Developmental Screening and Assessment: What Are We Thinking?

1
Developmental Screening and AssessmentWhat Are
We Thinking?

• Glen P. Aylward, Ph.D., ABPP
• Southern Illinois University
• School of Medicine
• Springfield, IL

2
(No Transcript)
3
(No Transcript)
4
Q 1 Is there a Gold Standard in Developmental
Evaluation?

• reference standard
• Flynn effect (.3-.5 pt/year)
• Bayley Scales (1969 1993 2006)
• BSIDgtBSID II (MDI 12 pts lower, PDI 7 points)
• BSID-II?BSID-III (mental 6 pts higher motor 8
  pts higher)
• Mean 7 pt increase comparability is limited
• Length/pragmatics

5
Q2 Is There Agreement as to What Qualifies as a
Developmental Delay?

• precision issue
• 20 delay ?
• 2 standard deviations below the mean for a
  reference group?
• Score compared to local norms?
• A ratio/criterion measure?
• Acceptance of psychometrically poor tests
• Recommend SD cutoffs

6
Q3 Does Development (DQ) Intelligence (IQ)?

• Neurologic?motor?sensorimotor? cognitive
• Skill?function?integrated functional unit?
  intelligence
• Complexity increases in concert with age
• Skill, function potential
• Different streams, different rates
• Younger than age 2 simple cognitive
  functionsonly after discrete functions are
  combined do we predict later intelligence

7
Canalized Behavior

• Species-specific, prewired, self-righting
• Fixed behavior pattern
• Not highly complex
• More canalized, less affected by adverse
  circumstances
• Less canalized, weaker self-righting, greater
  likelihood of disruption
• Sensorimotor behaviors are strongly canalized
• Impact on test results/prediction

8
Integrated Functions

• Individual developmental skill/ability is not
  most important
• Integration of abilities into functional units
  that control these abilities
• Ability to integrate functions?information
  processing, memory, discrimination, attention
• Musicians skills?section of orchestra
  function?integration of sections (conductor)?
  concert

9
IQ/DQ Ambiguity

• BSID-III Cognitive Composite
• Mullen Scales Early Learning Composite
• SB-V NVIQ, VIQ, FSIQ
• K-ABC/2 Mental Processing Composite Mental
  Processing Index
• WPPSI-III FSIQ
• MSCA General Cognitive Index (GCI)
• DAS General Cognitive Ability (GCA)
• Cattell IQ

10
Q4 Is a Ratio DQ useful?

• Ratio DQ MA/CA x 100
• Rate of development
• Not comparable at different age levels b/c the
  standard deviation (variance) of the ratios does
  not remain constant
• CIs vary tremendously
• Interpretation is difficult
• MA is totally dependent on test used
• Similar issues with developmental age
• Better to use 1.5, 2 SD lt average

11
Q5 Is Caretaker Report Sufficient for
Developmental Screening?

• AAP (2006) policy statement regarding
  surveillance and screening
• 1/3 of developmental screening instruments
  (excluding those targeting ASD) were parent
  completed
• Earlier, parent report considered a Stage I or
  prescreening technique
• Evolved to being considered comparable to
  hands-on screening
• ? Evidence-based use

12
Caretaker Report

• Little is known as to how parent completed
  questionnaires are affected by 1) child-related,
  or 2) environmental variables
• Accuracy depends on developmental area assessed,
  population
• ? Different tests for different populations
• How questions are answered (y/n, Likert, etc.)
• Considerations
• -- Length, detail
• -- Age range encompassed
• -- Presence/absence of examples of behavior
• -- Test behaviors or milestones

13
Caretaker Report

• Diamond Squires (1993) current behaviors,
  recognition (vs recall), behaviors should occur
  frequently, parents need skills to be able to
  complete questionnaire
• Screening risk status of infant most predictive
  of agreement lt 2-years at 2, race (marker of
  SES) predictive
• Camp (2007) spectrum bias better/worse
  identification depends on base rates of problems
• Items most predictive often are those with poorer
  agreement (puzzle board, stacks 6 cubes)

14
Q6 How Problematic Are Test Refusals?

• Behaviors have an impact frequently negative
• More pronounced with younger children
• Possibilities a) Declines to respond to any
  item b) specific types of items, or c) stops
  when items become too difficult
• Occasional refusals41 of young children
• State of arousal, affect, motivation,
  temperament, physiological issues
• Score refusals as failures, prorate scores, or
  consider testing to be invalid?

15
Test Refusals

• Potential causes
• --Reaction to poor underlying skills/attempt to
  avoid failure
• --Oppositional behavior
• --Shyness, anxiety
• --Temperament
• --Poor attentional skills/high activity level
• --Fatigue/malaise
• --Temper displays/crying
• --Parental behaviors

16
Test Refusals

• Verbal production tasks, gross motor activities,
  end of testing
• More in children born at biologic risk, low SES
• Those who refuse any aspect of testing differ
  from those who refuse some items or who refuse
  more difficult items
• High rates of refusal at one age associated with
  similar behaviors at later ages

17
Test Refusals--Implications

• Those who refuse to comply often have decreased
  scores in several areas of function--untestable
• Risk for lower test scores and higher rates of
  problems at ages 7-8 years in many areas
• Source of clinical information

18
Q7 Is There a Role for Qualitative Information?

• Not in place of quantitative rather, in
  conjunction with
• Causes for finding cognitive impairment,
  emerging LD, language dysfunction, environmental
  risk, testing issues, combination
• Clinicians vs. technicians
• Play-based assessment
• Examples form board naming pictures, stacking
  cubes
• Training to task

19
Quality ControlClinicians vs. Technicians

• Quality of assessment may be compromised because
  of the questionable proficiency of examiners
• Not clear who is qualified
• Conceptual and factual knowledge of normal
  development
• Awareness of significance of pathognomonic
  indicators
• Well versed in administration scoring
• (speed, best response, stop, eliciting report)

20
Q8What About Prediction?

• Prediction tells us if early alarm or reassurance
  has any basis
• Prediction is difficult because
• Rapid developmental change
• Intervening variables (environmental, biologic)
• Interventions (EI, medical, social)
• Testing itself has impact on developmental
  trajectory (observational effect)
• Emergent, latent, delayed, deficient, disordered
• Moving target
• Aspects of tests used at T1 T2 Tn
• Domain/area of development

21
Prediction

• Stable performance high riskgtlow riskgt moderate
  risk
• How does one define prediction (co-positivity/co-n
  egativity ORs, correlations)
• Time span/interval
• What predicts what?
• Single composite measure may not be appropriate
  sub-domains of function

22
Q9 Is There a Summary?

• Consider tests as reference standards be aware
  of psychometric issues
• Evaluation is a balance between concepts and
  pragmatics
• Percent delay is not accurate criterion based, gt
  1.5, 2, 3 SDs below average
• Consider what can be assessed at different ages
  (skillcapacity)
• Ratio DQs not accurate
• Serial screening/assessment

23
Summary

• We need to better understand strengths,
  weaknesses, and variables that affect caretaker
  report
• Consensus on test refusals should we include,
  prorate, or invalidate scores?
• Clinicians need to test
• Environment affects different skills and at
  different times
• Wear sunscreen and eat fiber