Screening for Autism Spectrum 
Disorders in Toddlers Using the 
M-CHAT 

Lauren Herlihy, M.A. 

Early Detection of Pervasive Developmental Disorders Study 

Department of Psychology 

University of Connecticut, Storrs, CT 

Webinar 

Date 

University of Louisville 

 


Funding 

•Development and revision of the M-CHAT and 
M-CHAT-Revised (Deborah Fein, PI) is 
supported by the National Institute of Child 
Health and Human Development. 


C:\Users\Lauren\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\MWRWULN3\MP900423023[1].jpg

Early Detection Study- Collaborators 

Georgia State University 

 Diana Robins, Ph.D. (first author of M-CHAT) 

Yale University 

Tammy Babitz, M.A. 

Katarzyna Chawarska, Ph.D. 

Ami Klin, Ph.D. 

Fred Volkmar, M.D. 

National Institutes of Health – Tokyo 

 Yoko Kamio, M.D. 

University of Washington 

Geraldine Dawson, Ph.D. (now at Autism Speaks) 

Karen Toth, Ph.D. 

Wendy Stone, Ph.D. 

 


UConn Collaborators 

•Faculty: 
–Marianne Barton, Ph.D., 
Director, Psychological 
Services Clinic 
–Thyde Dumont-Mathieu, M.D. 
–James Green, Ph.D. 
–Sarah Hodgson, Ph.D. 
–Jamie Kleinman, Ph.D. 




 


•Graduate Students: 
–Laura Brennan 
–Katelin Carr 
–Colby Chlebowski 
–Alex Hinnebusch 
–Kelley Knoch 
–Eva Troyb 




 

•Research Assistants: 
–Katie DeYeo 
–Sarah Hardy 
–Courtney Manning 





Abbreviations 

•M-CHAT= Modified Checklist for Autism in 
Toddlers (Robins et al., 2001) 
•ASD= Autism spectrum disorder, including: 
–Autistic Disorder 
–Asperger’s syndrome 
–Pervasive Developmental Disorder-Not Otherwise 
Specified (PDD-NOS) 


•EI= Early Intervention 
•AAP= American Academy of Pediatrics 



Why routinely screen for ASDs? 

•Screening improves 
detection of 
children at risk 
•Detection leads to 
diagnosis 
•Diagnosis leads to 
services 
•Services lead to 
improved outcome 


C:\Users\Lauren\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\5Z8SHGFG\MP900433196[1].jpg

Screening improves detection of children 
at risk 

•The use of more informal, non-validated 
screening strategies leads to an unacceptably 
low sensitivity of 20-30%. 
•Screening strategies can include the use of 
specific red flags during observation, as well 
as validated parent-report screeners. 



Detection leads to diagnosis, but 
is early diagnosis valid? 

•Diagnosis of ASD can be reliably made before 
age 2 (our sample is 16-30 months) 
•Clinical judgment is still the best, but ADOS 
and CARS work well at that age 
•Most diagnostic movement between age 2 
and 4 is toward improvement 
•Children with provisional diagnosis at age 2 
because of severe delays overwhelmingly 
meet criteria for ASD at age 4 


 


Diagnosis leads to services 

•Most EI programs 
require a diagnosis 
before autism 
intervention is begun 
•Some EI programs do 
their own diagnosis; 
many do not 


C:\Users\Lauren\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\MWRWULN3\MP900438799[1].jpg

Services lead to improved outcome 

 Early detection and early intervention can lead 
to substantially better prognosis, including: 

–improved language, 
–improved social relationships, 
–Improved adaptive functioning, 
–fewer maladaptive behaviors, 
–increased chance for successful inclusion 
–loss of diagnosis 




(Harris & Handleman, 2000; Howard et al., 2005; Lord & McGee, 2001; Myers, Johnson et 
al, 2007; Rogers, 1998; Sallows & Graupner, 2005) 


Nationwide Screening Recommendations 

•AAP (Gupta et al., 2007 in Pediatrics) recommends 
autism-specific screening at 18 and 24 months, in 
addition to developmental surveillance and broad-
band screening at multiple ages. 
•The Council on Children with Disabilities (Johnson & 
Myers, 2007) recommends routine ASD surveillance at 
every well-child visit complemented by ASD-specific 
screening at the 18- and 24-month visits. 


 

•At present, this represents the approach most likely to 
detect children at risk for ASD at the earliest times 



When should screening be done? 

•Early screens (<18 months) 
–May be less sensitive & specific 
–Miss children who regress between 18-24 months, or 
with later onset 




 

•Later screens 
–Delay diagnosis and intervention 




 

•Current Consensus 
–Screen twice 




C:\Users\Lauren\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\5Z8SHGFG\MP900202020[1].jpg

Comparison of Screening Instruments 

JADD Screeners.jpg
(Barton, Dumont-Mathieu, & Fein, in press, JADD) 


Definitions 

•Sensitivity: % of cases that screen positive 
•Specificity: % of noncases that screen negative 
•Positive predictive value: % of screen positives that are 
cases 
•Negative predictive value: % of screen negatives that are 
noncases 
•Sensitivity of a measure is of greatest concern to the 
primary care provider. 
–To detect the maximum number of children with the disorder, 
threshold for identification may be set low 
–This will result in the identification of more children with the 
disorder (high sensitivity), as well as a significant number of 
false positives (low specificity). 





M-CHAT Procedures 

–Children ages 16-30 months screened at pediatrician’s 
office or by EI provider 
–Researchers score M-CHAT 
•Positive screen (2 critical or 3 total items failed) 






 

–Scripted phone follow-up for positive (failed) screens 




 

–Clinical evaluation for those who continue to screen 
positive after phone follow-up 




 

–Rescreen entire sample at 4 




 

–Evaluate at 4 
•All those evaluated at 2, 
•New screen positives at 4, 
• Any child referred for, or diagnosed with, a PDD (asked on the 
4-year-old screen) 







Modified Checklist for Autism in Toddlers 
(Robins, Fein, Barton, Green, 2001) 

•Adapted from the Checklist for Autism in Toddlers 
(CHAT; Baron-Cohen, Allen, & Gillberg, 1992). 
•23-item yes/no parent-response checklist 
•Follow-up scripted interview for failed items reduces 
false positives 
•Tested on 16-30 month olds 
•About 18,000 children screened in our study 
•Translated into multiple languages, used in many 
countries 
•Available at www.mchatscreen.com (42 translations, 
scoring, free download) 



Psychometric Properties of the 
M-CHAT 

•Strong internal consistency (Chronbach’s 
alpha= .87) 
•Positive predictive value greatly increases with 
the use of the Follow-up Interview™ 
•Rationale for Scoring Criteria: 
–Odds of a child being diagnosed with ASD increases 
exponentially along with the number of items failed 
on the M-CHAT 
–Cut-off of 3 items failed reflects the greatest increase 
in ASD diagnosis from 2 to 3 failed items 





PPV for UConn and GSU data 

•Low-risk sample of N= 9,088 from Georgia, N= 
9,901 from UConn 
•Screened at either 18 or 24 months 
•M-CHAT found to have strong internal 
consistency (Cronbach’s alpha = .87) 


-N= over 18,000 children screened at both site, at either 18 or 24 months 
-Strong internal consistency (Chronbach’s alpha= .87) 


(Chlebowski et al., in prep) 


Current Issues in M-CHAT Screening 

•Is screening twice (at 18 and 24 months) 
necessary? 
•Are the 18 and 24 month screens equally 
effective? 
•Are there disparities in ASD screening 
practices? 



M-CHAT Items 
(Robins et al., 2001) 

1.Does your child enjoy being swung, bounced on your knee, etc.? 
2.Does your child take an interest in other children? 
3.Does your child like climbing on things, such as up stairs? 
4.Does your child enjoy playing peek-a-boo/ hide-and-seek? 
5.Does your child ever pretend, for example, to talk on the phone or take care of a doll or other pretend 
things? 
6.Does your child ever use his index finger to point, to ask for something? 
7.Does your child ever use his index finger to point, to indicate interest in something? 
8.Can your child play properly with small toys without just mouthing, fiddling or dropping them? 
9.Does your child ever bring objects over to you to show you something? 
10.Does your child look you in the eye for more than a second or two? 
11.Does your child ever seem oversensitive to noise, e.g. plugging ears? 
12.Does your child smile in response to your face or smile? 
13.Does your child imitate you? (e.g. you make a face - will your child imitate it?) 
14.Does your child respond to his/her name when you call? 
15.If you point at a toy across the room, does your child look at it? 
16.Does your child walk? 
17.Does your child look at things you are looking at? 
18.Does your child make unusual finger movements near his/her face? 
19.Does your child try to attract your attention to his/her own activity? 
20.Have you ever wondered if your child is deaf? 
21.Does your child understand what people say? 
22.Does your child sometimes stare at nothing or wander with no purpose? 
23.Does your child look at your face to check your reaction when faced with something unfamiliar? 



M-CHAT Items as Predictors of ASD 

Best discriminators of ASD from total 
sample: 

 

•Pointing to show interest 
•Responds to name 
•Interest in other children 
•Bring to show 
•Follows point 
•Imitation 
•Reciprocal smile 



M-CHAT Follow-up Interview™ 

•Tool to be used in conjunction with the M-CHAT 
to increase confidence in referral for ASD-related 
concerns 
•Administered over the phone or in person to 
probe individual M-CHAT items in greater depth 
•Provides greater explanation of items and 
multiple examples 
•Requires some clinical judgment; administered by 
trained staff 


 



Current Issues in M-CHAT Screening 

•Is screening twice (at 18 and 24 months) 
necessary? 
•Are the 18 and 24 month screens equally 
effective? 
•Are there disparities in ASD screening 
practices? 



Is Sceening Twice Necessary? 


•Younger (16-23 mo.) vs. older (24-30 mo.), high- and low-
risk toddlers 
•Parents of younger/low-risk children were most likely to 
refuse evaluation 
•Older and younger toddlers had similar symptomatology 
and developmental delays 
•For high-risk groups, and older/low-risk group, predictive 
power for ASD is adequate 
•For younger/low-risk group, predictive power for 
detecting a developmental disorder is good, but with less 
specificity for ASD – may be picking up more mild, 
transient delays, or other disorders 


 

Pandey, et al., 2008, in Autism: The International Journal of Research and 
Practice) 


Are the 18 and 24 month screens equally 
effective? 

•Dual screening will have the highest 
sensitivity. 
•Between the two ages, the 24 month screen 
appears to be more valid for the M-CHAT and 
may meet with less resistance. 
•Will providers be willing to do the 24 month 
screen if the 18 month screen is negative, or 
yields many false positives? 



Are there disparities in screening for 
ASD? 

•Referral, diagnosis, and treatment may be late or reduced for: 
–Minority children, particularly Hispanic children (CDC, 2006; Mandell et al, 2002; 
Mandell and Novack, 2005) 




 (Evidence for Black children is more mixed) 

–Low SES children 
–Children with less severe symptoms and delays (Liptak et al, 2008; Mandell, 
Novack, Zubritsky, 2005) 




 

•There is more data on age of diagnosis than age of screening 


 

•Screening is less consistent and later for children who are: 
–Ethnic minorities or Low SES, compared to children from middle class 
and non-minority families (Mandell, et al., 2002; Pinto-Martin, et al., 2005; 
Williams & Brayne, 2006; Zwaigenbaum et al., 2007, but see Wiggins, Baio, & Rice, 
2006). 




 

•Use of formal screening instruments increases sensitivity (Sices et 
al, 2003) and may reduce ethnic and SES bias. 



Racial/ethnic and Linguistic Differences in 
Caregiver Responses on the M-CHAT 

Screen positive children in 
Connecticut 

•Minority vs. White 
•No difference in M-CHAT 
score by ethnicity 
•No sig. difference in PPV of 
total score 
•3 items differed on M-CHAT 
but not Follow-up 
Interview™ 
–Unusual finger movements 
–Interest in peers 
–Eye contact 




Screen positive children in 
Atlanta 

•African-American vs. White 
•4 items endorsed more often 
by African-American parents 
after controlling for mother’s 
education: 
–Mouthing objects 
–Unusual finger movements 
–Does not attract caregiver’s 
attention 
–Wanders with no purpose 


•Follow-up Interview™ 
eliminated ethnic disparity 


(Dancel et al., 2008, presented at INS) 

(Oliver, Robins, & Hazzard, 2009, presented 
at IMFAR) 
Mean age of screening in mos. by ethnicity 

18.5 

19 

19.5 
20 

20.5 
21 

21.5 
22 

22.5 
Asian or Pacific Islander


Biracial


Black or African American


Hispanic/Latino
Other


White
Mean age (mos) 

Herlihy et al., unpublished data 


Interval Between Screening and 
Follow-up by Ethnicity 

C:\Users\Lauren\Pictures\Figure.jpg
(Herlihy et al., 2011, presented at INS) 


Conclusions 

•Disparity in initial screening age persists 


 

•PPV for White and Minority children was not sig. different. 


 

•There are item discrepancies which are largely resolved by Follow-
up Interview™ 


 

•There are disparities in time of screening, even when participating 
in a research study with prescribed protocol 


 

•There are increasing disparities after a positive screen, esp. with 
Hispanic/Latino families 


 

•Overall performance of M-CHAT is similar between groups 


 


M-CHAT and M-CHAT Follow-up 
Interview™ Administration 


Administering the M-CHAT 
Screening Questionnaire 


M-CHAT Items 

1.Does your child enjoy being swung, bounced 
on your knee, etc.? 


 

2. Does your child take an interest in other 
children? 

 

3. Does your child like climbing on things, such 
as up stairs? 


M-CHAT Items 

4. Does your child enjoy playing peek-a-boo/ 
hide-and-seek? 

 

5. Does your child ever pretend, for example, to 
talk on the phone or take care of a doll or 
pretend other things? 

 


M-CHAT Items 

6. Does your child ever use his/her index finger to point, 
to ask for something? 

 

7. Does your child ever use his/her index finger to point, 
to indicate interest in something? 

 

8. Can your child play properly with small toys (e.g. cars 
or bricks) without just mouthing, fiddling or dropping 
them? 

 

 


M-CHAT Items 

9. Does your child ever bring objects over to you 
(parent) to show you something? 

 

10. Does your child look you in the eye for more 
than a second or two? 


M-CHAT Items 

11. Does your child ever seem oversensitive to noise, 
e.g. plugging ears? 

 - Reverse-scored item (typical response is no) 

 

12. Does your child smile in response to your face or 
your smile? 


M-CHAT Items 

13. Does your child imitate you? (e.g. you make 
a face - will your child imitate it?) 

 

14. Does your child respond to his/her name 
when you call? 


M-CHAT Items 

15. If you point at a toy 
across the room, does 
your child look at it? 

 

16. Does your child walk? 

 

17. Does your child look at 
things you are looking 
at? 

 

C:\Users\Lauren\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\NPF6ONMF\MP900408980[1].jpg

M-CHAT Items 

18. Does your child make unusual finger 
movements near his/her face? 

 - Reverse-scored item (typical response is no) 

 

19. Does your child try to attract your attention to 
his/her own activity? 

 

20. Have you ever wondered if your child is deaf? 

 - Reverse-scored item (typical response is no) 

 

 


M-CHAT Items 

 

21. Does your child understand what people say? 

 

22.Does your child sometimes stare at nothing or 
wander with no purpose? 

–Reverse-scored item (typical response is no) 




 

23. Does your child look at your face to check your 
reaction when faced with something unfamiliar? 

 


M-CHAT Critical Items 

•Shows interest in other children (Item 2) 
•Pointing to indicate interest (Item 7) 
•Bringing objects to show (Item 9) 
•Imitation (Item 13) 
•Responding to name (Item 14) 
•Follows a point (Item 15) 



Scoring the M-CHAT 

•Passing responses for each 
item: 
1.Yes 
2.Yes 
3.Yes 
4.Yes 
5.Yes 
6.Yes 
7.Yes 
8.Yes 
9.Yes 
10.Yes 
11.NO 
12.Yes 
13.Yes 
14.Yes 
15.Yes 
16.Yes 
17.Yes 
18.NO 
19.Yes 
20.NO 
21.Yes 
22.NO 
23.Yes