Clinical Neurology and Neurosurgery 162 (2017) 72–79

Contents lists available at ScienceDirect

Clinical Neurology and Neurosurgery

journal homepage: www.elsevier.com/locate/clineuro

Towards reporting standards for neuropsychological study results: A MARK

proposal to minimize communication errors with standardized qualitative
descriptors for normalized test scores
⁎
Mike R. Schoenberg , Ruba S. Rum
Department of Neurosurgery and Brain Repair, University of South Florida Morsani College of Medicine, Tampa, FL, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Objective: Rapid, clear and efficient communication of neuropsychological results is essential to benefit patient
Neuropsychological assessment care. Errors in communication are a lead cause of medical errors; nevertheless, there remains a lack of con-
Ethics sistency in how neuropsychological scores are communicated. A major limitation in the communication of
Qualitative descriptors neuropsychological results is the inconsistent use of qualitative descriptors for standardized test scores and the
Reporting errors
use of vague terminology.
Practice standards
Patients and methods: PubMed search from 1 Jan 2007 to 1 Aug 2016 to identify guidelines or consensus
statements for the description and reporting of qualitative terms to communicate neuropsychological test scores
was conducted. The review found the use of confusing and overlapping terms to describe various ranges of
percentile standardized test scores.
Results: In response, we propose a simplified set of qualitative descriptors for normalized test scores (Q-Simple)
as a means to reduce errors in communicating test results. The Q-Simple qualitative terms are: ‘very superior’,
‘superior’, ‘high average’, ‘average’, ‘low average’, ‘borderline’ and ‘abnormal/impaired’. A case example illus-
trates the proposed Q-Simple qualitative classification system to communicate neuropsychological results for
neurosurgical planning.
Conclusions: The Q-Simple qualitative descriptor system is aimed as a means to improve and standardize
communication of standardized neuropsychological test scores. Research are needed to further evaluate neu-
ropsychological communication errors. Conveying the clinical implications of neuropsychological results in a
manner that minimizes risk for communication errors is a quintessential component of evidence-based practice.

1. Introduction standardized, reliable, and norm-based assessments to measure brain

Neuropsychological evaluations provide unique information to the
referring clinician and patient (and caregivers as appropriate), offering
answers to diagnostic and treatment-related questions in terms of brain-
behavior correlates, capacity, rehabilitation, and prognosis [1–5]. It is
essential the assessment results be communicated accurately, clearly
and efficiently to promote patient health, reduce costs, and minimize
risks for communication errors [6–9]. Ultimately, practice patterns that
diminish the efficient, understandable, and timely communication of
results adversely affect patient safety, reduce the quality of patient care
and increase costs [8,10].
1.1. Types of neuropsyhological data to convey
A central component of a neuropsychological evaluation consists of
function that can include attention/executive, language, memory/
learning, and/or visuospatial/constructional functions as well as pro-
cessing speed, academic skill development and fine motor speed/dex-
terity [2,11,12]. The tests are standardized, using normative data and
regimented administration procedures to allow for generalization of
measured behavior to established brain-behavior relationships for the
purposes of health care provision, rehabilitation, prognosis or capacity
decisions. In an ideal case, appropriate norms are used to increase the
specificity of test findings while simultaneously reducing the test var-
iance associated with other demographic variables such as age, gender,
and education level. A consequence of norm-referenced tests is that
cutoff values and the normative set used can dramatically influence: (1)
test sensitivity, (2) test specificity, and (3) diagnostic validity [13]. The
clinical interpretation is made particularly complex given that a single
neuropsychological assessment can include multiple (10 or more)

⁎
Corresponding author.
E-mail address: mschoenb@health.usf.edu (M.R. Schoenberg).

http://dx.doi.org/10.1016/j.clineuro.2017.07.010
Received 24 February 2015; Received in revised form 5 June 2017; Accepted 11 July 2017
Available online 25 July 2017
0303-8467/ © 2017 Elsevier B.V. All rights reserved.
M.R. Schoenberg, R.S. Rum
neuropsychological tests, each having an independent normative data
set from which standardized test scores are derived that can also differ
in extent demographic factors (age, education, gender, ethnicity, etc.)
are incorporated. Further, the same neuropsychological test may have
more than one (sometimes multiple) independent normative data sets a
clinician may use (see for example Trail Making Test normative data
sets [14,12,15]). In addition to normative test data, the neuropsycho-
logical study also includes history of symptoms or problems, medical/
psychiatric history, social/occupational/development history, mental
status, behavioral observations, and observations about study validity/
reliability. To provide holistic, personalized, and reliable medical care,
a neuropsychologist must be able to interpret and convey both quali-
tative aspects of the patient’s behavior and quantitative test data into
meaningful clinical judgment to answer referral question(s).
1.2. Conveying neuropsychological results
There is no agreed upon standard for a neuropsychological report
format, and neuropsychologists and users of the report may have dif-
fering perspectives of the ideal format and style [2,16–19,3–5,9]. The
existing recommendations for communicating neuropsychological re-
sults provide general guidance in what sections a written report should
have and calls for writers to clearly convey results/diagnosis and re-
commendations [2,9]. Furthermore, testing standards [6] and profes-
sional ethics [7] highlight the need to produce reports that: (1) support
the role of the neuropsychologist as a consultant by encouraging
communication of results, (2) is tailored to satisfy the need for time-
liness for communicating results, and (3) minimizes the risk for com-
munication errors [6,7,2,8,3–5,9]. The proposed suggestion for a neu-
ropsychology reporting guideline mirrors the recommendations
established by radiology [20–22] for reporting results such that the
format for communicating study results is less important than is it that
the report is: (a) timely and (b) efficiently conveys results that minimize
potential for communication errors. Unfortunately, there has been no
guidance or consistency in how standardized test scores are commu-
nicated.
2. Methods and materials
2.1. The need for consensus in communicating neuropsychological test
results
It is well understood across medicine that communication errors
adversely affect patient care, contribute to medical errors, and increase
costs [8,10]. “PubMed search from 1 Jan 2007 to 1 Aug 2016 for terms
related to ‘neuropsycholog*’, ‘reporting standards’, or 'consensus
statements' did not result in any practice guidelines to standardize the
communication of qualitative descriptors for neuropsychological nor-
mative test scores (see [2]”). At a minimum, the neuropsychological
report should clearly communicate if the study is: (1) abnormal and
related to known or suspected neurological (neurophysiological) dys-
function, (2) equivocal and the study could reflect normal variant or
mild abnormality, but are indeterminate from the neuropsychologist’s
opinion, or (3) normal (no brain dysfunction). Unfortunately, the
qualitative descriptors used to describe standardized test scores (i.e.,
the term(s) used to describe a standardized test score deviating from
‘normal’ or ‘abnormal’) are highly variable, which can obfuscate the
results and increase the potential for communication errors (e.g. [23]).
2.2. Test score qualitative descriptors
The term ‘test score qualitative descriptor’ refers to the terms au-
thors use to communicate how a patient performed on a norm-refer-
enced neuropsychological test and not to descriptors of mental status.
There are at least two inter-related problems in neuropsychological
practice related to the use of test score qualitative descriptors. First,
73
Clinical Neurology and Neurosurgery 162 (2017) 72–79
there is marked variability in the test score qualitative descriptors used
to describe neuropsychological standardized scores [23]. Second, there
is a lack of consensus in the range of standard scores/percentiles that
correspond to a particular test score qualitative descriptor (i.e., what
scores are “average”) [24,25,19,26,23,27,14,28,29]. Despite repeated
calls for more uniform use of the test score qualitative descriptors that
delineate the relative uniqueness or statistical probability of test scores
(e.g., ‘below average’, ‘mildly impaired’, ‘borderline’, ‘extremely low’,
etc.), there remains excessive variability and no consensus among
neuropsychologists [e.g. [2,17,23]]. Further adding to the confusion,
test score qualitative descriptors that sound similar (e.g., ‘low average’,
‘low normal’, ‘borderline’, ‘below average’) do not overlap in terms of
how rare or unusual a score is in its interpretation. Indeed, a survey of
110 neuropsychologists [23] found the index score of 70 (2nd percen-
tile) was described using 22 different test score qualitative terms, with 6
different terms using the word ‘impaired’ (e.g., ‘impaired’, ‘borderline
impaired’, ‘mildly impaired’, ‘moderately impaired’, ‘severely im-
paired’, and ‘significantly impaired’).
Currently, there are at least three different test score qualitative
classification schemas (with multiple permutations) that are generally
recognized [14,28,29]:
1. Clinical Classification system advocated by Heaton et al. [14]
2. Clinical classification advocated by Schretlen et al. [28]
3. The David Wechsler/Intelligence classification system [29]
The three commonly used qualitative classification systems above,
along with a fourth based on the Wechsler classification system that
uses different test score qualitative terms [26], are displayed in the first
four columns of Fig. 1. Surprisingly, there is lack of agreement for most
terms, including the most commonly used test score qualitative de-
scriptor, ‘average’. The Wechsler system [29] suggests scores falling in
the 25–74th (or 75th) percentiles are ‘average’ while the Heaton et al.
[14] classification system identifies scores falling between the 30–67th
percentiles as ‘average’. The Schretlen et al. [28] classification system
generally mirrors the Wechsler system and describes scores between
24–74th percentiles as ‘average’. There is even less consensus for per-
centile scores that fall outside of the ‘average’ range. A score falling at
the 9th percentile may be ‘below average’ using the Wechsler classifi-
cation system [29], ‘mildly impaired’ using the Heaton et al. [14]
classification system, or ‘low average’ using the Schretlen et al. [28]
rating system. Even more troubling is the use of similar test score
qualitative descriptors for different ranges of percentile scores between
the classification schemes (i.e., ‘low average’ versus ‘below average’).
For example, ‘low average’ describes scores falling between the 9–24th
percentiles using the Wechsler classification scheme [29], while “below
average” is used to describe scores ranging from 16–27th percentiles
using the Heaton et al. [14] system. The term ‘borderline’ equates to
scores ranging from the 2–8th percentiles using the Wechsler [29] and
Schretlen et al. [28] systems, but Heaton et al. [14] describes scores
from the 6th-15th percentiles as ‘mildly impaired’. Thus, Heaton et al.’s
[14] ‘mildly impaired’ scores reflect scores that are delineated as
‘borderline’ or ‘low average’ by Wechsler [29] and Schretlen et al. [28].
The confusion of terms and disagreement in score ranges lacks precision
and will contribute to errors in communicating results to health pro-
viders and patients [23]. Indeed, the lack of precision has contributed
to recommendations by neuropsychologists to include test scores in the
neuropsychological report itself [16,19].
To address the lack of consistency in test score qualitative de-
scriptors when communicating the results of neuropsychological as-
sessment, a simplified qualitative reporting system is recommended,
and is delineated in Table 1. This classification categorization is also
presented in reference to the four commonly used qualitative classifi-
cation systems in Fig. 1.
M.R. Schoenberg, R.S. Rum Clinical Neurology and Neurosurgery 162 (2017) 72–79

Fig. 1. Commonly Used Qualitative Descriptors with

Corresponding Standard Scores and Percentiles.

Table 1 deficits in the context of neuropathological patterns and appropriate

Proposed Simplified Clinical Test Score Qualitative Classification (Q-Simple) System. clinical history [24,25,30,31,27,32,11,33,34]. That is, test scores in this
Qualitative anchors for neuropsychological standard scores, T-scores and Percentile
range reflect performances below the 16th percentile that may be in-
equivalents.
terpreted as unusual or low by patients and clinicians alike, but are also
Reference Index Reference T- Reference commonly observed in studies evaluating the base rates of test scores
Scores Scores Percentile Scores among healthy individuals [24,25,30,31,27,32–34]. Thus, these test
scores may reflect a mild abnormality, a premorbid weakness or mea-
Suggested Qualitative Descriptor
Normal (scores are normal and/or above average):
surement error.
Very Superior: 130+ 70+ 98+ The next test score qualitative descriptor, ‘abnormal’, describes
Superior: 120–129 64–69 92–97 statistically rare scores (lowest 5th percentile of performances com-
High average: 110–119 57–63 75–91 pared to a normative peer group), and possibly associated with brain
Average: 90–109 44–56 25–74
dysfunction. While intuitively appealing, data are lacking to indicate
Low average: 85–89 40–43 16–24
that the communication of an “impaired” or “abnormal” neu-
Borderline/equivocal (scores may be normal or abnormal):
ropsychological test result are improved by providing subclassifications
Borderline: 77–84 35–39 6–15
(e.g., ‘mildly’ impaired, ‘mild-to-moderately’ impaired, ‘moderate-to-
Abnormal (scores are statistically rare and can reflect brain dysfunction): severely’ impaired, etc.) [23]. Further, there is a lack of consensus
Abnormal/ < 77 < 35 <6
Impaired
where “mildly abnormal” standardized test scores end and more severe
(“moderately abnormal/impaired”, “profoundly abnormal/impaired”)
standardized test score ranges begin across the current multiple quali-
3. Results tative classification systems. Another factor that complicates the com-
munication of standardized test scores with qualitative descriptors is
3.1. Proposal for a simplified clinical test score qualitative descriptor that the normative data used to derive standardized scores vary across
classification system neuropsychological tests with some scores being age-corrected, while
other tests have normative data that is age-, education- and other de-
Development of the proposed Simplified Clinical Test Score mographic (e.g., ethnicity and/or gender) based [14]. To avoid confu-
Qualitative Classification (Q-Simple) system was based on the in- sion, the proposed Q-Simple system advocates to describe all percentile
creasing demand to reduce potential for communication errors, reduce scores that are at or below the 5th% regardless of test/normative
health care costs, and promote evidence based care that has been sample used. Taken the limitations and complicating factors together,
coupled with an ongoing lack of consensus among clinicians to describe the Q-Simple system proposes to use a single descriptor (e.g., ‘ab-
standardized test scores. The Q-Simple development integrated several normal’) to describe a specific range of scores (< 5th percentile) that
factors thought to be important to its utilization: (a) adhere to his- are generally agreed upon as reflecting statistically rare scores that can
torically used psychometric principals, (b) be face validity to consumers denote neuropathology. The suggestion for a general descriptor as
of neuropsychological results, and (c) reduce the potential for com- ‘abnormal’ has intrinsic interpretive value, and can cut-across the
munication errors in the written description of neuropsychological test multitude of confusing qualitative descriptors to promote clear and
scores to patients and their health care providers. The test score qua- concise communication that an abnormality exists [44]; [7,2,8,10]).
litative descriptors for the upper end of the distribution of percentile
ranges are consistent with those of Wechsler [29] and Schretlen et al. 4. Discussion
[28] systems. The term ‘borderline’ is advocated for standardized test
scores from the 6th to 15th percentiles. These scores are often equivocal There have been much debate regarding what cut-off value should
with respect to clinical importance, and occur frequently as low scores be interpreted as ‘abnormal’ to best distinguish between brain pa-
in a test battery; however, these scores can also reflect early or mild thology and normal variability (e.g., scores < 15th percentile vs < 7th

74
M.R. Schoenberg, R.S. Rum
percentile vs. < 5th percentile or < 2nd percentile) (e.g.
[24,25,31,30,27,14,32,11,33,34]). Lezak [11] has advocated for dis-
tinguishing between clinically impaired (scores between 2nd to 7th%
iles or z-scores between −1.5 and −2.0) versus pathologically im-
paired (scores < 2nd%iles or z-scores < −2.0). However, Lezak’s [11]
distinction in clinically versus pathologically impaired is not utilized by
the four common qualitative classification systems [26,14,28,29].
Further, the psychometric distinction of clinically impaired versus pa-
thologically impaired based on a cut-off z-score of −2.0 is not stable
since the standardized score will be determined by the normative
sample used, the extent that a tests’ normative data are demo-
graphically corrected, and how normally distributed the neuropsycho-
logical function (attention, verbal memory, constructional praxis) is in
a healthy population (e.g. [31,30,11,12,15,35]). Indeed, neuropsycho-
logical performance is frequently not normally distributed, with nor-
mative distributions frequently being negatively skewed and kurtotic
(e.g., apraxias, recognition memory, confrontation naming) [11,12,15].
Thus, the use of clinically impaired versus pathologically impaired
based on a z-score cut-off of −2.0 (< 2nd percentile) may deter the
neuropsychological study from identifying early or mild neurological
disease or dysfunction [24,25,31,30,27,14,32,33,34].
The Q-Simple system distills the complexities of the evolving diag-
nostic science to a pragmatic clinical approach to rapidly communicate
health data. Score description and interpretation must incorporate
knowledge about the difference between statistically rare scores within
a normal distribution, the statistical differences between two or more
scores as they occur within a test battery, and the actual base rate of
observed differences in scores within a test battery
[24,25,30,31,27,32–34,15,36–38]. The Q-Simple system takes into ac-
count the extensive data establishing the low cognitive scores that
occur commonly in a battery of tests given to healthy individuals
[24,25,30,31,27,32–34]. As an example, 78 percent of the healthy
normative sample of the Neuropsychological Assessment Battery (NAB;
[38]) were observed to have 2 of the 36 scores below one standard
deviation (SD) of the mean (< 16th percentile or z-score = −1.0), and
21.8 percent of the healthy sample had two or more scores at or below
−2 SD below the mean (< 2nd percentile or z-scores < −2.0) [25].
Within the proposed system, scores at or below the 5th percentile
(scores < −1.55 SD of the mean) are considered to be statistically rare
and labeled ‘abnormal’, but not necessarily reflective of brain dys-
function. The implication that an abnormal score is interpreted as re-
flecting brain dysfunction versus some other factor (e.g., poor task
engagement, effects of fatigue, pain, etc.) is necessarily a clinical in-
terpretation. Thus, interpretation of performance on a neuropsycholo-
gical test battery does not rely solely on scores or standard deviations
below the mean, rather it incorporates other clinical information ob-
tained from the assessment. The proposed qualitative classification
system cut-off scores for ‘abnormal’ attempts to balance the relative
frequency of low scores with efforts to have adequate sensitivity and,
most importantly, have face validity to consumers.
It is recommended the neuropsychological report specify whether
scores were interpreted based on age-matched or demographic-adjusted
(age-, education-, and gender-matched) normative data [2], because of
the clear impact various normative data corrections have on a raw test
score’s interpretation. The problem of variations in normative data
samples and the extent test scores may be affected by age-, education-
and/or ethnicity/gender effects lead to the next suggestion to improve
transparency of neuropsychological reporting, including test scores in
the study report.
4.1. Test scores should be reported
The lack of consensus in test score qualitative descriptors and em-
phasis for evidence-based neuropsychology practice [16,39] both serve
as bases to extend previous calls for practice patterns to include neu-
ropsychological test scores in the report [7,14–16]. Furthermore,
75
Clinical Neurology and Neurosurgery 162 (2017) 72–79
reporting test scores serves as a transparent anchor for the interpreta-
tion of neuropsychological data and the qualitative descriptors used for
these data [16]. In cases when a neuropsychological test has multiple
normative data sets that maybe utilized, the raw test score is desirable
to report [15,12]. Including raw scores in a report also allows for im-
proved comparison over time, particularly if other normative data sets
are developed [16]. Of note, the Q-Simple system (and the other qua-
litative score descriptor schemas detailed in Fig. 1) are focused on de-
scribing standardized test score and not raw scores. Generally, perfor-
mance on tests reported in raw scores may be best described when
normative samples are markedly skewed, with metrics such as cumu-
lative percentage frequencies and/or bases to describe absolute vari-
ables that are remarkable (e.g., raw scores of zero or all correct).
4.2. Test score qualitative descriptor versus interpretation of brain
dysfunction or rehabilitation outcome
The issue of test score qualitative descriptors vs clinical inter-
pretation can be confused (e.g., report of ‘abnormal test score’ versus
interpretation of an abnormal/impaired score that reflects neuro-
pathology), and the Q-Simple classification system emphasizes that the
test score qualitative descriptors do not dictate the presence of disease
or brain dysfunction. Abnormal (impaired) scores can be obtained for
reasons unrelated to brain dysfunction such as measurement error/
statistical variability or insufficient task engagement/attention, among
others [24,25,31,30,40,32–34]. The likelihood that impaired score(s) is
(are) due to brain dysfunction and how the obtained score(s) answers
the referral question(s) must be determined by the clinician.
The qualitative description of test scores is similarly applicable to
treatment/rehabilitative applications and placement decision, as the Q-
Simple test score qualitative descriptor system delineates standardized
test scores and not the clinical interpretation of any change in test score
or functional skills as a result of treatment. By adopting the Q-Simple
qualitative descriptor system, it can provide a standardized reference
point other clinicians can use to allow for rapid determination of any
marked deterioration or improvement without having to wait for test
scores if these are not included or appended to the neuropsychological
report.
The importance of answering the referral question and/or concisely
describing rehabilitation programming with clear and concise language
cannot be understated. An increasing problem in communicating results
in neuropsychological consultations is the use of confusing terms that
have no empirical anchor (e.g. “cognitive inefficiency.”). Not only does
‘cognitive inefficiency’ not have a psychometric anchor in any quali-
tative classification scheme, it is also unclear whether this term is used
to describe brain dysfunction, refers to a psychiatric condition, or refers
to a normally occurring process for the individual. A list of selected
problematic terms are displayed in Table 2. We provide a case example
of a neuropsychological study to illustrate reporting using the Q-Simple
qualitative reporting system.
4.3. Case example
The patient is a 48 YO right-handed Caucasian female with 11 years
of education and a history of pharmacoresistent epilepsy with cognitive
complaints referred as part of a pre-surgical work-up. Her MRI study of
brain is illustrated in Fig. 2 and shows left mesial temporal sclerosis.
Neuropsychological Results:
Premorbid: intellectual function was estimated to be in the average
to high average range.
General Cognitive/Intelligence: Index of her general cognitive
function was borderline to low average compared to age-matched peers
(Full Scale = 84, 14th percentile). Index of her verbal general cognitive
function was borderline to low average (VCI = 81, 10th percentile)
[note the use of qualitative score descriptors ‘borderline’ and ‘low
average’ was used since the VCI score has a standard error of measure
M.R. Schoenberg, R.S. Rum Clinical Neurology and Neurosurgery 162 (2017) 72–79

Table 2
Summary of Ineffective and confusing qualitative descriptors to avoid in neuropsychological reports problem.

Term or phrase Setback Recommendation

Low average to superior range This is not a range; this signifies variability Scores in function x were variable, ranging from low average to the
superior range. This should be followed up with possible explanation
for variability (e.g., time vs. untimed tasks
Cognitive inefficiency Vague term Use more precise term. Slowness; inaccuracy, etc.
No anchor
Error-free Awkward accurate
Adequate Vague, subjective; relative term with varied implication across Recall was in the average range, consistent with XX’s general level of
patients. intellectual functioning.
No anchor
Recall was somewhat compromised Vague description Recall declined from 10 to 6 items (percentile)
No anchor
Immediate recall (9/12, 20%ile) Misleading; implies patient recalled more material after the e.g., Recall remained stable following a 30 min delay (9/12; average
improved after a 30-min delay (9/ delay. It is the neuropsychologists job to interpret normative range or 40 percentile). Or 30-min delayed recall revealed no
12, 40%ile) data and communicate appropriately forgetting (9/12, 40 percentile)
Below Average Term describing scores 16th to 27th percentile See Table 2 for proposed Q-Simple qualitative descriptors and/or
Technically, may also be used to describe scores below the reference%ile ranges and/or provide test percentile score.
average of the normative group (i.e., scores 49th percentile or
less)
Borderline Low Average Typically describing scores 6th to 15th percentile See Table 2 for proposed Q-Simple qualitative descriptors and/or
provide test percentile score.
Low Normal Typically, scores between 16th to 24th percentile See Table 2 for proposed Q-Simple qualitative descriptors an provide
test score percentile.
Unusually Low Scores between 3rd to 9th percentile See Table 2 for proposed Q-Simple qualitative descriptors and/or
provide test percentile score.
WNL or Within Normal Limits Can be scores falling within expectations for the individual or Can be appropriate in limited situations. Also see Table 2 for
equal to or above the 16th percentile (e.g., 16th percentile or proposed Q-Simple qualitative descriptors and/or provide reference
greater). percentile ranges and/or provide test(s) percentile score(s)

Memory: Scores varied from abnormal to average compared to age-

Fig. 2. MRI coronal image demonstrating left mesial temporal sclerosis for 48 YO female
with left temporal lobe epilepsy.
that could place it in either percentile score range]. She obtained an
index of nonverbal general cognitive function that was average
(PRI = 96, 39th percentile).
Attention/Executive: Basic attention intact. She repeated 6 digits
forward and 4 digits backward, which is low average. She scored ab-
normal to low average on tasks involving selective and/or divided at-
tention. No gross behavioral apathy or disinhibition. Compared to de-
mographically matched peers, verbal reasoning score was abnormal
while a visual reasoning score was average. Letter fluency was bor-
derline. Figural fluency scores were average. Rapid (timed) sequencing
and set-shifting score was average (Trail Making Test Part B = 67, 53rd
percentile). She scored low average on a complex problem solving,
sequencing and set-shifting task (WCST-64).
matched peers.
Immediate memory: Verbal immediate memory was abnormal for
stories and a word list. Visual immediate memory was low average to
average [note – this highlights score at 25th percentile, which falls
between low average and average].
Delayed memory: Verbal delayed memory was abnormal for stories
and a word list. Visual delayed memory was average for figures and a
complex figure. Verbal recognition cues did not improve recall. Visual
recognition memory was within normal limits.
Language: Receptive and expressive speech was functional with
adequate articulation and speech rate. Comprehension and repetition
were intact. Confrontation naming was borderline (BNT = 49/60) with
7 additional given phonemic cues. Semantic verbal fluency was ab-
normal. Letter fluency was borderline. No gross alexia or agraphia.
Visuoperceptual skills were intact. Visuoconstructional skills were
borderline. Complex figure copy was borderline with retained gestalt
but missing one detail. No gross spatial imprecision or distortion. Block
constructions were borderline. No constructional apraxia.
Fine Motor: manual dexterity was borderline with her right hand,
but average with her left hand.
Mood: She endorsed clinical symptoms of depression.
Neuropsychological Study Conclusions:
[Note. diagnoses and recommendations are not detailed below for
purposes of brevity]
Neuropsychological study was abnormal due to deficits in atten-
tion/executive, verbal memory, and language (confrontation naming
and semantic fluency) functions. The deficits in attention/executive and
language functions were mild to moderate while the deficits in verbal
memory were severe and indicate almost complete loss of declarative
verbal memory functions. There were clinical symptoms of depression.
Neuropsychological study results are consistent with left temporal
lobe dysfunction and known mesial temporal sclerosis assuming normal
functional neuroanatomical organization. Further clinical correlation
recommended for surgical planning.
The patient’s cognitive complaints are associated with brain dys-
function that is frequently associated with localization related epilepsy.

76
M.R. Schoenberg, R.S. Rum
Her learning and memory of verbal material was impaired, and ac-
counts for her difficulty at work and forgetfulness at home with lapses
in completing instrumental activities of daily. She was observed to close
her eyes when given verbal material to learn/remember, which she
described as an effort to visually picture verbal material given to her.
However, she found the rate that information was given to her was too
fast to compensate by this adaptation to verbal memory deficit.
Depressive disorder likely due to neurophysiological dysfunction
with noted exacerbation after a seizure for several hours to a day.
Pharmacological treatment is indicated.
Surgical Candidacy. From a neuropsychological standpoint, the pa-
tient is a good surgical candidate. She is at low risk for decline in
memory or language following a selective left temporal lobectomy.
Neuropsychological Data Summary Report
Note: Grooved Pegboard Test = G. Peg; Test of Premorbid
Intelligence = ToPF; Wechsler Adult Intelligence Scale – 4th
Ed. = WAIS-IV; Trail Making Test, Parts A and B = TMT Part A and
TMT Part B; Rey Auditory Verbal Learning Test = RAVLT; Wechsler
Memory Scale – 4th Ed. = WMS-IV; Boston Naming Test = BNT;
Verbal Fluency Tests (Semantic and Letter); Rey-Osterrieth Complex
Figure Test = ROCFT; Ruff Figural Fluency Test = RFFT; Wisconsin
Card Sorting Test-64 card version = WCST-64; Beck Depression
Inventory – 2nd Ed. = BDI-II.
77
Clinical Neurology and Neurosurgery 162 (2017) 72–79
4.4. Summary and limitations
Neuropsychological studies provide unique information for the
medical management of patients (whether diagnostic or rehabilitative
in purpose), and require that study results and recommendations/
treatments are conveyed in a manner that is concise and avoids com-
munication errors. Unfortunately, the test score qualitative descriptors
in neuropsychological reports are a source of confusion in commu-
nicating neuropsychological results due to: (a) lack of consensus of the
test score qualitative descriptors that describe standardized test scores,
(b) using similar test score qualitative descriptor terms (e.g., ‘below
average’ and ‘low average’) that refer to different ranges of standar-
dized scores, (c) the use of test score qualitative terms that have no
psychometric or commonly understood clinical foundation, and (d) lack
of uniformity in psychometric characterization of neuropsychological
assessments for different populations and conditions. The proposed Q-
Simple system aims to improve the communication of neuropsycholo-
gical study results, and minimize the potential for errors in information
transfer. The Q-Simple system uses test score qualitative descriptors
that are face valid, eliminates the use of overlapping and potentially
M.R. Schoenberg, R.S. Rum
confusing terms, and attempts to incorporate current research findings
and historical parametric statistical classifications.
A potential limitation of the Q-simple system is shared by other
qualitative systems, which is a lack of empirical support for the clas-
sifications. Why adopt the Q-simple system? The identified cut-off
scores for each qualitative descriptor of the Q-Simple system attempts
to integrate historical parametric statistics in describing rare scores
within a distribution along with Bayesian statistical analyses of the base
rate frequency of low scores in cognitive test batteries obtained by
healthy individuals [24,25,30,31,27,32–34]. Further, in lieu of a con-
sensus practice parameter for a qualitative descriptor reporting system,
the Q-Simple system can reduce the likelihood of communication errors
by minimizing overlapping and difficult to distinguish terms that have
been variously used to describe neuropsychological test scores (e.g.
[23]). The interpretation of test scores that fall within the confidence
interval between two test score qualitative descriptors (e.g., ‘low
average’ and ‘borderline’) require the clinician to interpret these scores
within the context of the patient and his/her presenting problem to
answer the referral question, and likely no qualitative score descriptor
system can avoid the confound of those scores that have a standard
error of measure that could fall within two qualitative descriptor do-
mains. While the use of subcategory test score qualitative descriptors
(e.g., ‘mild-to-moderately impaired’) have intuitive appeal, these dis-
tinctions have shown little reliability among clinicians [23], and there
are no data supporting that subclasification of a qualitative descriptors
reduces communication errors. The communication of subtleties or
profound level of deficits must be provided by the clinician in the Im-
pressions/Conclusions section where interpreting and integrating the
normative test scores with patient context, symptoms, medical history
and cultural factors to answer the referral question and make diagnostic
and treatment/placement decisions. The Q-simple system advocates
clinicians to use the same test score qualitative descriptor to commu-
nicate results regardless if the score was derived from age-matched
normative data or more detailed demographically matched normative
data. While conceptual and parametric limitations exist in comparing
normative scores derived from different samples and normative cor-
rections, the Q-simple system approach is proposed to be basic: using
the same qualitative descriptors for all standardized test scores, and the
psychometric limitations of this approach is then compensated by in-
cluding raw test scores in the report. Lastly, although the assessment
instructions are standardized, differences in scores and interpretations
can occur when using different test versions that use different proce-
dures (e.g., the Stroop Color-Word test) and when variations in ad-
ministration is required for unique patient/cultural factors. Generally,
it is likely desirable to have the test version and/or modifications to
procedures detailed when communicating results and how it affected
the study quality and interpretation. The essential point is to avoid
complicated language to describe standardized test scores to assure
assessment results are accurately conveyed and reduce the risk for
communication errors.
Clearly, a generally accepted test score qualitative classification
system is needed; however, any qualitative classification system pro-
posed that is not data driven by consumers of the evaluations them-
selves or based on a clinical outcome anchor point will lack relevance
beyond what is proposed here. The issue of including test scores in
reports has prompted controversy (e.g. [17,18,16,19,41,42]. Argu-
ments against including test scores in reports include the incorrect use
of scores by consumers and breach of confidentiality; however, in-
cluding test scores that are the basis for interpretation of a study is
advocated from an evidence-based practice standpoint to promote
transparency of interpretation and enhance communication of results
[2,17,18,16,19,41]. Combining the proposed Q-Simple system and re-
porting test scores can improve the communication of neuropsycholo-
gical results to patients and health providers, and promote transparency
of neuropsychological decision-making [43,16,39,5]. Consistent ad-
herence to improved practice standards stands to benefit patients,
78
Clinical Neurology and Neurosurgery 162 (2017) 72–79
referring health care providers and the field of neuropsychology.
Disclosure statement
No conflicts of interest to report.
Acknowledgements
The authors would like to thank Marla Hamberger, PhD for her
thoughtful insights and comments on the development of the proposed
reporting system.
References
[1] American Academy of Neurology, Assessment: neuropsychological testing of adults.
Considerations for neurologists, Neurology 47 (2) (1996) 592–599.
[2] Board of Directors, American Academy of Clinical Neuropsychology (AACN) prac-
tice guidelines for neuropsychological assessment and consultation, Clin.
Neuropsychol. 21 (2007) 209–231, http://dx.doi.org/10.1080/
13825580601025932.
[3] R.C. Hilsabeck, T.L. Hietpas, K.J.M. McCoy, Satisfaction of referring providers with
neuropsychological services within a Veterans Administration medical center, Arch.
Clin. Neuropsychol. 29 (2014) 131–140.
[4] R.O. Temple, J. Carvalho, G. Tremont, A national survey of physicians’ use and
satisfaction with neuropsychological services, Arch. Clin. Neuropsychol. 21 (2006)
371–382.
[5] G. Tremont, H.J. Westervelt, D.J. Javorsky, A. Podolanczuk, R.A. Stern, Referring
physicians’ perceptions of the neuropsychological evaluation: how are we doing?
Clin. Neuropsychol. 16 (2002) 551–554.
[6] American Educational Research Association, American Psychological Association,
National Council on Measurement in Education, Standards for Educational and
Psychological Testing Revised, 2014 edition, Author, Washington DC, 2014.
[7] American Psychological Association, Ethical principles of psychologists and code of
conduct, Am. Psychol. 57 (2002) 1060–1073.
[8] Committee on Quality of Health Care in America, To Err is Human: Building a Safer
Health System vol. 627, National Academies Press, Washington DC, 2000.
[9] P. Tzotzoli, A guide to neuropsychological report writing, Health (N. Y.) 4 (2012)
821–823.
[10] S. Kripalani, F. LeFevre, C.O. Phillips, M.V. Williams, P. Basaviah, D.W. Baker,
Deficits in communication and information transfer between hospital-based and
primary care physicians: implications for patient safety and continuity of care,
JAMA 297 (2007) 831–841.
[11] M.D. Lezak, Neuropsychological Assessment, 4th ed., Oxford University Press, New
York, 2004.
[12] M. Mitrushina, K.B. Boone, J. Razani, L.F. D’Elia, Handbook of Normative Data for
Neuropsychological Assessment, Oxford University Press, New York, 2005.
[13] O. Pedraza, D. Mungas, Measurement in cross-cultural neuropsychology,
Neuropsychol. Rev. 18 (3) (2008) 184–193.
[14] R.K. Heaton, S.W. Miller, M.J. Taylor, I. Grant, Revised Comprehensive Norms for
an Expanded Halstead-Reitan Battery: Demographically Adjusted
Neuropsychological Norms for African American and Caucasian Adults Professional
Manual, Psychological Assessment Resources, Lutz, 2004.
[15] E. Strauss, E.M.S. Sherman, O. Spreen, A Compendium of Neuropsychological Tests:
Administration, Norms, and Commentary, 3rd ed., Oxford University Press, New
York, 2006.
[16] G. Chelune, Evidence-based research and practice in clinical neuropsychology, Clin.
Neuropsychol. 24 (3) (2010) 454–467.
[17] J. Donders, A survey of report writing by neuropsychologists: I: general char-
acteristics and consent, Clin. Neuropsychol. 15 (2001) 137–149.
[18] J. Donders, A survey of report writing by neuropsychologists, II: test data, report
format: and document length, Clin. Neuropsychol. 15 (2001) 150–161.
[19] D. Freides, Proposed standard of professional practice: neuropsychological reports
display all quantitative data, Clin. Neuropsychol. 7 (1993) 234–235.
[20] American College of Radiology, ACR Practice Guideline for Communication of
Diagnostic Imaging Findings, American College of Radiology, Reston, VA, 2005.
[21] Board of the Faculty of Clinical Radiology: The Royal College of Radiologists,
Standards for the Reporting and Interpretation of Imaging Investigations, Royal
College of Radiologists, London, UK, 2006.
[22] D.C. Kushner, L.L. Lucey, American College of Radiology, Diagnostic radiology
reporting and communication: the ACR guideline, J. Am. Coll. Radiol. 2 (2005)
15–21.
[23] T.J. Guilmette, L.D. Hagan, A.J. Giuliano, Assigning qualitative descriptions to test
scores in neuropsychology: forensic implications, Clin. Neuropsychol. 22 (2008)
122–139.
[24] B.N. Axelrod, J.R. Wall, Expectancy of impaired neuropsychological test scores in a
non-clinical sample, Int. J. Neurosci. 117 (11) (2007) 1591–1602.
[25] L.M. Binder, G.L. Iverson, B.L. Brooks, To err is human: abnormal neuropsycholo-
gical scores and variability are common in healthy adults, Arch. Clin. Neuropsychol.
24 (2009) 31–46.
[26] G. Groth-Marnant, Handbook of Psychological Assessment, 5th ed., John
Wiley & Sons, Inc, Hoboken, New Jersey, 2009.
M.R. Schoenberg, R.S. Rum
[27] R.K. Heaton, I. Grant, C.G. Matthews, Comprehensive Norms for an Extended
Halstead-Reitan Battery: Demographic Corrections, Research Findings, and Clinical
Applications, Psychological Assessment Resources, Inc, Odessa, 1991.
[28] D.J. Schretlen, S.M. Testa, G.D. Pearlson, Hopkins Neuropsychological Normative
System Professional Manual, Psychological Assessment Resources, Inc, Lutz, FL,
2010.
[29] D. Wechsler, Wechsler Adult Intelligence Scale-revised, The Psychological
Corporation, New York, 1981.
[30] B.L. Brooks, G.L. Iverson, J.A. Holdnack, H.H. Feldman, The potential for mis-
classification of mild cognitive impairment: a study of memory scores on the
Wechsler Memory Scale-III in healthy older adults, J. Int. Neuropsychol. Soc. 14 (3)
(2008) 463–478.
[31] B.L. Brooks, G.L. Iverson, T. White, Substantial risk of accidental MCI in healthy
older adults: base rates of low memory scores in neuropsychological assessment, J.
Int. Neuropsychol. Soc. 13 (3) (2007) 490–500.
[32] L.J. Ingraham, C.B. Aiken, An empirical approach to determining criteria for ab-
normality in test batteries with multiple measures, Neuropsychology 10 (1996)
120–124.
[33] B.W. Palmer, K.B. Boone, I.M. Lesser, M.A. Wohl, Base rates of impaired neu-
ropsychological test performance among healthy older adults, Arch. Clin.
Neuropsychol. 13 (6) (1998) 503–511.
[34] D.J. Schretlen, S.M. Testa, J.M. Winicki, G.D. Pearlson, B. Gordon, Frequency and
bases of abnormal performance by healthy adults on neuropsychological testing, J.
Int. Neuropsychol. Soc. 14 (3) (2008) 436–445.
[35] D. Wechsler, Wechsler Adult Intelligence Scale, 4th ed., Psychological Corporation,
79
Clinical Neurology and Neurosurgery 162 (2017) 72–79
San Antonia, TX, 2008.
[36] D. Wechsler, Wechsler Adult Intelligence Scale, 3rd ed., Psychological Corporation,
San Antonio, 1997.
[37] D. Wechsler, Wechsler Memory Scale, 3rd ed., Psychological Corporation, San
Antonio, TX, 1997.
[38] T. White, R.A. Stern, Neuropsychological Assessment Battery: Psychometric and
Technical Manual, Psychological Assessment Resources, Lutz, 2003.
[39] D.L. Sackett, S.E. Straus, W.S. Richardson, W. Rosenberg, R.B. Haynes, Evidence-
based Medicine: How to Practice and Teach EBM, 2nd ed., Churchill Livingstone,
New York City, 2000.
[40] P. Green, M.L. Rohling, P.R. Lees-Haley, L.M. Allen, Effort has a greater effect on
test scores than severe brain injury in compensation claimants, Brain Inj. 15 (2001)
1045–1060, http://dx.doi.org/10.1080/02699050110088254.
[41] R.G. Matarazzo, Psychological report standards in neuropsychology, Clin.
Neuropsychol. 9 (1995) 249–250.
[42] R.I. Naugle, A.J. McSweeney, On the practice of routinely appending neu-
ropsychologial data to reports, Clin. Neuropsychol. 9 (1995) 245–247.
[43] S.C. Bowden, E.J. Harrison, D.W. Loring, Evaluating research for clinical sig-
nificance: using critically appraised topics to enhance evidence-based neu-
ropsychology, Clin. Neuropsychol. 28 (4) (2014) 653–668.
[44] American Psychological Association, National Council on Measurement in
Education, American Educational Research Association, Standards for educational
and psychological testing, 1st ed., American Educational Research Association,
1999.