This paper appeared in the Journal of Developmental and Learning Disorders, 2000,

4, 3-30.

The early identification of autism:

the Checklist for Autism in Toddlers (CHAT)

Simon Baron-Cohen+, Sally Wheelwright+, Antony Cox*, Gillian Baird*,

Tony Charman@, John Swettenham**, Auriol Drew*, and Peter Doehring#.

*Newcomen Centre, UMDS, Guy's Hospital,

St. Thomas Street, London SE1 9RT, UK.

+ University of Cambridge, Departments of Experimental Psychology and

Psychiatry,
Downing Street, Cambridge, CB2 3EB, UK.

@ Behavioural Sciences Unit, Institute of Child Health, London University,

30 Guilford Street, London WC1N 1EH, UK.

** Department of Human Communication and Science, University College London,

Chandler House, 2 Wakefield Street, London, WC1N 1PG, UK.

#Director, Delaware Autistic Program,

144 Brennan Drive, Newark DE 19713, USA.

Acknowledgements

1
We are grateful to the MRC for support for this work through 3 successive project
grants to SBC, AC, and GB. Carol Brayne gave us valuable feedback on an earlier
draft of this paper.

2
The CHecklist for Autism in Toddlers (CHAT) is a screening instrument which

identifies children aged 18 months who are at risk for autism. This article explains

how the CHAT was developed, how the CHAT should be used, and provides a brief

introduction to autism.

What is autism?

Autism, first described by Kanner in 1943 (Kanner, 1943), is one of a family of

Pervasive Developmental Disorders (APA, 1994). It is considered to be the most

severe childhood neuropsychiatric condition and is characterised by a ‘triad’ of

impairments in socialisation, communication and flexible behaviour. The exact cause

of autism remains unclear, though family and twin studies suggest a genetic basis

(Bailey et al., 1995; Folstein & Rutter, 1977; Folstein & Rutter, 1988). Molecular

genetic studies are underway (Bailey, Bolton & Rutter, 1998). Neural abnormality is

evident in a number of different brain regions, including the medial prefrontal cortex

(Happe et al., 1996) and the amygdala (Abell et al., 1999; Baron-Cohen et al., 1999;

Bauman & Kemper, 1988). It occurs at a rate of about 1 per 1000 (Wing & Gould,

1979).

The autistic spectrum

It is widely accepted that there is a spectrum of autistic conditions (listed in Box 1).

Classic autism lies at the extreme of this spectrum. In DSM-IV this is referred to as

Autistic Disorder, and in ICD-10 as Childhood Autism. (See Appendix 1 for the full

diagnostic criteria for autistic disorder.) To receive this diagnosis, the onset of the

3
difficulties in social interaction, communication, and flexible behaviour must be

before the age of three years (see Box 2). Atypical autism or Pervasive

Developmental Disorder not otherwise specified (PDD-NOS) also lie on the autistic

spectrum, but children with these conditions do not meet criteria for autism because

of late age of onset, atypical symptoms, symptoms which are not very severe, or all of

these. Asperger Syndrome (AS) is thought to be another condition on the autistic

spectrum. Individuals with Asperger Syndrome have the social interaction difficulties

and restricted patterns of behaviour and interests, but have a normal IQ and no

general delay in language. The criteria for AS are shown in Appendix 2. A final

subtype are individuals with High Functioning Autism (HFA), who are diagnosed

when all the signs of AS are present, but where the child had a history of language

delay. Here, language delay is defined as not using single words by 2 years old, or

phrase speech by 3 years old.

insert Boxes 1 and 2 here

Late detection and the importance of early detection

Until recently, autism was rarely detected before the age of 3 years. This is not

surprising as autism is a relatively uncommon condition with subtle symptoms. In

addition, no specialized screening tool has existed, most primary health care

professionals receive limited training in the detection of autism in toddlers, and may

not have a link to specialist diagnostic clinics. However, the earlier a diagnosis can be

4
made, the earlier intervention can be implemented and family stress reduced, and

intervention has been shown to improve outcome (Lovaas & Smith, 1988). In

addition, early professional recognition of parental concerns may prevent secondary

difficulties developing. The problem is to determine what counts as a cost-effective

method of detecting the early signs of this condition.

Which behaviours might be important?

Parents of children with autism often report that they first suspected that their child

was not developing normally around the age of 18 months (Wing, 1997). At this age,

there are certain behaviours which are present in the normally developing child which

researchers have found to be lacking or limited in older children with autism. Two of

these are joint attention (Baron-Cohen, 1989; Sigman, Mundy, Ungerer & Sherman,

1986) and pretend play (Baron-Cohen, 1987; Wing, Gould, Yeates & Brierley, 1977).

insert Box 3 here

Joint attention refers to the ability to respond or initiate a shared focus of attention

with another person via pointing, showing or gaze monitoring (e.g. glancing back and

forth between an adult’s face and an object of interest or an event) (Bruner, 1983).

5
Joint attention allows children to learn through others - both learning what words

refer to (Baldwin, 1995; Tomasello & Barton, 1994), and what to pay attention to in

the environment (‘social referencing’ (Feinman, 1982)). Joint attention is seen as the

earliest expression of the infant’s ‘mind-reading’ capacity, in that the child shows a

sensitivity to what another person might be interested in or attending to (Baron-

Cohen, 1991). In Box 3, pointing to share interest (or declarative pointing) is

distinguished from a simpler form of pointing (pointing to request, or imperative

pointing). This distinction comes from child language research (Bates, Benigni,

Bretherton, Camaioni & Volterra, 1979). It is the declarative form which is of

particular importance simply because in this type of pointing mind-reading is the

driving force (‘Look at that! Do you see what I see?), whereas in imperative pointing

this may not be required (‘I want that! Get me that!’).

Pretend play is a second behaviour to be distinguished. It involves the attribution of

imaginary features to people, objects or events (Leslie, 1987). Some theorists view it

as signalling the emergence of symbolic ability (Piaget, 1962) as well as mind-

reading. Pretence is symbolic in that one object is treated as if it represents something

different, and it involves mind-reading by virtue of requiring the child to appreciate

that the person pretending (oneself or another person) is imagining something in their

mind. Generally, pretend play is distinguished from simpler forms of play (functional,

where the child uses objects appropriately, and sensorimotor, where the child just

explores objects for their physical qualities).

The CHAT

6
The CHecklist for Autism in Toddlers (CHAT) is a screening instrument that was

devised to test the prediction that those children not exhibiting joint attention and

pretend play by the age of 18 months will be at risk for receiving a later diagnosis of

autism. The CHAT is shown in Figure 1. The CHAT takes just 5-10 minutes to

administer and is simple to score. The order of the questions avoids a YES/NO bias.

The nine questions in section A are answered by the parent whilst the Health Visitor

or General Practitioner completes the five items in section B. There are five ‘key

items’ and these are concerned with joint attention and pretend play. The key items in

section B are included in order to validate (by cross-checking) the parents’ answers to

the key items in section A. The remainder (‘non-key’) items provide additional

information so as to distinguish an autism-specific profile from one of more global

developmental delay (see Box 4). The non-key items also provide opportunity for all

parents to answer ‘yes’ to some questions.

insert Figure 1 and Box 4 here

Those children who fail all five key items (A5, A7, Bii, Biii, and Biv) are predicted to

be at the greatest risk for autism. In Box 5, we call this the high risk for autism group.

Children who fail both items measuring protodeclarative pointing, but are not in the

high risk for autism group, are predicted to be in the medium risk for autism group.

Children who do not fit either of these profiles are in the low risk for autism group.

7
 insert Box 5 here

High-risk (sibling) study

Our first study tested the effectiveness of the CHAT as a screening instrument in a

high-risk sample (Baron-Cohen, Allen & Gillberg, 1992). We studied a group of 50

unselected 18 month olds (Group A) and a group of forty-one 18 month old siblings

of children with autism (Group B). The sibling group was selected because they have

a raised genetic risk for autism compared to individuals in the general population.

Even if we take the most generous estimate of the prevalence of autism spectrum

conditions in the general population - 0.34% (Ehlers & Gillberg, 1993) - this is still

at least 10 times less than the recurrence risk rate among siblings of children with

autism (3%) (Folstein & Rutter, 1977). So the likelihood of finding cases of

undiagnosed autism in the sibling group was much higher than in the control group.

The toddlers in both groups were assessed using the CHAT. None of the children in

Group A failed all 5 key items whilst four of the children in Group B failed all 5 key

items. A year later, when the children were 30 months old, a follow-up was carried

out. None of the children in Group A had autism. The four children in Group B, who

had failed the five key items, were diagnosed with autism. This strongly confirmed

the prediction that absence of joint attention and pretend play at 18 months of age is a

marker that a child is highly likely to go on to receive a diagnosis of autism.

Population screening study

8
After the preliminary success of the CHAT in detecting children at risk for autism in

the sibling group, a more stringent test of the CHAT was set up in a population

screening study (Baron-Cohen et al., 1996). 16,235 children aged 18 months were

screened with the CHAT from April 1992 to April 1993. These were all children

born in the South Thames Region of the UK.

Following this first administration of the CHAT, 38 children matched the high risk for

autism profile, 369 the medium risk profile with the remainder fitting the low risk

profile according to the criteria described in Box 5 above. One month later, all 38 of

the high risk for autism group were re-screened and 12 continued to meet the profile.

Limited resources meant that only about half of the medium risk group could be re-

screened. 22 met the criteria on the second CHAT, two of whom did not continue to

participate in the project. 16 children were selected at random from the low risk

group to receive a second CHAT and continued to meet his profile. Thus 12 children

in the high risk for autism group, 20 children in the medium risk for autism group and

16 children in the low risk for autism group were assessed clinically at 20 months and

42 months. The diagnoses made at 20 months of age were considered to be

provisional since this is earlier than the age at which children are usually seen for

diagnostic assessment and there is little or no evidence about the accuracy and

stability of childhood autism and PDD diagnoses made in infancy. Confirmatory

diagnoses were made at the 42 month clinical assessment. On the whole, we were

able to diagnose children with childhood autism reasonably accurately at 20 months

of age, in that most were thought to have either autism or PDD at that time point.

9
By 42 months, 10 of the 12 children in the high risk for autism group had received a

diagnosis on the autistic spectrum. The eleventh child was clinically normal and the

twelfth child had language delay. In the medium risk for autism group, half the

children were diagnosed with autism spectrum conditions (Childhood Autism,

Asperger Syndrome, or PDD), two received no diagnosis and the rest had language or

learning difficulties. In the low risk for autism group, although 1 child was diagnosed

with language delay, the other 15 were normal. Box 6 gives a summary of how the

diagnoses in each group changed between 20 months and 42 months. A more detailed

picture of these changing diagnoses is provided in Appendices 3-5. Full details of

diagnostic methods can be found elsewhere (Cox et al., 1999).

insert Box 6 here

Key issues in any screening program

It is useful to summarise the key issues in all screening. The following is based on

the standard account (Hennekens & Buring, 1987). Screening refers to the

application of a test to people who are as yet asymptomatic. It is essential to

remember that the screening procedure itself is not diagnostic. Those individuals who

test ‘positive’ are sent on for further evaluation using a subsequent diagnostic

procedure to determine whether they do in fact have the condition.

10
There are standard criteria for whether it is even worth screening for a disease: To be

appropriate for screening, a disease should meet the following criteria: (1) It should

be serious; (2) treatment given early (before symptoms are fully developed) should be

more beneficial (in terms of reducing morbidity or mortality) than treatment given

later; and (3) the prevalence of the disease should be high among the population

screened. Autism meets all three of these criteria in that (a) it is considered to be the

most serious of all childhood psychiatric conditions; (b) early intervention improves

prognosis; and (c) the prevalence of the spectrum of autistic conditions is high

(around 1 in 300).

For a screening program to be successful, a suitable screening test must be available.

A screening test should ideally have three characteristics: it should be (i) inexpensive;

(ii) easy to administer; and (iii) impose minimal discomfort on the patients. The

CHAT meets all three of these criteria. In addition, the results of a screening test

must be valid, reliable, and reproducible.

The validity of a screening test is measured by its ability to do what it is supposed to

do, that is, correctly categorize persons who have the condition as test-positive and

those without the condition as test-negative. Table 1 summarizes the results of any

screening test and the actual presence of the condition as determined by an

appropriate subsequent diagnostic test. In this table,

a = The number of individuals for whom the screening test is positive and the

individual actually has the condition (true positive).

11
b = The number for whom the screening test is positive but the individual does not

have the condition (false positive).

c = The number for whom the screening test is negative but the individual does

have the condition (false negative).

d = The number for whom the screening test is negative and the individual does not

have the condition (true negative).

insert table 1 here

Sensitivity and specificity are two measures of the validity of a screening test.

Sensitivity is defined as the probability of being test positive and truly having the

condition and is calculated by a/(a + c). As the sensitivity of a test increases, the

number of people with the condition who are missed by being incorrectly classified as

test-negative (false negatives) will decrease. Specificity is defined as the probability

of being screen negative and truly not having the condition and is calculated by d/(b +

d).

insert table 2

Obviously, it is desirable to have a screening test that is both highly sensitive and

highly specific. Usually that it is not possible, and there is a trade-off between the

sensitivity and specificity of a given screening test. This trade-off is due to the fact

12
that for many clinical tests, there are some people who are clearly normal, some

clearly abnormal, and some who fall into the grey zone between the two.

Any decision regarding specific criteria for acceptable levels of sensitivity and

specificity in a given situation involves weighing the consequences of leaving cases

undetected (false negatives) against erroneously classifying healthy persons as having

the condition (false positives). Sensitivity should be increased at the expense of

specificity when the penalty associated with missing a case is high, such as when the

disease is serious and definitive treatment exists (e.g., PKU), when the disease can be

spread (e.g., syphilis or gonorrhea), or when subsequent diagnostic evaluations of

positive screening tests are associated with minimal cost and risk, such as a further

series of blood pressure readings to ascertain hypertension. On the other hand,

specificity should be increased relative to sensitivity when the costs or risks

associated with further diagnostic techniques are substantial, such as with breast

cancer.

Regarding the number of cases detected by a screening program, one measure that is

commonly considered is the predictive value of the screening test. Predictive value

measures whether or not an individual actually has the condition, given the results of

the screening test. Predictive value positive (PV+ ) is the probability that a person

actually has the disease given that he or she tests positive and is calculated (using the

notation in Table 1) as:

PV+ = a
_____
a+b

13
Analogously, predictive value negative (PV-) is the probability that an individual is

truly disease-free given a negative screening test and is calculated as follows:

PV- = d
_____
c+d

So is the CHAT a good screening instrument?

Table 2 shows the data from the CHAT population study, and for comparison, Table 3

shows data from a screening study of breast cancer. This shows the CHAT has

excellent specificity, low sensitivity, and good predictive value positive. For a

condition of this kind, these are acceptable properties of the CHAT.

How many children with autism did the CHAT miss?

Whereas Table 2 above shows how many cases of autism alone in the high-risk group

were identified and missed in the six years following administration of the CHAT at

18 months, a series of follow-up screening and surveillance procedures were

conducted with the aim of identifying all the children from the population with an

autism spectrum condition, so as to test the properties of the CHAT in terms of

identifying false negatives among other autism spectrum subgroups. Using these

methods, a total of 50 children (47 boys, 3 girls) were found who met ICD-10 criteria

for childhood autism, and 44 children (36 boys, 8 girls) with other pervasive

developmental disorders were also identified. Using the two stage administration of

the CHAT, a grand total of 74 children who went on to receive some sort of autism

spectrum diagnosis were not identified as being at risk. The sensitivity, specificity,

14
positive predictive value and negative predictive value of the CHAT for all PDDs are

shown in Box 7. A comprehensive examination of these data is presented in a six year

follow-up paper (Baird et al., submitted).

insert Box 7 here

Just as for autism more narrowly, it is clear that for all PDD’s the CHAT also has

excellent specificity, whilst the sensitivity is not as impressive. In other words,

consistently failing the CHAT on two administrations means that it is highly likely

that a child will go on to receive a diagnosis on the autism spectrum. Although with a

condition like autism spectrum, having a high false negative rate is not a serious

problem (because it is not life-threatening, etc.) it would be good to understand why

the rate of false negatives is so high. It may be because some parents are

understandably answering questions in Section A in such a way as to put their child in

the best possible light (“Yes, of course he points and pretends!”). Since to ‘fail’ on

the CHAT a child needs to fail on both Sections A and B, our team did not look

closely at children who only failed on Section B. A second possible reason for the

high rate of false negatives is that to fail on the CHAT a child must have never

produced the behaviour (“Has your child ever pointed/pretended?”). This is clearly

only going to pick up the severe or extreme cases and would miss the milder cases

who may simply show a reduced rate of pointing or pretending, rather than a

complete absence of this. A third reason which could lead to a false negative is if a

child had “late onset” autism (Volkmar & Cohen, 1989).

15
Who should use the CHAT and when?

The CHAT is designed to be administered by any primary health care workers or

clinicians in children’s services. As a screening tool, it is convenient to administer at

the 18 month developmental check up. Administration of the CHAT to children

younger than 18 months is not recommended because of the increased risk of false

positives (i.e. identifying children at risk for developmental problems who do not go

on to receive a diagnosis). Administration of the CHAT to children older than 18

months is possible, since if a child is still showing a high-risk profile at this age, this

is very likely to be a sign of an autism-spectrum condition. Some regions only screen

at 24 months, hence us considering this possibility. However, we do not yet know of

any data on the use of the CHAT at this age, and suspect that the false negative rate

will also be high.

What happens if a child fails the CHAT?

In the population screening study (Baron-Cohen et al., 1996), the first CHAT was

administered in the routine 18 month old check-up. Those children who failed this

CHAT were re-screened approximately one month later with the same questionnaire.

As with any screening, a second CHAT is advisable so as to check that a ‘fail’ on a

key item occurs for valid reasons. Thus, a child might fail on the first administration

of the CHAT simply because they are slightly delayed or because of transient factors

(‘a bad day’). Any child failing the CHAT for a second time should however be

referred to a specialist clinic for diagnosis. This underlines that the CHAT itself is not

16
a diagnostic tool. It is important to note that more than a half of children who fail on

the first administration of the CHAT lose their risk status following the second

CHAT; and the risk group a child is assigned to does not represent a statement of

diagnosis. See Box 8.

insert Box 8 here

Summary

The CHAT is primarily a screening tool for clinical use. It has a low false positive

rate but a high false negative rate. In other words, if a child is identified by the CHAT

as being at risk for receiving a diagnosis on the autistic spectrum, they are likely to

receive one. However, many children will go on to receive a diagnosis who were not

identified by the CHAT. If a child meets criteria for the high-risk group, they will

almost certainly have an autism or PDD diagnosis. If they meet the criteria for the

medium risk group, about half of these will have a diagnosis of autism or PDD, whilst

most of the others will have other developmental delay conditions. As far as we

know, apart from the CHAT, there is currently no other screening instrument

available which has been fully evaluated in this way and which has been

demonstrated to be able to identify toddlers who are at risk for autism spectrum

conditions. The CHAT is easy and convenient to use and the low false positive rate

means that few parents will be unnecessarily alarmed. The CHAT is therefore a very

17
useful clinical tool, especially given the demonstrated effectiveness of early

intervention (Lovaas, 1993).

18
Figure 1: THE CHAT
To be used by GPs or Health Visitors during the 18 month developmental check-up.
Child’s name:................................. Date of birth:............... Age:................
Child’s address:................................................. Phone number: ................
SECTION A: ASK PARENT:
1. Does your child enjoy being swung, bounced on your knee, etc.? YES NO
2. Does your child take an interest in other children? YES NO
3. Does your child like climbing on things, such as up stairs? YES NO
4. Does your child enjoy playing peek-a-boo/hide-and-seek? YES NO
5. Does your child ever PRETEND, for example, to make a cup of YES NO
tea using a toy cup and teapot, or pretend other things?
6. Does your child ever use his/her index finger to point, to ASK YES NO
for something?
7. Does your child ever use his/her index finger to point, to YES NO
indicate INTEREST in something?
8. Can your child play properly with small toys (e.g. cars or bricks) YES NO
without just mouthing, fiddling or dropping them?
9. Does your child ever bring objects over to you (parent) to YES NO
SHOW you something?

SECTION B: GP OR HV OBSERVATION:
i. During the appointment, has the child made eye contact with you? YES NO
ii. Get child’s attention, then point across the room at an interesting YES NO*
object and say “Oh look! There’s a (name of toy)!” Watch child’s
face. Does the child look across to see what you are pointing at?
iii. Get the child’s attention, then give child a miniature toy cup and YES NO**
teapot and say “Can you make a cup of tea?” Does the child pretend
to pour out tea, drink it, etc.?
iv. Say to the child “Where’s the light?”, or “Show me the light”. YES NO**
Does the child POINT with his/her index finger at the light? *
v. Can the child build a tower of bricks? (If so how many?) YES NO
(Number of bricks:.............)

19
* (To record YES on this item, ensure the child has not simply looked at your hand, but has actually
looked at the object you are pointing at.)
** (If you can elicit an example of pretending in some other game, score a YES on this item.)
*** (Repeat this with “Where’s the teddy?” or some other unreachable object, if child does not
understand the word “light”. To record YES on this item, the child must have looked up at your face
around the time of pointing.)  MRC/SBC 1995

20
Box 1

Related conditions
• Classic autism
• Atypical autism
• Pervasive Developmental
Disorder not otherwise specified
(PDD-NOS)
• High-functioning autism
• Asperger’s syndrome

21
Box 2

Diagnostic Criteria for “Autistic

Disorder”
• qualitative impairment in social
interaction
• qualitative impairments in
communication
• restricted repetitive and
stereotyped patterns of
behaviour, interests and activities

with onset prior to 3 years of age

22
Box 3

Behaviours we can look for in 18 month old infants

1. Pointing 2. Play
a) to request (imperative) a) sensorimotor
b) to comment or share b) functional
interest (declarative) c) pretend

23
Box 4

CHAT Key Items CHAT Non-key Items

Section A Section A
A5: Pretend play A1: Rough and tumble play
A7: Protodeclarative pointing A2: Social interest
A3: Motor development
A4: Social play
A6: Protoimperative pointing
A8: Functional play
A9: Showing
Section B Section B
Bii: Following a point Bi: Eye contact
Biii: Pretending Bv: Tower of bricks
Biv: Producing a point

24
Box 5

High risk for autism group: fail A5, A7, Bii, Biii, Biv

Medium risk for autism group: fail A7, Biv (but not in maximum
risk group)

Low risk for autism group: not in other two risk groups

25
Box 6

Summary of Changing Diagnosis

18 months 20 months 42 months

High 9 CA 9 CA
Risk for 12 2 PDD 1 PDD
Autism 1 Lang 1 Lang
Group 1N

Medium 1 PDD 1 CA
Risk 20 8 Lang 9 PDD
for 6 DD/LD 6 Lang
Autism 5N 2 DD/LD
Group 2N

Low
Risk for 16 1 Lang 1 Lang
Autism 15 N 15 N
Group

Abbreviations
A childhood autism
DD/LD developmental delay/learning difficulties
Lang language disorder
N normal
PDD pervasive developmental disorder

26
Box 7

For all PDDs and combining the medium & high risk groups,

Sensitivity: 20/94 (21.3%)

Specificity: 16127/16141 (99.9%)

Positive predictive value: 20/34 (58.8%)

27
Box 8

Administering and scoring the CHAT

Procedure for administering the CHAT

Administer CHAT1 at 18
month old developmental pass no further
check-up CHAT1 action

fail CHAT1

Administer CHAT2 pass no further

by telephone CHAT2 action

fail CHAT2

Refer to specialist clinic

for diagnostic assessment

28
Appendix 1

The diagnostic criteria for autism as described in the Diagnostic and Statistical

Manual of Mental Disorders, 4th Edition (APA, 1987).

Diagnostic Criteria for Autistic Disorder

A. A total of six or more items from (1), (2) and (3), with at least two from (1), and

one each from (2) And (3):

(1) qualitative impairment in social interaction, as manifested by at least two

of the following:

(a) marked impairment in the use of multiple nonverbal behaviours such as eye-to-

eye gaze, facial expression, body postures, and gestures to regulate social interaction

(b) failure to develop peer relationships appropriate to developmental level

(c) a lack of spontaneous seeking to share enjoyment, interests, or achievements with

other people (e.g., by a lack of showing, bringing or pointing out objects of interest)

(d) lack of social or emotional reciprocity

(2) qualitative impairments in communication as manifested by at least one of the

following:

(a) delay in, or total lack of, the development of spoken language (not accompanied

by an attempt to compensate through alternative modes of communication such as

gestures or mime)

(b) in individuals with adequate speech, marked impairment in the ability to initiate

or sustain a conversation with others

29
(c) stereotyped and repetitive use of language or idiosyncratic language

(d) lack of varied, spontaneous make believe play or social imitative play

inappropriate to developmental level

(3) restricted repetitive and stereotyped patterns of behaviour, interests, and

activities, as manifested by at least one of the following:

(a) encompassing preoccupation with one or more stereotyped and restricted patterns

of interest that is abnormal either in intensity or focus

(b) apparently inflexible adherence to specific, nonfunctional routines or rituals

(c) stereotyped and repetitive motor mannerisms (e.g., hand or finger flapping or

twisting, or complex whole-body movements)

(d) persistent preoccupation with parts of objects

B. Delays or abnormal functioning in at least one of the following areas, with onset

prior to age 3 years: (1) social interaction, (2) language as used in social

communication, or (3) symbolic or imaginative play.

C. The disturbance is not better accounted for by Rett’s disorder or Childhood

Disintegrative Disorder.

30
Appendix 2

In the population screening study, those children who failed the 5 key items on the
CHAT were assigned to the high risk for autism group. The number of children and
their changing diagnosis are shown in the diagram. The abbreviations for this and the
following two appendices are:-

CA childhood autism
PDD pervasive developmental disorder
Lang language disorder
DD/LD developmental delay/learning difficulties
N normal

31
 CHANGING DIAGNOSIS IN DETAIL: HIGH RISK FOR AUTISM GROUP

12

20m 9 CA 2 PDD 1 Lang

7 2 1 1 1

42m 9 CA 1 PDD 1N 1 Lang

32
Appendix 3

The medium risk for autism group in the population screening study is defined as
those children who fail protodeclarative pointing but are not in the maximum risk
group. Again, the number of children and their changing diagnosis are shown.

33
 CHANGING DIAGNOSIS IN DETAIL: MEDIUM RISK FOR AUTISM GROUP

20

20m 1 PDD 8 Lang 6 DD/LD 5N

3 1 4

3 1 2

2 1 2

42m 9 PDD 1 CA 6 Lang 2 DD/LD 2N

34
Appendix 4

Children who did not fit the profile for the maximum or medium risk groups were
placed in the low risk group. The vast majority of children in the population
screening study were assigned to this group, but just a very small sample was
assessed.

35
 CHANGING DIAGNOSIS IN DETAIL: LOW RISK FOR AUTISM GROUP

16

20m 1 Lang 15 N

42m 1 Lang 15 N

36
Table 1: Results of any screening test, and derived measures of validity of the
screen. [Adapted (Hennekens & Buring, 1987)].

Disease status

Results of screening test Positive Negative Total

Positive a b a+b

Negative c d c+d

Total a+c b+d

Sensitivity a
a+c

Specificity d
b+d

PV+ a
a+b

PV- d
c+d

37
Table 2: Data from the CHAT population study (for childhood autism in the
high risk group only)

Disease status

Results of screening test Positive Negative Total

Positive 9 3 12

Negative 41 16 182 16 223

Total 50 16 185 16 235

Sensitivity a = 9 = 18%
a+c 50

Specificity d = 16 182 = 100%

b+d 16 185

PV+ a = 9 = 75%
a+b 12

PV- d = 16 182 = 99.7%

c+d 16 223

38
Table 3: Data from S.Shapiro, J.D. Goldberg and G.B. Hutchison. Lead time in
breast cancer detection and implications for periodicity of screening. Am.J.
Epidemiol. 100:357-366, 1974. [Adapted (Hennekens & Buring, 1987)]

Disease status

Results of screening test Positive Negative Total

Positive 132 983 1 115

Negative 45 63 650 63 695

Total 177 64 633 64 810

Sensitivity a = 132 = 74.6%

a+c 177

Specificity d = 63 650 = 98.5%

b+d 64 633

PV+ a = 132 = 11.8%

a+b 1115

PV- d = 63 650 = 99.9%

c+d 63 695

39
References

Abell, F., Krams, M., Ashburner, J., Passingham, R., Friston, K., Frackowiak,

R., Happe, F., Frith, C., & Frith, U. (1999). The neuranatomy of autism: a voxel-

based whole brain analysis of structural scans. Cognitive Neuroscience, 10, 1647-

1651.

APA. (1987). Diagnostic and Statistical Manual of Mental Disorders, 3rd

Edition. Washington DC: American Psychiatric Association.

APA. (1994). DSM-IV Diagnostic and Statistical Manual of Mental

Disorders, 4th Edition. Washington DC: American Psychiatric Association.

Bailey, A., Bolton, P., & Rutter, M. (1998). A full genome screen for autism

with evidence for linkage to a region on chromosome 7q. Human Molecular Genetics,

7, 571-578.

Bailey, T., Le Couteur, A., Gottesman, I., Bolton, P., Simonoff, E., Yuzda, E.,

& Rutter, M. (1995). Autism as a strongly genetic disorder: evidence from a British

twin study. Psychological Medicine, 25, 63-77.

Baird, G., Cox, A., Charman, T., Baron-Cohen, S., Swettenham, J., Drew, A.,

& Nightingale, N. (submitted). Screening a large population of 18 month olds with

the CHAT (Checklist for Autism in Toddlers): A six year follow up study. .

Baldwin, D. (1995). Understanding the link between joint attention and

language acquisition. In C. Moore & P. Dunham (Eds.), Joint attention: its origins and

role in development, . Hillsdale, New Jersey: Lawrence Erlbaum Associates.

Baron-Cohen, S. (1987). Autism and symbolic play. British Journal of

Developmental Psychology, 5, 139-148.

40
 Baron-Cohen, S. (1989). Perceptual role-taking and protodeclarative pointing

in autism. British Journal of Developmental Psychology., 7, 113-127.

Baron-Cohen, S. (1991). Precursors to a theory of mind: Understanding

attention in others. In A. Whiten (Ed.), Natural theories of mind, . Oxford:: Basil

Blackwell.

Baron-Cohen, S., Allen, J., & Gillberg, C. (1992). Can autism be detected at

18 months? The needle, the haystack, and the CHAT. British Journal of Psychiatry,

161, 839-843.

Baron-Cohen, S., Cox, A., Baird, G., Swettenham, J., Drew, A., Nightingale,

N., Morgan, K., & Charman, T. (1996). Psychological markers of autism at 18 months

of age in a large population. British Journal of Psychiatry, 168, 158-163.

Baron-Cohen, S., Ring, H., Wheelwright, S., Bullmore, E., Brammer, M.,

Simmons, A., & Williams, S. (1999). Social intelligence in the normal and autistic

brain: an fMRI study. European Journal of Neuroscience, 11, 1891-1898.

Bates, E., Benigni, L., Bretherton, I., Camaioni, L., & Volterra, V. (1979).

Cognition and communication from 9-13 months: correlational findings. In E. Bates

(Ed.), The emergence of symbols: cognition and communication in infancy, . New

York: Academic Press.

Bauman, M., & Kemper, T. (1988). Limbic and cerebellar abnormalities:

consistent findings in infantile autism. Journal of Neuropathology and Experimental

Neurology, 47, 369.

Bruner, J. (1983). Child's talk: learning to use language. Oxford: Oxford

University Press.

Cox, A., Klein, K., Charman, T., Baird, G., Baron-Cohen, S., Swettenham, J.,

Drew, A., & Wheelwright, S. (1999). Autism spectrum disorders at 20 and 42 months

41
of age: stability of clinical and ADI-R diagnosis. Journal of Child Psychology and

Psychiatry, 40, 719-732.

Ehlers, S., & Gillberg, C. (1993). The epidemiology of Asperger syndrome. A

total population study. Journal of Child Psychology and Psychiatry, 34, 1327-1350.

Feinman, S. (1982). Social referencing in infancy. Merrill-Palmer Quarterly,

28, 445-470.

Folstein, S., & Rutter, M. (1977). Infantile Autism: A Genetic Study of 21

Twin Pairs. Journal of Child Psycholology and Psychiatry, 18, 297-321.

Folstein, S., & Rutter, M. (1988). Autism: familial aggregation and genetic

implications. Journal of Autism and Developmental Disorders, 18, 3-30.

Happe, F., Ehlers, S., Fletcher, P., Frith, U., Johansson, M., Gillberg, C.,

Dolan, R., Frackowiak, R., & Frith, C. (1996). "Theory of mind" in the brain.

Evidence from a PET scan study of Asperger Syndrome. NeuroReport, 8, 197-201.

Hennekens, C., & Buring, J. (1987). Epidemiology in medicine. Boston:

Little, Brown.

Kanner, L. (1943). Autistic disturbance of affective contact. Nervous Child, 2,

217-250.

Leslie, A. M. (1987). Pretence and representation: the origins of "theory of

mind". Psychological Review, 94, 412-426.

Lovaas, O. (1993). The development of a treatment-research project for

developmentally disabled and autistic children. Journal of Applied Behaviour

Analysis, 26, 617-630.

Lovaas, O., & Smith, T. (1988). Intensive behavioural treatment for young

autistic children. In B. Lahey & A. Kazdin (Eds.), Advances in Clinical Child

Psychology, (Vol. II, ): Plenum Publishing.

42
 Piaget, J. (1962). Dreams, play and imitation in childhood. London: Routledge

and Kegan Paul.

Sigman, M., Mundy, P., Ungerer, J., & Sherman, T. (1986). Social

interactions of autistic, mentally retarded, and normal children and their caregivers.

Journal of Child Psychology and Psychiatry, 27, 647-656.

Tomasello, M., & Barton, M. (1994). Learning words in nonostensive

contexts. Developmental Psychology, 30, 639-650.

Volkmar, F., & Cohen, D. (1989). Disintegrative disorder or `late onset'

autism? Journal of Child Psychology and Psychiatry, 30, 717-724.

Wing, L. (1997). The Autistic Spectrum: Pergamon.

Wing, L., & Gould, J. (1979). Severe impairments of social interaction and

associated abnormalities in children: epidemiology and classification. Journal of

Autism and Developmental Disorders, 9, 11-29.

Wing, L., Gould, J., Yeates, S. R., & Brierley, L. M. (1977). Symbolic play in

severely mentally retarded and in autistic children. Journal of Child Psychology and

psychiatry, 18, 167-178.

43