BRIEF REPORT: CLINICAL SCIENCE

Comparing Two Measures of Psychomotor Performance in Patients With HIV: The Coin Rotation Test and the Modied HIV Dementia Screen
Kyle S. Minor, MA,* Glenn N. Jones, PhD,* Diana W. Stewart, MA,* Benjamin D. Hill, PhD, and Magdalena Kulesza, MA*
is more prevalent than ever before,3 as survival time has increased after the development of highly active antiretroviral therapy (HAART).5 HAART is most effective when administered soon after HAD diagnosis, making early detection of HAD and the implementation of screening procedures to detect psychomotor slowing vitally important.5 HAD also affects executive resources, memory, attention, and concentration.6,7 These domains can be assessed using the HIV Dementia Scale (HDS)6 and Modied HIV Dementia Scale (MHDS),8 two brief clinical neuropsychological screening instruments for HAD. Psychomotor speed is measured on both screens by having patients write the alphabet (an over-learned behavior in most individuals) as quickly as possible. However, patients with low literacy or educational attainment may experience difculty writing or recalling the entire alphabet. Recall and production difculties unrelated to psychomotor speed may invalidate this subtest, yielding false positives. Supporting this, recent evidence suggests that the HDS may be inuenced by education level.9,10 Utilizing a measure of psychomotor speed unaffected by education and literacy may provide a more accurate assessment of cognitive functioning. The Coin Rotation Test (CRT),11 a measure of ne motor dexterity and psychomotor speed, is a possible alternative to the psychomotor performance subtest of the MHDS (MHDS-PS) that is unaffected by literacy or educational attainment. The CRT measures how many times individuals can rotate a US quarter in 10 seconds. It has high convergent validity with wellvalidated measures of psychomotor speed.12 Thus, the CRT seems to be a promising measure that is simple, fast, requires minimal equipment, and may be free of confounding educational factors. To date, CRT performance has been assessed in medical outpatient samples,11,12 but not specically in HIV+ patients. The present study examined the validity of the CRT for assessing psychomotor speed in HIV+ participants. It was hypothesized that (1) the CRT would have high convergent validity with MHDS-PS performance and overall MHDS score and (2) the MHDS-PS would be related to reading level and educational attainment, whereas the CRT would not.

Background: Psychomotor slowing is an early symptom of HIVassociated dementia. The Modied HIV Dementia Scale (MHDS) is a fast, reliable measure of HIV-associated dementia. However, literacy may inuence the psychomotor speed subscale (MHDS-PS). Psychomotor speed might be more accurately assessed using the Coin Rotation Test (CRT). Objective: This study examined the CRTs validity in assessing psychomotor speed in HIV+ patients by comparing it with the MHDS-PS. We also investigated whether demographic, education, or HIV-related variables were associated with either measure. Methods: The sample consisted of 204 HIV+ patients at a community health clinic. All participants completed the MHDS and CRT as part of admission to the clinic. Demographic, education, and disease data were also obtained. Results: Correlations revealed good convergent validity between the
CRT and MHDS-PS. Gender did not signicantly affect CRT performance but did affect MHDS-PS performance. CRT performance was less affected by education than MHDS-PS performance.

Conclusions: The CRT is a valid measure of psychomotor speed in HIV+ patients and may be more accurate than the MHDS-PS, particularly in individuals with lower levels of education.
Key Words: gender, HIV-associated dementia, psychomotor, HIV Dementia Scale, literacy (J Acquir Immune Dec Syndr 2010;55:225227)

sychomotor slowing is one of the earliest and most easily detectable symptoms of HIV-associated dementia (HAD) in those with HIV+ and AIDS.1,2 Research suggests that HAD is present in 3%15% of HIV+ cases.3,4 Those affected have increased mortality and reduced medication adherence.2 HAD

Received for publication September 3, 2009; accepted March 1, 2010. From the *Earl K. Long Medical Center, LSUHSC, Baton Rouge, LA; Department of Psychology, Louisiana State University, Baton Rouge, LA; and Department of Psychiatry and Human Behavior, Alpert Medical School of Brown University/Rhode Island Hospital, Providence, RI. Correspondence to: Glenn N. Jones, PhD, Adult Psychology, Earl K. Long Medical Center, 5825 Airline Highway, Baton Rouge, LA 70805 (e-mail: gjones@lsuhsc.edu). Copyright 2010 by Lippincott Williams & Wilkins

METHODS
Participants were 204 HIV+ patients at an urban HIV clinic serving primarily minority patients who are uninsured or underinsured. Most patients were not on HAART. The MHDS
www.jaids.com |

J Acquir Immune Dec Syndr  Volume 55, Number 2, October 1, 2010

225

Minor et al

J Acquir Immune Dec Syndr  Volume 55, Number 2, October 1, 2010

and the CRT are part of a clinical protocol used to screen for HAD. Demographic, clinical (ie, viral load, HIV stage, CD4 count, months since HIV diagnosis), and education data were also collected. Reading level was assessed in a subset of patients (n = 99) using the North American Adult Reading Test (NAART).13 The NAART consists of reading and pronouncing 61 irregularly spelled words and has demonstrated convergent validity with vocabulary tests.14 Patients completed 2 measuresthe MHDS and the CRT. The MHDS8 contains three subtests: (1) recall of 4 words after 5 minutes (MHDS-recall), (2) timed copy of a Necker cube (MHDS-construction), and (3) timed alphabet writing (MHDS-PS). Psychomotor speed was also assessed using the CRT.11 The CRT has participants rotate a US quarter (0.955 inch diameter coin) 180 as many times as possible in 10 seconds (dominant hand rst, followed by nondominant hand). Results for the total number of coin rotations are presented as dominant, nondominant, and total CRT scores had very similar patterns of relationships; data available upon request. Both the MHDS-PS and CRT have been validated against the Grooved Pegboard Test,15 the gold standard measure of psychomotor speed.16 Administration time is approximately 10 minutes for the MHDS and 1 minute for the CRT.

TABLE 1. Correlations of CRT and MHDS-PS With MHDS and MHDS Subscales, Potential Confounds and Disease Characteristics
Mean (6SD) MHDS MHDS: construction MHDS: recall MHDS: psychomotor speed CRTtotal Age (yrs) Months since HIV diagnosis Viral load log10 CD4 count Education (yrs) NAART
*P , 0.001. P , 0.01.

MHDS-PS 0.89* 0.21 0.36* 1.00 0.45* 20.24* 20.10 0.01 0.12 0.31* 0.35*

CRT 0.51* 0.30* 0.33* 0.45* 1.00 20.28* 20.05 20.12 0.12 0.22 0.29

7.9 0.84 2.89 4.2 20.3 38.0 58.6 4.12 281.4 11.6 13.4

(63.2) (60.93) (61.00) (62.3) (65.7) (611.0) (662.1) (61.27) (6229.7) (62.1) (610.6)

Statistical Methods
Correlations were used to examine the relations between the measures of psychomotor speed (CRT and MHDS-PS), the MHDS, MHDS subscale scores, and continuous demographic and clinical variables. T tests were used to examine whether gender inuenced the psychomotor test scores. As the sample was predominately African American, the effects of ethnicity were not examined. To investigate whether the CRT would yield a comparable overall MHDS score if it were substituted for the psychomotor subscale, we computed a scaled score for the CRT using regression techniques preserving the original MHDS scoring format. Alpha was set at P , 0.05. Descriptive statistics are presented as means, standard deviations (M 6 SD), and percentages.

CRT (P , 0.63), where females (20.44 6 5.35) and males (20.05 6 6.03) performed similarly. The additional point for females on the MHDS-PS translated to an additional point on the MHDS, which also had a gender effect (P , 0.01). The CRT can be substituted for the MHDS-PS by the formula: MHDS-PS (estimated) = 20.5 + 0.23*CRT, rounded to the nearest integer between 0 and 6. When the CRT was substituted, no gender effect was observed (P , 0.75) for the MHDS estimated total. A ROC curve was used to determine cutoff scores on the CRT predictive of a positive screen on the overall HDS. The highest combination of sensitivity (0.72) and specicity (0.61) was obtained using a CRT total cutoff score of 20 to predict a positive screen on the MHDS.

DISCUSSION
This is the rst study to validate the CRT in an HIV+ cohort. Previous studies examined the CRT in medical settings and found that the CRT demonstrates good convergent validity with other psychomotor measures.11,12 As psychomotor decline is one of the earliest signs of HAD,1 a brief, valid instrument to assess psychomotor performance in HIV+ patients is especially important. The present results suggest that the CRT meets these criteria, as it demonstrated high convergent validity with both the MHDS-PS subtest and overall MHDS. In addition, we calculated substitution scores for the CRT to correspond with the MHDS and determined that a cut score of ,20 rotations is predictive of a positive MHDS screen. One advantage of substituting the CRT for the MHDS-PS subtest is that it does not rely on assumptions about education, such as knowledge of the alphabet that may not be met in disadvantaged populations. In many settings, it is important to utilize a measure that is less dependent on a minimal level of literacy. Our results suggest that using the CRT either in place of or in addition to the MHDS-PS subtest is a way of minimizing education-related confounds. Although CRT performance was inuenced by education and reading level, these factors seem to have more inuence on the MHDS-PS. An unexpected gender
q 2010 Lippincott Williams & Wilkins

RESULTS
The sample was balanced with regard to gender (male = 52%, female = 48%) and primarily African American (90.7%). CDC Stages were: 46% A, 18.6% B, and 35.4% C. A disproportionate number of patients met criteria for AIDS (56.7%). As detailed in Table 1, level of education was generally low. The two psychomotor speed measures were signicantly correlated with each other (Table 1). CRT performance was signicantly correlated with the MHDS and its subscales. The MHDS-PS was also signicantly correlated with the MHDS score (inated by part-whole correlation) and MHDS subscales. The pattern of relations with the MHDS subscales was similar for the CRT and the MHDS-PS. Both psychomotor speed tests were similarly inversely related to age, and both were signicantly correlated with education and the NAART, although the MHDS-PS correlations were consistently larger in this respect. There were strong gender effects for the MHDS-PS (P , 0.001). Females (4.73 6 2.06) scored more than a point higher than males (3.65 6 2.34). This was not the case for the

226

| www.jaids.com

J Acquir Immune Dec Syndr  Volume 55, Number 2, October 1, 2010

The CRT and the MHDS

effect was also observed on the MHDS-PS (and MHDS as a result), as females performed signicantly better than males on this measure. In contrast, gender was not related to CRT performance, again suggesting the CRT is potentially a more accurate measure of psychomotor speed. When the CRT was substituted as the psychomotor scale on the MHDS, no signicant gender effect was observed. Our results suggest that the CRT is a useful screen for psychomotor slowing among HIV+ patients. However, further work exploring its usefulness in screening poorly educated and economically disadvantaged populations is needed. Future research should also address limitations in the current study. One limitation is the lack of a gold standard workup for HAD. The HDS has good sensitivity and specicity in accurately identifying HAD,6,9 but future work should determine whether substituting the CRT improves the accuracy of the MHDS in screening for HAD. A second limitation is that both the MHDSPS and CRT were inversely correlated with age. As HAD is more common in older patients,17,18 this nding may reect true differences across the lifespan, rather than confounds with either measure. More broadly, neither the CRT nor MHDS-PS is an appropriate measure of psychomotor speed when there are conditions that reduce manual dexterity (eg, arthritis). The results suggest that the CRT is a fast, reliable, costefcient, and valid measure of psychomotor speed in HIV+ patients. It is free from gender effects and less inuenced by education when compared with the MHDS-PS, suggesting that the CRT may be a more accurate method of identifying individuals with psychomotor speed impairments. The CRT is especially useful in clinics that have patients with low literacy and may help to reduce false positives. ACKNOWLEDGMENTS The authors wish to thank the patients, staff, and physicians of the Early Intervention Clinic, Baton Rouge, who made this project possible. We would also like to acknowledge the following clinicians who contributed to this project: Claire Adams, Arlene Gordon, Gina Najolia, Erin Ruether, and Julia Vigna.

REFERENCES
1. Cysique LA, Maruff P, Brew BJ. The neuropsychological prole of symptomatic AIDS and ADC patients in the pre-HAART era: a metaanalysis. J Int Neuropsychol Soc. 2006;12:368382. 2. Sacktor NC, Bacellar H, Hoover DR, et al. Psychomotor slowing in HIV infection. A predictor of dementia. J Neurovirol. 1996;2:404410. 3. McArthur JC, Haughey N, Gartner S, et al. Human Immunodeciency Virus-associated dementia: an evolving disease. J Neurovirol. 2003;9: 205221. 4. Sacktor N. The epidemiology of human immunodeciency virusassociated neurological disease in the era of highly active antiretroviral therapy. J Neurovirol. 2002;8:115121. 5. Dore GJ, McDonald A, Li Y, et al. Marked improvement in survival following AIDS dementia complex in the era of highly active antiretroviral therapy. AIDS. 2003;17:15391545. 6. Power C, Selnes OA, Grim JA, et al. HIV Dementia Scale: a rapid screening test. J AIDS. 1995;8:273278. 7. Antinori A, Arendt G, Becker JT, et al. Updated research nosology for HIV-associated neurocognitive disorders. Neurology. 2007;69: 17891799. 8. Davis HF, Skolasky RL, Selnes OA, et al. Assessing HIV-associated dementia: Modied HIV Dementia Scale versus the Grooved Pegboard. AIDS Read. 2002;12:2931, 38. 9. von Giesen HJ, Haslinger BA, Rohe S, et al. HIV Dementia Scale and psychomotor slowingthe best methods in screening for neuro-AIDS. J Neuropsychiatry Clin Neurosci. 2005;17:185191. 10. Berghuis JP, Uldall KK, Lalonde B. Validity of two scales in identifying HIV-associated dementia. J AIDS. 1999;21:134140. 11. Barkemeyer CA, Santa Maria MP, Browndyke JN, et al. The Coin Rotation Task: a convenient and sensitive measure of ne motor control. Arch Clin Neuropsych. 1998;13:18. Abstract. 12. Hill BD, Barkemeyer CA, Jones GN, et al. Validation of the Coin Rotation Test: a simple, inexpensive, and convenient screening tool for impaired psychomotor processing speed. The Neurologist. In Press. 13. Blair JR, Spreen O. Predicting premorbid IQ: a revision of the National Adult Reading Test. Clin Neuropsychol. 1989;3:129136. 14. Uttl B. North American Adult Reading Test: age norms, reliability, and validity. J Clin Exp Neuropsychol. 2002;8:11231137. 15. Klove H. Clinical neuropsychology. In: Forster FM, ed. The Medical Clinics of North America. New York, NY: WB Saunders; 1963:110125. 16. Strauss E, Sherman EMS, Spreen O. A Compendium of Neuropsychological Tests. New York, NY: Oxford University Press; 2006. 17. Wong MH, Robertson K, Nakasujja N, et al. Frequency and risk factors for HIV dementia in an HIV clinic in sub-Saharan Africa. Neurology. 2007;30:350355. 18. Grant I, Marcotte TD, Heaton RK. Neurocognitive complications of HIV disease. Psychol Sci. 1999;10:191195.

q 2010 Lippincott Williams & Wilkins

www.jaids.com |

227