ADHD assessment: about the Fifty percent referred to a child specialist.

The article by Jason Lanham (1) describing the results of a survey which addresses the evaluation of attention deficit hyperactivity disorder (ADHD) by family medicine residents and staff from 17 different programs was heartening. Over 90% of respondents used information not only from parent and child but also from the teacher to make the diagnosis. Even though only about a fifth of physicians responding were familiar specifically with the six point American Association of Pediatrics (AAP) clinical practice guidelines for assessment of ADHD, far more included many of the recommendations in their clinical practice.

Reading the article and the practice guidelines, I could not help but compare this with our own protocol for ADHD assessment taught to medical students and general psychiatry residents in our Child Psychiatry Outpatient Clinic. I do not think that our protocol is practical in most primary care settings. What follows should make clear the difference between assessment by family practice physicians in a teaching program, as described in the article by Lanham, and assessment when referral (50%, according to the article) is made to a child specialist.

Our initial assessment is scheduled for 45 minutes and includes interviewing both parent(s) and child. Any follow-up appointments are scheduled for 20 minutes. Additional time is scheduled for psychological testing by the physician.

We do screen for possible comorbid conditions, including the other disruptive behavior disorders (conduct and oppositional defiant disorder) and anxiety and depressive spectrum disorders. The first group of comorbid disorders is common and probably accounts for parental confusion over whether the medicine is helping or not and for the many requests to change dose or try a different medication. It is uncommon for the latter two disorders to account for the complaints addressed by the AAP guidelines.

In the past we used the computer-assisted version of the Diagnostic Interview Schedule for Children to screen for comorbid conditions in the hopes of decreasing physician time, but learned that screening by the physician during the interview is both quicker and more accurate. Many of our parents do not have sufficient computer skills or reading skills to complete the validated computer battery, which takes 45 to 90 minutes.

An especially important screen is one for academic ability. For this we use either the Peabody Picture Vocabulary Test (PPVT) or the Kaufman Brief Intelligence Test (K-BIT) to gain an idea of how much academic effort the child must put into studying. (2,7) The test takes 20 to 30 minutes to administer and score. Most children in the lower 10th or 15th percentile will have academic problems. This is the Borderline Intellectual Function range. Teachers often observe that paying attention is a problem for students with scores in this range and suggest evaluation for a possible inattentive type of ADHD. Those whose standard score is extremely low (usually below 40 on the PPVT or 55 on the K-BIT) need assessment for possible mental retardation. For the younger school age child from this group, we follow up with a developmental assessment by administering the Vineland Adaptive Behavior Scales. Administration and scoring time is 60 minutes for the Interview Edition. (3)

There is no quick screen for learning disabilities. A specific learning disability is diagnosed by showing a significant discrepancy between achievement (based upon the results of an individually administered and age-normed test such as the Wechsler Individual Achievement Test) and ability (using an intelligence test which is individually administered and age normed, such as the Wechsler Intelligence Scale for Children) in a specific area of academic processing, ie, reading comprehension or math computation. The testing battery takes about three hours to administer and score and is usually available through the school if the child meets education system criteria for concern. Usually the child with a learning disability is having problems with a specific academic processing skill rather than more global academic problems with low grades for "everything." About half of the children with ADHD have a comorbid specific learning disability. Without addressing this, hyperactive behavior may respond to medication, yet academic underachievement continues.

The mainstream classroom teacher is asked to write a narrative describing behaviors of concern and to complete a standardized and normed behavior rating scale to quantitate behavior which might improve with treatment. This scale is not used to make a diagnosis. It is used to quantitate baseline behavior and medication response. Many are available. We use the Connors' Teacher Rating Scale, a 28-item version, and from this, the Hyperkinesis Index, which is sensitive to medication response (it is sensitive to change in behavior for other reasons also and because of this, an improvement in score is not diagnostic of ADHD). (4) We also look for attention processing deficits with a continuous performance test which assesses ability to stay on task over a period of time. (5) Again, this is not a diagnostic instrument because children without ADHD can have impairment. Choose a version that will generate an index of impairment which will improve with medication use. We use the computer administered Connors' Continuous Performance Test and follow the Overall Index to quantitate medication responses. (6) Each administration takes about 25 minutes of physician time.

With the exception of the test battery for specific learning disabilities, most postgraduate physicians in our training program learn to administer, score, and interpret the aforementioned instruments. Although none of these instruments are absolutely necessary for diagnosing ADHD, they are necessary to assess comorbid and unrelated problems that may interfere with recognition of medication response.

It takes more time to discuss with parents and children the behaviors of concern that medicine will not help than it does to cover the specific target symptoms for ADHD medication use, ie, fights are not an essential or a defining feature of ADHD; neither is defiance or homework refusal. Medication does not make a child study if the child defiantly chooses to put in insufficient effort. We follow the patient weekly with 20 minute visits for 3 or 4 weeks until a medication response is clearly documented by repeat teacher-completed rating scales and a continuous performance test administered 30 minutes after using a short-acting stimulant. We find parent-completed rating scales to be of limited use in documenting medication response.

Follow up is at 4 week intervals thereafter. Each school year is started without medication so we can reassess with a classroom behavior screen and a continuous performance test for the presence of continuing symptoms (indexes) to target with medication. Once medication response is documented via the behavior scales and/or continuous performance testing, we next talk about the comorbid behaviors that must be addressed by the patient. This is a shift into psychotherapy and behavioral management of the comorbid problems once medication efficacy is established and is necessary to maintain medication compliance and avoid discontinuation of a medication that is helping.

The article by Lanham reports that half of children with ADHD are not sent to a child specialist. This suggests that an assessment as outlined here is not going to be necessary for all. Also, the situation that I describe is a teaching facility, and just as with any guidelines developed by a professional society, practitioners must change what is done in a teaching setting to what is indicated for their level of expertise, the patient population and the facility within which they practice.

References

1. Lanham JS. The evaluation of attention deficit/hyperactivity disorder in family medicine residency programs. South Med J 2006;99:802-805.

2. Kaufman AS, Kaufman NL. Kaufman Brief Intelligence Test Manual, 2nd Edition. Circle Pines, American Guidance Service Publishing, 2004.

3. Sparrow SS, Balla DA, Cicchetti DV. Vineland Adaptive Behavior Scales, Interview Edition, Expanded Form Manual. Circle Pines, American Guidance Service Publishing, 1984.

4. Connors CK. A teacher rating scale for use in drug studies with children. Am J Psychiatry 1969;126:152-156.

5. Riccio CA, Reynolds CR, Lowe PA. Clinical Applications of Continuous Performance Tests: Measuring Attention and Impulsive Responding in Children and Adults. New York, John Wiley & Sons, 2001.

6. Connors CK. Connors' Continuous Performance Test User's Manual. Circle Pines, Multi-Health Systems, 1994.

7. Robertson GJ, Eisenberg JL. Peabody Picture Vocabulary Test Technical Supplement (Forms L and M). American Guidance Service Publishing, 1981.

It's a job that's never started that takes the longest to finish.
--J.R.R. Tolkien

Manuel Cepeda, MD

From the Department of Psychiatry, University of South Alabama College of Medicine, Mobile, AL.

Reprint requests to Manuel Cepeda, MD, Department of Psychiatry, 2451 Fillingim Street, MCSB 1610, Mobile, AL 36617. Email: [e-mail]

Accepted April 12, 2006.