a weekly basis.The procedures do not entail the modification of any elements in the routine of semiannual reassessments of each patient under care; we only required the speech-language pathologists to fill out the checklist and to verify the occurrence frequency of each communicative function and their means of expression.With the results obtained, we sought to statistically compare the answers provided on the checklist to the performance of each child, revealed by means of the Functional Communication Profile.The purpose was to verify a potential equivalency in the application of both instruments to investigate communication in children and teenagers within the autism spectrum.

METHODS
This study was approved by the institution's Ethics Committee, under report number 228-11.The participants were 50 children aged between 3 and 12 years who were being systematically assisted on a weekly basis for at least six months (at least three months with the same speech-language pathologist), with diagnoses included in the autism spectrum.
The inclusion criteria established for the study were: signature of the Informed Consent by one of the legal guardians of each participant, diagnosis included in the autism spectrum, and age under 12 years.
The materials used to register the data were: digital camcorders, media (short DVDs), Functional Communication Profile protocols (Appendix 1) and Communicative Functions checklist (Appendix 2).
Each individual was recorded between April and May of 2012, with the purpose of transcribing the footage data onto the Functional Communication Profile protocol.Before the data were transcribed, the therapists (speech-language pathologists) were asked to fill out the Communication Functionality checklist.After the instrument had been handed in to the researcher, the data obtained from the footage were transcribed.Following the transcription of footage data to the Functional Communication Profile, these were also given to the researcher, with the purpose of compiling the information in a table, so that we could compare the data using a statistical method.

RESULTS
The 50 Communicative Functionality checklists were filled out, and 50 footage pieces were transcribed onto Functional Communication protocols.All answers on the Communicative Functionality checklists were tabulated, as well as the data from the Functional Communication Profile protocols.
Initially, it was necessary to equate the manner of classifying the results presented in both tests.Thus, we used the concept of quartiles for the answers on the Functional Communication Profile, and it was established that, in case a certain function was not observed, it would be classified as "never occurring"; if it happened between one and four times, it would be classified as "rarely occurring"; if it happened between five and 12 times, it would be classified as "occurring many times"; and if it happened 13 or more times, it would be classified as "frequently occurring".
With the results of both tests following the same pattern, it was necessary to find a numerical parameter for the answers.Based on the Likert scale, we attributed values from zero to three to the answers: zero for "never", one for "rarely", two for "many times", and three for "always".Lastly, we used Kruskal-Wallis' test to find similarities or differences in the occurrence of functions in both tests.
The results indicated that, when the occurrence of each function was compared, different results were found in nine out of the 20 communicative functions (Table 1), namely: information request, protest, commentary, self-regulatory function, performative function, targeted, game, narrative, and protest expression, and in nine of the 50 individuals (Table 2).
does not characterize them; in other words, it is possible to know that they have different pragmatic profiles, but it is not possible to identify the frequency of occurrence of the functions.
The results also suggest that the checklist can be utilized as an alternative form of assessment that complements the Functional Communication Profile.

CONCLUSION
The possibility of thoroughly investigating the ability to utilize functional communication in individuals of the autism spectrum, based on the use of simple instruments applicable to family members and speech-language pathologists, represents an important alternative to the speech-language assessment of this population.The checklist can be used as a complementary instrument in periodical pragmatic language assessments of patients within the autism spectrum, because it stresses each individual's specific issues.The use of the Functional Communication Profile is indispensable to an all-encompassing knowledge of the functions used by the patient.
*MAN elaborated the initial study proposal, organized and applied the procedures, analyzed the results, and wrote this paper in collaboration with FDMF, who organized the project and collaborated with data analysis and the discussion of the results.

DISCUSSION
Since its first description, autism, its diagnosis and the type of communication functions used by autistic children have been studied with the purpose of finding increasingly more complete information on the topic and on individuals within the spectrum.It has been widely registered in the literature (12) that the search for efficient alternative diagnostic methods that consider information provided by the parents of children in the autistic spectrum has been increasing rapidly.
Although there are many studies that compare the Functional Communication Profile to diagnostic tests (13) and identify the communication functions most used by autistic children, this is the first study to present a comparative analysis between the Functional Communication Profile and the Communicative Functionality checklist.
The results of this study indicate an incompatibility between the tests in nine of the 20 communication functions, and in nine of the 50 individuals.This indicates that the checklist is efficient to differentiate a group of individuals from another group, but it

Table 1 .
Significant differences in the occurrence of functions on the checklist and on the Functional Communication Profile, found through Kruskal-Wallis' test, for all individuals