Multiple Sclerosis Journal

Multiplehttp://msj.sagepub.com/ Sclerosis Journal

Using a highly abbreviated California Verbal Learning Test-II to detect verbal memory deficits Elizabeth S Gromisch, Vance Zemon, Ralph HB Benedict, Nancy D Chiaravalloti, John DeLuca, Mary A Picone, Sonya Kim and Frederick W Foley Mult Scler published online 17 July 2012 DOI: 10.1177/1352458512454347 The online version of this article can be found at: http://msj.sagepub.com/content/early/2012/07/17/1352458512454347

Published by: http://www.sagepublications.com

Additional services and information for Multiple Sclerosis Journal can be found at: Email Alerts: http://msj.sagepub.com/cgi/alerts Subscriptions: http://msj.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> OnlineFirst Version of Record - Jul 17, 2012 What is This?

Downloaded from msj.sagepub.com at ALBERT EINSTEIN COLG OF MED on January 22, 2013

454347 2012

MSJ0010.1177/1352458512454347Multiple Sclerosis JournalGromisch et al.

MULTIPLE SCLEROSIS MSJ JOURNAL

Short Report

Using a highly abbreviated California Verbal Learning Test-II to detect verbal memory deficits

Multiple Sclerosis Journal 0(0) 1–4 © The Author(s) 2012 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1352458512454347 msj.sagepub.com

Elizabeth S Gromisch1, Vance Zemon1, Ralph HB Benedict2, Nancy D Chiaravalloti3,5, John DeLuca3,5, Mary A Picone4, Sonya Kim4 and Frederick W Foley1,4

Abstract Multiple sclerosis (MS) is frequently accompanied by changes in verbal memory. W e hypothesized that administering an abbreviated California Verbal Learning Test-II (CVLT-II) would detect verbal memory problems in MS accurately, thus serving as a potential screening tool.We performed receiver-operating-characteristic (ROC) analyses of three trials (trial 1, trial 2, and trial 1+2 combined) for raw data against standardized total scores. The results showed that at 1.5 standard deviations (SD) from the mean, the first two trials were 96.3% accurate, while at 2 SD from the mean, the first two trials combined were 97.5% accurate. We conclude that this study demonstrates than an abbreviated CVLT-II is a valid screening tool for verbal memory impairments. Keywords Multiple sclerosis, cognition, verbal learning, memory, screening Date received: 5th April 2012; revised: 6th June 2012; accepted: 13th June 2012

Introduction Cognitive dysfunction is common in multiple sclerosis (MS)1,2 and can include impairments in executive functioning, visuospatial processing, information processing speed, and memory. Even with mild cognitive impairment, patients can have significant functional disabilities.3 Due to the impact of cognitive dysfunction, screening for cognitive impairment is important. Although valid assessments of cognitive function in MS patients are well established,4 screening of patients for potential cognitive problems is more problematic. Some have used the MiniMental State Examination (MMSE), a standard for dementia screening, to screen for cognitive function in MS. While the test is brief, it is not an accurate test in the MS population: the sensitivity ranges from 28% to 36%.5 One widely utilized battery for assessing cognitive functioning in the MS population is the Minimal Assessment of Cognitive Function in MS (MACFIMS).4 While the administration of the MACFIMS, or more comprehensive neuropsychological batteries, is feasible in settings with access to neuropsychologists it does not help the neurologist decide which patients to refer for testing, and

neuropsychological batteries may be too long for routine monitoring in many settings. The verbal memory test within the MACFIMS is the California Verbal Learning Test-II (CVLT-II) which is a reliable and valid measure of verbal learning and memory in MS.6,7 Research has indicated that acquisition difficulties are the primary source of memory impairment in the MS population, rather than retrieval from long term storage.8 Thus, we hypothesize that an abbreviated version of the CVLT-II may serve as a screen for verbal memory 1Ferkauf

Graduate School of Psychology,Yeshiva University, USA. University of New York at Buffalo, NY, USA. 3Kessler Foundation Research Center, USA. 4Holy Name Medical Center, Multiple Sclerosis Center, USA. 5University of Medicine and Dentistry of New Jersey, New Jersey Medical School, USA. 2State

Corresponding author: Frederick W Foley, Yeshiva University, Ferkauf Graduate School of Psychology, 1300 Morris Park Avenue, Bronx, NY 10461 USA. Email: [email protected]


2

Multiple Sclerosis Journal 0(0)

Figure 1. Receiver-operating-characteristic (ROC) analyses at (a) 1.5 standard deviations (SD) and (b) 2 SD from the mean. CVLT-II: California Verbal Learning Test-II.

acquisition problems in MS. The study objective was to examine whether fewer administrations of the CVLT-II still produced sensitive and specific findings of verbal memory deficits, making it feasible to utilize as a screen for memory impairments in the clinic setting.

Patients and methods Participants (n=572) from three different MS clinics in the US were included. The majority were women (n=419). The mean age was 46.65±9.35 years. The mean length of education was 14.57±2.48 years. The mean MS duration was 10.82±8.22 years. Receiver-operating-characteristic (ROC) curves were used to analyze the accuracy of an abbreviated CVLT-II, and these plot true positive rate versus false positive rate (sensitivity versus 1-specificity). ROC analyses were run at both 1.5 and 2.0 standard deviations (SD) from the mean, as both are used as determinants of impairment in neuropsychological evaluations. For each cut-off for impairment, three analyses were run: trial 1, trial 2, and the first two trials combined. The raw data from the CVLT-II trials 1 and 2 were run against the total learning (trials 1–5). T-scores were generated from administration of five trials of the

CVLT-II. The raw data was chosen as these scores would be available to clinicians immediately after administration.

Results At 1.5 SD from the mean, administration of the first trial was 88.4% accurate (Figure 1). Using four correct answers as a cut-off point for impairment during the first trial captures 69.7% of the impaired population and overestimates by 12.6% (Table 1). The administration of the second trial of the CVLT-II was 95.0% accurate. Using seven correct answers as a cut-off point for impairment during the second trial captures 97.4% of the impaired population and overestimates by 18.9%. Giving both the first and second trial of the CVLT-II at 1.5 SD from the mean is 96.3% accurate, and a cut-off of 12 correct answers detects 98.7% of patients with verbal memory acquisition problems and overestimates by 17.9%. At 2 SD from the mean, administration of just the first trial was 89.4% accurate. Using four correct answers as a cut-off point for impairment during the first trial captures 77.5% of the impaired population and overestimates by 15.9%. The administration of the second trial of the CVLT-II was 97.1% accurate. Using six correct answers as a cut-off


3

Gromisch et al. Table 1. Coordinates of the curve of the receiver-operating-characteristic (ROC) analyses at 1.5 and 2 standard deviations (SD) from the mean. Test result variable(s)

Positive if less than or equal to

Sensitivity (1.5 SD)

Sensitivity (2 SD)

1 – Specificity (1.5 SD)

1 – Specificity (2 SD)

CVLT Trial 1 raw CVLT Trial 2 raw CVLT Trial 1+Trial 2 raw

4.5000 5.5000 6.5000 5.5000 6.5000 7.5000 8.5000 9.5000 6.5000 7.5000 8.5000 9.5000 10.5000 11.5000 12.5000 13.5000 14.5000

0.697 0.947 1.000 0.539 0.803 0.974 0.987 1.000 0.092 0.158 0.342 0.566 0.697 0.882 0.987 0.987 1.000

0.775 0.975 1.000 0.775 0.975 1.000 1.000 1.000 0.175 0.250 0.550 0.775 0.900 1.000 1.000 1.000 1.000

0.126 0.335 0.591 0.018 0.089 0.189 0.374 0.577 0.000 0.000 0.004 0.014 0.039 0.093 0.179 0.285 0.423

0.159 0.376 0.620 0.036 0.125 0.241 0.416 0.605 0.000 0.004 0.011 0.036 0.068 0.139 0.233 0.332 0.463

CVLT: California Verbal Learning Test.

point for impairment during the second trial captures 97.5% of the impaired population and overestimates by 12.5%. Giving both the first and second trial of the CVLT-II at 2 SD from the mean is 97.5% accurate, and a cut-off of 11 correct answers detects 100% of patients with verbal memory acquisition problems and overestimates by 13.9% .

Discussion At both 1.5 SD and 2 SD from the mean, administration of up to two trials of the CVLT-II detected verbal memory problems accurately. This abbreviated version cuts the administration time from 20 minutes to approximately four minutes, making it feasible for the abbreviated version to be incorporated into a clinician’s mental status exam. For both cut-offs (1.5 and 2.0 SD), the first two trials yielded better sensitivity and specificity than using the first trial alone. The decision of which cut-off to use will depend on the purpose of the exam. If detecting modest memory changes with maximum sensitivity is desired, utilizing a cut-off of 12 correct responses for Trials 1+2 will detect almost all of the true cases at 1.5 SD (98.7%), but yield 17.9% false positives. A cut-off of 11 correct responses will drop the false positive rate to 9.3%, and detect 88.2% of those with mild (1.5 SD) impairment. Utilizing the latter cut-off might be more feasible in a busy clinical setting, to minimize referring excessive numbers of patients for further evaluation, who will be less likely to have a memory impairment. If the purpose of the exam is to detect or monitor patients for the presence of more than mild impairments in verbal memory, utilizing a cut-off of 10 for Trials 1+2

will detect 90.0% of those patients, yielding a false positive rate of 6.8% (2 SD). Some limitations of the current project include the fact that the abbreviated version of the CVLT-II does not test some of the domains that the full CVLT-II does, such as delayed recall. Thus, it is extremely important that clinicians distinguish between brief tests used for screening purposes versus more comprehensive neuropsychological tests that are used for characterizing the full array of a person’s cognitive strengths and deficits. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest The authors declare no conflicts of interest in preparing this article.

References 1. Peyser JM, Rao SM, LaRocca NG et al. Guidelines for neuropsychological research in multiple sclerosis. Arch Neurol 1990; 47: 94–97. 2. Rao SM, Leo GJ, Bernadin L et al. Cognitive dysfunction in multiple sclerosis. I. Frequency, patterns and predictions. Neurology 1991; 41: 685–691. 3. Rogers J and Panegyres PK. Cognitive impairment in multiple sclerosis: evidence-based analysis and recommendations. J Clin Neurosci 2007; 14: 919–927. 4. Benedict RH, Fischer JS, Archibald CJ et al. Minimal neuropsychological assessment of MS patients: a consensus approach. Clin Neuropsychol 2002; 16: 381–397.


4

Multiple Sclerosis Journal 0(0)

5. Amato MP, Zipolo V and Portaccio E. Cognitive changes in multiple sclerosis. Expert Rev Neurother 2008; 8: 1585–1596. 6. Benedict RHB. Effects of same- versus alternate-form memory tests during short -interval repeated assessments in multiple sclerosis. J Int Neuropsychol Soc 2005; 11: 727–736.

7. Stegen S, Stepanov I, Cookfair D et al. Validity of the California Verbal Learning Test-II in multiple sclerosis. Clin Neuropsychol 2010; 24: 189–202. 8. Chiaravalloti ND and DeLuca J. Cognitive impairment in multiple sclerosis. Lancet Neurol 2008; 7: 1139–1151.
