Discrepancy Between Parent Report and Clinician Observation of Symptoms in Children With Autism Spectrum Disorders
Abstract
The Center for Disease Control and Prevention defines Autism Spectrum Disorders (ASD) as pervasive developmental disabilities in which individuals have language impairment, social impairment and stereotyped behaviors. At this time there are no required assessments for a diagnosis. Diagnosis is often made with almost exclusive reliance on parent report of symptoms. The purpose of this research is to examine the consistencies between parent report and clinician observation of symptoms using the commonly used measures to diagnose ASD (ADOS & ADI-R). Previous research has shown diagnosis is more reliable and valid when using both ADOS (clinician observation) and ADI-R (parent interview). It is hypothesized that parents will score their children as less affected in the domains of social deficits and communication and more affected on restricted interests and stereotypies than clinicians. This discrepancy is expected based on the unique relationship between parents and their children and the resulting difference in social interactions and communication parents have with their children compared to an individual with whom the children are less familiar. This research included 12 individuals diagnosed with an ASD who ranged between the ages of 9 to 22 years old (mean age = 15.8). Results from the present study indicate a discrepancy between parent and clinician report on percent agreement on comparable items from the ADI-R and the ADOS. Such findings are crucial as they can help determine how different sources should be considered during the process of diagnosis and creating treatment plans for individuals with ASD.
Introduction
Parents and professionals provide crucial information during the diagnostic and treatment planning processes for individuals with special needs. Parents typically spend the most time with their child and are the main informants regarding their child's behaviors and needs. Parents provide valuable information about social behavior, obsessions, compulsions, self-injury, sensory needs, and other atypical behaviors. These observations however can lack objectivity, which becomes especially relevant when the child is being formally assessed for a developmental disability. Autism Spectrum Disorders are a group of pervasive developmental disorders characterized by difficulties in nonverbal and verbal communication, social interaction, and repetitive behaviors or restricted interests (Rice 2010). The Center for Disease Control reports that 1 out of every 110 children in the United States has an ASD (Rice, 2010). Some disabilities, including ASD, do not have a known genetic foundation or physical features and therefore diagnosis relies heavily on the identification by parents and professionals of atypical behaviors. Sources of data need to be examined for the best quality of assessment for ASD, this is now particularly important as the number of children being diagnosed with ASD is rising.
According to the American Psychiatric Association in the Diagnostic and Statistical Manual IV (DSM IV) in order to be given a diagnosis of ASD, a child must have impairments in all three areas and symptoms must appear in the first three years of life (American Psychiatric Association, 2000). Many people with ASD have language delay but are able to learn spoken language with therapy, or sign language; others are nonverbal. Poor eye contact, disinterest or discomfort in social situations and lack of theory of mind are social challenges faced by individuals with ASD. This disorder is pervasive in that it impacts children in many ability areas, and individuals are often unable to live independently as adults. With a proper diagnosis at the earliest signs, children can receive the most effective services and parents can become educated on the disorder. Because ASD is such a prevalent disorder and symptoms of it are detectable when a child is young, investigating parents' abilities to report behavior is important.
Physicians regularly check all pediatric patients for meeting developmental milestones by looking for abilities like eye contact, responding to their name, and language use. When a child is suspected of having an ASD, the child's pediatrician evaluates the child personally or refers them to a child psychologist. If the child is not developing typically, a developmental psychologist or pediatrician specializing in development will make a formal diagnosis based on the specific behaviors of the child using the DSM IV criteria for ASD. No test, survey or interview is required for diagnosis.. There are several standardized surveys and assessments that can be used when diagnosing and evaluating a person with an ASD. The results from these assessments are also used later when developing an Individualized Education Plan (IEP) for the child's education. The most commonly used assessments for a child with an ASD are the Autism Diagnostic Interview-Revised (ADI-R) (Rutter, Le Couteur & Lord, 2005), The Child Behavior Checklist (CBCL) (Achenbach, 2000), and the Social Communication Questionnaire (SCQ) (Rutter, Bailey & Lord, 2003). All three assessments are based on? parent interviews but content differs on each. Because there is no policy requiring the use of certain assessments, there are inconsistencies for individual children and for people with ASD as a group in the information that is provided for their health care, schooling and therapeutic providers. It is rare that children have a standardized clinical assessment and observation paired with a comprehensive parent interview. Because ASD is a disorder in which individuals have a variety of needs and symptoms, a detailed assessment is of the utmost importance. Over the last decade clinicians have been investigating the reliability and validity of commonly used diagnostic measures and behavioral assessments. The currently regarded "gold standard" for diagnosing a child with an ASD is the Autism Diagnostic Observation Schedule (ADOS) (Lord, Rutter, DiLavore & Risi, 2006), which is performed by specially trained objective professionals in a semi-structured environment (Bishop, 2002). This test uses direct observation from the tester in a standardized format of play in which they observe interaction, communication and any atypical repetitive behaviors (Sikora, 2008). Research has been performed to identify the reliability and validity of parent questionnaires and surveys (Lord, 2006; Sikora, 2008;). However, research is lacking in investigating any potential discrepancy between parent-report and objective clinician findings using standardized measures and the possible causes of such discrepancies. Parents often think their children are more or less capable than they may in fact be as they aren't with their children in school or therapy settings. In contrast, clinicians only see the children in clinical settings which limits the behaviors they may observe. It is likely that a discrepancy exists given these differences.
Research suggests that standardized parent questionnaires alongside objective clinical assessment lead to the most stable diagnosis in ASD (Risi, 2006; Lord, 2006; Sikora, 2008). Risi, Lord, Gotham, Corsello, Chrysler, Szatmari et. al investigated multiple sources used in diagnosis of autism using the ADI-R and the ADOS (2006). Their data suggest that both the ADI-R and the ADOS should be used, but this study does not investigate the domains of autism and settles on using both instead of examining the reasoning behind possible discrepancies (Risi et. al, 2006). Bender, Auciello, Morrison, MacAllister and Zaroff (2008) found the same to be true for children with epilepsy. In a study using both the ADOS and ADI-R to examine the stability of the diagnoses of ASD at age two and age nine, Lord et. al found that clinicians were the group that had the higher percentage of agreement in accurate diagnosis as compared to parents (2006). De Bildt, Sytema, Ketelaars, Kraijer and Volkmar compared parent and clinician agreement. De Bildt et. al (2003) used the ADI-R, ADOS, Autism Behavior Checklist (ABC) (parent report), and the Scale of Pervasive Developmental Disorder in Mentally Retarded Persons (PDD-MRS) (clinician observation). The participants all had intellectual disability and the authors were assessing the developmental disorder component of the participants' disabilities. De Bildt et. al (2003) found that the two parent report measures had good agreement with each other. Clinician observation measures also had good agreement with each other, but the parent report with the clinician observation had poor agreement. De Bildt et. al (2003) did not further investigate the discrepancy between clinicians and parents however, and only noted that it might be the facility and environment that may create these differences. The results from these studies suggest that parents and clinicians are reporting different behaviors.
The ADI-R and ADOS are very expensive diagnostic measures, as they require trained professionals, are expensive, and require large amounts of time compared with other diagnostic measures. In an effort to determine if diagnosis can become less expensive and time consuming, newer diagnostic tools have been developed. One such measure is a shortened version of the ADI-R, the developmental, diagnostic and dimensional interview (3Di) designed by Santosh, Mandy, Puura, Kaartinen, Warrington and Skuse (2009). Santosh et. al's (2009) found their 3Di had a strong agreement with the ADI-R. This is an important study that compares two parent-report diagnostic measures but includes no clinician evaluation thus providing no comparison to clinician observation.
In contrast, a study that used both parent and clinician assessment investigated the validity of the Children's Social Behavior Questionnaire (CSBQ)(Rutter, Bailey, Lord, 2008), a parent questionnaire for children with intellectual disability (de Bildt, 2009). In comparison to the ADOS and ADI-R, de Bildt et. al found much higher correlations of the various subscales with the CSBQ and the ADIR than with the ADOS (2009). The authors attribute this to the ADOS test requiring a short amount of time so that only a limited number of behaviors may be observed. Other research performed by Bishop and Norbury (2002) compared the Social Communication Questionnaire (SCQ), another questionnaire for parents, with the ADOS and the ADI-R and had similar results; a good agreement between the SCQ and the ADI-R but poor agreement between both of those and the ADOS (Bishop, 2002). The authors also noted that categorization on the ADI-R is much more closely related to the diagnostic information from school records than the ADOS and stated that the ADI-R is unlikely to be informed by the school staff and is strictly a parent report (Bishop, 2002). However, this literature does not discuss the fact that the parents communicate with the school staff and unlike the objective clinician, the parents may be biased on that communication. As the literature lacks this discussion, the ADI-R may be representing more than just home life and parents could be influenced by information from the school that the parent may not have observed first-hand.
Educators are providing different information than parents as they do not see the child at home and also do not have the same ASD specific training as clinicians who perform the ADOS. Comparing teacher and parent reports of communication skills in children with ASD, Bishop (2001) found significant differences in the parent versus teacher reports, p<0.05 for speech scores, p<0.01 for inappropriate initiation, stereotyped language, and social relationships, p<0.001 for pragmatic composite scores. Teachers were more accurate in diagnosis than parents when compared with the official diagnosis on record. These results indicate that teachers significantly scored more accurately than parents (Bishop, 2001). Additionally, parental ratings of the child's social rapport had significant differences within the parent group unlike the professional ratings which were consistent (Bishop, 2001). The findings show that parents do not give the same responses as trained professionals, however the results may have been confounded since some of the teachers included mainstream general education teachers with no special education training.
In an effort to improve the reliability of ASD diagnoses, Tomanik, Pearson, Loveland, Lane and Shaw (2007) used the ADI-R, ADOS and the Vineland Adaptive Behavior Scales (VABS)(Sparrow 2005) to examine the importance of parent-report. VABS is a parent interview which includes questions about daily living skills, socialization and adaptive functioning (Tomanik et. al, 2007). Results indicated a concordance rate of diagnoses of ASD of 75% between the ADI-R and the ADOS and including the VABS the accuracy improved to 84% (Tomanik et. al, 2007). This study provides insight on the value of parent report, as the addition of VABS further improved the accuracy.Continued on Next Page »