Acknowledgements

1. Update of Autism Evaluation, Treatment, and ResearchZachary Warren, PhD,Vanderbilt Kennedy Center Treatment and Research Institute for Autism Spectrum Disorders Vanderbilt University

2. Acknowledgements VKC / TRIAD Jeremy Veenstra-Vander Weele Vanderbilt EPC Team Melissa McPheeters Nila Sathe Vanderbilt LEND Terri Urbano & Tyler Reimschisel Financial Disclosure: AHRQ, Autism Speaks, HRSA, MARI, MCHB, NIMH, NICHD, NSF, Simons Foundation

3. Outline Background and current context of ASD Brief history – the disorders and intervention Presentation of systematic review Early Intensive Intervention Medical treatments Complementary and alternative approaches What do we do for the children with ASD that we serve?

4. Defining:Autism Spectrum Disorders 1943 – Leo Kanner – Infantile autism 1944 – Hans Asperger 1960s – Separation from schizophrenia 1970s – Biology / genetic underpinnings 1980 – DSM-III – Pervasive Developmental Disorders 1987 – DSM-III-R - Autistic Disorder / PDD-NOS 1994 – DSM-IV – Asperger’s Disorder

5. Our Context:What once was rare… Old estimate for autism: ~ 1/2500 (1985) Recent estimates for autism: ~ 1/500 (1995) Newest estimates for ASD: 1/150 (CDC, 2007) 1/110 (CDC, 2009) 1/91 (NSCH, 2009) What will it be in 2012?

6. Does this child have autism? Develop ranking for all children according to likelihood of an ASD diagnosis: #1 (likely has autism) to #7 (no way s/he has autism) NO TIES ITS ORDINAL RANKING Assign an ASD “present” or “absent” diagnosis for each child

7. Origins of the disorder(s) Neurodevelopmental disorder: Core neurogenetic vulnerabilities Complex environmental X neurogenetic interactions? Evidence: Maleness (3:1 to 4:1) Familial loading/risk Monozygotic twins: 60-95% Sibling inheritance: 12-25% (18.7) Population level: around 1%

8. A behaviorally based diagnosis Diagnosis is based on a pattern of symptoms in 3 domains: social impairments language/communication impairments restricted activities and interests Symptom expression is variable across children and over time

9. DSM-IV Pervasive Developmental Disorders a = one of these features must be presentb = at least 2 of these features must be present

10. Towards DSM-V Distinctions can be difficult both within the spectrum and across other disorders Autistic D/O Asperger’s PDD-NOS Autism Spectrum Disorder

11. DSM-V Deficits in social communication (all 3): Nonverbal and verbal communication deficits Lack of social reciprocity Failure to develop peer relationships Restricted, repetitive patterns of behavior, interest, and activities (2) Stereotyped motor or verbal behavior; unusual sensory behavior Excessive adherence to routines and ritualized bhx Restricted, fixated interests Symptoms present in early childhood (manifest when social demands exceed capabilities)

12. Why are numbers increasing: The addition of regions of diagnosis? Atypical behaviors/interests Social Reciprocity Autism Asperger’s PDD Language/Communication

13. Why are numbers increasing: Broadening spectrum? Constantino, JAACAP, 2003

14. New diagnostic measures Autism Diagnostic Interview (1989, 1994) Autism Diagnostic Observation Schedule (1989, 2000) Screening tools/algorithms/instruments in wide use: MCHAT / AAP guidelines SCQ SRS Why are numbers increasing: Better tools?

15. Diagnostic changes Categories Broadening Better tools and identification process Awareness Mental health providers, pediatricians, schools Media, parents Other factors: Previousunderestimates Methodology for obtaining epidemiological data What else??? Why are numbers increasing: Lots of reasons?

16. Developmental Impact of Intervention:Prevention of Cascading Effects?

17. The American Academy of Pediatrics recommends developmental surveillance at every well-child visit and developmental screening using formal, validated tools at 9-, 18-, and 24- or 30-month visits or whenever a parent or provider concern is expressed Reprinted from Pediatrics, 2006; 117:404-419.

18. In addition, autism-specific screenings are recommended at the 18- and 24-month visits Reprinted from Pediatrics, 2006; 117:404-419.

19. What happens after a positive screens? Ideally: Referred for evaluation of specialized team Accessibility: Resources not present Availability: Long waits (e.g., 6 - 12 months) Ideally: Appropriately intensive specialized services are rapidly initiated regardless of diagnostic status Reality: Positive screens lead can lead to long waits for diagnosis and intensive service delivery

20. Components of ‘Gold Standard’ Assessment Medical assessment Autism Diagnostic Interview – Revised (ADI-R) Structured clinical interview 90-180 minutes Autism Diagnostic Observation Schedule (ADOS) Semi-structured interaction 45 minutes / 15 minutes scoring Developmental assessment Cognitive/IQ Language Adaptive behavior

21. Autism Symptoms:Assessment Framework • Negative symptoms • absenceof expected behaviors • relative to developmental norms • Positive symptoms • presence of unusual behaviors • due to nature, intensity, interference • with family activities

22. Autism Symptoms:Assessment Framework • Negative symptoms • Social behaviors • Communication • Positive symptoms • Atypical body, sensory, play behavior

23. Social behavior is notall-or-nothing • Social behaviors are not completely absent • Children with autism do show social behaviors • (e.g., eye contact, imitation, attachment) • Social behaviors occur less consistently • across people and settings (e.g., at different • times, requiring greater effort)

24. Social Behavior Concerns Less responsiveness to social overtures Less participation in back-and-forth play Less “showing off” for attention Less imitation of the actions of others Less interest in other children Less functional play, doll play, imaginative play

25. Communication is not all-or-nothing Autism as a disorder of social communication Communication vs. Language vs. Speech Young children communicate for many reasons: To request things they want To protest about events that displease them To share their enjoyment with adults To direct adult’s attention to objects or events of interest

26. Communication Vulnerabilities Less sharing enjoyment with adults Less directing adult attention to interests Differences in use of gestures and gaze Differences in terms of social communicative responses (e.g., name press, response to joint attention) Delayed language

27. Limits of available tools • Broad-based screeners • CBCL, BASC • Level 1 Screening (Developmental risk vs. no) Young children: MCHAT, CHAT, PDDST School aged children: CAST • Level 2 (ASD vs. other developmental concerns) • ABC, ASDS, ASSQ, GARS/GADS, SCQ • Diagnostic Instruments: • ADOS, ADI-R, CARS / CARS-2 • Limits of parent report: categorical and likert ratings • How do tools work for younger children • Are these instruments sufficient for diagnostic consultation?

28. The Challenge Reduce waits between screening concerns and diagnosis/service delivery Meet time demands / restrictions Accurately identify children with/out ASD Link children with appropriate early intervention services Be adequately reimbursed Create a framework for performing ASD diagnostic consultations in specialty and potentially community pediatric practices that can:

29. Characteristics of STAT Screening Tool for Autism in Toddlers and Young Children (STAT) Designed as a Stage 2 screening tool (for referral populations) Consists of 12 play-based items assessing key social-communicative behaviors Takes about 15-20 minutes to administer Designed for children between 24 & 36 months Data / modifications for above and below Simple scoring system Published psychometric data

30. Agreement 14 out of the 19 children (74%), initially classified ASD • 1 of 2 (50%) non-ASD received an independent diagnosis of ASD • Overall 15 out of 21 cases (71%); provider range 57-100%. • Disagreements • Other delays clearly evident • Diagnostic certainty lower

31. Initial thoughts regarding potential impact • 2008: 85,000 live birth @ 1/150 = 567 • Even this has changed • Impact of 20 trained physicians in state • 1 assessment per week: 1040 • 1 assessment every other week: 470 • Population level impact? • Expedited entry into TEIS services • Reduced waits at specialty clinics for children in need of this service. • Long-term public health impact? • Serious concerns about false positives as well as under-identification

32. Replication and Practice Change • 27 providers • 20 Pediatricians, 3 PNPs, 3 DBPs, 1 Neurologist • M = 16 years of experience (SD=11.5, Range 2-36) • 5 regional trainings in state • 8-EAST , 13 -MID, 5 - WEST • 3.5 years

33. Replication and Practice Change • Confirmation sample • STAT certification • Provisions for diagnostic ambiguity • Independent evaluation • Assessment of practice change (1.61 years later)

34. Confirmation Sample 6 of 8 providers from mid-TN • 14 of 16 referred children independently evaluated • 13 boys, 1 girl • M = 31.71 months; Range = 24 to 42 months Agreement: 85.7% - forced model (6/7 ASD & 6/7 non-ASD) 92.9% - non-forced model

35. Practice Change - Training • 10 of 13 providers participated in validation • Embedded fidelity and implementation • Diagnostic support • 51.9% (14/27) total completion of fidelity/training • 4 additional STAT certified providers Regional differences: • 85.7% Mid • 40% West • 0% East

36. Practice Change - Survey

37. Practice Change - Survey • Clinically and statistically significant shifts (1 to 5): • Comfort discussing ASD • 2.59 to 4.24 • Appropriate for child to receive diagnosis from PCP • 2.35 to 4.24 • No provider provided a final rating below 3 • Diagnostic practice • 60% increase in diagnosis under 3 • Pre to last year of practice (4.29 to 7.5). • Several providers routinely performing • Hospital settings • Mega-practices

38. Reported Barriers to Change

39. Should we do this? • Who does this work for? • Providers • Families • Does it facilitate/expedite services? • Concerns • Misclassification • Family stress • Education/engagement • Should we be talking about “risk”

40. The Importance of Effective Early Diagnosis and Treatment:A public health perspective Earlier diagnosis = More intervention opportunities More opportunities = Optimal intervention benefit Core features: social communication / atypical behaviors Cognitive and adaptive functioning Fully integrated classroom placements Potentially promoting optimal adaptive independence Potentially reducing considerable lifetime cost and service system demands associated with ASD and related care

41. Costs of ASD • Most quoted total lifetime costs = $3.2 million (Ganz, 2007) • Hard to know if assumptions hold / included costs • Recent estimates of annual incremental costs (see Amendah et al., 2011) • $2,100 – $11,200 medical expenditures • $13,000 educational costs • $40,000-$60,000 intensive behavioral tx • $60,000-$128,000 residential costs for adults w/ASD • Productivity loss, overall lifetime costs, quantification of impact of early intervention programs

42. Science in Context Partners Friends Providers Pediatrician Noticing & Developing Concerns Discussing Concerns with Others Accurate Diagnosis Social Media Autism Speaks DAN AAP TV Google Blogosphere CDC

43. Why Are We Doing This?Our fundamental assumption… Accurate early identification of a specific common neurodevelopmental disorder in childhood should help us connect to specific intervention and treatment options that optimize functioning for children and families

44. Treatments and Therapies78,300,000 results (0.08 seconds) Auditory Integration Sensory Integration ABA Discrete Trial Training Lovaas/UCLA Intervention Early Start Denver Model Holding Therapy Dolphin Assisted Therapy Facilitated Communication Augmentative Communication Vision Therapy Vitamins Oxygen treatments Psychopharmacological treatments Floortime Music Therapy Social Skills Training Incidental Teaching TEACCH PECS Pivotal Response Therapy Son-Rise RDI Chelation Diets Drugs Supplements

45. Historical Perspective Not far removed from an “untreatable” era Rutter (1970): <2% functioning “normally” 60% requiring institutional placement/support Lovaas (1987): UCLA Young Autism Project Intensive ABA = 9 / 19 (47%) “recovered” or “normal fx” A breakthrough with major concerns

46. Which treatments work for specific children? Autism treatments vary widely: Scope, Focus, Intensity, Duration, Methodology Packages and combinations Children with ASD vary widely Core features Associated characteristics - IQ / challenging behavior Additional behavioral characteristics and medical vulnerabilities Families, communities, systems of care Different causes – different disorders Rapid discovery of CNVs What do we want to know?

47. AHRQ Systematic Reviews • Topic proposed to AHRQ • Centers for Medicare and Medicaid Services • Autism Speaks • Refined developed through discussions with key informants, TEP, public comment, review • Employed comprehensive search strategy • Extracted data and assessed study quality

48. Key Questions 48 • KQ1. Clinical outcome: What are the the short- and long-term effects of available behavioral, medical, allied health, or CAM treatment approaches? • Core features • Associated characteristics • KQ2. Modifiers: What are the modifiers of outcome for different treatments or approaches? • KQ3. Early Response to Treatment: Are there any identifiable changes early in the treatment phase that predict treatment outcomes? • KQ4. Functional Outcomes: What is the evidence that effects measured at the end of the treatment phase predict long-term functional outcomes?

49. Key Questions 49 KQ5. Generalization: What is the evidence that specific intervention effects measured in the treatment context generalize to other contexts? KQ6. What Drives Treatment: What evidence supports specific components of treatment as driving outcomes, either within a single treatment or across treatments? KQ7. Earliest Intervention: What evidence supports the use of a specific treatment approach in children under the age of two who are at high risk of developing autism based upon behavioral, medical, or genetic risk factors?

50. Methods 50 • MEDLINE®, ERIC and PsycInfo®, reference lists • Included studies • January 2000 – May 2010 • DSM-IV-TR, ADOS, ADI • n>30 medical / n> 10 behavioral • Dual review of abstracts • Extracted clinical data to standardized forms • Two reviewers quality assessed each study • Studies rated as good, fair, or poor • Overall strength of evidence assessed using EPC Methods Guide for Effectiveness and Comparative Effectiveness Reviews • Strength of evidence presented as insufficient, low, moderate, or high