Research in Developmental Disabilities 33 (2012) 1200–1210

Contents lists available at SciVerse ScienceDirect

Research in Developmental Disabilities

A new screening programme for autism in a general population of

Swedish toddlers
Gudrun Nygren a,c,*, Eva Sandberg b, Fredrik Gillstedt a,c, Gunnar Ekeroth a,
Thomas Arvidsson c, Christopher Gillberg a
a
Gillberg Neuropsychiatry Centre, Sahlgrenska Academy, University of Gothenburg, Sweden
b
Department of Speech and Language Pathology, Queen Silvia Childreńs Hospital, Gothenburg, Sweden
c
The Central Child Health Authority, Gothenburg, Sweden

A R T I C L E I N F O A B S T R A C T

Article history: The evidence from early intervention studies of autism has emphasised the need for early
Received 13 December 2011 diagnosis. Insight into the early presentation of autism is crucial for early recognition, and
Received in revised form 18 February 2012 routine screening can optimise the possibility for early diagnosis. General population
Accepted 20 February 2012 screening was conducted for 2.5-year-old children at child health centres in Gothenburg,
Available online 22 March 2012 Sweden, and the efficacy of the screening instruments in predicting a clinical diagnosis of
autism was studied. The tools used for autism screening comprised the Modified Checklist
Keywords: for Autism in Children (M-CHAT) and an observation made by trained nurses of the child’s
Autism spectrum disorders (ASD)
joint attention abilities (JA-OBS). From the new screening procedure a ‘‘definitive’’
Early symptoms
suspicion of autism spectrum disorder (ASD) was raised in 64 individuals in the study
Diagnosis
Screening
population of 3999 young children. Fifty-four of these were clinically assessed in detail.
Joint attention Forty-eight children had a confirmed diagnosis of ASD, three had severe language disorder,
M-CHAT and three (6%) were classified as having typical development. The Positive predictive Value
JA-OBS (PPV) for the combination of M-CHAT and the JA-OBS was 90%. The combination of
instruments used showed promise for early detection of autism as a routine in the
developmental programme at child health centres. Trained medical staff is a basic
requirement and enables earlier detection and the use of screening tools beyond routine
population screening regardless of the age at which a suspicion of autism is raised.
ß 2012 Elsevier Ltd. All rights reserved.

1. Introduction

In autism, in the vast majority of cases, symptoms are present from the first two years of life, but there is still often a
considerable delay before the diagnosis of autism is made. The rationale for early autism screening is the evidence that early
intervention is helpful in improving developmental functioning and reducing symptoms (Dawson et al., 2010; Howlin,
Magiati, & Charman, 2009; Ospina et al., 2008; Rogers & Vismara, 2008). Although there is much controversy regarding
different intervention approaches, and no specific treatment for all symptoms and for all children exists, the available
evidence underscores the importance of early detection and intervention for toddlers with autism.
In most western countries there are developmental child health programmes for children from their first months of life. In
Gothenburg, a metropolitan area on the Swedish west-coast, 95–99% of children are followed up at the child health centres
(CHCs) during their first years (Arvidsson, Holmberg, Reuter, & Strömbom, 2010). Despite regular health check-ups, in recent

* Corresponding author at: Gillberg Neuropsychiatry Centre, Kungsgatan 12, 411 19 Gothenburg, Sweden. Tel.: +46 31 3425970; fax: +46 31 3425979.
E-mail address: gudrun.nygren@gnc.gu.se (G. Nygren).

0891-4222/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ridd.2012.02.018
 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210 1201

years, the symptoms of autism have not been noticed, or if noticed not led to referral for autism diagnostic assessment until
several years later (Nygren et al., 2011). The observed delay in diagnosis was the background for a collaborative project for
early detection, diagnosis and early interventions for young children with autism spectrum disorders in Gothenburg. In the
following we will present methods for screening and early detection of autism.
The knowledge about early autism symptoms comes from at least three sources; (1) retrospective studies of symptoms in
children who later got the diagnosis of autism and (2) prospective studies of high-risk children (siblings of children with
autism) who have been followed up from birth and (3) combined retrospective prospective controlled studies comparing
infancy video-tapes of individuals with and without a later diagnosis of autism, atypical autism or Asperger syndrome
(Baron-Cohen, Allen, & Gillberg, 1992; Gillberg et al., 1990; Robins, Fein, Barton, & Green, 2001; Werner, Dawson, Munson, &
Osterling, 2005). The symptoms that have been noted have often been related to lack or deficiency of joint attention skills, an
ability that is crucial for development of social communication (Lawton & Kasari, 2011; Mundy, Sullivan, & Mastergeorge,
2009). Other early possible autism symptoms are motor difficulties, perceptual-sensory problems, stereotyped play with
objects and fixation on objects. Early sleeping and feeding problems are also common (Dahlgren & Gillberg, 1989; Dawson,
2011; Teitelbaum, Teitelbaum, Nye, Fryman, & Maurer, 1998). Some children exhibit autism symptoms early during the first
year, but more consistently around 12 months. Although there is individual variability at many levels in the heterogeneous
autism spectrum, several symptoms are often present by the age of 18 months (Bryson et al., 2007; Kleinman, Robins, et al.,
2008).
The choice of point for population screening is related to knowledge about the validity of neurodevelopmental diagnosis in
young children. Children with autism share many symptoms with children with other types of developmental delays,
including those with general developmental delay and developmental language delay. There is often also an overlap (or
‘‘comorbidity’’) between diagnoses such as autism and general developmental delay in young children. The social
characteristics in young children with ASD have been compared to those of children with other developmental problems
(Ventola et al., 2007). Children with ASD have been found to be more impaired in joint attention, imitation skills, pointing to
express interest, interest in other children, empathic responding and in range of facial expressions (Lord, 1995; Rogers,
Hepburn, & Wehner, 2003; Trillingsgaard, Ulsted Sorensen, Nemec, & Jorgensen, 2005). Children with autism also appear to
have a specific pattern of impairments in communication, including fewer conventional gestures, and they may be less likely
to initiate and respond to verbal communication (Lord, 1995; Trillingsgaard et al., 2005). Other studies have shown that
children with autism have significantly less pretend play than children with other disabilities (Cox et al., 1999; Noterdaeme,
Sitter, Mildenberger, & Amorosa, 2000). Several longitudinal studies have shown that the diagnosis of autism made around
the age of 2–4 years is relatively stable over time (Gillberg et al., 1990; Kleinman, Ventola, et al., 2008; Lord, 1995; Lord, Risi,
DiLavore, Shulman, Thurm, & Pickles, 2006). There is, internationally, a quest for very early screening, including during the
first year. Still, the evidence for population screening of children under age 2 years remains meager (Al-Qabandi, Gorter, &
Rosenbaum, 2011; Baron-Cohen et al., 1992; Pandey et al., 2008; Willemsen-Swinkels, Buitelaar, & van Engeland, 2001).

2. Aim

The primary aim of the Gothenburg study was to develop and review an autism screening procedure within the existing
developmental surveillance programme at the CHCs. A secondary aim included assessment of the psychometric properties of
the screening tools. It is this secondary aim that is the focus of the present paper.

3. Material and methods

The autism population screening was conducted at all the CHCs in Gothenburg during the entire year of 2010. The
screening was the first part of a three-wave collaborative ‘‘Autism: Detection and Intervention in the Early years (AUDIE)’’
project which had started in 2009 in Gothenburg. The second wave of the project comprised diagnostic assessments and
diagnostics at the Child Neuropsychiatry Clinic (CNC) at the Queen Silvia Children’s Hospital, serving the whole of
Gothenburg, and to which the children were referred if screening had raised any suspicion of ASD (see Nygren et al., 2011 for
details about the extremely detailed assessments, comprising the Autism Diagnostic Schedule (ADOS) (Lord et al., 2000). The
Diagnostic Interview for Social and Communication disorders (DISCO-11) (Nygren, Hagberg, Billstedt, Skoglund, Gillberg, &
Johansson, 2009; Wing, Leekam, Libby, Gould, & Larcombe, 2002) and cognitive assessment among other things in
accordance with a structural protocol in all cases). The ASD category comprised DSM-IV/ICD-10 autistic disorder/childhood
autism and DSM-IV/ICD-10 PDD-NOS/atypical autism. In the third wave of the project all parents of children with a
confirmed diagnosis of autism spectrum disorders were offered early interventions for their children at the habilitation
service centres. Clear routines were established as to how to guarantee care efficacy. The goal was to complete the diagnostic
assessment within 4 months from the date of raising the suspicion of ASD at the CHC and for multimodal interventions to be
provided after the diagnosis without delay.

3.1. Study population

The total population in the city of Gothenburg was 513 751 on December 31 2010. The study population consisted of
children, born in 2007 or 2008, coming for their 2.5-year-old check-up at the CHCs at any time during 2010. The 2.5-year-old
1202 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210

check-up was part of the routine developmental surveillance programme that is always offered at the Gothenburg CHCs. The
total population of 2.5-year-old children in 2010 was estimated at 6220 based on the numbers of the two birth years 2007
(6022) and 2008 (6418) at the end of 2010 (Statistics Sweden). According to statistics from the CHC authority (Arvidsson
et al., 2010) 80% of the 2.5-year-old children were screened for autism in 2010. There were no indications that the non-
screened population differed from the 80% who were screened. Thus, the total 2.5-year-old population available for the study
was approximately 5000 children who were actually screened. The ratio of parents born in Sweden (both parents born in
Sweden) to parents born abroad (one or both parents born in another country) for this age group was 2.6:1 (72.1%/27.9%)
(Statistics Sweden, 2010).
The nurses were instructed to use the screening instruments (see below) for all 2.5-year-old children and for all other
children (regardless of age) raising any suspicion of ASD. These later children, older and younger than 2.5 year were also
referred to the CNC for diagnosis and those with an ASD diagnosis were included in the present study of psychometric
properties of the screening instruments. A prevalence study based on a subset of the present study group (children aged 24–
36 months) has already been published (Nygren et al., 2011). The present study comprises a small subset of the total sample
of children younger or older than 24–36 months (see below in Section 5). Written informed consent for participation in the
study was received by the parents of 3999 children (2087 boys, 1912 girls) (i.e. 80% of the 80% of the whole population) and
these children were then the focus of further analyses in the present context.

3.2. Screening instruments

Some general requirements in relation to assessment tools used for screening were fundamental in planning this study.
The assessment instrument would be brief and low cost and designed for easy use in primary care at the CHCs. Parent check-
lists are easy to administer. Observations made by professionals, on the other hand, can be related to knowledge of typical
child development and, thus, provide more objective information (Dumont-Mathieu & Fein, 2005; Oosterling et al., 2010). A
combination of the two types of instrument was chosen for the present study; the Modified Checklist for Autism in Toddlers
(M-CHAT) and a joint attention observation (JA-OBS). One reason for including the M-CHAT was the fact that the original
CHAT had been developed by our group and used in a number of previous studies by us (Baron-Cohen et al., 1992; Carlsson,
Gillberg, Lannerö, & Blennow, 2010). Another major reason was the fact that at the time of planning the study, the M-CHAT
was the autism infant screening scale for level 1 screening (general population) supported by most published evidence.
Translations into many languages were already available. Studies published in the last few years have shown good
psychometric properties for another screening instrument, the Baby and Infant Screen for Children with aUtIsm Traits
(BISCUIT) (Matson, Wilkins, Sharp, Knight, Sevin, & Boisjoli, 2009), but this was not widely known at the time. A recent study
comparing the use of M-CHAT and the BISCUIT suggests that the target populations for the instruments differ (Kozlowski,
Matson, Worley, Sipes, & Horovitz, 2011). The BISCUIT appears to be a more sensitive instrument for screening children
already suspected of having developmental delays (level 2 screening), the brief M-CHAT is possibly more suitable for general
population screening.
Other demands often made on instruments used for population screening are clear cut-off scores, and validation against
clinical diagnosis and standard diagnostic tests. The screening tool’s specificity must have been shown in population studies
to have an acceptable level. Two psychometric properties for screening instruments are generally accepted as providing the
best basis for deciding whether or not a particular instrument could be used for population screening purposes. They are
Positive Predictive Value, PPV (of all children in the sample whom the screen identified as positive – the percentage who
actually was ‘‘positive’’ i.e. had a diagnosis of ASD), and sensitivity (of all children in a sample with diagnosis-the percentage
that the screen identified). Negative Predictive Value, NPV (of all children in the sample whom the screen identified as
typically developing – the percentage who actually were typically developing, i.e., did not have a diagnosis of ASD), and
specificity (of all children in the sample who were typically developing – the percentage that the screen identified as typically
developing) are also considered important, but cannot usually be adequately analysed in large transectional population
cohorts, as full diagnostic assessment of several thousand screen negative individuals is not within the resources of studies of
this type.

3.3. The modified checklist for autism in toddlers (M-CHAT)

This instrument was developed in the US on the basis of the pioneer (British–Swedish) screening tool, the Checklist for
Autism in Toddlers (CHAT) (Baron-Cohen, Allen, & Gillberg, 1992), with a view to improving the sensitivity of the instrument.
The M-CHAT comprises a 23 items yes/no parent report and a follow-up interview (Robins & Dumont-Mathieu, 2006). The
parent report was validated in the US in 2001 and proved to have promising psychometric properties, including good
internal reliability (Robins, Fein, Barton, & Green, 2001). Six of the 23 items (2, 7, 9, 13, 14, 15) pertaining to social relatedness
and communication were found to have the best discriminating ability for ASD/non ASD. Failure on the screening was
defined as ‘‘failure’’ on any three of the 23 items or on any two of the 6 critical items failed (see below for definition of
preliminary and definitive failure). In the first low-risk screening study using the M-CHAT the false positive rate was
relatively high. For this reason a structured follow-up interview was developed by the authors to reduce the false positive. In
cases identified with risk from the parent questionnaire a structured follow-up interview (originally a telephone-interview)
with the parents was conducted (Kleinman, Robins, et al., 2008; Robins & Dumont-Mathieu, 2006; Robins et al., 2001) which
 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210 1203

improved the PPV. Depending on screened population (high-risk, low-risk, age-range, whether interview performed or not)
the PPV of the M-CHAT has been reported with variation in the range of 28–80% (Kleinman, Robins, et al., 2008; Pandey et al.,
2008; Robins, 2008; Robins et al., 2001). The M-CHAT has been used in many studies and is currently the most accepted
instrument for early autism screening in the general population. Translations into many languages exist (in accordance with
guidelines from the originators and general recommendations for translations of instruments (Banville, Desrosiers, & Genet-
Volet, 2000)). The M-CHAT, including the interview, can easily be downloaded from the website M-CHATTM information
(Robins) in different languages.
The instrument was originally developed and used for children aged 16–30 months. Recently, Yama and colleagues
suggested, on the basis of results obtained in a study using the instrument in a low-risk sample, that the M-CHAT could be
appropriately administered in children aged 20–48 months (Yama, Freeman, Graves, Yuan, & Karen Campbell, 2012). The
study also emphasised the importance of clinician’s expertise in verifying questionnaire positive M-CHAT screens.
The recommended two-step procedure, M-CHAT + follow-up interview was chosen for the Gothenburg study. Both the
M-CHAT questionnaire, and the M-CHAT follow-up interview, were translated into Swedish (and back-translated into
English) by our group (Almgren, Nygren, Gillberg, & Gillberg, 2008), in accordance with guidelines provided by the
originators. After the translation process, a few minor adjustments of the Swedish language were done in order to ensure that
the integrity of the instrument was not lost through translation. In the present study a failure identified (according to failure
criteria as described above) on the parent questionnaire was referred to as ‘‘preliminary failure’’, and if failure was confirmed
at the follow-up interview it was referred to as ‘‘definitive failure’’.

3.4. A five-item nurse observation of the child’s ability as regards ‘‘joint attention’’ (JA-OBS)

A five-item CHC nurse observation schedule of the child’s ability as regards joint attention was the other tool used for
autism screening. The schedule (Fig. 1), the Joint Attention Observation schedule (JA-OBS), was developed by the first author
of the present study on the basis of results obtained in studies of early symptoms which related to lack or deficiency in the
ability to initiate/engage in joint attention (Baranek, 1999; Baron-Cohen et al., 1992; Osterling, Dawson, & Munson, 2002;
Werner, Dawson, Osterling, & Dinno, 2000). After the piloting of the JA-OBS in 2009, it was decided that screen positivity for
autism on the JA-OBS would be defined as failure on any two or more of the five items. The JA-OBS requires the nurse (who
should have attended two educational seminars, see below under Screening procedure) to check five observable behaviours
in the child (see Fig. 1).

3.4.1. A brief speech–language screen

A brief language screen (Mattsson, Marild, & Pehrsson, 2001), consisting of a parent questionnaire and an assessment
made by the nurse, was also used in parallel with the specific autism screening. This screen has been part of CHC routines in
many parts of Sweden for several years, and had been administratively decided to become routine as part of the 2.5-year-old
check-up in Gothenburg from 2008. The procedure, including the questionnaire and the observation schedule, has been
researched in the Gothenburg area and has adequate psychometric properties, including good-excellent reliability and
validity (for language disorder) (Mattsson et al., 2001; Miniscalco, Westerlund, & Lohmander, 2005). In accordance with
Miniscalco’s suggested guideline, failure on this screen was defined as one or more of the following (according to either
parent or nurse or both): (1) fewer than 25 single words, (2) lack of 2-word utterances, (3) poor verbal comprehension or (4)
parental concern for the child’s language and communicative ability.

3.5. Screening procedure

The 134 nurses at 49 different CHCs were the key professionals responsible for the 2.5-year-old check-up. Before the
population screening was introduced in 2009 the nurses attended educational seminars (one full-day and one half-day) held
by the same two trained researchers (the first and second author, who are both experienced clinicians, one a pediatrician/
child and adolescent psychiatrist and the other a speech and language pathology therapist) about children’s typical
developmental milestones, particularly focusing on social interaction and communication, early symptoms of autism, and on
the autism screening per se. The nurses were specifically trained in the use of the screening instruments. Clear instructions
were given as how to act with children of any age if suspicion of ASD was raised including in children under age 2.5 years. The
nurses were encouraged to listen to the parents and any concern they might have about their child’s development. It was

Does the child –

1. react to own name (turns to person addressing the child)?
2. try to establish eye-contact with you?
3. gaze at something that you point to further away in the room?
4. use his/her index-finger to point at something (e.g. in a book)?
5. interact with you or parent in pretend play (e.g. during feeding a doll ,or putting the
doll to bed; does the child use eye contact to monitor that you are watching)?

Fig. 1. JA-OBS (Joint Attention Observation of Toddlers).

1204 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210

further emphasised that the screening tools could be used also if a suspicion about autism was raised at any time other than
the 2.5-year-old check-up. It was also made clear, that if there was a suspicion of ASD either raised by medical staff or
parents, the parents would be offered further child assessment even in cases not screening positive on the screening
instruments. An instruction booklet for the nurses that had been produced specifically for the project was delivered by the
researchers in connection with the seminars. The doctors who were responsible for the CHCs (pediatricians and general
practitioners), were all invited to separate seminars and information meetings about autism, the screening at the 2.5-year-
old check-up and the routines for further evaluation and interventions.
About two weeks in advance of the actual check-up at the CHC, parents of all 2.5-year-olds in Gothenburg were sent a
letter of invitation to their child’s health check-up including information about the special focus at this visit on language
and the child’s communication and social interaction. The M-CHAT as well as the parental questionnaire concerning
language development was attached with the letter and the parents were asked to bring them completed to the visit. There
was also information about the autism screening study and the parents were asked for written consent for their child’s
participation in that part of the study. More information about the study could be given person-to person at the CHC visit. In
some cases the information was given with the help of an interpreter because of parent language problems. The M-CHAT
and the language questions took about 5 min to respond to. The 2.5-year-old visit took about 45–60 min including a full
explanation of the study and the procedure (after which a number of families opted out, see below), and the full
developmental check-up. About 10 min of this time was needed for the nurse JA-OBS and for the language assessment, 1–
5 min for the scoring of the M-CHAT. Sixty minutes was the recommended nurse time for all 2.5-year-old visits. This time
was estimated to be necessary for children where difficulties were observed, when M-CHAT follow-up interview was
indicated and also in all cases where an interpreter was needed. The M-CHAT interview was performed directly at the CHC
by the nurse (not as a phone interview). If a ‘‘preliminary failure’’ was observed according to the parent report, the nurse
completed the interview during the CHC visit. If ‘‘screen positivity’’ was confirmed by the interview (any three of the 23
items failed or any two of the 6 critical items failed) that case was ‘‘definitively screen positive’’. Screen positivity for autism
could be either a (i) definitive failure on the M-CHAT or (ii) failure on two or more of the items of the JA-OBS (or, of course,
both).
If the child was screen positive for autism, or, if for other reasons, there was a suspicion of autism, there was a plan for a
second visit (within a month) to the CHC for an examination made by the paediatrician or general practitioner together
with the nurse. At this visit developmental and family history was taken including a screen for other possible medical
problems. If, after this brief clinical examination, a clinical ASD suspicion remained, information about this suspicion was
given to the parents who were informed about the need for further evaluations of the child, and a referral to the specialist
clinic (CNC, please note that this is not organised by the CHC, but a different highly specialised clinic for children with
neuropsychiatric disorders) for assessment was made. If the parents needed more time or refused to go with their child for
further evaluations there was also the possibility for a first evaluation by the team at the CHC (psychologist, speech and
language therapist, nurse and doctor) for possible later referral to the CNC for evaluation. Children, who failed the language
screening, but not the autism screening, were referred by the nurse to auditory examination and to a speech and language
therapist (SLP). The data from all children, with a suspicion of ASD, was sent by the nurses to the Central Child Health
Authority for the study. Data for those who screened negative (even if there had been a preliminary failure on M-CHAT) was
not collected.

3.6. Diagnostic procedure at the neuropsychiatric specialist clinic (CNC)

A comprehensive neuropsychiatric assessment was conducted in every referred screen positive case including clinical
psychiatric, psychological (including developmental assessment with the Griffiths (Norberg, Tingwall, & Ahlin-Åkerman,
1980)), and general medical examinations made by experienced neuropsychiatrists, neuropediatricians (four in total) and
neuropsychologists (two in total) with expertise in autism. The Vineland Adaptive Behavior Scales (Sparrow, Balla, &
Cicchetti, 1984) was completed at the interview with the parent. The Autism Diagnostic Observation Schedule (ADOS) (Lord
et al., 2000) was performed (and video-recorded) by experienced ADOS-trained clinicians. In two thirds of the cases the full
2–4 h version of the Diagnostic Interview for Social and COmmunication Disorders (DISCO-11) (Nygren et al., 2009; Wing
et al., 2002) was performed with a parent (this interview was too cumbersome for some parents with limited Swedish or
English skills). Language assessments were done by an experienced speech and language therapist. The evaluation also
included a 1-h observation of the child at preschool by an experienced special education teacher from the team. All these
assessments were done independently of each other and results were not revealed until the final case conference at which
final DSM-IV diagnosis (‘‘best estimate clinical research diagnosis’’) was based on all available evidence obtained at the
assessment and from medical records. The results on the screening measures used at the CHC did not contribute in any way
to the clinical research diagnosis.

4. Ethics

The study was approved by the Regional Ethics Committee in Gothenburg. Parents signed written informed participation
consent forms both in connection with the screening study at the CHC (see above), and at the specialist clinic (CNC) before
the comprehensive autism diagnostic assessment was started.
 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210 1205

4.1. Statistical methods used

Sensitivity was defined as percent of diagnosed children who also screened positive on a particular measure. PPV was
defined as percent of children who screened positive on a particular measure and who had a diagnosis of ASD confirmed. As a
measure of inter-rater reliability per cent of agreement was calculated. Kappa values were in several cases not applicable due
to missing values either in an entire column or an entire row. The Confidence Intervals (CI) were calculated assuming
binominal distribution and presented with 95% confidence limits.

5. Results

Suspicion of ASD within the developmental programme at the CHCs in relation to further evaluations and diagnostic work
at the specialist CNC

5.1. Suspicion of ASD leading to referral to specialist CNC assessment

The suspicion of ASD was raised in 64 individuals (9 girls, 55 boys; 62/64 at the CHCs directly and two from SPLs after
referral to them for suspected language delay) (Fig. 2). The majority of the children were 2.5 years old, 30–35 months (75%).
In 16 of the 64 cases the suspicion was raised before or after the routine 2.5-year visit (five children were younger than
24 months, seven were 24–29 months and four children were three years old, 36–46 months).

5.2. Attrition

Ten of the 64 children have not yet been further evaluated at the specialist CNC. The reasons for this are as follows:
parents refused to come for in-depth assessment (n = 4), parents wanted to wait a further year before diagnostic assessment
(n = 2), family had moved abroad (n = 1), and unknown reason (n = 3).

5.3. Diagnosis of the children who were comprehensively psychiatrically assessed

Fifty-four children received the comprehensive neuropsychiatric autism diagnostic assessment. Forty-eight (8 girls, 40
boys) of these were confirmed to have a diagnosis of ASD after the assessment at the specialist CNC. The mean age of these

Fig. 2. Procedure for ASD screening and diagnostic assessment.

1206 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210

60

50

40

percent
30

20

10

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
MCHAT items

Fig. 3. M-CHAT definitive failure (%) on the different items (1–23) in children with confirmed ASD (n = 33).

children was 2.5 years (SD 0.5). In 27 individuals the criteria for autistic disorder/childhood autism according to DSM-IV/ICD-
10 were met, and in 21 individuals the criteria for PPD-NOS/atypical autism were met. Three of the 54 children were
diagnosed with language disorder, of whom one also had autistic features, and three (6%) were considered to have
development within the ‘‘typical’’ range.

5.4. Parental concerns before referral to CNC

In 35 (of 48) cases the parents had already expressed concern for their child’s development before the screening
procedure at the CHC. In twelve of these cases there was a known family history of ASD already noted by the medical staff at
the CHC in the referral letter to the CNC.

5.5. Migrant status of parents

In 24 (50%) cases with ASD both parents were of non-Swedish descent, in 8 cases one parent had not been born in Sweden,
and in 16 cases both parents were of Swedish descent.

5.6. Language screen positivity followed by autism suspicion within one year

Four children, with no initial failure on the autism screen, but on the language screen, have later been referred from
speech and language pathologists (SLP) for neuropsychiatric evaluations at the CNC. The neuropsychiatric assessments have
not yet been completed.

5.7. Psychometric properties of the instruments for autism screening

5.7.1. M-CHAT
The PPV, calculated from the proportion of children with definitive failure on the M-CHAT (+interview) and diagnosed
with ASD, was 91.7% (95% CI 77.5–98.2). Thirty-three of 36 children with definitive failure on the M-CHAT were from the
neuropsychiatric assessment at the CNC confirmed to have a diagnosis of ASD. The sensitivity of the M-CHAT (+interview)
was 76.7% (95% CI 61.4–88.2) (Table 2). Only three (2, 7, 13) of the 6 most discriminating items (endorsed by more than 50%
of the sample) in the present study) ‘‘overlapped’’ with the 6 critical items identified by Robins et al. (2001) (Fig. 3). Most
notable was the low endorsement rate for item 14 (‘‘Does your child respond to his/her name when you call?’’) and the
relatively high rate for items 21, 22 and 23.

5.7.2. JA-OBS
The PPV of the nurse observation, the JA-OBS, was 92.5% (95% CI 79.6–98.4) and the sensitivity for the instrument was
86.0% (95% CI 72.1–94.7) for the used algorithm of failure in 2 items (Table 2). Thirty-seven of 40 JA-OBS positive cases were
confirmed with an ASD diagnosis after the comprehensive neuropsychiatric assessments at the CNC. In cases with a
confirmed ASD (n = 37) a failure had been observed at the CHC on three to five items in 23 individuals and observed on two
items in 14 individuals (Table 1). The ‘‘sensitivity’’ for different items is illustrated in Fig. 4. Unexpectedly, but consistently
with the results from item 14 on the M-CHAT, item 1 (‘‘Reacts to own name’’) is not the most sensitive.
A preliminary inter-rater reliability test of the instrument was performed as follows: Per cent agreement was studied in 14
cases observed by two nurses independently at the same 2.5-year check-up at the CHC. In one of the 14 cases ASD suspicion
was raised at the CHC visit, the remaining 13 were regarded as typically developing. Complete agreement for total scores
applied in 93% of the cases (individual item agreement 86–100%). In another sample per cent agreement was studied
 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210 1207

Table 1
Failure on JA-OBS items in children with a confirmed ASD diagnosis (n = 37).

1 2 3 4 5 Frequency Cumulative frequency

+ + 2 2
+ + 1 3
+ + 1 4
+ + + 2 6
+ + 6 12
+ + 1 13
+ + + 1 14
+ + 2 16
+ + + + 3 19
+ + 1 20
+ + + + 2 22
+ + + + 1 23
+ + + + + 14 37

Cut-off defined as a failure in two or more items and a failure marked in the table with +

90
80
70
60
Percent

50
40
30
20
10
0
1 2 3 4 5
Joint Aenon items

Fig. 4. JA-OBS failure on items 1–5 (%) in children with ASD (n = 37).

between JA-OBS performed by the nurse at the CHC and the observation made 2–3 months later by a clinician at the CNC
when the ADOS was performed. Agreement was obtained in 92% of 12 cases (individual item agreement 58–100%) with later
confirmed ASD. (Use of kappa statistic was not possible due to values missing in one column or row.)

5.8. Combination of instruments M-CHAT and JA-OBS

The combination of instruments had been used in 51 of the 54 further evaluated children and in 45 of the 48 children with
a confirmed diagnosis of ASD. Forty-three of these had screened positive in one or both of the instruments. Three cases were
clearly false positives (typically developing) and two were ‘‘false positive’’ for autism (language disorder). The PPV for the
combination of M-CHAT and JA-OBS (the proportion of children with a definitive failure in M-CHAT (+interview) or in JA-OBS
or in both and diagnosed with an ASD) was 89.6% (95% CI 77.3–96.5) and the sensitivity was 95.6% (95% CI 84.9–99.5) (Table
2).

5.9. Language screen

The language screen had been used in 42 (40 screen positive, two screen negative) of the 54 cases which had been further
evaluated with the suspicion of ASD. Thirty-four of the screen positive cases were confirmed to have a diagnosis of ASD, three
individuals were found to have a severe language delay and three individuals were classified as typically developing. Two
children with negative language screen at 2.5 year were diagnosed with autism at the CNC.
Seventy-eight 2.5-year-old children (2%) from the study population had a failure in the language screen, but not in the
autism screen (Fig. 1). They were referred from the CHCs for assessments by a speech and language pathologist (SLP). Two of
the 78 children were directly sent by the SLP on to the CNC with the suspicion of ASD and a diagnosis of atypical autism was
confirmed in both of them. Fifty-six children were considered to be in need of a scheduled follow-up for their language delay/
disorder and four of these children have recently been referred to the CNC for neuropsychiatric evaluations with the

Table 2
PPV and Sensitivity for M-CHAT and JA-OBS and for the combination of instruments.

Psychometric property M-CHAT (n = 49) JA-OBS (n = 48) M-CHAT and/or JA-OBS (n = 51)

PPV (95% CI) 91.7% (77.5–98.2) 92.5% (79.6–98.4) 89.6% (77.3–96.5)

Sensitivity (95% CI) 76.7% (61.4–88.2) 86.0% (72.1–94.7) 95.6% (84.9–99.5)
1208 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210

Fig. 5. M-CHAT, item 23: ‘‘Does your child look at your face to check your reaction when faced with something unfamiliar.’’ The Japanese illustration (http://
www2.gsu.edu/cpsydlr/Diana_L.Robins._Ph.D.html).

suspicion of ASD or other neuropsychiatric disorder. Twenty-two children no longer have contact with the SLP (in 14 cases
there was no further need for interventions and eight families dropped out).

6. Discussion

The purpose of the current study was to validate new methods for autism screening that had been introduced as part of
the developmental programme for all children at the 2.5-year check-up (and also for all other children whenever a suspicion
of ASD was raised at the CHCs). The PPV for the combined M-CHAT/JA-OBS was excellent (90%).
In the 2.5-year-old children included in the study both the instruments for autism screening (M-CHAT and JA-OBS) had
been used. In cases where a suspicion of ASD had been raised beyond the 2.5-year-old check-up (n = 12) the screening
instruments had not been used consistently, but in most cases at least one of them had been applied. The language screen
was only used for the 2.5-year-old children.
The completed M-CHAT was brought by the parents to the CHC in the vast majority of cases. About 10% of the parents
completed it in the CHC waiting room. The recommended two-step use of the M-CHAT (questionnaire followed up by
interview in screen positive cases) proved to be necessary for several reasons. Many parents had difficulties with the
questionnaire; some items were difficult to understand. Three items, 11 (‘‘reaction to noise’’), 22 (‘‘staring, wondering’’) and
23 (‘‘look at your face’’) in particular, often regardless of parental language, needed clarification from the nurse. In cases with
preliminary failure on the M-CHAT questionnaire, the follow-up interview was performed even though not always with the
exact use of the M-CHAT interview questions but with focus on clarifying items with the parents directly at the CHC visit.
Non-Swedish parents were offered questionnaires both in their own language (if their native language was one of the ones
found from the website) and in Swedish, but there was still usually a need for further explanations, and some of the follow-
up interviews had to be performed with the help of an interpreter (illustrations might be needed in the future development
of the M-CHAT. The Japanese M-CHAT form with illustrations shows an example of how illustration might help improve
‘‘readability’’ (Fig. 5)). However, even though a few of the items on the M-CHAT were a bit unclear, and the nurse interview
not always easy to perform and interpret, the majority of the nurses reported positive experiences from the use of M-CHAT.
The fact that the parents were actively involved in the developmental check-up of their child was regarded as particularly
positive.
Surprisingly, the sensitivity for item 14 in the M-CHAT (‘‘response to name’’) was not high as expected (Fig. 4) from original
studies of the instrument and from a number of studies of symptoms in young children with autism. Only a few of the
original six ‘‘key items’’ were among the best single items predictive of a confirmed ASD diagnosis in the present study. The
chosen screening procedure makes it impossible to draw definite conclusions regarding the sensitivity of particular items. In
cases with screen positivity according to parent questionnaire the nurses performed follow-up interview, but this interview
did not necessarily confirm all the parent screen positive items; a failure on any two key items or on any three of the 23 items
was considered definitive. The unexpected results regarding some of the items (particularly M-CHAT item 14 ‘‘response to
name’’, and JA-OBS item 1 ‘‘response to name’’) call for further studies, perhaps even of a cross-cultural nature. The majority
of the individuals in this study were 30–35 months old, but there were also some children detected with ASD beyond the 2.5-
year-visit which may have implications for the results. The psychometric properties for the instruments (including the
sensitivity for different items in the instruments) need to be studied separately for different age groups in further studies.
The five-item JA-OBS clearly raised nurse autism awareness and screening skills. In at least three of the 48 confirmed ASD
cases, the parents had not reported any difficulties on the M-CHAT questionnaire, but the nurse observation at the 2.5-year-
visit started a process of dawning awareness in the parents as problems became obvious in this part of the screening. In a
couple of months after the CHC visit two parents raised their concern over several M-CHAT items that they had originally not
endorsed once they had completed the questionnaire a second time around. The inter-rater reliability for the JA-OBS was
calculated as per cent agreement for the different items in two limited samples. The agreement was generally high, but
further studies are needed.
 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210 1209

The PPV for the combination of autism screening instruments, failure on M-CHAT (+interview) and/or on JA-OBS, was 90%
and the sensitivity was 96%. The PPV value for the instruments separately analysed was 92% for M-CHAT (+interview) and
92.5% for JA-OBS, while sensitivity was 77% and 86%. It is not possible to determine how many children with ASD were missed
at this point of time (the screen negative have not been evaluated or followed up in registers). Under the assumption that all
obvious ASD cases in the population would be known at the CNC, we estimated that the false negative severe ASD cases at
autism screening are few. Four children with a failure on the language screen, but not on the autism screen, have later (in
2011) been referred from the SLP to the CNC for neuropsychiatric evaluations with some suspicion of ASD. It is, however, also
likely that the instruments used, designed to screen children with ASD symptoms detectable around the age of 2–3 years,
will miss a number of cases with mild symptoms (and more obvious symptoms later in childhood, including those with
typical Asperger syndrome). The ‘‘true specificity’’ and the ‘‘true NPV’’ for the methods used would not be possible to estimate
until long-term follow-up studies have been performed. In summary, however, the most important psychometric index of
any autism instrument used for population screening in primary care is a high PPV. The PPVs obtained for the methods used
in the present study were actually higher than in previously published studies (Kleinman, Robins et al., 2008; Robins, 2008).
It appears that the two instruments complement each other and optimise early detection of autism. All the early
symptoms seen in autism may be part of other early onset developmental disorders, ‘‘ESSENCE’’ (Gillberg, 2010), and some
symptoms might be variants of typical development. Trained medical staff at the CHCs, responsible for the screening, was
essential for the success of the screening programme as were the routines established a priori for further evaluations
whenever developmental problems were noted. A positive language screen at 2.5 year can be considered as an ‘‘indicator’’ of
other possible neurodevelopmental symptoms, but is not specific for autism (Gillberg, 2010; Miniscalco, Nygren, Hagberg,
Kadesjö, & Gillberg, 2006).
Albeit there is increasing knowledge concerning the very early signs (<2 years) of ASD, the scientific support for population
screening around the age of two-three years as part of the developmental programmes for young children in primary care is
much stronger. In Gothenburg there were practical circumstances to take into account in planning for screening as part of the
developmental surveillance programme. The given setting was the already existing 2.5-year-visit, but no routine check up at
2 years. A population screening at the age of 2 years, half a year earlier, would probably have been optimal.
The ‘‘ASD’’ trained nurses and doctors have learnt to use the instruments also beyond the routine screening. From the year
when the screening was introduced several children, both younger and older than 2.5 years, have been referred without
delay from the CHCs for further evaluations and have been diagnosed as having ASD. The increased knowledge among the
medical staff at the CHC s from the training courses and the introduction of autism screening have raised awareness and
skills for recognition of different symptoms in the autism spectrum. The introduced methods for autism screening in
combination with new routines for further assessment without delay have enabled early diagnosis and interventions for a
very considerable number of young children with autism in Gothenburg. In 2005, before the start of the screening 2 children
<3 years of age were referred and diagnosed with ASD (Nygren et al., 2011). In 2010 the corresponding number was 48. The
new screening programme for autism, reported here, is now established as part of the developmental programme.

Acknowledgements

We would like to express our sincere gratitude to all the parents, their children, and, not least, all the devoted child health
nurses and doctors who helped us make this study possible. In addition, we are grateful to Björn Kadesjö, head of the CNC,
who made the study possible in the first place. Mats Cederlund and Carina Gillberg helped in the collection of the data. This
study was supported by the Queen Silvia Jubilee Foundation, the Wilhelm and Martina Lundgren Foundation, the Swedish
Science Council (grant no B41-f 1883/09), and a grant for CG under the ALF-agreement.

References

Almgren, I., Nygren, G., Gillberg, C., & Gillberg, T. (2008). M-CHAT and M-CHAT with follow-up interview http://www2.gsu.edu/psydlr/Diana_L._
Robins,_Ph.D.html.
Al-Qabandi, M., Gorter, J. W., & Rosenbaum, P. (2011). Early autism detection: Are we ready for routine screening? Pediatrics, 128(1), e211–e217. doi:10.1542/
peds. 2010-1881.
Arvidsson, T., Holmberg, L., Reuter, A., & Strömbom, A. (2010). Barnhälsovårdsrapport verksamhetsåret 2010.
Baranek, G. T. (1999). Autism during infancy: A retrospective video analysis of sensory-motor and social behaviors at 9–12 months of age. Journal of Autism and
Developmental Disorders, 29(3), 213–224.
Banville, D., Desrosiers, P., & Genet-Volet, Y. (2000). Translating questionnaires and inventories using a cross-cultural translation technique. Journal of Teaching in
Physical Education, 19(3), 374–387.
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.
Bryson, S. E., Zwaigenbaum, L., Brian, J., Roberts, W., Szatmari, P., Rombough, V., et al. (2007). A prospective case series of high-risk infants who developed autism.
Journal of Autism and Developmental Disorders, 37(1), 12–24. doi:10.1007/s10803-006-0328-2.
Cox, A., Klein, K., Charman, T., Baird, G., Baron-Cohen, S., Swettenham, J., et al. (1999). Autism spectrum disorders at 20 and 42 months of age: Stability of clinical
and ADI-R diagnosis. Journal of Child Psychology and Psychiatry, 40(5), 719–732.
Dahlgren, S. O., & Gillberg, C. (1989). Symptoms in the first two years of life. A preliminary population study of infantile autism. European Archives of Psychiatry and
Neurologic l Sciences, 238(3), 169–174.
Dawson, G. (2011). Coming closer to describing the variable onset patterns in autism. Journal of the American Academy of Child and Adolescent Psychiatry, 50(8),
744–746. doi:10.1016/j.jaac.2011.04.011.
Dawson, G., Rogers, S., Munson, J., Smith, M., Winter, J., Greenson, J., et al. (2010). Randomized, controlled trial of an intervention for toddlers with autism: The
Early Start Denver Model. Pediatrics, 125(1), e17–e23. doi:10.1542/peds. 2009-0958.
1210 G. Nygren et al. / Research in Developmental Disabilities 33 (2012) 1200–1210

Dumont-Mathieu, T., & Fein, D. (2005). Screening for autism in young children: The Modified Checklist for Autism in Toddlers (M-CHAT) and other measures.
Mental Retardation and Developmental Disabilities Research Review, 11(3), 253–262. doi:10.1002/mrdd.20072.
Gillberg, C. (2010). The ESSENCE in child psychiatry: Early Symptomatic Syndromes Eliciting Neurodevelopmental Clinical Examinations. Research in Devel-
opmental Disabilities, 31(6), 1543–1551. doi:10.1016/j.ridd.2010.06.002.
Gillberg, C., Ehlers, S., Schaumann, H., Jakobsson, G., Dahlgren, S. O., Lindblom, R., et al. (1990). Autism under age 3 years: A clinical study of 28 cases referred for
autistic symptoms in infancy. Journal of Child Psychology and Psychiatry, 31(6), 921–934.
Howlin, P., Magiati, I., & Charman, T. (2009). Systematic review of early intensive behavioral interventions for children with autism. American Journal of Intellectual
and Developmental Disabilities, 114(1), 23–41. doi:10.1352/2009.114:23;nd41.
Höglund Carlsson, L., Gillberg, C., Lannerö, E., & Blennow, M. (2010). Autism: Screening toddlers with CHAT in a child health care programme did not improve early
identification. Acta Paediatrica doi:10.1111/j. 1651-2227.2010.01958.x.
Kleinman, J. M., Robins, D. L., Ventola, P. E., Pandey, J., Boorstein, H. C., Esser, E. L., et al. (2008). The modified checklist for autism in toddlers: A follow-up study
investigating the early detection of autism spectrum disorders. Journal of Autism and Developmental Disorders, 38(5), 827–839. doi:10.1007/s10803-007-
0450-9.
Kleinman, J. M., Ventola, P. E., Pandey, J., Verbalis, A. D., Barton, M., Hodgson, S., et al. (2008). Diagnostic stability in very young children with autism spectrum
disorders. Journal of Autism and Developmental Disorders, 38(4), 606–615. doi:10.1007/s10803-007-0427-8.
Kozlowski, A. M., Matson, J. L., Worley, J. A., Sipes, M., & Horovitz, M. (2011). Defining characteristics for young children meeting cutoff on the modified checklist
for autism in toddlers. Research in Autism Spectrum Disorders, 6, 472–479.
Lawton, K., & Kasari, C. (2011). Brief report: Longitudinal improvements in the quality of joint attention in preschool children with autism. Journal of Autism and
Developmental Disorders, 42(2), 307–312. doi:10.1007/s10803-011-1231-z.
Lord, C. (1995). Follow-up of two-year-olds referred for possible autism. Journal of Child Psychology and Psychiatry, 36(8), 1365–1382.
Lord, C., Risi, S., DiLavore, P. S., Shulman, C., Thurm, A., & Pickles, A. (2006). Autism from 2 to 9 years of age. Archives of General Psychiatry, 63(6), 694–701.
doi:10.1001/archpsyc.63.6.694.
Lord, C., Risi, S., Lambrecht, L., Cook, E. H., Jr., Leventhal, B. L., DiLavore, P. C., et al. (2000). The autism diagnostic observation schedule-generic: A standard measure
of social and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders, 30(3), 205–223.
Matson, J. L., Wilkins, J., Sharp, B., Knight, C., Sevin, J. A., & Boisjoli, J. A. (2009). Sensitivity and specificity of the Baby and Infant Screen for Children with aUtIsm
Traits (BISCUIT): Validity and cutoff scores for autism and PDD-NOS in toddlers. Research in Autism Spectrum Disorders, 3(4), 924–930.
Mattsson, C. M., Marild, S., & Pehrsson, N. G. (2001). Evaluation of a language-screening programme for 2.5-year-olds at Child Health Centres in Sweden. Acta
Paediatrica, 90(3), 339–344. doi:10.1111/j.1651-2227.2001.tb00315.x.
Miniscalco, C., Westerlund, M., & Lohmander, A. (2005). Language skills at age 6 years in Swedish children screened for language delay at 2 (½) years of age. Acta
Paediatrica, 94(12), 1798–1806. doi:10.1080/08035250500244242.
Miniscalco, C., Nygren, G., Hagberg, B., Kadesjo, B., & Gillberg, C. (2006). Neuropsychiatric and neurodevelopmental outcome of children at age 6 and 7 years who
screened positive for language problems at 30 months. Developmental Medicine and Child Neurology, 48(5), 361–366. doi:10.1017/s0012162206000788.
Mundy, P., Sullivan, L., & Mastergeorge, A. M. (2009). A parallel and distributed-processing model of joint attention, social cognition and autism. Autism Research,
2(1), 2–21. doi:10.1002/aur.61.
Norberg, L., Tingwall, V., & Ahlin-Åkerman, B. (1980). Standardisering av Griffiths utvecklingsskala för åldern 2–8 år. Stockholm: Högskolan för lärarutbildning,
Institutionen för pedagogik.
Noterdaeme, M., Sitter, S., Mildenberger, K., & Amorosa, H. (2000). Diagnostic assessment of communicative and interactive behaviours in children with autism
and receptive language disorder. European Child and Adolescent Psychiatry, 9(4), 295–300.
Nygren, G., Cederlund, M., Sandberg, E., Gillstedt, F., Arvidsson, T., Gillberg, I. C., et al. (2011). The prevalence of autism spectrum disorders in toddlers: A
population study of 2-year-old Swedish children. Journal of Autism and Developmental Disorders doi:10.1007/s10803-011-1391-x.
Nygren, G., Hagberg, B., Billstedt, E., Skoglund, A., Gillberg, C., & Johansson, M. (2009). The Swedish version of the Diagnostic Interview for Social and
Communication Disorders (DISCO-10). Psychometric properties. Journal of Autism and Developmental Disorders, 39(5), 730–741. doi:10.1007/s10803-008-
0678-z.
Oosterling, I. J., Wensing, M., Swinkels, S. H., van der Gaag, R. J., Visser, J. C., Wondenberg, T., et al. (2010). Advancing early detection of autism spectrum disorder by
applying an integrated two-stage screening approach. Journal of Child Psychology and Psychiatry, 51(3), 250–258. doi:10.1111/j. 1469-7610.2009.02150.x.
Ospina, M. B., Krebs Seida, J., Clark, B., Karkhaneh, M., Hartling, L., Tjosvold, L., et al. (2008). Behavioural and developmental interventions for autism spectrum
disorder: A clinical systematic review. PLoS One, 3(11), e3755. doi:10.1371/journal.pone.0003755.
Osterling, J. A., Dawson, G., & Munson, J. A. (2002). Early recognition of 1-year-old infants with autism spectrum disorder versus mental retardation. Development
and Psychopathology, 14(2), 239–251.
Pandey, J., Verbalis, A., Robins, D. L., Boorstein, H., Klin, A. M., Babitz, T., et al. (2008). Screening for autism in older and younger toddlers with the Modified
Checklist for Autism in Toddlers. Autism, 12(5), 513–535. doi:10.1177/1362361308094503.
Robins, D. L. (2008). Screening for autism spectrum disorders in primary care settings. Autism, 12(5), 537–556. doi:10.1177/1362361308094502.
Robins, D. L., & Dumont-Mathieu, T. M. (2006). Early screening for autism spectrum disorders: Update on the modified checklist for autism in toddlers and other
measures. Journal of Developmental and Behavioral Pediatrics, 27(2 Suppl.), S111–S119.
Robins, D. L., Fein, D., Barton, M. L., & Green, J. A. (2001). The Modified Checklist for Autism in Toddlers: An initial study investigating the early detection of autism
and pervasive developmental disorders. Journal of Autism and Developmental Disorders, 31(2), 131–144.
Rogers, S. J., Hepburn, S., & Wehner, E. (2003). Parent reports of sensory symptoms in toddlers with autism and those with other developmental disorders. Journal
of Autism and Developmental Disorders, 33(6), 631–642.
Rogers, S. J., & Vismara, L. A. (2008). Evidence-based comprehensive treatments for early autism. Journal of Clinical Child and Adolescent Psychology, 37(1), 8–38.
doi:10.1080/15374410701817808.
Sparrow, S., Balla, D., & Cicchetti, D. (1984). The Vineland Adaptive Behavior Scales: Circle Pines MN: American Guidance Service.
Statistics Sweden. (2010/2011). Statistiska centralbyrån/Statistics Sweden http://www.scb.se/.
Teitelbaum, P., Teitelbaum, O., Nye, J., Fryman, J., & Maurer, R. G. (1998). Movement analysis in infancy may be useful for early diagnosis of autism. Proceedings of
the National Academy of Sciences of the United States of America, 95(23), 13982–13987.
Trillingsgaard, A., Ulsted Sorensen, E., Nemec, G., & Jorgensen, M. (2005). What distinguishes autism spectrum disorders from other developmental disorders
before the age of four years? European Child and Adolescent Psychiatry, 14(2), 65–72. doi:10.1007/s00787-005-0433-3.
Ventola, P., Kleinman, J., Pandey, J., Wilson, L., Esser, E., Boorstein, H., et al. (2007). Differentiating between autism spectrum disorders and other developmental
disabilities in children who failed a screening instrument for ASD. Journal of Autism and Developmental Disorders, 37(3), 425–436. doi:10.1007/s10803-006-
0177-z.
Werner, E., Dawson, G., Munson, J., & Osterling, J. (2005). Variation in early developmental course in autism and its relation with behavioral outcome at 3-4 years
of age. Journal of Autism and Developmental Disorders, 35(3), 337–350.
Werner, E., Dawson, G., Osterling, J., & Dinno, N. (2000). Brief report: Recognition of autism spectrum disorder before one year of age: A retrospective study based
on home videotapes. Journal of Autism and Developmental Disorders, 30(2), 157–162.
Willemsen-Swinkels, S. H., Buitelaar, J. K., & van Engeland, H. (2001). Is 18 months too early for the CHAT? Journal of the American Academy of Child and Adolescent
Psychiatry, 40(7), 737–738. doi:10.1097/00004583-200107000-00002.
Wing, L., Leekam, S. R., Libby, S. J., Gould, J., & Larcombe, M. (2002). The Diagnostic Interview for Social and Communication Disorders: Background, inter-rater
reliability and clinical use. Journl of Child Psychology and Psychiatry, 43(3), 307–325.
Yama, B., Freeman, T., Graves, E., Yuan, S., & Karen Campbell, M. (2012). Examination of the Properties of the Modified Checklist for Autism in Toddlers (M-CHAT) in
a Population Sample. Journal of Autism and Developmental Disorders, 42, 23–24.