REVIEW Open Access

Clinical practice guidelines for diagnosisof autism spectrum disorder in adultsand children in the UK: a narrative reviewJennie Hayes* , Tamsin Ford, Hateem Rafeeque and Ginny Russell

Abstract

Background: Research suggests that diagnostic procedures for Autism Spectrum Disorder are not consistent acrosspractice and that diagnostic rates can be affected by contextual and social drivers. The purpose of this review wasto consider how the content of clinical practice guidelines shapes diagnoses of Autism Spectrum Disorder in theUK; and investigate where, within those guidelines, social factors and influences are considered.

Methods: We electronically searched multiple databases (NICE Evidence Base; TRIP; Social Policy and Practice; USNational Guidelines Clearinghouse; HMIC; The Cochrane Library; Embase; Global health; Ovid; PsychARTICLES;PsychINFO) and relevant web sources (government, professional and regional NHS websites) for clinical practiceguidelines. We extracted details of key diagnostic elements such as assessment process and diagnostic tools. Aqualitative narrative analysis was conducted to identify social factors and influences.

Results: Twenty-one documents were found and analysed. Guidelines varied in recommendations for use of diagnostictools and assessment procedures. Although multidisciplinary assessment was identified as the ‘ideal’ assessment, someguidelines suggested in practice one experienced healthcare professional was sufficient. Social factors in operational,interactional and contextual areas added complexity to guidelines but there were few concrete recommendations as tohow these factors should be operationalized for best diagnostic outcomes.

Conclusion: Although individual guidelines appeared to present a coherent and systematic assessment process, theyvaried enough in their recommendations to make the choices available to healthcare professionals particularly complexand confusing. We recommend a more explicit acknowledgement of social factors in clinical practice guidelines withadvice about how they should be managed and operationalised to enable more consistency of practice andtransparency for those coming for diagnosis.

Keywords: Autism spectrum disorder, Diagnosis, Clinical guideline, Narrative review, Social factors, Diagnosticuncertainty, Clinical judgement

BackgroundThe diagnosis of autism poses particular challenges forhealthcare professionals (HCPs) as, in common with otherneurodevelopmental disorders and most psychiatric disor-ders, there are no biomarkers utilised in clinical practice[1–3]. In addition, the condition is heterogeneous, withwide ranging levels of severity and symptom expressionand characteristics common to autism may occur inpeople with other conditions [4]. Those coming for

diagnosis may also have symptoms of other conditionssuch as epilepsy, learning disability or sleep disorders, forexample, complicating diagnosis further, with some argu-ing for a de-compartmentalisation of these conditions inyounger children [5]. The ‘gold standard’ of diagnosis isconsidered to be consensus agreement within amulti-agency team [6, 7]. However, negotiating consensusbetween HCPs with different training, professional roles,experience and knowledge can be challenging and timeconsuming. Finally, a review of the accuracy, reliability,validity and utility of reported diagnostic tools and assess-ments found that many diagnostic instruments for autism

* Correspondence: Jennie.Hayes@exeter.ac.ukUniversity of Exeter Medical School, St Luke’s Campus, University of Exeter,Exeter EX1 2LU, UK

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Hayes et al. BMC Psychiatry (2018) 18:222 https://doi.org/10.1186/s12888-018-1800-1

lack a high-quality independent evidence base [6]. For ex-ample, only three instruments - the Autism DiagnosticObservation Schedule (ADOS), Autism Diagnostic Inter-view Revised (ADI-R) and the Childhood Autism RatingScale (CARS) - had a strong supporting evidence base [6].Given the potential challenges, clinical practice guide-

lines (CPGs) perform an important role in informing HCPsof best practice. CPGs are ‘systematically developed state-ments to assist practitioner and patient decisions about ap-propriate health care for specific clinical circumstances’ [8].National CPGs in the UK help to provide evidence-basedrecommendations to support Autism Strategies and ActionPlans [9] and form the guidance framework for HCPsundertaking assessment and diagnosis of autism in the UK.In addition to CPGs produced by specialist, governmentsupported healthcare associations, for example, theScottish Intercollegiate Guidelines Network (SIGN) [10],professional clinical bodies also publish discipline-specificpractice parameters or position papers, for example, theRoyal College of Psychiatrists (RCPsych) [11].

Social factorsAlthough CPGs aim to inform diagnostic practice, re-search suggests that diagnostic and assessment proceduresvary in practice [9]. Diagnosis is dependent on observingsocially-based behaviours that are arguably not necessarilycharacteristic of the person under assessment but arisefrom two-way social relationships and social context.Assessment mechanisms include drawing informationfrom a range of sources, including clinician observation,reporting from family members and wider contexts suchas school or workplace. This means that assessments arecontextual and inter-relational and symptoms may changeaccording to context or interpersonal relationship, makingdifferent assessment sources potentially contradictory.Some studies show that social factors such as individ-

ual patient preference, availability of resources or localorganisational factors can shape diagnostic practice, in,for example, heart disease [12]. Studies in autism havealso shown how diagnostic rates can be affected bycontextual and social drivers, such as diagnostic re-sources [13] or diffusion of information about autismthrough social networks [14]. Where there is diagnosticuncertainty clinicians may ‘upgrade’ to a diagnosis ofautism if they believe it would be in the best interests ofthe patient; if the diagnosis would trigger appropriateservices and funding; or counteract the limitations ofdiagnostic tools, particularly in atypical presentations[15, 16]. It seems, in practice, clinicians may adopt apragmatic, practical or functional approach.

Socio-economic and cultural factorsResearch has shown that lower social economic status(SES) is associated with increased parent-reported

prevalence [17], contrasting with the US where higherSES and parental education is linked to increased likeli-hood of diagnosis [14, 18]. Research also suggests thatpeople with autism from Black, Asian and MinorityEthnic (BAME) communities are less likely to bediagnosed with autism or access appropriate services[19] despite research which shows that behavioursassociated with autism are likely to be consistent acrosscultures and countries [20].Prior to diagnosis, social factors can also determine

who comes forward for diagnosis and who is referred forfurther assessment. Research examining a longitudinalUK cohort study identified that with the severity of aut-istic traits held constant, younger mothers and mothersof first-born children were significantly less likely to havechildren diagnosed with ASD [21]. In addition, boyswere more likely to receive a diagnosis than girls, andmaternal depression was linked with a lack of diagnosis[21]. These findings suggest both cultural and economicinfluences impact the diagnostic pathway.

Biomarkers in autism diagnosisThere is a great deal of research that explores the under-lying neurobiological, genetic, chemical and cognitivefactors that may, in future, provide biomarkers whichcould be utilised in autism diagnosis (see [22] for a reviewof genetic, metabolic and brain focused biomarkers). Forexample, a recent research study has identified a link be-tween damage to proteins in blood plasma and autismsymptoms [23]; while another found shared brain activitybetween boys diagnosed with ASD and those withobsessive-compulsive disorder (OCD) which in turn dif-fered from a non-diagnosed control group [24]. However,it has been argued that the heterogeneous and interactivenature of autism symptoms makes the identification ofclinically useful biomarker tests problematic [25]. Further-more, findings from biomarker research have yet to beintegrated with clinical practice and none currently haveenough evidence to support routine clinical use [22]. Forthe foreseeable future, therefore, these developments areunlikely to change diagnostic practice [26].

Purpose of the reviewAlthough a few studies have begun to explore healthprofessionals’ views of autism diagnosis [16, 27, 28], toour knowledge there are few studies that examine howclinical guidelines may inform assessment. One excep-tion is a recent systematic review of English speakingguidelines undertaken by Penner et al. [29] whichreported that guidelines varied considerably in quality,content and recommendations but included guidelinesworking across incomparable health systems in differentcountries. We therefore carried out a focussed narrativereview of guidelines that impact on UK-based practice.

Hayes et al. BMC Psychiatry (2018) 18:222 Page 2 of 25

Penner et al. suggest that in the face of disparate clinicalguidance clinicians should ‘be mindful of local resourcesand wait times, eligibility requirements for ASD ser-vices…and the wishes of families when deciding on howbest to assess for ASD’ [29]. Our narrative review re-sponds to this call for a pragmatic approach by investi-gating where, within guidelines, social factors andinfluences such as those suggested are considered.

MethodScoping searchA scoping search was undertaken to check there was nosimilar review published. A search was made in the fol-lowing databases; PsychARTICLES; Embase; Globalhealth; HMIC; Ovid (books; medline; journals); Psy-chINFO; Social policy and practice. One relevant articlewas retrieved [29], as discussed above.

Inclusion and exclusion criteriaFull inclusion and exclusion criteria are in Table 1.Whilst we took a broad approach to CPGs, including,for example, journal articles summarizing national CPGsand the diagnostic process, as well as national CPGs, theresearchers acknowledge that each of these type ofguidelines have different purposes (see Table 2). How-ever, we argue that each may have an impact on HCP’s

process of diagnosis to a greater or lesser extent and forthe purposes of this study all were included under theterm clinical practice guidelines.

Identification of CPGsWe did not set out to undertake a comprehensive sys-tematic review, as it was not a requirement of our studythat we consider risk of bias either within or acrossstudies [30]. However, we took a PRISMA approach toour search strategy, borrowing from systematic reviewmethodology in terms of screening titles and abstractsand data extraction techniques [31]. A systematic searchwas conducted in June 2017 using the following data-bases: NICE Evidence Base; TRIP; Social Policy andPractice; US National Guidelines Clearinghouse; HMIC;The Cochrane Library. In addition, searches were madeof government related websites and relevant professionalbodies as well as NICE and SIGN. We used the follow-ing search terms to search all databases and websites:‘autism’, ‘diagnosis’, ‘guidance’, ‘statutory’, ‘clinical’, ‘practice’,‘guideline’, ‘protocol’, ‘strategy’, ‘policy’, ‘bill’, ‘act’, and ‘par-ameter’. A full search strategy is in Fig. 1.

Study selectionThe first reviewer (JH) removed duplicates and screenedtitles for relevance. Full text copies of the potentially rele-vant documents were downloaded for screening. The firstreviewer screened full text documents and excluded thosenot relevant. The remaining titles were independentlychecked by the clinical specialist (TF) using pre-specifiedinclusion/exclusion criteria (outlined in Table 1). Discrep-ancies were resolved by discussion, with involvement of athird reviewer (GR). Twenty-eight documents wereconsidered for analysis, with seven being withdrawn at fullanalysis stage. See Fig. 2 for full details.Guidelines from the International Classification of

Mental and Behavioural Disorders (Tenth edition)(ICD-10) [32] and the Diagnostic and Statistical Manualof Mental Disorders (Fifth edition) (DSM-5) [33] wereconsidered alongside UK relevant guidelines as they areconsidered authoritative sources for the definition ofsymptoms utilized in autism diagnosis, as well as otherneurological conditions.

Data extractionA data extraction framework was created to draw keycharacteristics from the guidelines (year, author, geo-graphical remit, target audience, age range, range ofdiagnoses covered, age at which symptoms are recog-nised, diagnostic criteria referred to); as well as key ele-ments in the diagnostic process (recommended tools,role and composition of the multidisciplinary team(MDT), who can diagnose, assessment targets and keyfeatures of assessment). This framework was piloted with

Table 1 Inclusion and Exclusion Criteria

Inclusion Criteria

Documents with guidance-based status for HCPs working in secondarycare in the UK; or were published papers, aimed at HCPs, with the aimof reviewing CPGs

Documents related to autism diagnosis and assessment for eitherchildren, adults or both

Documents produced either by or through government orprofessional clinical bodies or published in a journal aimed at HCPs

Documents related to diagnosis and assessment in UK (England,Scotland, Wales and N Ireland)

Documents dated from 2009 (reflecting publication of the first UKspecific Autism Act) or were the most recent CPG published by a keyprofessional body

Exclusion Criteria

Documents related solely to referral, treatment, prognosis or supportservices

Reviews of diagnostic criteria and other academic papers

Guidelines related to primary care as we were interested in diagnosisrather than referral

Narrative reviews, editorials and opinions

Documents related to parliament or legislature; national or regionalstrategies as they are not the primary source for clinicians

Local guidance

Guidance provided by private providers of diagnostic services

International professional body guidelines (other than ICD/DSM)

Hayes et al. BMC Psychiatry (2018) 18:222 Page 3 of 25

four reviewers (JH, GR, RW and DE) in a comparison ofanalysis of three guidelines. The framework wasamended accordingly and is included in Additional file 1.Data were independently extracted by two reviewers (JHand HR) from 21 CPGs and disagreements were resolvedby discussion and further checks. Data were tabulatedand analysed.

Analysis of social factorsA modified form of narrative review, as described byPopay et al. [34] and Ferrari [35], was adopted wherebydata extraction enabled synthesis of key data, whilst alsoallowing rich narrative description [35]. Narrative reviewwas selected as it enabled the telling of the ‘story’ ofCPGs, and consideration of how guidelines, as a set oftexts, shape diagnosis [34].A process of inductive analysis was undertaken based

on social factors and influences. These were defined, forthe purpose of this review, as contextual factors that in-fluence diagnosis but are not based on symptoms of aut-ism. We drew from the concept of a social model ofdiagnosis as developed by Jutel and Nettleton [36]. Thismodel considers how diagnostic classifications and med-ical diagnoses are socially created and how social forces– including technological, professional, cultural and

economic forces – contribute to shaping aspects of thediagnostic process including those related to classification,the consequences of diagnosis and the process of diagnosisitself [36]. Overall, a social model challenges the idea ofdiagnosis as ‘a moment of clinical purity’ [37] or as a waysimply to identify underlying biological problems. We in-cluded factors that were relevant to multidisciplinary work-ing or parental/family influence (the process of diagnosis);the potential outcomes of diagnosis for the patient and howHCPs may take this into account (the consequences ofdiagnosis); and how issues around classification shape thediagnostic process such as how borderline cases are dealtwith (diagnosis as a category). This was a dynamic processwhereby data extracts were considered in relation to eachother via conceptual mapping and clustering [34].

TerminologyFor the purposes of this review and in line with the Aut-ism Strategy [38] we use the term ‘autism’ throughout.

ResultsCharacteristics of guidelinesA total of 236 documents were retrieved, and 21 wereincluded in the final narrative review (see Table 3 for fulllist of included documents and guideline characteristics).

Table 2 Purpose of Diagnostic Guidelines

Type of guideline General purpose of type of guideline

Diagnostic Criteria To assist clinicians in the diagnosis of mental conditions by providing descriptions of the main clinical features ineach category

National Clinical Guidelines To offer best practice advice and guidance for professionals and service users and their families

Guidelines from ProfessionalBodies

To offer profession specific advice to clinicians and healthcare professionals in their specialist area

Journal Articles To summarise clinical guidelines in clinician-facing publications to keep clinicians up to date and/or alert them tochanges in good practice

Fig. 1 Full Search Strategy

Hayes et al. BMC Psychiatry (2018) 18:222 Page 4 of 25

The documents studied are grouped into four types: a)International Diagnostic Criteria (n = 2); b) NationalClinical Guidelines (n = 5); c) Journal articles thatsummarize National Clinical Guidelines and the diag-nostic process, published in key clinical journals (n = 10);d) Guidelines from professional bodies (n = 4). It shouldbe noted that journal articles, in some cases, are de-signed to give an update rather than a full guidelinetherefore the lack of detail in some areas should not ne-cessarily be seen as a weakness.Of the 21 guidelines considered, six dealt with diagno-

sis of adults, seven with children and eight with all ages.

Of those, two guidelines were international but key todiagnostic practice in the UK (ICD-10 and DSM-5),five related to the UK as a whole, five to England andWales, one to Scotland, two to Northern Ireland andone to outside the US and Canada (and therefore in-cluded the UK). Five guidelines did not specify a geo-graphical remit but were published in the UK inclinician-facing journals. All guidelines were aimed atHCPs, with six aimed at particular specialist rolesthat included psychiatrists, psychologists, speech andlanguage therapists, community practitioners andpaediatricians.

Fig. 2 Study selection flow diagram

Hayes et al. BMC Psychiatry (2018) 18:222 Page 5 of 25

Table 3 Key characteristics of guidelines

Title Year Author(s) Publisher/Journal

Geographicalremit

Targetaudience

Agerange

Range ofdiagnosescovered

Diagnosticcriteriareferred to

Age at whichsymptoms arerecognised

DIAGNOSTIC CRITERIA

The ICD-10Classificationof Mental andBehaviouralDisorders:clinical de-scriptions anddiagnosticguidelines[32]

1993 N/A World HealthOrganisation

International Clinical,educationaland serviceuse

All ages Pervasivedevelopmentdisorders

N/A Before age of3 years(childhoodautism); after age3 (atypicalautism).

Diagnosticand StatisticalManual ofMentalDisorders(Fifth Edition)[33]

2013 N/A AmericanPsychiatricAssociation

International Clinicians,students,practitioners,researchers

All ages AutismSpectrumDisorder

N/A During 2nd yearof life (12–24 months) orearlier than12 months ifdevelopmentaldelays are severe

NATIONAL CLINICAL GUIDELINES

NICE Autismin under 19 s:recognition,referral anddiagnosis(NICE CG128)[39]

2011 NationalCollaboratingCentre forWomen’s andChildren’sHealth

NationalInstitute forHealth andCareExcellence(NICE)

England andWales

Healthcareprofessionals

Frombirth upto19 years

Pervasivedevelopmentaldisorder (PDD)

ICD-10 orDSM-IV

May beuncertaintybefore24 months, orwithdevelopmentalage of less than18 months

Six Steps ofAutism Carefor childrenand youngpeople inNorthernIreland(RASDN) [44]

2011 RegionalAutisticDisorderNetwork forNorthernIreland

Health andSocial CareBoard

NorthernIreland

Health careand educationprofessionals,parents, carers,service usersand providers.

Up tothe ageof18 years

Autismspectrumdisorder

ICD-10, DSM-IV, NICE, SIGN,NZ Guide-lines, NHSMap ofMedicine

Pre-school.Language delayby the age oftwo years.

AutismSpectrumDisorder inadults:diagnosis andmanagement(NICE CG142)[9]

2012 NationalCollaboratingCentre forMental Health

NationalInstitute forHealth andCareExcellence(NICE)

England andWales

Health andsocial careproviders andcommissioners

Adultsaged 18and over

Autismspectrumdisorders

N/S*ICD-10specified infull version ofCG142 [62]

N/A

Autism AdultCare Pathway(RASDN) [54]

2013 RegionalAutisticSpectrumDisorderNetwork

Health andSocial CareBoard

NorthernIreland

Professionals,adults andfamilies

Adultsfrom age18

Autismspectrumdisorders

DSM-5 andICD-10, NICEguidanceCG142.

N/S

Assessment,diagnosis andinterventionsfor autismspectrumdisorders: Anationalclinicalguideline(SIGN 145)[10]

2016 N/A ScottishIntercollegiateGuidelinesNetwork

Scotland Healthcareprofessionals

Wholeagerange

Autismspectrumdisorder

ICD-10 andDSM-5

Autism can bereliablydiagnosedbetween theages of 2–3.

Hayes et al. BMC Psychiatry (2018) 18:222 Page 6 of 25

Table 3 Key characteristics of guidelines (Continued)

Title Year Author(s) Publisher/Journal

Geographicalremit

Targetaudience

Agerange

Range ofdiagnosescovered

Diagnosticcriteriareferred to

Age at whichsymptoms arerecognised

GUIDELINES FROM PROFESSIONAL BODIES

RCSLT (RoyalCollege ofSpeech andLanguageTherapistsClinicalGuidelines(Autism) [41]a

2005 N/A Royal Collegeof Speech andLanguageTherapists

UK Speech andlanguagetherapists

Childrenandadults

Autismspectrumdisorder

ICIDH-2 (forgeneralclinicalassessment)

N/S

Good practicein themanagementof autism(includingAspergersyndrome) inadults(RCPychCR191) [11]

2014 Royal CollegeofPsychiatrists

Royal Collegeof Psychiatrists

UK Psychiatristsworking withadults of atleast normalintellectualability

Adultsfrom age18

Autism ICD-10, DSM-5, NICE, 2012.

N/S

AutismSpectrumDisorders:Guidance forPsychologists(BPS) [40]b

2016 Stuart-Hamilton,Dillenburger,Hood &Austin

BritishPsychologicalSociety

UK Psychologists All ages AutismSpectrumDisorder

ICD-10 andDSM-5, NICE,2011.

Both diagnosticmanuals considerASD indicators tobe present by theage of 36 monthsalthough somechildren can beidentified underthe age of24 months.

BMJ BestPracticeonlineresource [43]

2017 Parr&Woodbury-Smith

British MedicalJournal

Outside USand Canada

MedicalPractitioners

All ages AutismSpectrumDisorder

DSM-IV, DSM-5 & ICD-10.NICE, SIGN,AACAP, AAP,NZ ASDguideline,AAN

More than 80%of children withASD show clearbehavioural signsby the age of24 months, someindicators in 12–18 months

JOURNAL ARTICLES

Diagnosis andmanagementof autism inchildhood[47]

2011 Blenner,Reddy &Augustyn

British MedicalJournal

N/S Generalclinicians

Children AutismSpectrumDisorder

DSM-IV TR orICD-10

N/S

Diagnosis andassessment inautismspectrumdisorders [48]

2012 Carpenter Advances inMental HealthandIntellectualdisabilities

N/S Thosedesigning andprovidingdiagnosticservices

All ages AutismSpectrumDisorder

DSM-IV TR orICD-10. Gill-berg’s for AS.There areothers butfew use them(Kopra et al.,2008; Chiap-pedi et al.,2010).

N/S

Autismspectrumdisorder inadults: clinicalfeatures andthe role ofthepsychiatrist[49]

2013 Garland,O’Rourke &Robertson

Advances inPsychiatricTreatment

UK Psychiatrists Adults AutismSpectrumDisorders

ICD-10 andDSM-5, NICE

To satisfy ICD-10criteria for child-hood autism, im-pairments mustmanifest beforethe age of3 years

Hayes et al. BMC Psychiatry (2018) 18:222 Page 7 of 25

Guidelines acknowledged that there is variation in ratesof identification, assessment criteria and practice [9]; thatthere is increasing demand for diagnostic services [39]; andthat increased awareness of autism is likely to lead to a risein people presenting for assessment [40].

Definitions of autismDefinitions of autism in ICD-10 and DSM-5 differed.ICD-10 took a categorical approach with a definition ofPervasive Development Disorders that includedsub-diagnoses within it; whilst DSM-5 used the overarching

Table 3 Key characteristics of guidelines (Continued)

Title Year Author(s) Publisher/Journal

Geographicalremit

Targetaudience

Agerange

Range ofdiagnosescovered

Diagnosticcriteriareferred to

Age at whichsymptoms arerecognised

Recognising,referring anddiagnosingautism [45]

2012 Howlett &Richman

Every ChildJournal

England andWales

Professionalsworking withchildren andyoung people

Childrenandyoungpeople

Autism NICE The core autismbehaviours aretypically presentin earlychildhood; butfeatures canappear differentwith age orchange withcircumstances

Autism [50] 2013 Lai,Lombardo &Baron-Cohen

The Lancet N/S N/S All ages Autism or theautismspectrum

DSM-5, ICD-10

N/S

Autism [51] 2009 Levy, Mandell& Schultz

The Lancet N/S N/S N/S butprimarilytalksaboutchildren

AutismSpectrumDisorder

DSM-IV andICD-10

Parents oftenaware from age18 months, adiagnosis is oftennot made until2 years after theinitial expressionof parentalconcern.

Autismspectrumdisorder:diagnosis andmanagement[53]

2009 O’Hare Archives ofDisease inChildhood:Education andPracticeEdition

N/S butrelatesprimarily toSIGNguidelines

Paediatricians Childrenandyoungpeople

AutismSpectrumDisorder

ICD-10 andDSM-IV, SIGN

N/S

Recognition,referral,diagnosis,andmanagementof adults withautism:summary ofNICEguidance [58]

2012 Pilling, Baron-Cohen,Megnin-Viggars, Lee &Taylor

British MedicalJournal

England andWales

N/S Adults Autism N/S N/S

AutismSpectrumDisorders inchildhood: aclinicalupdate [46]

2011 Reynolds CommunityPractitioner

UK Communitypractitioners

Children AutismSpectrumDisorder

ICD-10, DSM-IV

N/S

The NICEguideline onrecognition,referral,diagnosis andmanagementof adults onthe autismspectrum [52]

2014 Wilson,Roberts,Gillan, Ohlsen,Robertson &Zinkstok

Advances inMental HealthandIntellectualDisabilities

England andWales

Health careprofessionals,servicemanagers,service users,practitioners

All adults Autismspectrumdisorder

N/S N/S

aPre 2009 but constitutes current guideline in use from RCSLTbCurrently under review but represents the most recent published guideline from BPS

Hayes et al. BMC Psychiatry (2018) 18:222 Page 8 of 25

dimensional concept of Autism Spectrum Disorder. Someinconsistencies were present related to the differences inclassification in ICD-10 and DSM-5, therefore, for example,Rett’s Syndrome and Asperger’s Syndrome weresub-diagnoses of Pervasive Development Disorders inICD-10, but were encompassed in the overarching diagno-sis of Autism Spectrum Disorder in DSM-5 [32, 33]. Defini-tions of autism in all other guidelines considered in thisstudy were broadly consistent with the idea of a ‘spectrum’.Most guidelines (n = 14) referred to symptom criteria

from both ICD-10 and the (then) current version ofDSM (DSM-IV up to 2012 and DSM-5 from 2013), witheight guidelines recommending that HCPs should usethe current version of DSM or ICD criteria for diagnosis.Exceptions were NICE CG142, which was based onICD-10, [9]; Royal College of Speech and LanguageTherapists (RCSLT) [41], which drew on the Inter-national Classification of Functioning, Disability andHealth (ICIDH-2) for general clinical assessment [42];and journal articles describing NICE guidelines whichmade no mention of DSM/ICD (n = 3).Overall, therefore, the guidelines were mixed in their

recommended sources for symptom criteria due to thecurrent differences in the two classification systems.

Narrative review of social factorsWe used three inter-related elements as an organisingframework to describe the social factors identified inclinical guidelines: operational, interactional and con-textual. These factors do not stand alone from eachother, indeed, they appear to have a dynamic andinter-dependent relationship, however, organising themprovides a way to map their range and scope (see Fig. 3).

Operational factorsOperational factors included how different assessmentprocesses impact on the diagnostic decision, such aswhich tools and processes are engaged and when; whatconstitutes an assessment; and whether the decisionstake place as part of diagnosis or formulation. Table 4outlines some of these operational factors.

The assessment processOne guideline suggested that clinical practice variesgreatly [43] and we found this to be mirrored in CPGswith a wide range of potential assessment processes in-cluded. DSM-5 recommended that a diagnostic assess-ment should include gathering multiple sources ofinformation from clinician’s observations, caregiver

Fig. 3 Social factors in clinical guidelines

Hayes et al. BMC Psychiatry (2018) 18:222 Page 9 of 25

history and self-report (where possible). National guide-lines, although providing far greater detail, tended toinclude these areas and additionally suggested various otherdetailed assessments such as gathering wider functional/as-sessment information [10]; using documentary evidence,assessing risks, and assessment of challenging behaviour[9]; assessing for co-conditions [9, 39]; physical examination[39]; comprehensive educational assessment [44]; assess-ment of communication, neuropsychological functioning,motor and sensory skills, and adaptive functioning [10].Professional guidelines added other factors such as compre-hensive cognitive assessment [40] and impact of individual’smental health [41], accounts of relationships in differentsettings [11] and observation in school or another setting[43]. Journal articles tended to reflect national guidelinesand varied in the level of detail outlined for assessment fac-tors. Two articles gave little detail of assessment processesbut one referred readers directly to NICE guidelines for fur-ther detail [45] and the other was aimed at communitypractitioners who would be more likely to be involved in re-ferral than diagnosis [46]. Articles also included assessmentof co-occuring conditions (e.g. [47–52]) and a physical ormedical examination (e.g. [47, 50]). Additional assessmentareas included assessment of specific domains such as fam-ily stressors and coping abilities [47]. In one guideline [48]it was suggested that some clinicians bypass ICD/DSM cri-teria and instead undertake:

‘…testing for specific underlying difficulties such aslack of theory of mind or lack of central coherenceand then using these to decide the presence of thebehavioural criteria’ [48].

The RCSLT guideline [41] differed from most by sug-gesting consideration of theories relating to the triad ofsocial impairments, such as executive functioning defi-cits, motivation, memory and central coherence, as wellas social interaction and communication. However, some(e.g. [40]) suggested cognitive or neuropsychologicaltesting whilst SIGN guidelines stated that such assess-ments are ‘useful for individual profiling but are notdiagnostic instruments’ [10]. This anomaly may reflectthe specialist role of SLTs in the diagnostic process.Overall, we would concur with a reflection in one

guideline, which noted how the HCP may be faced with‘possible uncertainty as to where to go next in their in-vestigation framework as this could be potentially enor-mous’ [53].

Diagnostic toolsRecommendations about the use of diagnostic tools weremixed. One third of the guidelines (n = 7) did not specifyany particular tool for diagnostic assessment. OtherCPGs tended to suggest the consideration of a range of

tools without specifically recommending any particularinstrument(s), although regular references were made toADOS (n = 13), ADI-R (n = 11), DISCO (n = 9) and 3di(n = 6). The NICE guideline for children and youngpeople emphasised use of DSM/ICD criteria rather thantools; the NICE guideline for adults did the opposite [9,39]. Overall, findings concurred with Penner et al. inthat guidelines varied substantially in their recommenda-tions for use of diagnostic tools [29].

Diagnosis and formulationThere were differences in the way guidelines describedthe relationship between, or referred to, diagnosis,assessment, profiling, needs assessment and widerformulation. All guidelines encompassed the concept ofa wider (needs related) assessment but few explicitly sep-arated out these processes or discussed how this relatedto a diagnostic assessment. One exception to this wasthe Regional Autistic Spectrum Disorder Network(RASDN) children’s guideline, which separated the diag-nostic from the formulation process, describing thelatter as including examination of the person’s widerenvironment:

‘The outcome of the formulation should be tounderstand an individual in a more global holistic wayrather than merely in terms of signs and symptoms,as in the case of diagnosis’ [44].

The RCPsych guideline suggested that diagnosis isonly one component of the wider multidisciplinary exer-cise [11]. Some guidelines did not mention formulationbut suggested a profile of strengths, abilities and weak-nesses should be carried out alongside a diagnosticassessment (e.g. [10, 39]). Adult guidelines from RASDNseparated out a diagnostic assessment from a full needsassessment [54]; NICE guidelines for adults consideredcomprehensive assessment to include diagnostic,needs and risk assessment [9]; whilst the full chil-dren’s guidelines similarly brought together the diag-nostic and needs elements under ‘autism diagnosticassessment’, explaining that:

‘..the label of autism does not constitute a completediagnostic assessment and a profile of the child oryoung person’s strengths and weaknesses is alsoessential. This requires a multidisciplinary team whichhas the skills to undertake the assessments necessaryfor profiling’ [55].

Operationally, therefore, there were contradictionsbetween guidelines about what constitutes the diagnosticprocess, how it should be structured and which diagnos-tic tools should be used.

Hayes et al. BMC Psychiatry (2018) 18:222 Page 10 of 25

Table

4Keydiagno

sticrecommen

datio

ns

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

DIAGNOST

ICCRITE

RIA

ICD-10

(1993)

[32]

N/S

N/S

N/S

N/S

Diagn

oseon

thebasisof

behaviou

ral

features

DSM

-5(2013)

[33]

Nospecifictool

N/S

N/S

N/S

Careful

clinicalhistory&summaryof

social,p

sycholog

ical&biolog

ical

factors.

Multip

lesourcesof

inform

ation:

•clinician’sob

servations

•caregiverhistory

•self-repo

rt(whe

repo

ssible)

Clinicaljudg

emen

t

NATIONALCLINICALGUIDELINES

NICE

CG128

(2011)

[39]

Nospecifictool

recommen

ded

Autism

team

mem

bersshou

ldcarry

outassessmen

t(sho

rtversion).A

diagno

siscanbe

madeby

asing

leexpe

rienced

HCP;profile

ofstreng

ths

&weaknessesisessential,and

requ

iresMDT[55]

(fullversion).

Autism

team

madeup

ofPaed

iatrician&/or

Child

&Ado

lescen

tPsychiatrist,SLT,Clinical&/or

EducationalP

sycholog

ist&access

topaed

iatrician/paed

iatricne

urolog

ist,

Child

&Ado

lescen

tPsychiatrist,

EducationalP

sycholog

ist,Clinical

Psycho

logist,O

T,ifno

tin

team

.Also

consider

specialisthe

alth

visitoror

nurse,specialistteache

ror

social

worker.

Starttheautism

diagno

stic

assessmen

twith

in3mon

ths

ofreferral.Follow

upappo

intm

entwith

in6weeks

ofassessmen

t.

Seek

repo

rtfro

mthepre-scho

olor

scho

ol;g

athe

radditio

nalh

ealth

orsocialcare

inform

ation.Includ

ein

everyautism

diagno

sticassessmen

t:•qu

estio

nsabou

tparent/carer/child’s

concerns

•de

tails

ofthechild’sexpe

riences

ofho

melife,ed

ucationandsocialcare

•de

velopm

entalh

istory,focusingon

developm

entaland

behaviou

ral

features

•assessmen

t(throu

ghinteraction

with

andob

servationof

thechild

oryoun

gpe

rson

)of

socialand

commun

icationskillsand

behaviou

rs•med

icalhistory,includ

ingpren

atal,

perin

atalandfamily

history,and

pastandcurren

the

alth

cond

ition

s•ph

ysicalexam

ination

•considerationof

thedifferential

diagno

sis

•system

aticassessmen

tfor

cond

ition

sthat

may

coexistwith

autism

•de

velopm

entof

aprofile

ofthe

child’sor

youn

gpe

rson

’sstreng

ths,

skills,im

pairm

entsandne

edsthat

canbe

used

tocreate

ane

eds-

basedmanagem

entplan,taking

into

accoun

tfamily

anded

ucational

context

•commun

icationof

assessmen

tfinding

sto

theparent/carer/child

Hayes et al. BMC Psychiatry (2018) 18:222 Page 11 of 25

Table

4Keydiagno

sticrecommen

datio

ns(Con

tinued)

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

RASD

N(2011)

[44]

Nospecifictool

Theuseof

MDTapproach

isne

cessary

Involvingat

leasttw

odisciplines:

paed

iatrician;child

psychiatrist;SLT,

OT,clinicalpsycho

logist;spe

cialist

health

visitor;men

talh

ealth

practitione

r(CAMHS);socialw

orker;

nurse;ed

.psych.Teacher;other

traine

dprofession

als

Referralscreen

edwith

in5days.Infoprovided

with

in4weeks.13weeks

tofirst

appo

intm

ent.Feed

back

with

in4weeks,rep

ortwith

in6weeks

ofform

ulation.

Step

one:Initialdirected

conversatio

n.Step

two:Integrated

multid

isciplinary

team

assessmen

t(leadsto

diagno

sis/

non-diagno

sis)includ

es:

•med

icalhistoryinc:birthhistory,

family

history,&ge

neralm

edical

concerns

•de

velopm

entalh

istory

focusing

onde

velopm

ental&

behaviou

ral

concerns

•ob

servationalassessm

entof

the

child/you

ngpe

rson

•furthe

rassessmen

t/ob

servations

inanothe

rsetting(schoo

l/hom

e)•ph

ysicalexam

insomegrou

ps•specificassessmen

tsmay

berequ

ired,

e.g.

SLTassessmen

t•ed

ucationalassessm

ent

Step

three:Integrated

MDT

form

ulation(leadsto

wider

unde

rstand

ingof

difficulties)

Step

four:fam

ilyfeed

back

andcare

planning

NICE

CG142

(2012)

[9]

Iden

tification:Con

side

rAQ-10(with

-ou

tLD

);Briefassessmen

t(with

LD).

Diagn

osisandassessmen

t:AAAin-

clud

ingAQandEQ

;ADI-R;A

DOS-G;

ASD

I;RA

ADS-R(with

outLD

).ADOS-G;

ADI-R

(with

LD);DISCO,A

DOS-G,A

DI-

R

Com

preh

ensive

assessmen

tshou

ldbe

team

based(sho

rtversion).A

ta

minim

umby

aqu

alified

clinician

usually

aclinicalpsycho

logist,

psychiatristor

neurolog

ist[62]

(full

version).

Specialistautism

team

madeup

of:

ClinicalPsycho

logists,Nurses,OTs,

Psychiatrists,SocialW

orkers,SLTs,

Supp

ortStaff

N/S

Duringacompreh

ensive

assessmen

t,en

quire

abou

tandassess

the

following:

•core

autism

sign

sandsymptom

sthat

have

been

presen

tin

childho

odandcontinuing

into

adulthoo

d•early

developm

entalh

istory,w

here

possible

•be

haviou

ralp

roblem

s•functio

ning

atho

me,in

education

orin

employmen

t•pastandcurren

tph

ysicaland

men

tald

isorde

rs•othe

rne

urod

evelop

men

tal

cond

ition

s•hype

r-and/or

hypo

-sen

sory

sensitiv-

ities

andattentionto

detail.

Direct

observationof

core

autism

sign

sandsymptom

sespe

ciallyin

socialsituations.

Assessforpo

ssibledifferential

diagno

sesandcoexistin

gdisorders

Assessrisks;D

evelop

care

plan,

providehe

alth

passpo

rt,con

side

r

Hayes et al. BMC Psychiatry (2018) 18:222 Page 12 of 25

Table

4Keydiagno

sticrecommen

datio

ns(Con

tinued)

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

24hcrisismanagem

entplan;A

ssess

challeng

ingbe

haviou

rCon

side

rfurthe

rinvestigations

onindividu

albasis

RASD

N(2013)

[54]

Screen

ing:

GADS,GARS-2,A

ASQ

,ASA

S,NAS,AQ-10History:A

DI-R,

DISCO,A

SDI,RA

ADS-R;

Direct

assessmen

t:ASIT,HSST,SSQ,

Observatio

n:ADOS-G

Diagn

osismustbe

team

based&

draw

onarang

eof

profession

als.

Atleasttw

oof:clinicalpsycho

logy

(core),p

sychiatry,SLT,LD

/MH

nursing;

OT,othe

rapprop

riately

traine

dprofession

als.

Finalrep

ortto

beprovided

with

in6weeks

ofassessmen

t.

Asan

absolute

minim

um,elemen

ts2,

3&4mustbe

includ

edin

the

assessmen

t.1.Neurode

velopm

entalh

istory,

corrob

orated

viarelative/family;

2.Direct

autism

specificassessmen

twith

individu

al;

3.Observatio

nalrecording

ofassessmen

tsessions;

4.Clinicaljudg

emen

t.May

also

includ

e;standardized

measure

ofadaptivefunctio

ning

;assessmen

tof

lang

uage

&commun

icationskills;functio

nal

assessmen

tof

prob

lematicbe

haviou

r;fullne

edsassessmen

t

SIGN145

(2016)

[10]

Iden

tification:AQ-10Diagn

osisand

Assessm

ent:E.g.

ADI-R,D

ISCO,3di,

CARS,C

ARS-2,A

DOS-G.N

APC

and

RCPsychgu

ides.

MDT…

shou

ldbe

considered

asthe

optim

umapproach

Expe

rienced

profession

als

N/S

•History

taking

(inform

antinterview):

pren

atal,p

erinatal&de

velopm

ental

history;de

scrip

tionof

thecurren

tprob

lemsexpe

rienced

;fam

ilyhistory;de

scrip

tionof

who

isin

family;coe

xistingcond

ition

sand

differentiald

iagn

oses

•Clinicalob

servation/assessmen

t(individu

alassessmen

t/interview):

directlyob

serve&assess

the

individu

al’ssocial&commun

ication

skillsandbe

haviou

r•Con

textualand

functio

nal

inform

ationfro

mavariety

ofsettings

andpe

ople

•Profile

oftheindividu

al’sstreng

ths

anddifficulties:com

mun

ication,

cogn

itive,neuropsycho

logicaland

adaptivefunctio

ning

;motor

and

sensoryskills

•Biom

edicalinvestigations

onan

individu

albasiswhe

nclinically

relevant

•Assessm

entof

men

talh

ealth

need

s,wellbeing

andriskshou

ldbe

considered

Hayes et al. BMC Psychiatry (2018) 18:222 Page 13 of 25

Table

4Keydiagno

sticrecommen

datio

ns(Con

tinued)

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

GUIDELINES

FROM

PROFE

SSIONALBODIES

RCSLT

(2005)

[41]

N/S

Shou

ldalwaysbe

multid

isciplinary&

multi-agen

cyto

achieveop

timum

bene

fit.

Thismay

includ

eSLT,child

psycho

logy,child

psychiatry,clinical

psycho

logy,p

aediatrician,EdPsych.,

OT&teache

r

N/S

Duringassessmen

t,consideration

mustbe

givento

thetriadof

social

impairm

ents,aswellastheo

ries

relatin

gto

thetriad,

forexam

ple

sensorysensitivity

andintegration;

intersub

jectivity;executive

functio

ning

deficits;m

otivation;

mem

oryandcentralcoh

eren

ce.

•Jointattention

•Readiness&ability

tofocus&shift

attention

•Socialinteraction

•Use

ofcommun

icativestrategies

•Evaluatio

nof

child’splay

•Info

abou

tlearning

potential

•Im

pact

ofindividu

al’smen

talh

ealth

RCPsych

(2014)

[11]

Iden

tification:AQ,RAADS-R.RPsych

Guide

.Questionn

aires:ASA

S,GARS,

GARS-2,SCQ,SRS-2,A

Q,A

Q-10,

RAADS-R,SC

DS,ABC

.Diagn

ostic

in-

terviews:ADI-R,A

DOS-2,DISCO,3Di,

AAA,RPsychGuide

,PDD-M

RS,A

SDI,

CARS-2,H

BS,W

ADIC

Assessm

entfor

associated

devdisabilities:AQ,EQ,

SQ,Faces

test,eyestest,FauxPas

Recogn

ition

Test,SSQ

,Dew

ey’sSocial

Stories,Adu

lt/Ado

lescen

tsensory

profile

NICEadvocatesmultid

isciplinary

exercise,b

utpsychiatristsmight

beexpe

cted

todiagno

sestraightforw

ardcases&be

alertto

indicatio

nsforamorespecialist

assessmen

t.

MDTusually

includ

espsycho

logy

&nu

rsingas

core

mem

bership

N/S

•Speakwith

inform

ant

•Take

neurod

evelop

men

talh

istory

•Con

side

rob

tainingearly

health

records

Might

includ

eassessmen

tfor;

cogn

itive

ability,fun

ctionalability,

coexistent

neurod

evelop

men

tal

disabilities,coexistent

psychiatric

disorders,men

talcapacity,riskof

harm

/offend

ing,

med

icalprob

lems

Whe

reverpo

ssible,itisessentialthat

thecliniciange

tsaccurate

accoun

tsof

relatio

nships

indifferent

settings

(e.g.atwork&at

home),p

articularly

whe

rethey

might

bemore

demanding

forthat

individu

al.

BPS(2016)

[40]

e.g.ADOS,ADI,DISCO,A

DI-R

Itisrecommen

dedthat

assessmen

tismultid

isciplinary.

Atleaston

epsycho

logist,inadditio

nto

othe

rrelevant

person

nel,such

asOTs,m

entalh

ealth

workersetc.

Itisrecommen

dedthat

assessmen

tistim

ely.

Thetaking

ofade

velopm

ental

historywith

carersas

wellas

observationacross

different

settings.

Inform

ationfro

marang

eof

sources.

Psycho

logistscontrib

utionto

iden

tificationandassessmen

tmay

includ

e:•Assessm

entof

protectivefactors,

streng

thsandabilities

•Assessm

entof

associated

men

tal

health

issues

•Com

preh

ensive

developm

entaland

family

history

•Assessm

entof

learning

styles

•Assessm

entof

streng

thsandof

barriersto

learning

Hayes et al. BMC Psychiatry (2018) 18:222 Page 14 of 25

Table

4Keydiagno

sticrecommen

datio

ns(Con

tinued)

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

•Assessm

entof

environm

ental

cond

ition

sforlearning

•Functio

nalb

ehaviouralassessmen

t•Assessm

entof

social

commun

icationstyle

•Assessm

entof

thene

edsof

families.

•Com

preh

ensive

cogn

itive

assessmen

t,which

may

includ

epsycho

metricsifde

emed

necessary

BMJ

(2017)

[43]

Screen

ing:

CHAT,M-CHAT

Parentalqu

estio

nnaires:SC

Q,C

AST,

CARS;for

adults,the

SRS,ASQ

.Diagn

osisandAssessm

ent:eg

ADOS-

G,A

DI-R;3di;D

ISCO

Diagn

osisshou

ldbe

confirm

edor

madeby

anapprop

riatelytraine

dprofession

al,ide

allyworking

aspart

ofMDT

Paed

iatricians,child

psychiatrists,

adultpsychiatristsor

psycho

logists,&

othe

rprofession

als

N/S

Acombinatio

nof:

•ne

urod

evelop

men

talh

istory

•standardised

interview,&

•ob

servationalassessm

ent

Gathe

rinform

ationabou

tfunctio

ning

inmorethan

oneen

vironm

ent;Afull

neurolog

icalexam

inationinclud

ing

measuremen

tof

head

circum

ference

isroutinelype

rform

edin

allchildren.

JOURN

ALART

ICLES

Blen

neret

al(2011)

[47]

Screen

ing:

CHAT,PD

DST,STA

T,CHAT-23,M

-CHAT,ITC,SCQ.

Diagn

osis:A

DOS.

Paed

iatricne

urolog

ists,

developm

ental&

behaviou

ral

paed

iatricians,child

psychiatristsor

psycho

logists,or,ide

ally,M

DT.

N/S

N/S

Com

preh

ensive

evaluatio

nthat

includ

es•lifetim

e&family

history

•review

ofmed

ical&ed

ucational

records

•be

haviou

ralo

bservatio

n•ph

ysicalexam

ination

•administrationof

standardised

instrumen

tssuch

astheautism

diagno

sticob

servationsche

dule

•cogn

itive

&adaptiveassessmen

t•review

ofestablishe

dDSM

orICD

diagno

sticcriteria

•Assessm

entof

specificdo

mains,

such

ascommun

icationskills,

sensoryandmotor

prob

lems,and

family

stressorsandcoping

abilities

•Look

forcauses

&co-occuringcon-

ditio

ns(fu

rthe

rtests)

Carpe

nter

(2012)

[48]

Screen

ing:

ASD

ASQ

,AQandEQ

,AAA.A

Q-10,RA

ADS-R.RC

Psych

guide.

Observatio

n:PD

D-M

RS(with

ID);

ADOS-G.

Interview:A

DI-R,D

ISCO,3Di.

AAAto

providestructure.

Diagn

osiscanbe

madeby

one

clinician.Wider

assessmen

trequ

ires

ateam

.Avariety

ofprofession

alscan

diagno

se.

N/S

Labo

urintensive-up

to8h

tomake&do

cumen

tdiagno

sis.

Threeelem

ents(judg

edagainst

criteria

ofICD-10or

DSM

-4):

•interview

with

person

•ob

servation

•interview

with

aninform

ant

Someclinicians

bypass

thecriteria

&test,for

exam

ple,theo

ryof

mind,

centralcoh

eren

ce.

Con

side

rpo

ssibleco-m

orbidities

Hayes et al. BMC Psychiatry (2018) 18:222 Page 15 of 25

Table

4Keydiagno

sticrecommen

datio

ns(Con

tinued)

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

Holistic

assessmen

tsne

edsto

bestructured

arou

nd:

•Needforsocialsupp

ortandforhe

lpwith

employmen

t•Sensoryandprocessing

difficulties

•Med

icalissues

•Neuro-psychiatriccond

ition

s•Practicalskills,includ

ingmotor

difficulties

•Socialinteractionskills

•Em

otionalu

nderstanding

(ofself

andothe

rs)andpe

rson

alcoping

strategies

•Interestsandpreo

ccup

ations

•Sexualinterestsandfuture

desires

•Insigh

tandfuture

desiresand

motivation

•Psychiatric

concerns

•Other

behaviou

rsthat

may

get

person

into

contactwith

thelaw

•Supp

ortforcarers

Garland

etal.(2013)

[49]

Screen

ing:

AQ-50,AQ-10

Diagn

osis:A

DI-R,A

DOS=G,RCPsych

Diagn

ostic

Interview

Guide

Whe

nmen

talh

ealth

difficulties

also

exist,theexpe

rtiseof

thewider

MDT

islikelyto

been

gage

d.

Outlines

psychiatrist’s

role.

Enou

ghtim

eshou

ldbe

set

aside

•History

ofpresen

tingcomplaint

•Psychiatric

history

•Family

history

•Med

icalhistory

•Develop

men

talh

istory

•Person

al&socialhistory

•Men

talstate

exam

ination

•Assessforcomorbiddisordersinc.

neurod

evelop

men

tdisorders

•Ph

ysicalassessmen

t•Functio

nallevelassessmen

t•Assessrisk

•Assessm

entof

care

&supp

ortne

eds

•Con

side

ratio

nof

need

inareasof

education&em

ploymen

t

How

lett&

Richman

(2011)

[45]

Nospecifictool

Ifthelocalautism

team

does

not

have

theskillsto

assess

these

childrenthem

selves,the

yshou

ldliaisewith

profession

alswho

areable

todo

so

Minim

um,p

aediatrician&/or

child

&adolescent

psychiatrist,SLT&clinical

&/or

Ed.Psych.O

ther

profession

als…

specialisthe

alth

visitor,nu

rse,

specialistteache

r,socialworker

Timely&approp

riate.Follow

upappo

intm

entwith

insix

weeks

ofassessmen

t

Shou

ldprovidede

tailed

developm

entalp

rofile.Basedon

NICE

guidance.

Laiet

al.......

(2013)

[50]

Screen

ing:

CHAT,ESAT,M-CHAT,ITC,

Q-CHAT,STAT(fo

ryoun

gchildren);

SCQ,SRS,SRS-2,C

AST,A

SSQ,A

Q(fo

rolde

rchildrenandadolescents);A

Q,

RAADS-R(FORADULTS).D

iagn

osis

andassessmen

t:ADI-R,D

ISCO,3Di

(forstructured

interview);ADOS,

Assessm

entne

edsto

bemultid

isciplinary

N/S

N/S

•Interview

with

theparent

orcaregiver

•Interactionwith

theindividu

al•Collectionof

inform

ationabou

tbe

haviou

rin

commun

itysettings

•Cog

nitiveassessmen

ts•Med

icalexam

ination

•Co-occurringcond

ition

s

Hayes et al. BMC Psychiatry (2018) 18:222 Page 16 of 25

Table

4Keydiagno

sticrecommen

datio

ns(Con

tinued)

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

ADOS-2,CARS,C

ARS-2

(observatio

nal

measure).

Levy

etal

(2009)

[51]

SCREEN

ING:Q

-CHAT,M-CHAT,FYI,

ECI-4,C

SI-4,SCQ

,ASD

S,KA

DI,AQ-

Child,A

(AUTISM

)ABC

(autism),

PDDRS,PDD-M

RS,D

BC,D

BC-ES,

PDDBI,A

BC(abe

rrant),

CCC,

SRS,RBS-

R,SC

DC.D

iagn

osisandassessmen

t:PIA-CV,DISCO,A

DI-R,3Di.CHAT,

STAT,AOSI,A

DOS,CARS

Theseassessmen

tsshou

ldbe

multid

isciplinary

TheMDTshou

ldinclud

eclinicians

skilled

inspeech

&lang

uage

therapy,

occupatio

nalthe

rapy,edu

catio

n,psycho

logy,&

socialwork.

•Use

ICDor

DSM

criteria

•Coreandcomorbidsymptom

s,cogn

ition

,langu

age,&adaptive,

sensory,&motor

skills.

•Review

ofcaregiverconcerns,

descrip

tions

ofbe

haviou

r,med

ical

history,&qu

estio

nnaires.

•Includ

estage1data.

•Observatio

nsacross

settings

•Cog

nitive,commun

ication,&ASD

-specificassessmen

t•Med

icalassessmen

t•Differen

tiald

iagn

osis

O’Hare

(2009)

[53]

Screen

ing:

M-CHAT,NAPC

Che

cklist

Diagn

osis:A

DOS-G,SRS

Amultid

isciplinarydiagno

stic

approach

isrecommen

ded

Paed

iatricians

areessentialm

embe

rs.

N/S

•Direct

clinicalstructured

observations

•Criticalthat

inform

ationisgathered

from

different

settings,outwith

the

clinic–therearestructured

questio

nnairesforparents/teache

rs•Ph

ysicalexam

andothe

rspecialist

testsas

requ

ired

Pilling

etal.(2012)

[58]

Iden

tification:AQ-10.

N/S

N/S

N/S

Inqu

ireabou

t&assess

thefollowing:

•Coreautism

sign

s&symptom

s•Early

developm

entalh

istory

•Behaviou

ralp

roblem

s•Functio

ning

atho

me,ed

ucation,

employmen

t•Past&curren

tph

ysical&men

tal

disorders

•Other

neurod

evelop

men

tal

cond

ition

s•Neurologicald

isorde

rs(fo

rexam

ple,

epilepsy)

•Com

mun

icationdifficulties

•Hypersensory&/or

hypo

sensory

sensitivities&attentionto

detail

•Carry

outdirect

observationof

core

autism

sign

s&symptom

sespe

cially

insocialsituations

•Functio

nalanalysis

Reynolds

(2011)

[46]

Nospecifictool

N/S

N/S

N/S

Observedbe

haviou

rswith

patient

presen

tingsymptom

sfro

m‘Triadof

Impairm

ents’:socialinteraction,social

commun

ication,socialim

agination

Iden

tification:AQ-10

Shou

ldbe

carriedou

tby

MDT

consistin

gof

profession

alswho

have

N/S

N/S

Hayes et al. BMC Psychiatry (2018) 18:222 Page 17 of 25

Table

4Keydiagno

sticrecommen

datio

ns(Con

tinued)

CPG

Recommen

dedtools

MDTrecommen

ded

MDTmem

bership

Assessm

enttargets

Keyfeatures

ofassessmen

t

Wilson

etal(2013)

[52]

Diagn

osisandassessmen

t:ADI-R;

ADOS-G.A

AA,A

DI-R,A

DOS-G,A

SDI,

RAADS-R(with

outID).ADI-R

and

ADOS-G(with

ID).DISCO,A

DI-R,or

ADOS-G.

expe

riencein

diagno

sing

autism

(from

NICE).

Acompreh

ensive

assessmen

tof

autism

shou

ldinvolvean

assessmen

tof •core

autism

sign

sandsymptom

s•early

developm

entalh

istory,w

here

possible,and

intheabsenceof

aninform

antwritteninform

ation,such

asscho

olrepo

rtsmay

beused

•be

haviou

ralp

roblem

s•functio

ning

atho

me,in

education,

orin

employmen

t•pastandcurren

tph

ysicaland

men

tald

isorde

rs•othe

rne

urod

evelop

men

tal

cond

ition

s•ne

urolog

icaldisorders(e.g.epilepsy)

•sensoryprocessing

andsensory

sensitivity

issues

Assesscoexistin

gmen

talh

ealth

disorders.Risk

assessmen

t.Functio

nal

analysisforchalleng

ingbe

haviou

r

KeyOTOccup

ationa

lThe

rapist,SLT

Speech

andLang

uage

Therap

ist,HCP

Health

care

profession

al,M

DTMultid

isciplinaryteam

,Ed.PsychEd

ucationa

lPsycholog

ist

Hayes et al. BMC Psychiatry (2018) 18:222 Page 18 of 25

Interactional factorsInteractional factors related to how the dialogue betweenHCPs and between HCPs and families impacts on theassessment process. These include how consensus isreached, how disagreement is resolved and how theviews of the person and family are integrated into thedecision-making process.

Multidisciplinary assessment versus single practitionerassessmentWhere specified, all guidelines advocated for diagnosisto take place within a multidisciplinary setting with vari-ous guidelines suggesting this was ‘necessary’ [44], the‘optimum approach’ [10] or ‘ideal’ [43] (See Table 4).Some suggested (n = 4) that an appropriately trained andexperienced single professional is sufficient to diagnosein particular cases, but to be alert for indications for amore specialist assessment [11] and with access tomultidisciplinary support if required [48].Despite this almost universal recommendation, the ex-

tended version of NICE children’s guidelines (and citedby SIGN [10] and Carpenter [48]) questioned the evi-dence base for multidisciplinary assessment reporting astudy [56] that showed moderate agreement between anindividual HCP and an MDT in making a diagnosis, butstating that it was a low quality study [55]. These guide-lines also suggested in practice that a diagnosis can bemade by a single experienced HCP but that a compre-hensive profile of the patient requires a multidisciplinaryapproach [55]. SIGN guidelines also cited research [57]which demonstrates that parents value a multidisciplin-ary assessment [10].None of the guidelines in this review dealt with how

HCPs come to a consensus within a multi-disciplinarycontext, although Northern Ireland guidelines recom-mended that training should include the promotion ofcollaborative and innovative working [44] and that clini-cians must understand the profession specific roles andresponsibilities of the overall team [44, 54].Therefore, most guidelines referred to MDTs as best

practice, but lacked recommendations about how roleswithin MDTs are negotiated, how disagreement is re-solved (other than second opinion outside the team); orhow teams should work together, a factor that is ac-knowledged by NICE adults guideline [9].

Interaction with the person and their familyMany guidelines (n = 9) outlined the importance of keep-ing the person/family informed and involved throughoutthe process or recommended a person-centred approach.Some described the relationship with the person comingfor diagnosis and their family as a partnership (e.g. NICEadult guideline [9]) or as person-centred (e.g. RASDN adultguideline [54]). Some guidelines (n = 6) acknowledged that

the person or family may disagree with or be reluctant toaccept a diagnosis or, alternatively desire one [46] and bedetermined on a particular outcome, which can lead tomisleading results [11]. Carpenter asserted that somepeople may begin to see diagnosis as a desirable outcomeand pre-prepare answers based on structured interviewspublished on the internet [48]. The potential for disagree-ment or desire for diagnosis, therefore, may impact on theinteraction with the person or their family. So, although therelationship with the patient/family is considered withinCPGs, there is little guidance as to how HCPs might dealwith patient/family desire or disagreement.

Contextual factorsThere were factors related to the way in which HCPs inter-pret symptoms in different settings, how diagnostic thresh-olds are judged against criteria and included considerationsaround the impact and consequences of a diagnosis.

Interpreting needsAll national guidelines (n = 5) outlined the requirement toconsider the needs, preferences and values of the individ-ual and their family and/or support them to communicatetheir needs and concerns. Most guidelines (n = 17) de-scribed elements of diagnosis that relate to either familyenvironment, family needs and concerns, circumstances,relationships, functioning, experiences in different set-tings, contextual information or level of support needs.Many guidelines reflected the need to consider assessmentof support required. Enquiries should be made about howsymptoms impact on function within the family, at home,school or work [9, 39, 47, 54, 58]. Overall, therefore, therewas a focus not only on the assessment of symptoms, andthe way in which these affect the daily life of the personand their family, but the wider environmental and socialcontext of the person and the way in which they are sup-ported, or not, by that context.

Masking and social contextSome guidelines (n = 6) reported the difficulties of diag-nosing autism when compensation strategies may ‘mask’difficulties in some contexts, particularly as an adult[33], and in girls [50] where autism may go unrecog-nised. Some suggested that individuals may come for-ward for diagnosis when their circumstances changeand/or stressors increase (e.g. [10, 45, 54]). Some guide-lines (n = 5) noted that cultural differences will exist innorms for social interaction or that cultural variationscan deliver misleading signs and symptoms. DSM-5 sug-gested that the boundaries between normality and path-ology differ between cultures and the level at whichexperience may become problematic may differ [33].SIGN suggested that those with autism may not have

met ‘normal’ adult milestones in work, relationships or

Hayes et al. BMC Psychiatry (2018) 18:222 Page 19 of 25

independence and contained extensive information onhow females can present with a different symptom profile[10]. Others warned that behaviours might be the result ofdisruptive home experiences, carer illness [39] or complexpsychosocial or child protection backgrounds [53].Despite research showing links between diagnostic

rates and SES, there was very little mention of the im-pact of SES in CPGs. DSM-5 stated that cultural and so-cioeconomic factors may affect age at recognition ordiagnosis [33] but generally guidelines failed to considerhow this might be considered in practice, other than tobe aware that ‘cultural variations can deliver misleadingsigns and symptoms’ [45] or that autism is ‘not restrictedto particular ethnic or economic backgrounds’ [40].RCSLT guidelines considered assessment of bilingual in-dividuals [41] and some suggested that ethnicity maydelay engagement in the diagnostic process [11] or mayincrease difficulty in accessing services [54].Overall, guidelines suggested it was the responsibility

of the HCP to make a judgement about which behav-iours appear to be ‘normal’ in complex social and familycircumstances, as well as against norms for behaviour.

Diagnostic uncertainty, thresholds and the role of clinicaljudgementOverall the general focus of guidelines was to outline aframework to find the best way to decide whether autismis present or not around a threshold of symptom severity.However, many guidelines problematized this, for ex-ample, one guideline discussed how definitions of autismhave changed with DMS-5 [11] and others suggested thatsocial factors, such as an upbringing characterized by lackof boundaries [45] or symptoms amplified by distress [11]may cause diagnostic difficulties.All national guidelines considered uncertainties

around diagnosis, particularly with very young chil-dren or those with co-existing disorders [39]; whenthere may be disagreement within the diagnosing team orbetween the team and the patient or family, or when thereis a lack of local expertise [9]. Many warned of diagnosticdifficulties, or ‘obscuring’ [11, 53] that can take place ifthere is an intellectual disability or other complex coexist-ing condition and several considered the difficulties ofoverlapping diagnostic criteria [33, 50, 51, 53]. Further un-certainty was outlined when individuals may not reach thediagnostic threshold [39] or when children with autismscore below the cut-off as determined by the diagnosticinstrument [43].Despite this uncertainty, CPGs generally proposed a

systematic approach to diagnosis and, in some cases,asserted that progress has been achieved in establish-ing consensus around a behavioural definition andestablished systematic clinical assessments (e.g. [50])

even whilst recognizing that the ‘boundaries betweendisorders are more porous than originally perceived’ [33].Eight guidelines stressed the key role of clinical

judgement in the diagnostic process. DSM-5 outlinedthat the use of diagnostic criteria should be informedby clinical judgement [33] and ICD-10 suggested thatguidelines should be used flexibly in clinical work[32]. The full version of NICE children’s guidelinerecommended: ‘Use information from all sources, to-gether with clinical judgement, to diagnose autismbased on ICD-10 or DSM-IV criteria’ [55]. One guide-line suggested that clinicians may depend on the ‘feel’of the interaction with the patient for diagnosis [48].The RCPsych guideline stated that:’…much will de-pend on the extent of the clinician’s experience, theirrigour in applying standard criteria and their abilityto recognise alternative diagnoses’ [11]. Uncertainty,clinical judgement and clinician experience, therefore,were all identified as important factors in the diag-nostic process.

Pragmatic outcomes and diagnostic valueMost guidelines (n = 17) discussed the need for HCPsto have knowledge of local support and resourcesavailable to deliver appropriate advice when required.The value of the diagnosis was generally described asa way to provide appropriate support, interventionand resources. NICE guidelines for children andyoung people clarified this:

‘Diagnosis and the assessment of needs …can opendoors to support and services…all of these canimprove the lives of the child or young person andtheir family’ [39].

However, NICE guidelines for adults acknowledgedthat adults who are diagnosed may receive no supportdue to lack of services [9] and Pilling et al. stated thatwhilst care for children and young people is generallywell coordinated, this is not always the case for adultservices [58].Although some guidelines acknowledged that people

may not want a diagnosis and the label it brings with it(e.g. [44, 54]) or that it can be stigmatising or damagingto career plans [11], generally guidelines described thebenefits of a diagnosis primarily as relating to improvedquality of life, creating an opportunity to have needsmet, greater understanding and reassurance about one’sown situation and access to interventions and services.Some guidelines considered that diagnosis can providerelief, understanding or an opportunity to move on withincreased support [11, 39, 45].Many guidelines stressed the importance of early diag-

nosis as this enables early intervention which leads to

Hayes et al. BMC Psychiatry (2018) 18:222 Page 20 of 25

improved health outcomes (e.g. [41, 44, 47, 50]). How-ever, the BMJ guideline asserted that the, ‘…efficacy ofearly intervention varies from child to child’ [43], andthat ‘consideration of the direct financial costs, indirectcosts… and the impact on relationships within the fam-ily… must be balanced against likely and possible im-provements in outcome for the person with ASD’ [43],bringing uncertainty into the benefits of diagnosis. Fur-thermore, O’Hare asserted that it is difficult to provethat earlier intervention is more effective [53].Overall, guidelines reflected a concern about the

potential impact or benefits on the child or adult receiv-ing a diagnosis and considered positive factors such asaccess to support and intervention, increased under-standing or relief; as well as potential negative impactssuch as stigma. Carpenter, however, questioned the rela-tionship between need and diagnosis, by asking whetherdiagnosis is influenced by what intervention the personneeds or ‘…explicitly determined by the person’s need tohave the label to access a service… rather than theirfitting strict diagnostic criteria?’ [48].To conclude, whilst CPGs appeared to frame a meth-

odical and clinical diagnostic process, they alsorehearsed a number of subjective dilemmas that HCPshave to negotiate along the way. Some CPGs themselvesdrew attention to social issues that muddle the process:the difficulties of establishing a clear threshold in a con-dition where symptoms are impacted by the stressors ofenvironment [11]; the problem of relying on mechanisticassessments or algorithms [11, 48]; the crucial role ofclinical judgement [54]; the possibility of diagnostic un-certainty through disagreement, lack of local expertise orwhen a complex coexisting condition is present [9]; thecomplexity caused by interaction with co-occurring con-ditions; masking of autism by comorbid conditions insecondary care [58]; the impact of good (or poor) socialsupport and coping strategies on how symptoms present[33], to name a few.

DiscussionWe found that CPGs varied in how they described thediagnostic process in relation to use of diagnostic tools,key elements and structure of the diagnostic process (forexample how diagnosis related to wider needs assess-ment) and how autism was classified, defined either bycurrent versions of DSM or ICD. In addition, whilstsome recommendations were clear and universal, forexample, recommendations for multidisciplinary work-ing, there was little guidance as to how this should workin practice.In addition, we found that uncertainty was central to

many diagnostic decisions, placing a great emphasis onclinical judgement. This uncertainty included questionsaround the benefits of early intervention, the shifting

nature of the diagnostic threshold, the difficulties ofinterpreting needs in different social contexts, the prob-lems of interpreting ‘masking’ or coping strategies, thedifferences in presentation across age and breadth ofsymptoms, the inter-relationship with co-conditions andsharing of symptoms, the impact of stressors on symp-toms as well as interpretation of symptoms and needs indifferent cultural contexts.Overall, therefore, our narrative review found that al-

though individual guidelines appeared to present a co-herent and systematic assessment process, they variedenough in their recommendations to make the choicesavailable to healthcare professionals particularly complexand confusing; and presented a context of uncertaintywhich appeared to be central to the diagnosis of autism.We argue that clinical guidelines for autism diagnosisilluminate the process of diagnosis as social rather thanstraightforwardly clinical, and that judgement is requiredto consider a number of sometimes contradictory andcomplex social factors.

Social factors in CPGsOrganising the narrative review findings in relation tooperational, interactional and contextual factors enabledconsideration of the influence of social factors through-out the diagnostic process.In the wide range of inter-related assessment processes

that HCPs negotiate in order to make the diagnosticdecision, the factors considered appear to be both socialand medical. Social factors include: how the category of‘autism’ is defined and boundaried; operational andinteractional factors present in the process of diagnosis;to the consequences of diagnosis including how diagno-sis is valued (see Fig. 3). Each of these factors had aplace in clinical guidelines to a greater or lesser extentbut in many cases they were not operationalized toenable a clear and transparent framework. For example,although there were many references to individualsmasking symptoms, family ‘scaffolding’ of social impair-ment and coping strategies, there was little guidanceabout how HCPs can judge the impact of these on need,behavioural symptoms or functioning.CPGs, therefore, tended to mask (whilst paradoxically

acknowledging) the existence of social factors in thediagnostic process. A more explicit acknowledgement ofsocial factors and how to manage them mightproblematize the nature of autism diagnosis altogether:if all these factors have a place in diagnosis, how do theyrelate to clinical factors and what does it mean fordescriptions of symptoms? Whilst it is not our intentionto undermine the utility of diagnostic categories in rela-tion to access to resources or support, there appears tobe a need for balance in CPGs between a clinicalapproach which both recognises and acknowledges the

Hayes et al. BMC Psychiatry (2018) 18:222 Page 21 of 25

uncertainty of the diagnostic threshold; and a pragmaticor functional approach which responds to individual andwider needs and takes account of social factors.

Diagnostic tools and processClinical guidelines for autism varied in aspects of theirkey recommendations in operational factors. Ambigu-ities around which tools to use, the key elements in thediagnostic process and the relationship between diagno-sis, assessment and formulation suggest that local prac-tice may be shaped by other factors, such as availableresources, experience and professional roles. Whichtools are used, whether different elements of the processare considered together, sequentially or inconsistently,and the specific aims of each part of the assessmentprocess may have an impact on diagnostic outcomes. Aclearer framework would help HCPs to consider whichelements of the process are relevant and when.

MDT working and views of the familyGuidance about how HCPs can reach a consensus withothers in a multidisciplinary context or deal withpatient/family disagreement or desire was lacking, leav-ing interactional factors as key to the process but largelyunexplained. Whilst it might not appear to be in theremit of CPGs to make specific recommendations abouthow teams are organised and configured, particularlyacross different health systems, we argue that team func-tioning as a key shaping factor in diagnosis requiresmore attention in CPGs, to ensure clarity of roles andtransparency for those coming for diagnosis. Similarly,as acknowledged by some CPGs, desire of the patient/family can influence the diagnostic process, thereforeCPGs should offer guidance about how that might bemanaged.

Diagnostic uncertainty and judgementUncertainty about diagnostic thresholds and differencesin diagnostic criteria make clinical judgement key to thediagnostic process and yet how this comes about wasnot clearly defined. The extent to which diagnosisshould be based on underlying symptoms versus con-textual factors such as wider needs or circumstances ofthe individual was unclear. In addition, how HCPs con-sider the consequences of the diagnosis for the patientand their family was unclear, although there was a stronglink described between diagnosis and access to support.Ambiguities in CPGs suggest that guidelines have limi-

tations in how far they are able to promote consistencyacross practice especially given the lack of a biomarkerfor autism, the reliance on observed behaviour and fam-ily narratives for diagnosis, and the differences acrosshealth systems. However, adults, children and familiescoming for diagnosis might expect a consistent process

of assessment in keeping with a framework outlined inCPGs, as CPGs become a fixed reference point both forHCPs and the lay public. There is, therefore, a tensionbetween potential expectations of those coming for diag-nosis that there should be a uniform process; and theflexibility HCPs require to respond to individual need.Given the social nature of diagnosis as argued in this

article, biomarker use in clinical practice, if and when itis successfully developed, is likely to remain only one as-pect of an interactive diagnostic process, and thereforemay not necessarily alleviate some of the difficulties andcomplexities of diagnosis that we describe. However, asbiomarker research develops, it is likely that it will pro-duce important evidence to be considered in the devel-opment of future CPGs.

Building on previous workWhilst our narrative review differed in purpose to thesystematic review undertaken by Penner et al. [29], therewere some similar findings across the two studies. Wefound, as did the authors of this previous review, thatguidelines were inconsistent in their recommendationsaround diagnostic assessment. For example, whilstguidelines generally recommended MDT assessment,some suggested that a single experienced clinician coulddiagnose [11, 39, 48] and there was little cited evidencefor the efficacy of MDT assessment. In addition, CPGsdid not provide guidance as to how waiting times (wherespecified) would be achieved and we would add that theyprovided little operational guidance as to how MDTdecision-making should operate to be most effective. Wefound, as did Penner et al., that guidelines varied sub-stantially in recommended tools and personnel; and thatnone of the professional guidelines provided targetwaiting times for assessment (See Table 4). Whilst wedid not assess guidelines for quality, we agree that thereare multiple guidelines that HCPs might access, and thatthey vary in their level of detail and theirrecommendations.We built on Penner et al.’s findings in a number of

ways. Our review of the range of assessment processesthat HCPs involved in autism diagnosis may undertake(See Table 4) suggested a wide range of choices in as-sessment processes. We also found that using differentclassification criteria (ICD-10 and DSM-5) further in-creases complexity in CPGs. Finally, we found that con-sideration of factors such as interaction with the patientand family, how needs might be defined and assessed,and issues of masking, social context, uncertainty andclinical judgement highlighted the way in which socialprocesses and factors might impact on diagnosticdecision-making. We also found that, despite the CPGsin our study operating within comparable health systemsacross the UK, CPGs did not make consistent

Hayes et al. BMC Psychiatry (2018) 18:222 Page 22 of 25

recommendations around how diagnosis might releasepost-diagnostic resources, and what that means for theprocess of diagnosis itself.Overall we agree with Penner et al.’s findings that

CPGs should incorporate flexibility to ensure that indi-vidual needs are met. Additionally, we suggest thatguidelines should acknowledge more explicitly the socialframing of diagnosis and support clinicians with aframework which enables them to act pragmaticallyin the best interests of the patient. We would arguethat inconsistencies and lack of operational guidancearound social factors in CPGs suggests that local fac-tors such as access to resources and HCP expertiseare likely to shape diagnosis more than is explicitlyoutlined in CPGs.Unlike Penner et al., we do not think that a formal ap-

proach to decision-making such as the Delphi methodwould help HCPs in the assessment process; rather itmight simply add another layer of complexity to aprocess which is already challenging. Our experience isthat HCPs already struggle to find time to meet togetherin the context of an ever-increasing workload; an extraadministrative burden may make this even moredifficult.Finally, unlike Penner et al., we included in our review

CPGs for adult diagnosis and children over 6 years old,which enabled us to consider factors common acrossage groups. Whilst we did not specifically look for differ-ences between children’s and adult’s CPGs we are awarethat the different pathways for children’s and adult’sassessment [3, 59], may well impact on an individual’sability to access diagnostic services, the process of assess-ment itself as well as potential support post-diagnosis. Wewould consider these differences as social organisationalfactors that may impact the assessment process and meritfurther consideration in the development of future CPGs.Guidelines, therefore, appear to offer a relatively linear

and straightforward pathway towards a diagnostic deci-sion in their presentation, with DSM-5 asserting thatcriteria facilitate an objective assessment of symptompresentations in a variety of settings [20]. However,comparing individual guidelines suggests inconsistenciesin this framework and close analysis reveals a more fluidprocess, disrupting the apparent clinical purity ofdiagnosis [37].

ConclusionOverall, there was a bewildering range of options forHCPs in the assessment process, and a number of differ-ent emphases in guidelines which might lead a clinicalteam one way or another. Navigating this framework inpractice is, therefore, likely to be less systematic than theguidelines might suggest, allowing for, as it must, socialand contextual influences. In reality, the clinical pathway

for autism diagnosis differs across health systems andtrusts across the UK [3] leading to the potential for agreat deal of variation in diagnostic decision-making.

Strengths and limitationsAlthough there has been a recent systematic review ofclinical guidelines [29], we consider our narrativeapproach to be helpful to understand the complex andsometimes contradictory nature of the diagnosticprocess. Methodologically, we undertook a systematicsearch and included a transparent but pragmatic selec-tion of documents. This is, to our knowledge, the firstreview which strives to consider where social factors areconsidered in clinical guidelines for autism diagnosis.One limit was that as it was a review of current guide-lines, changes through time were not exposed. Ourreview was limited to the UK context because healthcare settings vary widely in international contexts. Inaddition, we only examined the content of guidelinesrather than how they are used. Whilst CPGs areintended to assist clinical decision-making by improvingeffectiveness and decreasing variations in clinical prac-tice [60], one review of guidelines for psychiatric diagno-ses suggested that CPGs are not implemented enough inclinical practice due to either lack of agreement or ambi-guity between guidelines [61]. It is likely that there iswide variation in how CPGs are used in practice in aut-ism diagnosis and we plan further studies to considerthis.

Implications and recommendations for future researchSocial factors were not only explicit in guidelines, butwere central to them. However, an observer might beforgiven for assuming these are subsidiary factors indiagnosis, with the more ‘medical’ ‘symptom checklist’ atits core. HCPs are expected, as outlined in DSM-5, to in-tegrate the social, psychological and biological in caseformulation, however, greater clarity about how thisshould operate would be helpful. Our findings suggestthat more detail about how clinical judgement shouldconsider social factors in diagnosis would provide amore transparent guideline for HCPs.We would not recommend greater rigidity within

CPGs when evidence for best diagnostic practice is in-consistent (e.g. use of diagnostic tools), and which mayrestrict HCPs in making decisions that are in the bestinterest of the person coming for diagnosis. Rather werecommend a more explicit acknowledgement of socialfactors in CPGs with advice about how they should bemanaged and operationalised to enable more consistencyof practice and transparency for those coming fordiagnosis.Specifically, greater clarification is required related to

the sequence and timing of the diagnostic, assessment

Hayes et al. BMC Psychiatry (2018) 18:222 Page 23 of 25

and formulation processes. The recognition and assess-ment of needs is both part of the assessment processand inextricably linked to the consequences of diagnosis;guidelines might attempt to consider how these mightbe reconciled. A greater acknowledgement of the activerole of the patient, client or patient’s family in the diag-nostic process would help to place potentially competingnarratives into context. It would be useful to considerwhether guidelines are culturally specific to health ser-vices and setting and we would recommend that furthernarrative reviews should be conducted to examine CPGsin other countries. In addition, greater clarity is requiredaround how multidisciplinary interaction might operateto support consensus decision-making. Further researchcreating an evidence base on best practice for multidis-ciplinary decision-making and the use of different diag-nostic tools in practice is required, taking into accountthe complexity of social factors in diagnosis.

Additional file

Additional file 1: Data Extraction Framework. (PDF 298 kb)

Abbreviations3di: The Developmental, Dimensional and Diagnostic Interview; ADI-R: Autism Diagnostic Interview Revised; ADOS: Autism DiagnosticObservation Schedule; BMJ: British Medical Journal; BPS: The BritishPsychological Society; CPG: Clinical Practice Guideline; DISCO: The DiagnosticInterview for Social and Communication Disorders; DSM-5: Diagnostic andStatistical Manual of Mental Disorders (Fifth Edition); HCP: HealthcareProfessional; HMIC: The Healthcare Management Information Consortium;ICD-10: International Classification of Mental and Behavioural Disorders(Tenth edition) ICIDH-2International Classification of Functioning, Disabilityand Health; MDT: Multidisciplinary Team; NICE: The National Institute forHealth and Care Excellence; RASDN: The Regional Autistic Spectrum DisorderNetwork; RCPsych: Royal College of Psychiatrists; RCSLT: Royal College ofSpeech and Language Therapists; SIGN: The Scottish IntercollegiateGuidelines Network; SLT: Speech and Language Therapist

AcknowledgementsThe authors thank Daisy Elliott and Rhianna White for their help with pilotingthis project.

FundingThe study is part of the Exploring Diagnosis project made possible by aWellcome Trust Investigator Award http://blogs.exeter.ac.uk/exploringdiagnosis/

Availability of data and materialsThe datasets used and/or analysed during the current study are availablefrom the corresponding author on reasonable request.

Authors’ contributionsJH undertook the systematic search, analyzed and interpreted the data. TFand GR were major contributors in conception and design, reviewed thedata selection and revised the manuscript for important intellectual content.HR supported with data collection and initial analysis. All authors read andapproved the final manuscript.

Ethics approval and consent to participateNot applicable.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Received: 7 March 2018 Accepted: 26 June 2018

References1. Kobeissy F, Alawieh A, Mondello S, Boustany RM, Gold MS. Biomarkers in

psychiatry: how close are we? Frontiers in Psychiatry. 2013;32. Klin A, Lang J, Cichetti D, Volkmar F. Brief Report: Interrater reliability of

clinical diagnosis and DSM-IV criteria for autistic disorder: results of theDSM-IV Autism field trial. J Autism Dev Disord. 2000;30:163–7.

3. Vllasaliu L, Jensen K, Hoss S, Landenberger M, Menze M, Schütz M, et al.Diagnostic instruments for autism spectrum disorder (ASD) (protocol). 2011.

4. Huerta M, Lord C. Diagnostic evaluation of autism spectrum disorders.Pediatr Clin N Am. 2012;59:103–11.

5. Gillberg C. The ESSENCE in child psychiatry: Early Symptomatic SyndromesEliciting Neurodevelopmental Clinical Examinations. Res Dev Disabil. 31:1543–51. https://doi.org/10.1016/j.ridd.2010.06.002.

6. Falkmer T, Anderson K, Falkmer M, Horlin C. Diagnostic procedures inautism spectrum disorders: a systematic literature review. Eur Child AdolescPsychiatry. 2013;22:329–40. https://doi.org/10.1007/s00787-013-0375-0.

7. Woolfenden S, Sarkozy V, Ridley G, Williams K. A systematic review of thediagnostic stability of Autism Spectrum disorder. Res Autism Spectr Disord.2011;6:345–54.

8. Field M, Lohr K. Guidelines for clinical practice: from development to use.Washington DC: Institute of Medicine. Washington, D.C: National AcademyPress; 1992. https://www.ncbi.nlm.nih.gov/books/NBK234504/#ddd00035

9. NICE. Autism Spectrum disorder in adults: diagnosis and management.2012. https://www.nice.org.uk/guidance/cg142/resources/autism-spectrum-disorder-in-adults-diagnosis-and-management-pdf-35109567475909.

10. Scottish Intercollegiate Guidelines Network. SIGN 145: assessment, diagnosisand interventions for autism spectrum disorders: a national clinicalguideline. 2016. http://www.sign.ac.uk/assets/sign145.pdf. Accessed 27 Mar2017.

11. Royal College of Psychiatrists. Good practice in the management of autism(including Asperger syndrome) in adults: College Report CR191. 2014.http://www.rcpsych.ac.uk/files/pdfversion/CR191.pdf. Accessed 7 Jun 2017.

12. Fuat A, Hungin APS, Murphy JJ. Barriers to accurate diagnosis and effectivemanagement of heart failure in primary care: qualitative study. Br Med J.2003;326:196–201. https://doi.org/10.1136/bmj.326.7382.196.

13. Mazumdar S, Winter A, Liu KY, Bearman P. Spatial clusters of autism birthsand diagnoses point to contextual drivers ofincreased prevalence. Soc SciMed. 2013;95:87–96.

14. Liu K-Y, King M, Bearman PS. Social influence and the autism epidemic. AmJ Sociol. 2010;115:1387–434.

15. Skellern C, Schluter P, McDowell M. From complexity to category:responding to diagnostic uncertainties of autistic spectrum disorders. JPaediatr Child Health. 2005;41:407–12. https://doi.org/10.1111/j.1440-1754.2005.00634.x.

16. Rogers CL, Goddard L, Hill EL, Henry LA, Crane L. Experiences of diagnosingautism spectrum disorder: a survey of professionals in the United Kingdom.Autism. 2016;20:820–31. https://doi.org/10.1177/1362361315611109.

17. Russell G, Rodgers LR, Ukoumunne OC, Ford T. Prevalence of parent-reported ASD and ADHD in the UK: findings from the millennium cohortstudy. J Autism Dev Disord. 2014;44:31–40. https://doi.org/10.1007/s10803-013-1849-0.

18. King M, Bearman P. Socioeconomic status and the increased prevalence ofAutism in California. Am Sociol Rev. 2011;76:320–46. https://doi.org/10.1177/0003122411399389.

19. The National Autistic Society. Diverse perspectives: the challenges forfamilies affected by autism from black. London: Asian and Minority EthnicCommunities; 2014.

20. Norbury CF, Sparks A. Difference or disorder? Cultural issues inunderstanding neurodevelopmental disorders. Dev Psychol. 2013;49:45–58.

Hayes et al. BMC Psychiatry (2018) 18:222 Page 24 of 25

21. Russell G, Steer C, Golding J. Social and demographic factors that influencethe diagnosis of autistic spectrum disorders. Soc Psychol PsychiatrEpidemiol. 2011;46:1283–93.

22. Goldani AAS, Downs SR, Widjaja F, Lawton B, Hendren RL. Biomarkersin autism. Front psychiatry. 2014;5:100. https://doi.org/10.3389/fpsyt.2014.00100.

23. Anwar A, Abruzzo PM, Pasha S, Rajpoot K, Bolotta A, Ghezzo A, et al.Advanced glycation endproducts, dityrosine and arginine transporterdysfunction in autism - a source of biomarkers for clinical diagnosis. MolAutism. 2018;9 https://doi.org/10.1186/s13229-017-0183-3.

24. Carlisi CO, Norman L, Murphy CM, Christakou A, Chantiluke K, Giampietro V,et al. Disorder-specific and shared brain abnormalities during vigilance inAutism and obsessive-compulsive disorder. Biol Psychiatry Cogn NeurosciNeuroimaging. 2017;2:644–54. https://doi.org/10.1016/j.bpsc.2016.12.005.

25. Anderson GM. Autism biomarkers: challenges, pitfalls and possibilities.J Autism Dev Disord. 2015;45:1103–13. https://doi.org/10.1007/s10803-014-2225-4.

26. NHS Choices. New blood test for Autism a long way off. 2018. https://www.nhs.uk/news/pregnancy-and-child/new-blood-test-autism-long-way/.Accessed 3 May 2018.

27. Imran N, Chaudry MR, Azeem MW, Bhatti MR, Choudhary ZI, Cheema MA. Asurvey of Autism knowledge and attitudes among the healthcareprofessionals in Lahore, Pakistan. BMC Pediatr. 2011;11:107. https://doi.org/10.1186/1471-2431-11-107.

28. Taylor LJ, Eapen V, Maybery MT, Midford S, Paynter J, Quarmby L, et al.Diagnostic evaluation for autism spectrum disorder: a survey of healthprofessionals in Australia. BMJ Open. 2016;6:e012517. https://doi.org/10.1136/bmjopen-2016-012517.

29. Penner M, Anagnostou E, Andoni LY, Ungar WJ. Systematic review of clinicalguidance documents for autism spectrum disorder diagnostic assessmentin select regions. Autism. 2017:136236131668587. https://doi.org/10.1177/1362361316685879.

30. Moher D, Liberati A, Tetzlaff J, Altman DG, Group TP. Preferred reportingitems for systematic reviews and meta-analyses: the PRISMA statement(reprinted from annals of internal medicine). Phys Ther. 2009;89:873–80.https://doi.org/10.1371/journal.pmed.1000097.

31. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JPA, et al.The PRISMA statement for reporting systematic reviews and meta-analyses ofstudies that evaluate health care interventions: explanation and elaboration.PLoS Med. 2009;6 https://doi.org/10.1371/journal.pmed.1000100.

32. World Health Organization. The ICD-10 Classification of Mental andBehavioural Disorders: diagnostic criteria for research. 1993.

33. APA. Diagnostic and statistical manual of mental disorders. Fifth ed.Washington DC: American Psychiatric Association; 2013.

34. Popay J, Roberts H, Sowden A, Petticrew M, Arai L, Rodgers M, et al.Guidance on the Conduct of Narrative Synthesis in Systematic Reviews: AProduct from the ESRC Methods Programme. 2006. https://www.researchgate.net/profile/Lisa_Arai/publication/242311393_Guidance_on_the_conduct_of_narrative_synthesis_in_systematic_reviews_a_comparison_of_guidance-led_narrative_synthesis_versus_meta-analysis/links/5532159f0cf2f2a588ad67fd.pdf. Accessed 20 Jun 2017.

35. Ferrari R. Writing narrative style literature reviews. Med Writ. 2015;24:230–5.https://doi.org/10.1179/2047480615Z.000000000329.

36. Jutel A, Nettleton S. Towards a sociology of diagnosis: reflections andopportunities. Soc Sci Med. 2011;73:793–800.

37. Latimer J. The gene, the clinic and the family: diagnosing dysmorphology,reviving medical dominance. Oxfordshire: Routledge; 2013.

38. Department of Health. “Fulfilling and rewarding lives”: The strategy foradults with autism in England. 2010. http://webarchive.nationalarchives.gov.uk/20130107105354/http://www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@en/@ps/documents/digitalasset/dh_113405.pdf.Accessed 19 Jun 2017.

39. NICE. Autism spectrum disorder in under 19s: recognition, referral anddiagnosis. 2011. https://www.nice.org.uk/guidance/cg128/resources/autism-spectrum-disorder-in-under-19s-recognition-referral-and-diagnosis-pdf-35109456621253. Accessed 2 Nov 2016.

40. British Psychological Society. Autistic Spectrum Disorders: Guidance forPsychologists. Br Psychol Soc. 2016; www.bps.org.uk. Accessed 27 Mar 2017Available at https://blogs.exeter.ac.uk/exploringdiagnosis/.

41. RCSLT. Royal College of Speech & Language Therapists Clinical Guidelines.Oxon: Speechmark; 2005. p. 63–71.

42. World Health Organisation. International classification of functioning,Disability and health. 2001.

43. Parr J, Woodbury-Smith M. Autism spectrum disorder. BMJ Best Pract. 2017:59–83. https://bestpractice.bmj.com/topics/en-gb/379.

44. RASDN. Six Steps of Autism Care. 2011. http://www.belfasttrust.hscni.net/pdf/Six_Steps_of_Autism_Care_Pathway_Report.pdf. Accessed 25 Mar 2018.

45. Howlett D, Richman S. Recognising, referring and diagnosing autism. EveryChild J. 2011;2:44–9.

46. Reynolds KE. Autism spectrum disorders in childhood: a clinical update.Community Pr. 2011;84:36–8.

47. Blenner S, Reddy A, Augustyn M. Diagnosis and management of autism inchildhood. BMJ. 2011; https://doi.org/10.1136/bmj.d6238.

48. Carpenter P. Diagnosis and assessment in autism spectrum disorders. AdvMent Heal Intellect Disabil. 2012;6:121–9. https://doi.org/10.1108/20441281211227184.

49. Garland J, O’Rourke L, Robertson D. Autism spectrum disorder in adults:clinical features and the role of the psychiatrist. Adv Psychiatr Treat. 2013;19:378–91. https://doi.org/10.1192/apt.bp.112.010439.

50. Lai M-C, Lombardo MV, Baron-Cohen S. Autism. Lancet. 2014;383:896–910.https://doi.org/10.1016/S0140-6736(13)61539-1.

51. Levy SE, Mandell DS, Schultz RT. Autism. Lancet. 2009;374:1627–38. https://doi.org/10.1016/S0140-6736(09)61376-3.

52. Wilson CE, Roberts G, Gillan N, Ohlsen C, Robertson D, Zinkstok J. The NICEguideline on recognition, referral, diagnosis and management of adults onthe autism spectrum. Adv Ment Heal Intellect Disabil. 2014;8:3–14. https://doi.org/10.1108/AMHID-05-2013-0035.

53. O’Hare A. Autism spectrum disorder: diagnosis and management. Arch DisChild - Educ Pract. 2009;94:161–8. https://doi.org/10.1136/adc.2008.150490.

54. Regional Autistic Spectrum Disorder Network. Autism Adult Care Pathway.2013. http://www.hscbusiness.hscni.net/pdf/Adult_Care_PAthway_Reviewed_August_2013.pdf. Accessed 21 Jun 2017.

55. National Collaborating Centre for Women’s and Children’s Health. Autism:recognition, referral and diagnosis of children and young people on theautism spectrum: NICE Clinical Guideline. 2011. https://www.ncbi.nlm.nih.gov/pubmed/22624178. Accessed 17 Feb 2017.

56. Mahoney WJ, Szatmari P, MacLean JE, Bryson SE, Bartolucci G, Walter SD,Jones MB, Zwaigenbaum L. Reliability and accuracy of differentiatingpervasive developmental disorder subtypes. J Am Acad Child AdolescPsychiatry. 1998;37(3):278–85.

57. Moore V, Titcomb J, Johnson C, Cronk E, Baker S, Thysson L, et al.Developing an autism assessment service II: analysis of the first 81 casesseen. Child Psychol Psychiatr Rev. 1998;3(3):121–7.

58. Pilling S, Baron-Cohen S, Megnin-Viggars O, Lee R, Taylor C.Recognition, referral, diagnosis, and management of adults with autism:summary of NICE guidance. BMJ. 2012;344(jun27 1):e4082. https://doi.org/10.1136/bmj.e4082.

59. Baird G, Douglas HR, Murphy MS. Recognising and diagnosing autismin children and young people: summary of NICE guidance. BMJ. 2011;343:–d6360.

60. Kredo T, Bernhardsson S, Machingaidze S, Young T, Louw Q, Ochodo E, etal. Guide to clinical practice guidelines: the current state of play. Int J QualHeal Care. 2016;28:122–8. https://doi.org/10.1093/intqhc/mzv115.

61. Saddichha S, Chaturvedi SK. Clinical practice guidelines in psychiatry:more confusion than clarity? A critical review and recommendation ofa unified guideline. ISRN Psychiatry. 2014;2014:828917. https://doi.org/10.1155/2014/828917.

62. National Collaborating Centre for Mental Health. Autism: the NICE guidelineon recognition, referral, diagnosis and management of adults on the autismspectrum. In: The British Psychological Society and the Royal College ofPsychiatrists; 2012. https://www.nice.org.uk/guidance/cg142/evidence/full-guideline-186587677.

Hayes et al. BMC Psychiatry (2018) 18:222 Page 25 of 25