6 October 2003
This week, I want to take a closer look at an aspect of so-called 'ADHD' that's seemed to me from quite early on to be
something of an anomaly.
'ADHD', as we know, is generally presented to the public as some kind of 'malfunction' of the brain. A 'neurobiological
disorder' or 'brain disease'.
What, then, usually happens next if someone should suspect that a child has a neurobiological disorder - a brain
Well, let's take a look at an example of something that, as far as I'm aware, nobody disputes is a GENUINE
neurobiological disorder or brain disease. Enchephalitis.
According to information obtained at the 'Neurology Channel' website:
"Encephalitis is inflammation (irritation and swelling) of the brain. It often coexists with meningitis (inflammation
of the covering of the brain and spinal cord) and most cases are caused by viral infection. Encephalitis ranges from mild
to severe and may result in permanent neurological damage and death."
This is typically how encephalitis is diagnosed:
"Diagnosis of encephalitis is based on the following:
A neurological exam is performed to evaluate mental status, detect neurological problems such as motor dysfunction and
seizures, and help determine which area of the brain is affected.
- Medical history (including recent exposure to insects, travel, personality changes, and contact with unusual
animals or illnesses)
- Neurological examination
- Blood and urine tests
- Imaging tests (e.g., CT scan, MRI scan, EEG)
- Spinal tap
Blood and urine tests are used to isolate and identify viruses. Enzyme-linked immunosorbent assays (ELISA),
including IgM-capture ELISA (MAC-ELISA) and IgG ELISA, can identify viruses that cause encephalitis soon after
infection. Polymerase chain reaction (PCR) can identify small amounts of viral DNA.
CT scan (computer tomography) and MRI scan (magnetic resonance imaging) produce computer images of the brain and
are used to detect abnormalities such as swelling (edema) and bleeding (hemorrhage). MRI is able to detect
abnormalities earlier in the course of the infection. EEG(electroencephalogram) involves placing electrodes on the
scalp to record and analyze electrical activity in the brain. Wave patterns can suggest seizure disorder or a
specific viral infection, such as herpesvirus.
Spinal tap, or lumbar puncture, is performed to detect signs of infection in cerebrospinal fluid and help make a
diagnosis. In this procedure, a needle is inserted between two lumbar (lower spine) vertebrae, cerebrospinal fluid is
collected, and the fluid is analyzed for elevated white blood cell counts, blood, and the presence of virus.
That's pretty impressive - note how much effort is put into establishing a correct diagnosis.
'ADHD', the public is being told, is also a 'brain disease'. So, you'd expect it to be diagnosed using similar methods
and with similar effort and care, wouldn't you?
In fact, it transpires that 'ADHD' is typically diagnosed using something called the 'Conners' Rating Scales'.
They're essentially questionnaires.
The Conners Rating Scales were devised by C. Keith Conners, Ph.D., in 1970.
Here's something about them from the sales page of a company called Wide Range:
"The new Conners' Rating Scales Revised (1997) incorporate many new enhancements to a set of measures that have long
been the standard for assessing attention-deficit/hyperactivity disorder (ADHD) in children and adolescents.
Normative data for the revised forms comes from a large community based sample of children and adolescents collected
throughout the United States and Canada. Norms are available for children and adolescents aged 3 to 17 on the parent and
teacher rating forms.
The scales correspond with symptoms used in the DSM-IV(tm) as criteria for ADHD. They also contain a new empirically
created index for assessing children and adolescents at risk for a diagnosis of ADHD.
Conners' Parent and Teacher Rating Scales are available in short and long versions and offer alternate measures with
varying content and psychometric properties. The CTRS-R: L and the CPRS-R: L include a ColorPlot(tm) Profile Form to
profile scale scores. The ColorPlot(tm) Forms are ideal for presenting assessment results to parents."
The 'Complete Parent/Teacher Kit' costs $168.00 (that's about $250 Australian).
Here's an article about the Conners' Rating Scales that's probably a little more neutral. It's from the 'Clinical
Psychology Review', Volume 21, Issue 7, October 2001 (Pages 1061-1093)
THE CONNERS' PARENT RATING SCALES: A CRITICAL REVIEW OF THE LITERATURE
William J. Gianarris, Charles J. Golden and Lorie Greene
Center for Psychological Studies, Nova Southeastern University, 3301 College Avenue, Fort Lauderdale, FL 33314, USA
The Conners' Parent Rating Scales (CPRS) have undergone a considerable amount of scrutiny-and subsequent refining,
reshaping, and revising-since their development in 1970. While such longitudinal scrutiny has ultimately led to a
more reliable, valid assessment tool, it has left behind a wake of literature filled with misinformation and ambiguity.
Multiple versions of the Conners' Rating Scales (CRS), their misuse, and inaccurate reporting by researchers have created
a body of literature that is difficult to interpret and misleading to both researchers and clinicians. This review
is aimed at clarifying issues regarding the proper use of the CPRS as both a diagnostic instrument and a research
Before the specific properties of the CPRS can be discussed, it is important to understand the body of literature from
which the data were drawn. A total of 108 studies spanning nearly three decades were reviewed, with 51 of these being
rejected for a variety of reasons (e.g., statistical inadequacy, inaccurate reporting, etc.). Of the remaining
57 studies, few specifically targeted the CPRS. Instead, the diagnostic and psychometric properties were taken as a
given, and the scales were used to diagnose, track therapeutic progress, validate new measures, and so on.
The general willingness to use the CPRS at face value has created 30 years worth of research with little specific
exploration into the validity and reliability of a now widely used assessment tool. This paucity of targeted
research has made a retrospective analysis of the CPRS difficult at best. Much of the data presented in this
review have been extracted from studies in which the CPRS were used in an ancillary or supportive role--leaving
conclusions about the scales themselves to be pieced together from an assortment of disconnected investigations.
In addition, many researchers failed to accurately report which of the CRS were used, how the test was scored, or
which factors were included or omitted in the analyses. Although these omissions were likely due to the fact that
the research was not focused specifically on the CPRS, often times the data were rendered useless in furthering the
understanding of the CPRS. Despite this lack of cohesion and focus, some relatively well-supported conclusions regarding
the use of the CPRS have been extracted from this disjointed body of literature.
As the CPRS were originally introduced as instruments to assess hyperactivity in children, it seems logical to first
discuss how well the literature supports this claim. In the nearly 30 years since its development, research has shown
the CPRS to be not only sensitive to hyperactivity, but to a wide variety of external (e.g., natural disasters, parental
involvement) and internal (e.g., Fragile X syndrome, brain development) influences. The CPRS appear to be a reliable
and valid tool in assessing general psychopathology, but seems to fall short in its ability to discriminate along diagnostic lines.
Studies showed that the CPRS consistently distinguishes ADHDs from normals (e.g., Ackerman and Plomin). However,
numerous studies also show that the CPRS is equally effective at distinguishing normals from many other
psychiatric controls such as conduct disorder (Barkley, 1984), learning-disabled (Kuehne et al., 1987), and children
who had hydrocephalus as infants (Fernell et al., 1991).
Only a single, unreplicated study reported that the CPRS is able to separate subjects within a group of mixed
psychiatric diagnoses (Kuehne et al., 1987). Zelko (1991) found that the ASQ-P/T was able to distinguish between ADHD,
mixed psychiatric controls, and normals, but was unable to effectively separate the psychiatric controls by diagnostic
grouping. Many studies revealed that the CPRS is unable to differentiate ADHD from other disorders (e.g., Stein and
Zelko) and no support was found that any specific scale was effective at identifying its corresponding disorder (i.e.,
Learning Problems Scale effectively discriminating Learning-Disabled subjects).
It is important to realize that the CPRS is not a direct measure of the child's behavior, but a reflection of
parental perception. In fact, research shows that the CPRS correlates poorly with laboratory testing (e.g., Plomin &
Foch, 1981) with only a few exceptions, such as selected scores on CPTs (Seidel & Joschko, 1991), and activity level
(Reichenbach et al., 1992). As a measure of perception, the CPRS is subject to influence by any number of external
variables such as parental mood and ability, time spent with the child, motivation for seeking treatment, and the locus
of behavior of most concern to the rater.
Rapoport and Benoit (1975) found that while mothers' diaries of their children's behaviors correlated well with direct
observation, CPRS scores did not. This may indicate that parents have a difficult time translating behavior into
clinically differential classifications such as "Pretty much" and "Very much." Numerous studies showed that factors
such as parents' mental health (Frick and Schaughency), number of children in the household (Conger et al., 1984),
marital discord (Smith & Jenkins, 1992), and time spent with the child (Fitzgerald et al., 1994) all have a significant
effect on CPRS scores.
CPRS scores can also vary according to which area of behavior the rater judges to be most important. Chelune et
al. (1986) found that parent ratings of hyperactivity correlated well with physician ratings of medication
response, while teacher ratings were inversely related, suggesting that perhaps the parents and teachers were using
entirely separate criteria in evaluating behavior, or perhaps viewed similar behavior in opposing ways. Lastly, a
parent's ratings may be elevated due to some need for the child to be sick (e.g., litigation) or ratings may be
deflated secondary to denial or a desire for the child to be well. It is of utmost importance when using the CPRS to keep
in mind the myriad of potential influences that may significantly alter scores.
While the CPRS has a number of empirically and conceptually sound qualities, it is also limited by its scope and
indirect nature. The CPRS was designed as a global measure intended to provide an overall picture of a child's pattern
of behavior. It is not a situation-specific measure that narrowly focuses clinical attention on any one particular
facet of a child's life. When considering the CPRS as a diagnostic tool, it is crucial to remember that its
usefulness is limited by the manner in which it assesses behavior. Its questions tap into a broad range of behaviors,
with no focused attention given to any particular area.
While the CPRS's lack of situation specificity may rule it out as the test-of-choice when the goal is a detailed
assessment of a specific problem area (e.g., peer relations, anger management), its broad scope is more likely to
identify the existence of a wide range of problematic areas. These problem areas can then be targeted for further
scrutiny and/or therapeutic attention. Therefore, the CPRS is ideally suited for an initial evaluation or an intake
assessment where a panoramic view is desired.
While broad in nature, the CPRS can also be seen as narrow in scope, for it extracts information from only one source-
parents. Due to this limited scope, the CPRS cannot be used as the sole instrument on which to base a diagnosis.
Research consistently shows the interrater reliability of the CRS to be low to moderate. Reliability between mother
and father has been found to be between 0.59 and 0.35 (Conners and Fitzgerald), while reliability between parents
and teachers is even lower, from 0.33 (Barkley, 1988) to 0.03 (Chelune et al., 1986).
Such potential discrepancy in perception among raters, while demonstrating the CRS's vulnerability to rater bias, adds
significant value to the scales diagnostic utility. Comparisons between mother and father or between parents and
teachers may lend valuable diagnostic information regarding consistency of behaviors and how they are perceived across
situations and observers. The latest revision offers a self-report scale to add to the diagnostic battery. In addition,
a clinical interview, laboratory testing, and direct observation may all complement the CRS to produce a well-
rounded, comprehensive picture. The consensus found in the literature seems to support the idea that the CPRS is an
effective instrument at discerning normal children from those with significant problems but is empirically unable
to function as a stand-alone diagnostic tool.
If a screening device is needed, there can be little doubt that the CPRS is an excellent choice. It is time-and cost-
effective, easily distinguishes normals from a psychiatric population, covers a broad range of behaviors, and is
available in an abbreviated form that is as equally effective as the long form in identifying deviant behavior
(e.g., Zelko, 1991). As a research tool, the CPRS's strengths lie in its brevity, ease of administration,
reliability across time, and sensitivity to treatment. While the CPRS can be useful in identifying subjects for inclusion
in a normal control group, it is crucial to remember that its discriminative power is not high enough to warrant its
use in selecting a homogenous experimental psychiatric group (i.e., ADHD, Conduct Disorder, etc.). The use of the CPRS as
the sole basis for group inclusion is likely to produce invalid results that will not bear up to the scrutiny of
Despite its limited diagnostic utility, the CPRS appear to be useful for monitoring treatment outcome. Well-controlled
studies have shown the Conners' Scales to be sensitive to a wide array of treatments (e.g., psychopharmacological,
behavioral, dietary) among varied populations (e.g., hyperactive, autistic, cancer patients). The overwhelming
consensus among the literature points to the abbreviated forms of the Conners' Scales as the ideal choice for
measuring treatment outcome (e.g., Conners and Pollock). In addition to the empirical backing in the literature, the
use of the abbreviated forms is further supported by the ease in which they are administered, completed, and scored.
Despite the misleading title (i.e., the HI), the abbreviated forms represent the items found to be most sensitive to
treatment effects and are therefore the obvious choice when deciding on a treatment outcome measure.
Two important points must be considered when using the Conners' Scales to monitor therapeutic progress. First,
there is well-documented evidence of a significant drop in scores from the first to the second administration of the
CRS (e.g., Conners; DiTraglia and Fischer). Such a drop could produce the erroneous impression that a treatment is
effective if only the first and second administrations are considered (e.g., baseline vs. posttreatment). For this
reason, the Conners' Scales should be administered twice before treatment is initiated, with comparisons being drawn
between the second and subsequent administrations.
Second, research has shown that as a measure of global behavior, the CRS are subject to rater bias, which may
influence how a particular treatment is perceived. For example, DiTraglia (1991) found that while teachers
considered the therapeutic effects of methylphenidate to be significant, parents did not, suggesting that perhaps the
effects that one rater considered beneficial (i.e., overall dampening of behavior), the other considered deleterious.
Conners et al. (1976) found a high rate of agreement between parents and teachers when a nonpharmacological treatment was
used, providing further evidence that the Conners' Scales may be particularly susceptible to rater perception.
In the almost 30 years since its development, the CPRS has become a standard in the assessment of ADD with
Hyperactivity. Unfortunately, its use has not been so well standardized. The misuse of the CPRS seems to stem from the
scale's earliest years of development when both researchers and clinicians applied the CRS in the absence of solid
empirical backing and without the necessary guidance of an officially published manual. As a result of this early
misuse, the CPRS continues to be applied outside the scope of its proven effectiveness and to be misrepresented in the
Conners designed the ratings scales to be brief, easy to score, and easy to administer. These qualities are
particularly useful when multiple administrations are necessary or if time and expense are necessary
considerations. The likelihood that an important area is overlooked on initial assessment is significantly reduced
due to the CPRS's broad scope and standardized administration. This highly standardized nature of the
scales also allows for more relevant comparisons of scores to the normative base that is now an integral part of the
solid empirical foundation behind the CPRS.
Thus, research identifies three primary uses for the CPRS: a general screen for psychopathology, an ancillary diagnostic
aid, and a general treatment outcome measure. While the CPRS easily identifies children manifesting problematic behavior,
it fails to consistently separate those children along diagnostic lines. Due to this shortcoming and its
sensitivity to deviant behavior, the CPRS should not be used to diagnose, but rather as a screening device to help
target children potentially in need of medical or psychological treatment. Even though the CPRS should not be
used alone to diagnose, its comprehensive approach and standardized administration provide a solid complement to
any diagnostic evaluation. Lastly, treatment outcome can be easily monitored through the use [of] the abbreviated scales
contained within the CPRS. By design these abbreviated scales, in addition to providing simplicity and increasing
compliance, are especially effective in detecting behavior change in response to treatment.
As the CRS have firmly entrenched themselves over the last 30 years as a standard assessment tool in the diagnosis of
attentional disorders, it is unlikely that their use will diminish. On the contrary, with the release of the most
current edition of the CRS, there appears to be a concerted effort to reintroduce the scales as a "new family of
diagnostic rating scales for children and adolescents" (Conners, 1997), to refine the utility of the scales, and
to provide a more solid empirical foundation for their use through an expanded normative population. The current body
of literature, as stated previously, is lacking in studies directly targeting the CPRS. If the use of the scales is to
appropriately continue, further research aimed specifically at the CRS is needed to more precisely define the role that
the scales should, and are able to, play.
The latest revision of the CRS offers researchers an opportunity to more systematically and accurately explore
the properties of the CRS. Though the most recent revision of the CRS does not differ drastically from the previous
revisions, it does contain additional items (i.e., ADHD Index, DSM-IV Symptoms subscale) that need to be empirically
validated. Due to the latest revision's similarity to previous editions, the opportunity also exists to reevaluate
the scales' psychometric properties that have been accepted (almost blindly) by the psychological community. By focusing
future research efforts on the most recent edition of the CRS while bearing in mind the limitations revealed in
previous studies, perhaps strides can be made towards a greater understanding of this widely accepted but poorly
I noticed two comments in particular in the above article:
"It is of utmost importance when using the CPRS to keep in mind the myriad of potential influences that may
significantly alter scores."
"While the CPRS easily identifies children manifesting problematic behavior [Excuse me? Simple observation, does
that, too - and it doesn't cost $168 a go!], it fails to consistently separate those children along diagnostic lines.
Due to this shortcoming and its sensitivity to deviant behavior, the CPRS should not be used to diagnose, but
rather as a screening device to help target children potentially in need of medical or psychological treatment."
Those comments suggest to me that, firstly, the Conners' Rating Scales are about as scientific as a pop survey in a
teen magazine; secondly, they clearly shouldn't be used to diagnose 'ADHD'!
But, I know that they ARE used to diagnose 'ADHD' - from what I've read, they're the most popular method used for
that purpose. When I posted a question about this at the discussion forum of Texans For Safe Education, this is the
reply I got from one of its members:
The points are interesting you brought up. Most of the parents I have been involved with... I keep hearing from the
parents the same story....In most cases the Neurologist or Peds doc....takes a gander at the scale....may ask a few
questions read rapidly off a mini "IQ" test and then reaches for the prescription pad to write a script for a
psychotropic such as MPH. So in essence the doc has used that tool in the article to diagnose. Just the opposite
[of] what the article says. Some parents have been fortunate to have some doctor do at least a BP and listen
to the heart and stretching it looking at the eyes. But most appointments have lasted no more than a half hour at
the most.....I had one that lasted 12 minutes."
So, there we have it. A child with the symptoms of the brain disease encephalitis will typically be diagnosed using a
combination of the child's medical history, a neurological examination, blood and urine tests, brain imaging tests and
a spinal tap. A child with the 'symptoms' of the 'brain disease' 'ADHD' will typically be 'diagnosed' by ticking
boxes on a questionnaire.
This has apparently been going on for many years - to the extent that there are now millions of children diagnosed
with 'ADHD'. Parents have been informed that their child may have a 'neurobiological disorder' - a 'brain disease'
- yet a 'diagnosis' of that allegedly serious affliction has then been achieved (sometimes, it seems, in a matter of
minutes!) using little more than a pop survey!!
Why have the parents not noticed the highly suspicious and rather obvious discrepancy?