MENTAL RETARDATION AND DEVELOPMENTAL DISABILITIES RESEARCH REVIEWS 11: 253262 (2005)

SCREENING FOR AUTISM IN YOUNG CHILDREN: THE MODIFIED CHECKLIST FOR AUTISM IN TODDLERS (M-CHAT) AND OTHER MEASURES
Thyde Dumont-Mathieu1,2* and Deborah Fein2
1

Department of Pediatrics, University of Connecticut School of Medicine, Farmington, Connecticut

2

Department of Psychology, University of Connecticut, Storrs, Connecticut

The literature on the importance of early identication and early intervention for children with developmental disabilities such as autism continues to grow. The increased prevalence of autistic spectrum disorders has fostered research efforts on the development and validation of autismspecic screening instruments for use with young children. There are currently several such autism-specic screening tools meant to be used with young children in various stages of development. Data from a few of these screening instruments have been published, and they include the Checklist for Autism in Toddlers (CHAT), Pervasive Developmental Disorders Screening Test (PDDST), Screening Tool for Autism in Two year olds (STAT), Checklist for Autism in Toddlers-23 (CHAT-23), and the Modied Checklist for Autism in Toddlers (M-CHAT). In this review, these ve tools designed for use with children under three years old will be highlighted. In particular, the Modied Checklist for Autism in Toddlers (M-CHAT) will be discussed.

2005 Wiley-Liss, Inc.

MRDD Research Reviews 2005;11:253262.

Key Words: autism; pediatric screening; early identication; early intervention

BACKGROUND utism is a neuro-developmental disorder rst described by Leo Kanner in 1943. It is one of a group of disorders known as Pervasive Developmental Disorders (PDDs), now more commonly referred to as Autistic Spectrum Disorders (ASDs). This group of disorders includes Autistic Disorder, Pervasive Developmental Disorder-Not Otherwise Specied (PDD-NOS), Aspergers syndrome, Rett syndrome, and Childhood Disintegrative Disorder (CDD) (American Psychiatric Association (APA), 1994). Autistic Spectrum Disorders are currently estimated to affect somewhere between 1 in 166 and 1 in 1000 children [Baird et al., 2000; Bertrand et al., 2001; Chakrabarti and Fombonne, 2001; Yeargin-Alsopp et al., 2003 Volkmar et al., 2004; Barbaresi et al., 2005; Chakrabarti and Fombonne, 2005]. The diagnostic criteria for autistic disorder require impairment in three areas: reciprocal social interaction, communication, and specic patterns of behavior, interests, and activities [APA, 1994]. Research has documented areas of dysfunction related to these core categories of symptoms in autism: (1) social abilities (attachment, joint attention, social imitation, orienting to social stimuli, face and affect processing, expression of emotion, and symbolic play) [McEvoy et al., 1993; Dawson et al.,

1998; Klin et al., 2002; Klinger et al., 2003]; (2) communicative abilities (use of language, both verbal and non-verbal, quality of communication and play) [Travis et al., 2001; Klinger et al., 2003; Dawson et al., 2004] and (3) restricted or stereotyped behaviors; interests, and activities (non-functional routines or ritualistic behaviors, resistance to changes in the environment, repetitive motor mannerisms, unusual interests and preoccupations; and visual fascinations) [Volkmar et al., 1986; Szatmari et al., 1989; Campbell et al., 1990; Turner, 1999; Charman and Swettenham, 2001; Klinger et al., 2003]. There are additional symptoms that are sometimes present with ASDs, these include self injurious behaviors, functional disturbances such as difculties with sleeping and eating, abnormal fears, and abnormal responses to sensory stimuli [Klinger et al., 2003]. Cognitive functioning can range from severe mental retardation to above average intelligence quotient (IQ), with many children who have an ASD showing some degree of cognitive impairment. A recent study by Chakrabarti and Fombonne [2005] found that in a sample of four to six year olds diagnosed with an ASD, nearly a third had signicant cognitive impairment. Previous studies have found that regression in ASD occurs most often between 18 and 24 months of age, with the frequency of the regressive type of autism being somewhere between 10 and 50% [Tuchman and Rapin, 1997; Goldberg et al., 2003; Lord et al., 2004; Rogers, 2004]. Clinical experience of the present authors suggests that regression often occurs sooner, sometimes as early as 1218 months. It is possible that parents may recall 18 24 months as the age of regression for children who were diagnosed later, or that parents are experiencing recall bias [Siperstein and Volkmar, 2004]. It is possible that as the age at diagnosis drops with earlier identication, so will the reported age at regression.
Contract grant sponsor: National Institute of Child Health and Development (NICHD); Contract grant number: 5-R01 HD39961. *Correspondence to: Thyde Dumont-Mathieu, M.D., M.P.H., Department of Psychology, University of Connecticut, 406 Babbidge Rd., Storrs, CT 06269-1020. E-mail: tdumont@stfranciscare.org Received 11 July 2005; Accepted 12 July 2005 Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrdd.20072

2005 Wiley-Liss, Inc.

There are many important theoretical and clinical issues related to the subject of regression in autistic spectrum disorders. Do parents and providers mean the same thing when they discuss regression? What constitutes regression as opposed to developmental stagnation? Is the regression seen with autism similar mechanistically to that seen in other regressive disorders such as Landau-Kleffner syndrome [Ballaban-Gil and Tuchman, 2000; Robinson et al., 2001; Stefanatos et al., 2002 Trevathan, 2004], or to disorders within the autistic spectrum, Childhood Disintegrative Disorder [Volkmar and Rutter, 1995; Malhotra and Gupta, 1999 Goldberg et al., 2003; Manning-Courtney et al., 2003], and Rett syndrome [Glaze, 2004]? Is the outcome different for those children who have had a normal developmental trajectory prior to the onset of regression, as compared with those whose regression is limited only to language or those who were developmentally delayed even before the regression? Does the timing of the regression impact outcome? Does autism with regression differ in etiology, impact, and prognosis than autism without regression? Various mechanisms of regression have been suggested, including illness, sub-clinical seizures, and variations in postnatal brain development due either to the impact of genetic factors or postnatal events [Lainhart et al., 2002; Hrdlicka et al., 2004]. A discussion of these questions and issues is beyond the scope of this article and the readers are referred to a review of developmental regression published in this journal [Rogers, 2004]. However, there are two issues pertinent to this discussion of screening young children. Screening tools need to ask about regression in a way that is compatible with parent and provider perspectives. It is also important to not only ask about regression, but lack of developmental progression. Secondly, it is important to remember that autism screening will not capture regression if the screening precedes the regression. Retrospective studies have shown that children who are later diagnosed with autism exhibit symptoms as early as 8 12 months of age [Osterling and Dawson, 1994 Baranek, 1999; Werner et al., 2000; Osterling et al., 2002]. In general these studies reviewed videotapes of the children prior to or at their rst birthday and analyzed their behaviors, as compared with typically developing children or children with non-ASD diagnoses. Some children with ASDs seem to exhibit symptoms as early as 8 months, but 254 MRDD RESEARCH REVIEWS

more consistently around 12 months. The identied symptoms are fairly consistent and include delay or absence in pointing, showing objects, looking at others, and orienting to their name [Osterling and Dawson, 1994]. In addition to the difculties with aspects of social responsiveness (poor visual orientation/ attention, response to name) these children also seem to exhibit sensory-motor difculties (mouthing objects excessively, aversions to social touch) [Baranek, 1999]. Other symptoms that seem to indicate possible ASD include stereotyped play with objects, unusual posturing of body parts, looking at a camera less frequently, staring/xating on objects, and having less animated affective expressions [Baranek, 1999]. Lack of response to name seems to be the most consistent singular symptom at this young age [Osterling and Dawson, 1994; Baranek, 1999; Werner et al., 2000 Osterling et al., 2002]. The ndings from these studies, although retrospective, document some of the behavioral signs of autism in very young children, and suggest that early screening might be feasible. Ongoing research is needed to uncover proles of children with ASDs at young ages, since the signs may be different from those seen in older children. Prospective studies are needed with infants and toddlers to further investigate the best means of differentiating between variants of typical development and the earliest symptoms of autism. In the meantime, it would seem prudent to verify that children under 12 months old being seen for their well child care visit are looking at others and orienting to their name, since these two skills may be two of the earliest red ags for possible autism. Baranek [1999] states that the early markers for autism noted on videotape review preceded the voicing of concerns by parents. She also mentions that the parents in her study compensated for these markers, suggesting that these compensatory behaviors on the parents part may somehow be reective of the childs symptoms and a possible avenue for developing measures to help with the early identication of these children. Despite this delay in parents concerns relative to the early markers identied retrospectively, studies suggest that parental concerns regarding their childs development were typically expressed to pediatricians by the age of 1.5 years, and yet, a denitive diagnosis of autism is not made until approximately 4 years old. [De Giacomo and Fombonne, 1998; Siegel et al., 1988; Flannagan and Nuallain,
FOR

2001]. Simultaneously, the evidence suggesting that young children as young as 24 months old can be diagnosed with autism continues to grow [Dahlgren and Gillberg, 1989; Lord, 1995; Stone et al., 1999; Chakrabati and Fombonne, 2001; Zwaigenbaum, 2001; Charman and Baird, 2002; Dawson et al., 2002]. Furthermore; the neurobiological abnormalities noted in the child with ASD in the rst two years of life [Courchesne et al., 2004] support the idea that as early as 1 year of age, children with autism may show signs of developing quite differently from their typically developing peers. Although discussions continue as to whether and why autism seems to be on the upswing, awareness has certainly increased [Fombonne, 2003]. Several studies have demonstrated that early detection and early intervention do have a positive impact on outcomes for children with autistic spectrum disorders [Hoyson et al., 1984; Lovaas, 1987; Rogers and Lewis, 1989; Harris et al., 1991; Birnbrauer and Leach, 1993; McEachin et al., 1993; Lord, 1995; Dawson and Osterling, 1997; Smith et al., 1997; Jocelyn et al., 1998; Sheinkopf and Siegel, 1998; Smith and Lovaas, 1998; Harris and Handleman, 2000; Sonnander, 2000]. In fact research suggests that intervention before three or three and a half years of age has the greatest impact. [Harris and Handleman, 2000; Woods and Wetherby, 2003]. The ndings that early identication and intervention lead to improved outcome suggest that autism is an appropriate disorder for which screening should be undertaken. This realization has prompted a focus on the development of autism-specic screening tools appropriate for use with young children. Several issues regarding the screening for autism in young children need further clarication. What type of screening is best: should the screening be broad-based developmental screening, or autism-specic? What is the best age at which to screen? What is the best method for screening: should parent report be used, or is direct observation a better approach? WHAT TYPE OF SCREENING IS BEST? Since research suggests that the rst indication of a problem in children who are subsequently diagnosed with an ASD may occur as early as the rst year of life [Osterling and Dawson, 1994; Baranek, 1999; Werner et al., 2000; Osterling et al., 2002] and that in the U.S., early intervention is not only an accepted but

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN

mandated means of addressing the needs identied for children from birth to the age of three, screening all children once or twice during that timeframe would facilitate the timely identication of children who need early intervention and/or monitoring. There are generally two models used to screen for autism. The rst model is consistent with practice parameters endorsed by the American Academy of Pediatrics (AAP), as well as the American Academy of Neurology (AAN) [Filipek et al., 1999; Filipek et al., 2000AAP, 2001a, 2001b, 2001c]. It consists of ongoing developmental surveillance of children within the primary care setting. In Dworkin [1993], developmental surveillance is dened as:
a exible, continuous process whereby knowledgeable professionals perform skilled observations of children during the provision of health care. The components of developmental surveillance include eliciting and attending to parental concerns, obtaining a relevant developmental history, making accurate and informative observations of children, and sharing opinions and concerns with other relevant professionals (p. 533).

It is suggested that standardized general developmental screening tools be used as part of the developmental surveillance process. These tools can either be a parent-completed questionnaire or clinician-completed measure. If general screening raises concerns suggestive of possible autistic spectrum disorders, autism-specic screening could then be administered. The challenge of providerimplemented screening was highlighted by a recent study, which found that half of the surveyed physicians used a formal screening tool (Sices et al., 2003). However, a primary benet of using this model to screen for autism is that it incorporates autism screening into an already accepted practice within the eld of pediatrics. As such, autism-specic screening is less likely to represent an additional item to be included in the already overburdened well child care (WCC) visit. Another benet is that it allows the screening to be undertaken within the context of the childs medical home. The term medical home was rst used in 1967 [Sia et al., 2004]. The American Academy of Pediatrics (AAP) in a policy statement provided 7 dimensions (accessible, continuous, comprehensive, family-centered, compassionate, culturally effective, and coordinated with specialized services provided outside of the primary care setting) that should be found within a medical home [AAP, 2002; Bethell et al., 2004]. As a result of
MRDD RESEARCH REVIEWS

this type of care, the child is viewed as the multi-faceted being that he or she is. The fact that any identied concerns will be raised by a provider with whom the family and child already have a relationship may lend additional credibility to the screening process and ndings. There are, however, potential limitations with this model. It depends on the primary care provider performing developmental surveillance/screening, recognizing the red ags for autism, and pursuing the appropriate next steps in a timely manner. This possibility is especially problematic as the frequency of well-child care visits begins to decrease in the toddler years. Also, a potential challenge with this approach is the limited amount of time available to most providers for accomplishing a growing number of tasks during the WCC visit. Perhaps due in part to this, in some instances, the dimensions of the medical home are not fully implemented [Strickland et al., 2004]. Although there are currently no validated autism-specic screening instruments designed for children less than 18 months old, several groups are working to better understand how autism presents in that age group in a prospective manner. Zwaigenbaum et al. [2005] recently published data from their ongoing prospective study with siblings of children diagnosed with an ASD. They have designed an observational scale for assessing autism-specic behavior in children. This is a promising line of research whose ndings may lead to validated measures for screening and diagnosing children less than 18 months of age, or even less than 12 months of age. At the present time, providers should use validated general developmental surveillance/screening with their patients under 18 months old. The second model consists of the routine administration of autism-specic screens at multiple high-risk ages (e.g., 18 and 24 months of age) by primary care providers, regardless of the presence of symptoms or concerns suggestive of an ASD. This would be done in addition to developmental surveillance, including the use of general developmental screening. The primary benet of screening at several ages is that the likelihood of not recognizing a child with possible ASD is diminished. There are three main limitations to this approach. First, it would require already busy primary care providers to remember to administer an autismspecic screening instrument to all children, even those who they feel are developing appropriately. Providers may hesitate to raise the possibility of a disorFOR

der such as autism with every parent they encounter at a preset interval due to concerns that the mere administration of these screens will be anxiety provoking for some parents. Secondly, the question arises as to which instrument primary care providers should use. Screening instruments for children younger than 16 18 months have not yet been validated, although multiple research groups are working to identify reliable signs in infancy. Multiple versions of a screener would be needed for children of different ages. Lastly, as the number of screening instruments for different conditions increase, primary care providers may be unable to t them all into the limited time available for the well-child care visit. Research suggests that the age at diagnosis seems to be inuenced by socioeconomic status and/or race/ethnicity [Mandell et al., 2002]. Furthermore, missed and misdiagnosed ASDs seem to be more prevalent in certain populations [Mandell et al., 2002; Dyches et al., 2004; Pinto-Martin and Levy, 2004]. The cause for this is not yet known, but may be linked to several factors related to racial and ethnic disparities in the quality of health care [Smedley et al., 2003], including screening and referral practices. There is also the possibility that the symptoms of autism may differ across populations, not only in terms of which symptoms are present and the timing of presentation, but also which symptoms concern parents, and whether they share these concerns with pediatric providers. Providers need to ensure that all children in their care are being screened. Towards that end, whichever model is used to screen for autism should include predetermined steps. Empirical data is needed to determine which model is most effective. It is possible that different models will work best in different patient and provider populations, and at different points in development. WHAT IS THE BEST AGE TO SCREEN? There are both advantages and disadvantages to early screening. On one hand, earlier screening may yield a diagnosis and subsequent intervention at an earlier age, allowing the maximum time for intervention. On the other hand, autism-specic screening at less than 16 18 months old has not yet been validated. Additionally, children who regress or those with Aspergers syndrome and high functioning autism may be missed by early screening, as they often present

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN 255

later. Furthermore, health care providers may be reluctant to screen for autism at younger ages because of concerns that such screening will lead to parental anxiety. Perhaps contrary to this idea is the nding that signicant numbers of parents have concerns about how their young child is developing, but if they are not asked, they may not share these concerns with their childs primary care provider [Young et al., 1998; King and Glascoe, 2003; Bethell et al., 2004]. A core principle embedded in the concepts of developmental surveillance, the medical home, and early identication is eliciting and responding to parental concerns. Studies have demonstrated the importance of parental report in the early detection of developmental problems [Stone and Lemanek, 1990; Stone et al., 1994; Glascoe and Dworkin, 1995; Young et al., 2003]. It is important that providers respond to parental concerns and yet, a recent study found that expression of parental concerns did not lead to increased likelihood of the child being referred for the purpose of diagnosis or service provision [Sices et al., 2004]. Screening at later ages, such as screening after 18 24 months, may lead to missed opportunities for early intervention. However, this later screening will likely have better specicity and sensitivity. In addition, it may capture the children who develop autistic features at later ages, those who regress, or are higher functioning. Also, health care providers may feel more comfortable with the idea of screening older children. Another challenge to early diagnosis is the possibility that the provider may be concerned when the parents are not. Since participation in early intervention is voluntary, unless the parents agree to receive the services, the child will not have the opportunity to receive them. Some parents may even refuse to allow their child to undergo the evaluative process. In such instances, repeated screenings may increase the likelihood that the parents will be convinced that the concerns being identied are worth further investigation, and screening at an older age may be less likely to encounter parental denial of concern. In deciding on what age is most appropriate for autism-specic screening, it is important to remember that different items may be needed for screening at different ages. For example, some studies suggest that toddlers diagnosed with autism and those with developmental delays both have impaired pretend play [BaronCohen et al., 1996; Charman et al., 1998; Cox et al., 1999]. Yet, when preschool 256 MRDD RESEARCH REVIEWS

aged children with autism are compared with those with developmental delays, children diagnosed with autism are found to exhibit less pretend play skills than those with developmental disabilities [Lord et al., 1994; Cox et al., 1999; Noterdaeme et al., 2000]. To further complicate the issue, many typically developing young toddlers may show behaviors consistent with those seen in ASD, such as repetitive behaviors (e.g., opening and closing drawers and doors, turning lights on and off), repetitive motor behaviors such as hand apping when excited, and insistence on routines (and indeed a certain sense of consistency and routine is crucial to minimizing frustration in that age group). WHAT IS THE BEST METHOD FOR SCREENING? Several factors must be considered in contemplating the best approach to autism screening. These include the availability, cost, ease of administration of instruments with acceptable sensitivity/ specicity, timing of well-child care visits for children, the amount of time available for these visits, the general approach to developmental screening and use of nonphysician personnel (e.g., nurse practitioners, physical assistants, home visitors), cost-effectiveness, reimbursement, and the cultural expectations of the population served regarding screening and diagnosis both in terms of the process and the content. Two methods often used in developmental screening are parent reports and observations by trained clinicians; each has advantages and drawbacks. Parental report is encouraged by many primarily because parents know their children best. They spend the most time with them in a variety of settings. This, therefore, provides them with many opportunities to observe the child across multiple settings, interacting with various individuals while they are in a variety of moods. This method, then, does not rely on a brief and what may be atypical sample of behaviors, and allows for observation of behaviors that may not be observable in most primary care ofces, such as interest in peers. Well-developed parent checklists are easy to administer, and in the short term may be the most cost effective option available to primary pediatric providers. Developmental surveillance, which includes clinical judgment, is aided by the systematic elicitation of parental concerns across developmental domains, using validated parent questionFOR

naires.[Glascoe et al., 1989; Glascoe, 1994; Glascoe and Dworkin, 1995; Glascoe and Sandler, 1995; Glascoe, 1997]. Finally, parental concerns regarding the emotional, behavioral, cognitive, ne motor, and language development of their children are predictive of the existence of a true problem [Glascoe and Dworkin, 1995]. Efforts to increase public awareness of developmental disorders such as autism include the Centers for Disease Control and Prevention (CDC)s Learn the Signs campaign (www.cdc.gov/actearly; 1 800CDC-INFO), as well as a comprehensive parent-developed site (www.rstsigns.org). These campaigns are aimed at educating parents about child development and the warning signs of developmental disorders such as autism. As part of the campaign, both the CDC and First Signs have developed kits with information about screening and early development for pediatric providers. It is always possible that parents of typically developing children may unnecessarily become concerned as a result of public awareness campaigns such as Learn the Signs. In such instances, reassurance and parent education may be all that is required. More research is needed to ascertain whether the nature of the concerns raised by parents of children found to have developmental disorders differ from those raised by the parents of typically developing children. The advantages of professional observation include knowledge of the full range of what constitutes typical development. Clinicians are more objective than parents and may be less likely to over- or under-estimate a given childs skills or problem behaviors. The signicance of a given behavior may not be as clear to a parent as it might be to a trained observer. Finally, in many instances, the setting used for undertaking the observations can be standardized. GENERAL DEVELOPMENTAL SCREENING TOOLS There are many broad developmental screening tools that may have a role in the early identication process. Discussion of these many tools is beyond the scope of this article, and the reader is referred to a recent review article on the subject [Glascoe, 2000]. This review focuses briey on two of the tools widely used by pediatric providers and one that focuses on communication and symbolic behavior. The three measures are the Parents Evaluation of Developmental Status (PEDS), the Ages and Stages Questionnaire (ASQ), and the Communication and Symbolic Behavior Scales

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN

Developmental Prole (CSBS DP). As screens of general development some of these tools may not target the core signs and symptoms of autism, for example, the PEDS does not include questions about play and imitation. The PEDS is a ten item instrument designed to be self-administered by parents of children from birth to 8 years of age, and elicits parent concern in developmental domains such as ne and gross motor skills, receptive and expressive language, and self-help skills. The type of parental concerns leads to categorization of risk for developmental difculties: high, moderate, and low with corresponding responses such as refer, reassure, and offer developmental promotion and parent education. The PEDS has been standardized and validated and meets the recommended psychometric properties for a screening instrument (sensitivity and specicity greater than 0.70) [Glascoe, 2001; Glascoe, 2003; www.pedstest. com]. The Ages and Stages Questionnaire (ASQ) is developed for use with children 4 60 months old, parents are asked to respond to descriptions of a list of skills with the response options: not yet, sometimes, or yes [Bricker and Squires, 1999]. Each form is intended for use with a certain age group (e.g., 4 years old) and consists of thirty developmental items divided into ve areas: communication, gross motor, ne motor, problem solving, and personal-social. There is also a section for general parental concerns. Many of the items are accompanied by illustrations, and the reading level is sixth grade or below. The items were selected based on the likelihood of being observed or elicited by parents in the home setting. The ASQ reportedly has a range of classication agreement from 76 to 91% [Bricker and Squires, 1999], with developmental assessment tools such as the Bayley Scales of Infant Development [Bayley, 1993]. The Communication and Symbolic Behavior Scales Developmental Prole (CSBS DP, Wetherby and Prizant, 2002] includes three measures: the Infant-Toddler checklist, an expanded Caregiver Questionnaire, and a Behavior Sample. The CSBS DP is based on a tool previously designed by the same authors the CSBS [Wetherby and Prizant, 1993]. The CSBS DP is a screening and evaluation instrument designed to measure the communicative and symbolic abilities of children aged 1224 months. The measured skills form three composites: social (emotion, eye gaze, and communication), speech (sounds and words), and
MRDD RESEARCH REVIEWS

symbolic (understanding and object use) [Wetherby et al., 2004]. The 24 item Infant-Toddler checklist asks about developmental milestones (e.g., when you look at and point to a toy across the room; does your child look at it?; does your child point to objects?). Nineteen of the twenty four items have the answer choices: not yet, sometimes, and often, the remaining ve are how many questions. At the bottom of the form parents are also asked a yes/no question: do you have any concerns about your childs development? They are asked to describe any existing concerns they may have. It is designed to be completed by the parent or caregiver. The Behavior Sample is a videotaped evaluation of the child during an interaction with his/her parent and the clinician. Wetherby et al., [2004] propose the use of the Infant-Toddler Checklist as a rst level screen, with the Behavior Sample serving as a second level evaluation tool. The tool was not designed to be an autism-specic screening tool, but rather a tool to screen for delays in communication, including prelinguistic skills; as such, it may be sensitive to a range of social communicative delays, including autistic spectrum disorder. The PEDS, ASQ, and CSBS DP are not designed to selectively screen for autism, but they may be effective in detecting children whose developmental problems are consistent with autism. Screening a large sample of children from an unselected population will be needed to demonstrate their usefulness as primary screens for children with autism. SCREENING TOOLS FOR AUTISTIC SPECTRUM DISORDERS (ASD) There are several autism-specic tools currently under development, or revision, and as such it would be premature to recommend one. Many of these promising tools still need to undertake continued follow up of the original samples used in their development, as well as cross-validation with new samples. In addition to the Modied Checklist for Autism in Toddlers (MCHAT), which will be discussed in more detail later, four other autism-specic screening tools are the Checklist for Autism in Toddlers (CHAT), Pervasive Developmental Disorders Screening Test-II (PDDST-II), Screening Tool for Autism in Two year olds (STAT), and Checklist for Autism in Toddlers-23 (CHAT-23). CHAT The CHAT is a screening tool intended for use in the general population
FOR

at 18 months of age. It consists of nine parent report items (A19) and ve observation items (Bi-v) [Baron-Cohen et al., 2000; 1996; 1992]. When looked at together ve of the fourteen items have been found to be key indicators of possible autism based on the Diagnostic and Statistical Manual (DSM) criteria for autistic disorder: gaze monitoring (Bii); proto-declarative pointing (pointing to indicate interest rather than to request) (A7 Biv), and pretend play (A5 Biii). Baron-Cohen et al., [1996] found that if at 18 months the items gaze monitoring, proto-declarative pointing, and pretend play were failed, there was an 83.3% risk of being subsequently diagnosed with autism via a diagnostic evaluation. A follow up study [Baird et al., 2000] with 16,235 eighteen month old children using the CHAT found, however, that although the specicity of the CHAT was excellent, the sensitivity (20 38% depending on criteria used) was unacceptably low. This suggests that the CHAT may not be an appropriate tool to use as an exclusive screening tool for identifying children who may have an ASD, and is currently under revision by the authors. PDDST-II PDD ST-II [Siegel, 2004] consists of 3 stages designed to be used in three different clinical settings. The PDDST Stage 1 consists of 22 items with the response choices yes, usually true or no, usually not true. Its intended use is in the primary care setting with 12 48 month olds. A positive screen signals the need for further evaluation and is dened as 5 or more items being answered yes, usually true. The author reports a sensitivity of 0.92 and 0.91 based on a sample of 681 children at risk for ASD and 256 children with mild-to-moderate other developmental disorders. The other two stages of the PDDST are intended to be used in settings other than the primary care providers ofce. Stage 2 is a 14-item screening tool for use in developmental clinics. The cut-off is still 5 items with a reported sensitivity and specicity of 0.73 and 0.49 based on 490 children with conrmed ASD (Autism, PDD-NOS, or Aspergers syndrome) and 194 children who were evaluated for an ASD, but who did not receive a diagnosis on the autistic spectrum [Siegel, 2004]. Stage 3 is a 12 item screening tool designed to be administered in autism clinics with a cut-off score of 8. The reported sensitivity and specicity is 0.58 and 0.60, respectively [Siegel, 2004]. The sensitivity and spec

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN 257

icity reported for stage 1 are good. Those for stage 2 and 3 are not as good, particularly, the specicity for stage 2. The major statistical shortcoming of this instrument is that no sensitivity/specicity data are reported for large-scale screening of an unselected sample. STAT The STAT [Stone et al., 2000; Stone et al., 2004] is designed as a second level screening tool for autism. It is intended to be used by professionals in communities to identify children with possible autism, as opposed to other developmental disabilities. The items are administered within a twenty-minute long play-based interactive session. The ten scored items include play (2), imitation (4), directing attention (4), and do not require language comprehension. Two additional requesting items are included as a means of promoting interaction between the evaluator and child. Two samples were used to assess the validity of the STAT as a second level screening instrument. The sample used in the development phase consisted of 40 children (3 with autism, 33 without), the sample used for validation had 33 children (12 with autism, 21 without autism). The children in the two samples ranged in age from 24 to 35 months old. The children with autism had lower developmental ages than the children without autism. Data from a subsample of children with (12) and without (12) autism matched for developmental age was also analyzed. The two examiners administering the STAT were blind to the referral questions and the results of the diagnostic assessments. The parents of the participating children were blind to the results of the diagnostic evaluation until after the STAT screening was completed. The diagnoses were made based on DSM-IV criteria (APA, 1994) and the Childhood Autism Rating Scale (CARS; Schopler et al., 1988]. The sensitivity and specicity of the subsample matched for DA were each 0.83. The sensitivity and specicity for the validation sample were 0.83 and 0.86, respectively. The authors of the STAT highlight three areas needing additional research. Firstly, replication studies with larger samples from many different settings are needed before the ndings can be generalized. Secondly, future studies need to use standardized measures such as the Autism Diagnostic Observation Schedule Generic (ADOS-G) [Lord et al., 1999], in addition to clinical judgment. Finally, follow up studies with larger sample sizes need to match those children diagnosed 258 MRDD RESEARCH REVIEWS

with autism and those without for mental ages and assess the sensitivity and specicity of the STAT in that instance. Nevertheless, the STAT is a promising Level 2 screening tool for autism in children between two and three years old. The M-CHAT The M-CHAT [Robins et al., 2001] is a 23-item (yes/no) parent report checklist developed to screen children aged 16 30 months. As suggested by the name; the M-CHAT is a modication of CHAT [Baron-Cohen et al., 1992], described above. The M-CHAT expands upon the parent report portion of the CHAT, and seeks to identify children with a possible autistic spectrum disorder, including PDD-NOS, and not just strictly dened Autistic Disorder. The initial population screened by the M-CHAT consisted of 1,293 children aged 16 30 months [Robins et al., 2001]. Of the 1293 children screened 58 were evaluated and 39 were diagnosed with an autistic spectrum disorder. None of the children evaluated at that time were found to be typically developing. Although absolute sensitivity and specicity are pending completion of follow up of the initial sample, based on the discriminant function analysis (DFA) classication at the time of rst screening and evaluation, the M-CHAT has a sensitivity of 0.87, specicity of 0.99, positive predictive power of 0.80, and negative predictive power of 0.99 [Robins et al., 2001]. Excluded from the ongoing validation study are children who already have an ASD diagnosis prior to completion of the M-CHAT, and children younger than 16 months or older than 30 months at the time of M-CHAT completion. The sample is drawn from primary care practices, as well as early intervention sites. Children who fail the M-CHAT receive a conrmatory follow up telephone call at which time the completed M-CHAT is reviewed to clarify the selected answers and ascertain whether the child has in fact screened positive on the checklist. The completed forms are scored and the child is considered to have failed the initial screening if he/she fails any three of the twenty three items or two of the six critical items. The critical items on the M-CHAT are as follows: item 2 (interest in other children), item 7 (proto-declarative pointing), item 9 (bringing objects to show the parent), item 13 (imitating), item 14 (responding to name), and item 15 (following a point).
FOR

Those who fail in both the initial screening and the telephone follow up are evaluated by a team of evaluators, using the Autism Diagnostic InterviewRevised (ADI-R) [Lord et al., 1994]; ADOS-G [Lord et al., 2000; 1999]; CARS [Schopler et al., 1988]; Mullen Scales of Early Learning [Mullen, 1995]; Vineland Adaptive Behavior Scales [Sparrow et al., 1984], and the DSM-IV criteria [APA, 1994]. The ADOS is a semistructured instrument, which allows assessment of a childs communication and social interactions. The ADI is a parent report measure with questions on the childs communication, social relatedness, including play and interests and behaviors. The ADI and ADOS are currently considered the gold standard for diagnosing autism. The number screened at Time 1 (completed form between 16 and 30 months old) has been increased to 4,200 children. 236 children who failed the initial screening have been evaluated at a mean age of 27.63 months: 165 were found to have an ASD, 67 have been diagnosed with a non-ASD developmental disorder, and 4 have been assessed as typically developing. To date 1,937 children evaluated at Time 1 have reached eligibility for rescreening at age 4. Data has been collected from 940 (mean age 55.35 months). At rescreening, parents are asked to complete the M-CHAT, and are also asked whether in the interim the child has been referred for or diagnosed with an ASD. Six children have been identied as possible misses (failed the rescreening or referred for possible ASD): 2 of the 6 have been conrmed as misses, two are not misses, and two are classied as possible misses because they have not been able to be evaluated. Sixty three children evaluated at Time 1 have been reevaluated at Time 2. At Time 2, 38 of the 63 children have been diagnosed with an ASD, 17 have been diagnosed with a non-ASD developmental disorder, and 8 have been assessed as typically developing. Based on the M-CHAT score at Time 1 and the evaluation outcome at Time 2, the sensitivity and specicity are 0.85 and 0.93, respectively. (The sensitivity is as high as 0.95 if the possible misses are not included as misses). It is possible that the psychometric properties of the M-CHAT will be different when larger numbers from an unselected population are obtained. Although this is still a work in progress, based on the data collected to date, certain items appear to be helpful in distinguishing between children with ASD as compared to those with a non

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN

ASD developmental disorder [Dixon et al., in press]. In particular, item 7 point for interest seem to be the most potent discriminator between the two groups of children. The diagnoses made at 2 years old also appear to be relatively stable based on the available follow up data. Although in a few cases, children moved from a non-ASD diagnosis to the typical development category, most of the changes in diagnosis were from an ASD to a nonASD developmental disorder. Whether this change in diagnosis is due to misdiagnosis at Time 1 or is the result of early intervention and/or increasing maturity remains unclear at this time. Part of the Early Detection study is the screening of the siblings of children with ASDs. Recent analyses of data collected from younger siblings of children diagnosed with an ASD found an initial screen positive rate (before phone interview) of 43%. As the M-CHAT is a parent-report measure, it is possible that these gures are partially inated because of increased parental concern as a result of the previous ASD diagnosis of the older sibling. After the phone interview, 30% remained a positive screen. Twenty three percent of the younger siblings were evaluated and diagnosed with an ASD. This is a higher rate than what has been reported in the literature for siblings, which may be due to an ascertainment bias, the parents of the siblings in this study may have sought participation in the study as a result of concerning signs that they observed. The study also involves screening of children within low risk settings (primary care provider sites, Peds) and high risk settings (early intervention sites, EI). Based on 20 children screened through pediatric ofces, the items that discriminate the children with Peds-referred ASD from the EI-referred ASD appear to be highly similar. CHAT-23 Wong et al. [2004] used the MCHAT and CHAT to develop a screening tool for use with a sample of Chinese children, the CHAT-23. The changes made to the original M-CHAT were the translation of the questions into traditional chinese and the addition of a graded scoring system [never (0%), seldom (25%), usually (2550%), or often (50%)] for 22 of the 23 M-CHAT questions instead of the yes/no responses of the original M-CHAT. The graded system was subsequently collapsed into yes (usually/often) and no (never/seldom) grouped to dene fail/pass. In adMRDD RESEARCH REVIEWS

dition to the 23 item-questionnaire, the CHAT-23 has an observational section consisting of 5 direct observational items from the CHAT. When used with 212 children (87 with ASD and 125 nonASD) with the mental age range of 18 24 months, Wong et al. found that failing any 2 of 7 key questions (imitate, pretend, point for interest, check reaction of parent, bring objects to show, follow a point, and take interest in other children) or any 6 of the overall 23 questions in the parent administered questionnaire yielded a sensitivity of 0.931 and 0.839, respectively. The specicity was of 0.768 or 0.848, respectively. The authors suggest a two-level screening system beginning with the M-CHAT and proceeding to the use of the observational portion of the CHAT for those who failed the M-CHAT. If the observation is failed, the child would then be referred for an autism evaluation. The fact that the psychometric properties being found when the CHAT-23 is used with a sample of Chinese children so closely resemble what is being found with the M-CHAT supports the validity of these instruments. In Wong et al.s study, the chronologic ages of the enrolled children whose mental ages were 18 24 months old were 16 86 months old. It is important to note that when used with children whose mental ages are not known to be 18 24 months, the efcacy of the tool may be different. In addition, this range of chronologic ages is quite wide, future research should examine whether the screener is equally valid across this age range. CONCLUSION At this time, early intervention is the best response to Autistic Spectrum Disorders. Given this fact, the development of reliable tools for screening children under three years old is imperative. Early screening will lead to early diagnosis, early intervention, and the ultimate goalimproved outcomes. There remains the need for further research to elucidate many issues. When is the best age to screen? Are the diagnoses made in the toddler years stable? If the children subsequently improve, is it due to the effects of early intervention or was the early diagnosis incorrect? What is the best way to screen for autism? Is preliminary general developmental surveillance/screening best or is autism-specic screening needed for all children? Are there populations, such as the siblings of children with ASDs, for whom autism-specic tools are best?
FOR

Since many of the autism-specic tools for screening in young children are still under development, close attention needs to be paid to their psychometric properties in different SES and ethnocultural groups. The M-CHAT is one of a few tools, which show promise as a screening tool in different populations of unselected children. It has been translated into Turkish, Chinese, Japanese, and Spanish. It will be important to evaluate the ndings of studies using the M-CHAT in these different languages and cultures for similarities and differences in their ndings. Beyond the languages being used, different cultural and socio-economic groups may interpret the questions differently, yielding different results. Not only may the specic signs/ symptoms of autism vary across cultural groups, the meaning attributed to them by parents may as well. For example, the fact that a child does not point or make eye contact may be considered appropriate and acceptable if to point and make eye contact with adults is considered a sign of disrespect. Therefore, more work focused on understanding what concerns compatible with ASDs may be more universal than others is needed. Despite these unresolved issues, a few facts remain clear. First, primary care providers are the appropriate people to screen children between birth and the age of 5 years old, since they are the qualied professionals with whom parents have ongoing contact. Second, the well-child care visit provides only a limited amount of time for the screening of children, with each of the many disorders potentially requiring that their own screening tool and process be used. The tools that will be used by providers are likely to be those that accurately ag children for many developmental disorders. Third, in light of the data supporting the usefulness of eliciting parental concerns, any tool used will probably need to incorporate the parents perspectives. Fourth, data suggests that children of parents from minority racial/ethnic backgrounds regardless of SES and those with lower socio-economic status may not be receiving quality health care. This disparate care includes a lack of screening, evaluation/diagnosis, and service provision. This must be addressed promptly, since the literature supports the notion that the earlier the identication of and intervention for developmental disorders, the better the outcomes. ACKNOWLEDGMENTS We thank the Early Detection Team at the University of Connecticut (Mari

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN 259

anne Barton, Sarah Hodgson, James Green, Gail Marshia, Pam Dixon, Jamie Kleinman, Juhi Pandey, Hilary Boorstein, Emma Esser, Leandra Wilson, Michael Rosenthal, Saasha Sutera, and Alyssa Verbalis); our collaborators at the Yale University Child Study Center (Fred Volkmar, Ami Klin, Katarzyna Chawarska, Tammy Babbitz), the University of Washington (Geraldine Dawson, Karen Toth), and Diana Robins at Georgia State University. Most importantly, we thank the families and children who are participating in the ongoing Early Detection Study. Special thanks to the library staff at St. Francis Hospital in Hartford, CT for their ongoing assistance. REFERENCES
American Academy of Pediatrics. 2001a. Technical report: the pediatricians role in the diagnosis and management of autistic spectrum disorder in children. Pediatrics 107(5). American Academy of Pediatrics, Committee on Children with Disabilities. 2001b. Developmental surveillance and screening of infants and young children. Pediatrics 108:192196. American Academy of Pediatrics, Committee on Children with Disabilities. 2001c. The pediatricians role in the diagnosis and management of autistic spectrum disorder in children. Pediatrics 107:12211226. American Academy of Pediatrics, Medical Home Initiatives for Children with Special Needs Project Advisory Committee. 2002. The Medical home. Pediatrics 110:184 186. American Psychiatric Association (APA). 1994. Diagnostic and Statistical Manual of Mental disorders, 4th edition (DSM-IV). Washington, DC: American Psychiatric Association. Baird G, Charman T, Baron-Cohen S, et al. 2000. A screening instrument for autism at 18 months of age: a six-year follow-up study. J Am Acad Child Adolesc Psychiatry 39:694 702. Ballaban-Gil K, Tuchman R. 2000. Epilepsy and epileptiform EEG: association with autism and language disorders. Ment Retard Dev Disabil Res Rev 6:300 308. Baranek GT. 1999. Autism during infancy: a retrospective video analysis of sensory-motor and social behaviors. J Autism Dev Disord 29:213224. Barbaresi WJ, Katusic SK, Colligan RC, et al. 2005. The incidence of autism in Olmsted county, Minnesota, 1976 1997: results from a population-based study. Arch Pediatr Adolesc Med 159:37 44. Baron-Cohen S, Allen J, Gillberg C. 1992. Can autism be detected at 18 months? The needle, the haystack, and the CHAT. Br J Psychiatry 161:839 843. Baron-Cohen S, Cox A, Baird G, et al. 1996. Psychological markers in the detection of autism in infancy in a large population. Br J Psychiatry 168:158 163. Baron-Cohen S, Wheelwright S, Cox A, et al. 2000. Early identication of autism by the Checklist for Autism in Toddlers (CHAT). J R Soc Med 93:521525. Bayley N. 1993. The Bayley Scales of Infant Development, 2nd edition. San Antonio, TX: Harcourt Brace. Bertrand J, Mars A, Boyle C, et al. 2001. Prevalence of autism in a United States population:

the Brick Township, New Jersey, investigation. Pediatrics 108:11551161. Bethell C, Reuland CH, Halfon N, et al. 2004. Measuring the quality of preventive and developmental services for young children: national estimates and patterns of clinicians performance. Pediatrics 113:19731983. Birnbrauer JS, Leach DJ. 1993. The Murdoch early intervention program after 2 years. Behav Change 10: 6374. Bricker D, Squires J. 1999. Ages and Stages Questionnaires: A parent-completed childmonitoring system. 2nd ed. Baltimore: Paul H. Brookes Publishing Co. Campbell M, Locascio JJ, Choroco MC, et al. 1990. Stereotypies and tardive dyskinesia: abnormal movements in autistic children. Psychopharmacol Bull 26:260 266. CDCs Learn the Signs. Available at www.cdc.gov/ actearly; 1 800-CDC-INFO Chakrabarti S, Fombonne E. 2001. Pervasive developmental disorders in preschool children. JAMA 285:30933099. Chakrabarti S, Fombonne E. 2005. Pervasive developmental disorders in preschool children: conrmation of high prevalence. Am J Psychiatry 162:11331141. Charman T, Baird G. 2002. Practitioner review: diagnosis of autism spectrum disorder in 2and 3-year-old children. J Child Psychol Psychiatry 43:289 305. Charman T, Swettenham J. 2001. Repetitive behaviors and social-communicative impairments in autism: implications for developmental theory and diagnosis. In: Burack JA, Charman T, Yirmiya N, Zelazo PR, editors. The Development of Autism: Perspectives from Theory and Research. Mahwah, NJ: Erlbaum. p 325345. Charman T, Swettenham J, Baron-Cohen S, et al. 1998. An experimental investigation of socialcognitive abilities in infants with autism: clinical implications. Infant Ment Health J 192: 260 275. Courchesne E, Redcay E, Kennedy D. 2004. The autistic brain: birth through adulthood. Curr Opin Neurol 17:489 496. Cox A, Klein K, Charman T, et al. 1999. Autism spectrum disorders at 20 and 42 months of age: stability of clinical and ADI-R diagnosis. J Child Psychol Psychiatry 40:719 732. Dahlgren SO, Gillberg C. 1989. Symptoms in the rst two years of life: a preliminary population study of infantile autism. Eur Arch Psychiatry Neurol Sci 238:169 174. Dawson G, Osterling J. 1997. Early intervention in autism. In: Guralnick MJ, editor. The Effectiveness of Early Intervention. Baltimore, MD: Paul H. Brookes. p 307326. Dawson G, Meltzoff AN, Osterling J, et al. 1998. Children with autism fail to orient to naturally occurring social stimuli. J Autism Dev Disord 28:479 485. Dawson G, Munson J, Estes A, et al. 2002. Neurocognitive function and joint attention ability in young children with autism spectrum disorder versus developmental delay. Child Dev 73:345358. Dawson G, Toth K, Abbott R, et al. 2004. Early social attention impairments in autism: social orienting, joint attention, and attention to distress. Dev Psychol 40:271283. DeGiacomo A, Fombonne E. 1998. Parental recognition of developmental abnormalities in autism. Eur J Child Adolesc Psychiatr 7:131 136. Dixon P, Kleinman J, Pandey J, et al. (in press). Differentiating between autism spectrum disorders and other developmental disabilities in
FOR

children who failed a screening instrument for ASD. J Autism Dev Disord. Dworkin PH. 1993. Detection of behavioral, developmental and psychosocial problems in pediatric primary care practice. Curr Opin Pediatr 5:531536. Dyches TT, Wilder LK, Sudweeks RR, et al. 2004. Multicultural issues in autism. J Autism Dev Disord 34:211222. Filipek P, Accardo P, Ashwal S, et al. 2000. Practice parameter: screening and diagnosis of autism: a report of the quality standards committee of the American Academy of Neurology and the Child Neurology Society. Neurology 55:468 479. Filipek P, Accardo P, Baranek G, et al. 1999. The screening and diagnosis of autistic spectrum disorders. J Autism Dev Disord 29:439 484. First Signs. Available at www.rstsigns.org Flanagan O, Nuallain S. 2001. A study looking at the effectiveness of developmental screening in identifying learning disabilities in early childhood. Ir Med J 94:148 150. Fombonne E. 2003. Epidemiological surveys of autism and other pervasive developmental disorders: an update. J Autism Dev Disord 33:365382. Glascoe F. 1994. Its not what it seems. The relationship between parents concerns and children with global delays. Clin Pediatr 33:292 296. Glascoe F. 2001. Parents Evaluation of Developmental Status. Nashville, TN: Ellsworth and Vandermeer Press, LLC. www.pedstest.com Glascoe F. 2003. Parents evaluation of developmental status: how well do parents concerns identify children with behavioral and emotional problems? Clin Pediatr 42:133138. Glascoe FP. 1997. Parents concerns about childrens development: prescreening technique or screening test? Pediatrics 99:522528. Glascoe FP. 2000. Early detection of developmental and behavioral problems. Pediatr Rev 21:272280. Glascoe FP, Dworkin PH. 1995. The role of parents in the detection of developmental and behavioral problems. Pediatrics 95:829 836. Glascoe FP, Sandler H. 1995. Value of parents estimates of childrens developmental ages. Disabil Rehabil 127:831 835. Glascoe FP, Altemeier WA, MacLean WE. 1989. The importance of parents concerns about their childs development. Am J Dis Child 143:955958. Glaze D. 2004. Rett syndrome: of girls and mice lessons for regression in autism. Ment Retard Dev Disabil Res Rev 10:154 158. Goldberg WA, Osann K, Filipek PA, et al. 2003. Language and other regression: assessment and timing. J Autism Dev Disord 33:607 616. Harris S, Handleman J. 2000. Age and IQ at intake as predictors of placement for young children with autism: a four to six-year follow-up. J Autism Dev Disord 30:137142. Harris SL, Handleman JS, Gordon R, et al. 1991. Changes in cognitive and language functioning of preschool children with autism. J Autism Dev Disord 21:281290. Hoyson M, Jamisson B, Strain P. 1984. Individualized group instruction of normally developing and autistic-like children: the LEAP curriculum model. J Div Early Child 8:157172. Hrdlicka M, Komarek V, Propper L, et al. 2004. Not EEG abnormalities but epilepsy is associated with autistic regression and mental

260 MRDD RESEARCH REVIEWS

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN

functioning in childhood autism. Eur Child Adolesc Psychiatry 13:209 213. Jocelyn LJ, Casiro OG, Beattie D, et al. 1998. Treatment of children with autism: a randomized control trial to evaluate a caregiverbased intervention program in community day-care centers. J Dev Behav Pediatr 19: 326 334. King T, Glascoe F. 2003. Developmental surveillance of infants and young children in pediatric primary care. Curr Opin Pediatr 15: 624 629. Klin A, Jones W, Schultz R, et al. 2002. Dening and quantifying the social phenotype in autism. Am J Psychiatry 1596:895908. Klinger, Dawson & Renner. 2003. Autistic disorder. In: Mash EJ, Barkley RA, editors. Child Psychopathology. 2nd ed. New York: The Guilford Press. p 409 454. Lainhart JE, Ozonoff S, Coon H, et al. 2002. Autism, regression, and the broader autism phenotype. Am J Med Genet 113:231237. Lord C. 1995. Follow-up of two-year-olds referred for possible autism. J Child Adolesc Psychiatr 36:13651382. Lord C, Risi S, Lambrecht L, et al. 2000. The autism diagnostic observation schedule generic: a standard measure of social and communication decits associated with the spectrum of autism. J Autism Dev Disord 30:205 223. Lord C, Rutter M, DiLavore P, et al. 1999. Autism Diagnostic Observation Schedule-WPS edition. Los Angeles, CA: Western Psychological Services. Lord C, Rutter M, Le Couteur A. 1994. Autism diagnostic interview-revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord 24: 659 685. Lord C, Shulman C, DiLavore P. 2004. Regression and word loss in autistic spectrum disorders. J Child Psychol Psychiatr 45:936 955. Lovaas O. 1987. Behavioral treatment and normal educational and intellectual functioning in young autistic children. J Consult Clin Psychol 55:39. Malhotra S, Gupta N. 1999. Childhood disintegrative disorder. J Autism Dev Disord 29:491 498. Mandell DS, Listerud J, Levy SE, et al. 2002. Race differences in the age at diagnosis among Medicaid-eligible children with autism. J Am Acad Child Adolesc Psychiatr 41:14471453. Manning-Courtney P, Brown J, Molloy CA, et al. 2003. Diagnosis and treatment of autism spectrum disorders. Curr Probl Pediatr Adolesc Health Care 33:283304. McEachin J, Smith T, Lovaas O. 1993. Long-term outcome for children with autism who received early intensive behavioral treatment. Am J Ment Retard 97:359 372. McEvoy R, Rogers S, Pennington B. 1993. Executive function and social communication deficits in young autistic children. J Child Psychol Psychiatr 34:563578. Mullen E. 1995. Mullen Scales of Early Learning. Circle Pines, MN: American Guidance. Noterdaeme M, Sitter S, Mildenberger K, et al. 2000. Diagnostic assessment of communicative and interactive behaviours in children with autism and receptive language disorder. Eur Child Adolesc Psychiatr 9:295300. Osterling J, Dawson G. 1994. Early recognition of children with autism: a study of rst birthday home videotapes. J Autism Dev Disord 24: 247257.

Osterling J, Dawson G, Munson J. 2002. Early recognition of 1-year-old infants with autism spectrum disorder versus mental retardation. Dev Psychopathol 14:239 251. Pinto-Martin J, Levy S. 2004. Early diagnosis of autism spectrum disorders. Curr Treat Options Neurol 391 400. Robins DL, Fein D, Barton ML, et al. 2001. The modied checklist for autism in toddlers: an initial study investigating the early detection of autism and pervasive developmental disorders. J Autism Dev Disord 31:131144. Robinson R, Baird G, Robinson G, et al. 2001. Landau-Kleffner syndrome: course and correlates with outcome. Dev Med Child Neurol 43:243247. Rogers S. 2004. Developmental regression in autism spectrum disorders. Ment Retard Dev Disabil Res Rev 10:139 143. Rogers S, Lewis H. 1989. An effective day treatment model for young children with pervasive developmental disorders. J Am Acad Child Adolesc Psychiatr 28:207214. Schopler E, Reichler RJ, Renner BR. 1988. The Childhood Autism Rating Scale. Los Angeles: Western Psychological Services. Sheinkopf SH, Siegel B. 1998. Home-based behavioral treatment of young children with autism. J Dev Disord 28:1523. Sia C, Tonniges TF, Osterhus E, et al. 2004. History of the medical home concept. Pediatrics 113:14731478. Sices L, Feudtner C, McLaughlin J, et al. 2003. How do primary care physicians identify young children with developmental delays? A national survey. Dev Behav Pediatr 24:409 417. Sices L, Feudtner C, McLaughlin J, et al. 2004. How do primary care physicians manage children with possible developmental delays? A national survey with an experimental design. Pediatrics 113:274 281. Siegel B. 2004. Pervasive Developmental Disorders Screening Test II (PDDST-II): Early Childhood Screeners for Autistic Spectrum Disorders. Harcourt Assessment, Inc. Siegel B, Pliner C, Eschler J, et al. 1988. How children with autism are diagnosed: difculties in identication of children with multiple developmental delays. J Dev Behav Pediatr 9:199 204. Siperstein R, Volkmar F. 2004. Brief report: parental reporting of regression in children with pervasive developmental disorders. J Autism Dev Disord 34:731734. Smedley BD, Stith AY, Nelson AR. 2003. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. National Academies Press. Smith T, Eikeseth S, Klevstrand M, et al. 1997. Intensive behavioral treatment for preschoolers with severe mental retardation and mental retardation and pervasive developmental disorder. Am J Ment Retard 102:238 249. Smith T, Lovaas O. 1998. Intensive and early behavioral intervention with autism: the UCLA young autism project. Infants Young Child 10:6778. Sonnander K. 2000. Early identication of children with developmental disabilities. Acta Paediatr Suppl 89:1723. Sparrow S, Balla D, Cichetti D. 1984. The Vineland Adaptive Behavior Scales. Circles Pines, MN: American Guidance Services. Stefanatos G, Kinsbourne M, Wasserstein J. 2002. Acquired epileptiform aphasia: a dimensional view of Lansu-Kleffner syndrome and the relation to regressive autistic spectrum disoders.
FOR

Neuropsychol Dev Cogn C Child Neuropsychol 8:195228. Stone W, Coonrod E, Ousley O. 2000. Brief report: screening tool for autism in two-yearolds (STAT): development and preliminary data. J Autism Dev Disord 30:607 612. Stone WL, Coonrod EE, Turner LM, et al. 2004. Psychometric properties of the STAT for early autism screening. J Autism Dev Disord 34:691701. Stone W, Hoffman E, Lewis S, et al. 1994. Early recognition of autism: parental reports vs. clinical observation. Arch Pediatr Adolesc Med 148:74 179. Stone W, Lee E, Ashford L, et al. 1999. Can autism be diagnosed accurately in children under 3 years? J Child Psychol Psychiatry 40:219 226. Stone W, Lemanek K. 1990. Parental report of social behaviors in autistic preschoolers. J Autism Dev Disord 20:513522. Strickland B, McPherson M, Weissman G, et al. 2004. Access to the medical home: results of the National Survey of Children with special health care needs. Pediatrics 113:14851492. Szatmari P, Bartolucci G, Bremmer R. 1989. Aspergers syndrome and autism: comparison of early history and outcome. Dev Med Child Neurol 31:709 720. Travis L, Sigman M, Ruskin E. 2001. Links between social understanding and social behavior in verbally able children with autism. J Autism Dev Disord 31:119 130. Trevathan E. 2004. Seizures and epilepsy among children with language regression and autistic spectrum disorders. J Child Neurol 19:S49 S57. Tuchman R, Rapin I. 1997. Regression in pervasive developmental disorders: seizures and epileptiform electroencephalogram correlates. Pediatrics 99:560 566. Turner M. 1999. Annotation: repetitive behavior in autism: a review of psychological research. J Child Psychol Psychiatr 40:839 849. Volkmar F, Cohen D, Paul R. 1986. An evaluation of DSM-III criteria for infantile autism. J Am Acad Child Psychiatr 25:190 197. Volkmar F, Rutter M. 1995. Childhood disintegrative disorder: results of the DSM-IV autism FIELD trial. J Am Acad Child Adolesc Psychiatr 34:10921095. Volkmar FR, Lord C, Bailey A, et.al. 2004. Autism and pervasive developmental disorders. J Child Psychol Psychiatr 45(1): 135170. Werner E, Dawson G, Osterling J, et al. 2000. Brief report: recognition of autism spectrum disorder before one year of age: a retrospective study based on home videos. J Autism Dev Disord 30:157162. Wetherby A, Prizant B. 1993. The Communication and Symbolic Behavior Scale. New York, NY: Riverside. Wetherby A, Prizant B. 2002. Communication and Symbolic Behavior Scales Developmental ProleFirst Normed Edition. Baltimore, MD: Paul H. Brookes. Wetherby AM, Woods J, Allen L, et al. 2004. Early indicators of autism spectrum disorders in the second year of life. J Autism Dev Disord 34:473 493. Wong V, Hui L, Lee W, et al. 2004. A Modied Screening Tool for Autism (Checklist for Autism in Toddlers [CHAT-23] for Chinese children. Pediatrics 114(2):e166 e176. Woods J, Wetherby A. 2003. Early identication and intervention for infants and toddlers who are at risk for autism spectrum disorder. Language Speech Hearing Services Schools 34: 180 193.

MRDD RESEARCH REVIEWS

SCREENING

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN 261

Yeargin-Allsopp M, Rice C, Karapurkar T, et al. 2003. Prevalence of autism in a US metropolitan area. JAMA. 289(1):49 55. Young KT, Davis K, Schoen C, et al. 1998. Listening to parents. A national survey of parents

with young children. Arch Pediatr Adolesc Med 152:255262. Young R, Brewer N, Pattison C. 2003. Parental identication of early behavioural abnormalities in children with autistic disorder. Autism 2003 7:125143.

Zwaigenbaum L. 2001. Autistic spectrum disorders in preschool children. Can Fam Physician 47: 20372042. Zwaigenbaum L, Bryson S, Rogers T, et al. 2005. Behavioral manifestations of autism in the rst year of life. Int J Dev Neurosci 23:143152.

APPENDIX: M-CHAT* Please ll out the following about how your child usually is. Please try to answer every question. If the behavior is rare (e.g., you have seen it once or twice), please answer as if the child does not do it.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Does your child enjoy being swung, bounced on your knee, etc.? Does your child take an interest in other children? Does your child like climbing on things, such as up stairs? Does your child enjoy playing peek-a-boo/hide-and-seek? Does your child ever pretend, for example, to talk on the phone or take care of dolls, or pretend other things? Does your child ever use his/her index nger to point, to ask for something? Does your child ever use his/her index nger to point, to indicate interest in something? Can your child play properly with small toys (e.g., cars or bricks) without just mouthing, ddling, or dropping them? Does your child ever bring objects over to you (parent) to show you something? Does your child look you in the eye for more than a second or two? Does your child ever seem oversensitive to noise? (e.g., plugging ears) Does your child smile in response to your face or your smile? Does your child imitate you? (e.g., you make a facewill your child imitate it?) Does your child respond to his/her name when you call? If you point at a toy across the room, does your child look at it? Does your child walk? Does your child look at things you are looking at? Does your child make unusual nger movements near his/her face? Does your child try to attract your attention to his/her own activity? Have you ever wondered if your child is deaf? Does your child understand what people say? Does your child sometimes stare at nothing or wander with no purpose? Does your child look at your face to check your reaction when faced with something unfamiliar?

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

No No No No No No No No No No No No No No No No No No No No No No No

*1999 Diana Robins, Deborah Fein, and Marianne Barton

262 MRDD RESEARCH REVIEWS

SCREENING

FOR

AUTISM: THE M-CHAT

AND

OTHER MEASURES

DUMONT-MATHIEU & FEIN