EVIDENCE BASED

1
EVIDENCE BASED ASSESSMENT OF ADHD IN PRIMARY
CARE SETTINGS
AMERICAN ACADEMY OF PEDIATRICS 2007 NATIONAL
CONFERENCE AND EXHIBITION
Thomas K Pedigo Ed.D., NCSP
DISCLOSURE Dr. Pedigo is co-owner of Targeted
Testing, Inc., and co-developer of the PADDS
Program which is referenced in this presentation.
2

• RELEVANCE TO PRIMARY CARE
•   
• ADHD is the most commonly diagnosed childhood
  psychiatric disorder affecting school age
  children with estimates ranging from 3 to 12
  percent
• (American Psychiatric Association, 1994).
•  
• Concern has been expressed for the these large
  numbers coupled with reportedly wide variations
  in clinical practice and research approaches all
  point to the need to develop pragmatic assessment
  tools and approaches for use in the major systems
  of service entry.
• (American Academy of Pediatrics, 2000)
•  
• Researchers and authorities have pointed out the
  weak Negative Predictive Power (good performance
  actually rules out attention disorders) of CPTs
  and numerous neuropsychological measures
  traditionally used to evaluate ADHD in light of
  the base rate (conservative estimate of 4),
• (Barkley, R.A., Grodzinksi, G. M., 1994).,
• (Ellwood, R.W., 1993)., (Matier-Sharma, K.,
  et.al., 1995).

3
Other Comorbid conditions often occur with ADHD.
These conditions may include but are not limited
to Mood Disorders, Anxiety Disorders, Disruptive
Behavior Disorders and Learning Disorders.
Bipolar Disorder. Pliszka, S. R., Carlson, C. L.,
Swanson, J. M., (1999). ADHD with Comorbid
Disorders Clinical assessment and Management.
New York, N.Y. The Guilford Press.   PG ADHD/
ODD-CD 15 to 61 90 ODD-CD/ADHD 35 to
87 90 ADHD/Depression 0 to 38 127 Depression/
ADHD 0 to 57 127 ADHD/Anxiety 23 to
30 151 Anxiety/ADHD 9 to 35 151 ADHD/LD
7 to 60 192 (Across- Reading, Spelling,
Math) ADHD/OCD 6 to 33 214
4
RELEVANCE TO PRIMARY CARE  WHY THE PEDIATRIC ADD
SCREENER WAS DEVELOPED The Development and
Validation of Diagnostic Tools   During the 1998
NIH Consensus statements indicate a need to
develop more objective assessment tools, rating
scales and/or diagnostic interviews that map onto
basic underlying processes as well as a need to
supplement behavioral assessment tools with
improved cognitive and/or neuropsychological
measures.   Consequently, there is a great
need for the development of practical, reliable
and valid procedures to be used in primary care
settings to identify and manage ADHD symptoms, as
well as to distinguish appropriate referral
needs. (NIH conference, 1998)
5

• Recent developments within the field of ADHD have
  increasingly pointed to the need to evaluate the
  various executive operations and working memory
  of children suspected of Attention Disorders.
  (Biderman, J. et al 2004, Brown, T.E., 2002,
  2000,1999 Barkley, R.A. 1997,1998 Denckla M,
  1996.)
• Difficulties in these Executive Processes
• (planning, attending, organizing input, storing
  and retrieving information, modulating emotions
  and sustaining effort)
• exemplify the complaints of teachers and parents.
  

6

• Basic Demands of the Classroom
•  
• Attending to instruction
• Assimilating information
• Accommodating information
• Organizing, sequencing, manipulating information
• Monitoring emotional activity
• Formulating a plan of action
• Implementing the plan
•  
• Other Factors
• Working under time pressure
• Avoiding distraction
• Being adequately prepared
• PADDS TARGET TESTS OF EXECUTIVE FUNCTIONS WERE
  DESIGNED TO PRODUCE WORK DEMANDS SIMILAR TO THOSE
  OUTLINED ABOVE.

7
Clinical sample consists of 629 children age 6-12
(266 females and 367 males) evenly balanced
between ADHD and Non-ADHD/Typical peers. Data was
collected from 10 sites in 7 states with
Institutional Review Board (IRB) approval for the
overall project handled at Armstrong Atlantic
State University in Savannah, Georgia. Specific
sites included specialty ADHD assessment centers
in Illinois, Georgia, Idaho, New Jersey,
Tennessee, California, and Florida.
8
PEDIATRIC ADD SCREENING SYSTEM Target Tests of
Executive Functioning (TTEF) ASSESSES EXECUTIVE
FUNCTIONS  COMPARES TO ADHD TYPICAL PEERS  CAN
EFFECTIVELY RULE IN OUT ADHD  EFFECTIVELY CROSS
VALIDATES BEHAVIOR RATINGS 
9
Computer Administered/Scored Diagnostic Interview
(CADI)  EFFECTIVELY ASSESSES FOR
COMORBIDITY  ESTABLISHES A PRELIMINARY TREATMENT
PLAN  CAN PROVIDE DOCUMENTATION TO SUPPORT
REFERRALS AND OTHER TESTING REQUESTS
10
CLINICAL PRACTICAL UTILITY of (PADDS) This
Evidenced based format compares these results
in incremental fashion beginning with a base rate
of 4 percent. By using a conservative base rate
and multiple inputs the clinician is able to
conduct an evidence-based analysis that along
with clinical judgment can rule in or out a
diagnosis of ADHD in the office setting.        
This standardized evidence-based approach
efficiently provides a preliminary treatment plan
that can support a diagnosis when combined with
other procedures as deemed necessary by clinical
input and judgment. This screening can also
support decisions for further evaluation and
referrals.
11
Nomographic Evidenced Based Report Analysis
Positive and/or Negative Predictive Power is
developed through the combination of these lines
of evidence by calculating a likelihood ratio for
each measure. By applying the likelihood ratios
incrementally using a Nomogram, we have
effectively Summed the gathered evidence into a
single, valid factor called the Post-test
Probability. Using a nomogram for charting
likelihood ratios is the best method to properly
combine the results of different tests In the
following examples we combined information from
Parent ratings of DSM-IV ADHD diagnostic
criteria. Teacher ratings of DSM-IV ADHD
diagnostic criteria. Results from the three
Target Tests of Executive Functions.
12
Case Example 1 Prior to any input from the
PADDS system the ADHD base rate of 4 is equal
to the Post-test probability of 4 .
Input of a Parent rating meeting DSM-IV criteria
producing a likelihood ratio of 9 moves the
pre-test probability from 4 to a post-test
probability of 25 .
We re-set the Likelihood Ratio to 1 and Pre-test
Probability to 25
Input of a teacher rating meeting DSM-IV
criteria producing a likelihood ratio of 9 moves
the new pre-test probability of 25 to a
post-test probability of 74.
13
The combined values of the Parent and Teacher
Ratings alone calculates a probability of
74 There is clearly not sufficient evidence for
a diagnosis at this point using rating scales
alone.
This is where it becomes important to collect
further information along multiple lines of
evidence. Any measures with known values for
specificity and sensitivity may be used For this
example we are using a set of newly developed
cognitive tests called Target Tests of Executive
Functions, from our PADDS program
14
Input of the Target Recognition subtest
performance produces a likelihood ratio of 6
moves the new pre-test probability of 74 to a
post-test probability of 93.
Input of Target Sequencing subtest performance
producing a likelihood ratio of 8 moves the new
pre-test probability of 93 to a post-test
probability of 99.
Input of Target Tracking Subtest performance
producing a likelihood ratio of 1 maintains the
pre-test probability of 99 to a posttest
probability of 99.
15
This example illustrates how the Target Test
scores can be used to support and validate the
results of Behavioral Rating Scales. In this case
the Target Tests of Executive Functions scores
significantly modify the predictive index in a
direction supporting a clinical diagnosis.
16
Case Example 2 Prior to any input from the
PADDS system the ADHD base rate of 4 is equal
to the Post-test probability of 4 .
Input of a Parent rating meeting DSM-IV criteria
producing a likelihood ratio of 9 moves the
pre-test probability from 4 to a post-test
probability of 25 .
We re-set the Likelihood Ratio to 1 and Pre-test
Probability to 25
Input of a teacher rating meeting DSM-IV
criteria producing a likelihood ratio of 9 moves
the new pre-test probability of 25 to a
post-test probability of 74.
17
The combined values of the Parent and Teacher
Ratings again calculates a probability of
74 This time however, the test scores will
adjust the overall index in an entirely different
direction, also clearly illustrating the
weaknesses in using rating scales alone as the
basis for a diagnosis.
18
Input of the Target Recognition subtest
performance produces a likelihood ratio of 0.7
moves the new pre-test probability of 74 to a
post-test probability of 66.
Input of Target Sequencing subtest performance
producing a likelihood ratio of 1 moves the new
pre-test probability of 66 to a post-test
probability of 65.
Input of Target Tracking Subtest performance
producing a likelihood ratio of 0.5 adjusting the
pre-test probability of 65 to a post-test
probability of 49.
19
In this case the Target Tests of Executive
Functions scores significantly modify the
predictive index away from supporting a clinical
diagnosis. Thus, reviewing any other clinical
information obtained becomes critical. The risks
of over identification are evident when relying
on behavioral rating scales alone. This examples
subject had a hearing disability, not ADHD
20
The Benefits of using Evidence -
Based Assessments that Test Executive
Functions Are clearly superior To rating scales
alone
EVIDENCE
EVIDENCE
EVIDENCE
EVIDENCE
The way to Reduce ADHD Over-Identification is
with More Accurate Assessments, Based on more
lines of evidence
21
EVIDENCE BASED ASSESSMENT OF ADHD IN PRIMARY
CARE SETTINGS
AMERICAN ACADEMY OF PEDIATRICS 2007 NATIONAL
CONFERENCE AND EXHIBITION
Thomas K Pedigo Ed.D., NCSP
For more information on PADDS and Evidence Based
ADHD Assessment Please Visit http//www.targettest
.com