Assessment of autism in children with visual impairment


  • Carey A Matsuba

    1. Division of Developmental Pediatrics, BC Children's Hospital and Sunny Hill Health Centre, University of British Columbia, Vancouver, BC, Canada
    Search for more papers by this author


This commentary is on the original article by Williams et al. on pages 66–72 of this issue.

It is well recognized that the prevalence of autistic spectrum disorders (ASDs) has increased substantially.[1] However, the clinical tools to assist in the diagnosis of ASDs may not always be applicable to unique populations. Through a case series of nine children, Williams et al. have reported on their modified approach to the diagnosis of autism in children with visual impairment.[2] This paper adds to the discussion on what clinicians need to consider when diagnostic or assessment tools may not always be applicable to special populations. Several points should be highlighted.

First, the authors recognize a need to modify the diagnostic tools required. As many clinicians know, standard assessment tools may not be appropriate in specific populations. In this article, the authors assessed children with visual impairment, including eight with optic nerve disease and one with retinal disease. Unfortunately, this is far from a representative population. Of the nine children assessed, four children, all with optic nerve hypoplasia (ONH), were ‘diagnosed’ with autism. ONH is a spectrum condition that has etiological associations – some similar to autism – including systemic conditions and teratogenic agents. In addition, with ONH there are often significant associated health conditions including hypopituitarism, seizure disorders, and structural brain anomalies. Hence, the population being considered may not be truly representative and thus the modifications to the assessment tools may not apply to children with other visual impairment conditions.

Second, diagnostic tools have become more widely used in the evaluation of autism. In their case series, Williams et al. chose to modify aspects of the assessment tools. At face value, this seems to be appropriate. But we must consider the possible challenges with the modifications in the diagnostic tools in children with ONH as one of the etiological conditions of visual impairment under study. ONH has its own associated developmental disabilities including cognitive, motor, and speech impairments. For this reason, some of the proposed modifications, such as using Braille books to assist in the diagnosis, may not be meaningful. Braille relies on fine motor and cognitive skills. If other impairments were present, perhaps Braille would not be a good option. When considering whether items should be modified in the context of visual impairment, it may be important to consider comorbid associations.

Acquisition of developmental skills requires opportunity and experience. Many children with visual impairment may have had limited learning experiences. Braille, like many other developmental skills, relies on opportunity and experience. If these were limited, it is uncertain how modifications would be considered meaningful.

Third, the usefulness of diagnostic tools is to establish a degree of certainty about the condition that is being assessed. In the current article, the authors have chosen to modify diagnostic tools along with clinical evaluation. In doing the clinical evaluation, Williams et al. recognize that autism is more than a checkbox of language, social, and stereotypic abnormalities. Therefore, the modified tool should be able to assist us in making the ultimate diagnosis. To date, others who work with visual impairment have been able to modify assessment tools to make a diagnosis of autism in that context. So the question is whether these specific modifications were needed. There is still some uncertainty.

I applaud the efforts of Williams et al. in bringing this important issue to the forefront of developmental medicine. The developmental needs of children with visual impairment must not be underserved. When clinically indicated, the formal assessment of developmental disabilities should be undertaken in children, whether or not they have visual impairment. It must also be recognized that there will be a need to adjust or modify assessment tools to assist in the clinical diagnosis. How these assessments are undertaken remains, in my opinion, unclear, as it is dependent on the challenge of each presenting child. Thus, perhaps at present, the diagnosis of ASD in children with visual impairment should be based on clinical judgment.