Laterality Differences in Quantitative and Qualitative Hooper Performance

Transcription

1 Pergamon Archives of Clinical Neuropsyehology, Vol. 1 I, No. 3, pp , 1996 Copyright 1996 National Academy of Neuropsychology Printed in the USA. All rights reserved /96 $ SSDI (95) Laterality Differences in Quantitative and Qualitative Hooper Performance lodi D. Nadler Florida Hospital, Medical Psychology Section Janet Grace Desiree A. While University of California San Diego Meryl A. Butters Western Psychiatric Hospital and Clinic Paul F. Malloy The Hooper Visual Organization Test (HVOT) is a measure of visuospatial processing commonly employed in neuropsychological assessment. Despite the well-documented relationship between visuospatial abilities and right hemisphere function, the literature has not supported a right hemisphere association with HVOT performance. The current study was conducted to examine laterality differences in HVOT performance. Sixty-seven geriatric stroke patients (44 right CVAs, 23 left CVAs) were administered the HVOT and the Mini-Mental State Exam (MMS). Results revealed significant differences between CVA groups for total score, with right CVA patients performing more poorly. Qualitative error analyses revealed highest frequencies for part responses and don't know/no response errors. Between-group differences were seen for part and unformed~unassociated errors (higher right CVA rates), and language-based errors (higher left CVA rates). Findings are consistent with theories of brain lateralization and suggest that whereas A portion of this article was presented at the 13th annual meeting of the National Academy of Neuropsychology, Phoenix, AZ, Address correspondence to: Jodi Nadler, PhD, Medical Psychology, Florida Hospital, 5165 Adanson St., Orlando, FL

2 224 J. D. Nadler et al. HVOT performance predominantly involves right hemisphere functions, left hemisphere dysfunction may also lead to impaired performance, and the two can be discriminated by qualitative analysis of the errors. The Hooper Visual Organization Test (HVOT; Hooper, 1958, 1983) is a measure of visuospatial processing that is commonly employed in neuropsychological assessment. The test consists of 30 line drawings of common objects that have been dissassembled and rearranged on a page in a puzzlelike fashion (see Figure 1). The test requires the visual integration of the objects. Specifically, a patient must determine the identity of each object if the pieces were properly integrated. Research examining the clinical utility of the HVOT has demonstrated adequate discrimination of brain-damaged and neurologically normal individuals (Boyd, 1981; Hooper, 1983; Wang, 1977). However, these studies have not shown differences between fight and left hemisphere lesion patients. A recent study (Fitz, Conrad, Horn, & Sarff, 1992), which also failed to find differences in HVOT performance between fight and left hemisphere lesion patients, did demonstrate significantly poorer performance in patients with lesions that included the fight parietal lobe, compared to all other lesion sites. The general lack of literature support for laterality differences in HVOT performance is surprising given the well-documented association between visuospatial abilities and fight hemisphere function (Benton & Tranel, 1993; Kolb & Whishaw, 1990; Lezak, 1983). However, this outcome may be due to the failures of researchers to consider the multifactorial nature of the task. For example, HVOT performance involves language abilities (i.e., naming of a common object), as well as visual perception and synthesis. Therefore, a patient who is able to successfully visually integrate an object may still respond incorrectly because of an inability to provide the correct name of the object. In addition, a patient with self-regulation difficulties may perseverate on a previous item name for subsequent items, thereby providing an incorrect response. Hence, patients with adequate visuospatial integration may perform poorly on the HVOT for a variety of reasons. This suggests a need for qualitative analysis of HVOT performance, a notion that was discussed by Hooper (1983) but has not been adequately investigated to date. The current study was conducted to investigate laterality differences in HVOT performance from both a quantitative and qualitative perspective. Comparisons were made between FIGURE 1. This HVOT figure, properly integrated, depicts a "mouse" or "guinea pig."

3 Laterality Differences 225 right and left cerebrovascular accident (CVA) patients on overall HVOT performance. Further analyses examined and compared types of errors committed within each lesion group. It was hypothesized that right and left CVA patients would perform similarly on overall HVOT performance, but would commit different rates of qualitative errors. Specifically, it was believed that right CVA patients would produce more part response errors, whereas left eva patients would produce more language-based errors. METHOD Subjects Sixty-seven acute geriatric stroke patients consecutively referred to a medical rehabilitation unit participated in the study. Subjects had a mean age of years (SD = 8.31), and a mean education level of years (SD = 2.67). The majority of subjects were female (59%) and right handed (95%). Of the 67 subjects, 44 sustained right hemisphere CVA and 23 sustained left hemisphere CVA, as determined by two independent raters of CT and/or MRI neuroimaging. All subjects had discrete lesions and were 1 to 2 weeks poststroke. Subjects had a mean Functional Independence Measure (FIM) score of (15.92). Patients presenting with features suggestive of an acute confusional state were excluded, as were patients with significant substance abuse histories, psychiatric histories, or current affective disturbances. Also excluded were patients with sensory, motor, or linguistic impairments of sufficient degree to preclude valid psychometric assessment. Table 1 contains characteristics for the lesion groups. Procedure Subjects were administered the HVOT and the Mini-Mental State Exam (MMS; Folstein, Folstein, & McHugh, 1975) by an experienced neuropsychologist or predoctoral neuropsychology intern. HVOT administration involved presenting 30 pictures of disassembled common objects to a patient one at a time. Following each presentation, the patient attempted to identify and name the object represented, were the parts properly integrated. Data Analysis HVOT total scores were computed for each patient based on the standard quantitative scoring procedure (Hooper, 1983). In addition, a revised qualitative scoring system based on the procedure described by Hooper (1983) was developed for evaluating error types. Error types were defined as follows: 1. Part errors: (a) naming one part of the correct object, for example, "finger" for hand; (b) naming one part of the stimulus card, for example, "pipe" for mouse; (c) using only a part of the stimulus card to determine the whole response, for example, "it has TABLE 1 Means and Standard Deviations for Patient Characteristics by Lesion Groups Lesion Site N Age Educ MMS HVOT Right (8.24) (2.33) (2.15) (6.35) Left (8.60) (3.27) (5.53) (4.21) Total (8.31) (2.67) (3.74) (6.47)

4 226 J. D. Nadler et al. eyes so it must be an animal;" or (d) providing a response lacking integration of the parts, for example, "a broken saw." 2. Perseverative errors: (a) repeating a previous correct or incorrect response on a later item; (b) providing other responses within a category that are unrelated to the current stimulus item but are related to a previous item, for example, "pliers," "wrench," "drill," for subsequent items after successful identification of the hammer item. 3. Language-based errors: (a) providing an incorrect semantically related name for an item, for example, "clippers" for scissors; (b) providing an incorrect phonemically related name for an item, for example, "flow" for flower; (c) providing a circumlocutory response for an item, for example, "something to read with pages" for book; (d) providing a neologistic response, for example, "clof" for mouse; (e) providing an agrammatic response, for example, "put a truck apart" for truck. 4. Part/Language-based errors: (a) providing a response that meets criteria for both part and language-based responses, for example, "cutting blades" for scissors. 5. Unformed/unassociated errors: providing a response that appears unformed or unrelated to the stimulus item, for example, "shaving cream" for fish. 6. Don't know/no response errors: providing the response, "I don't know" or giving no response. Scoring of errors involved categorization of the last response given for each item. Ratings were made by two independent scorers (J.G. and D.W). In the event of a scoring discrepancy, a third independent rater (J.N.) was used to determine categorization of the response. Analyses of error types were conducted using Mann -- Whitney statistical tests. These nonparametric procedures were selected because the distributions of errors reflected significant violations of the assumptions for parametric statistics. RESULTS No significant differences were found between right and left CVA groups for age or education (p >.05). Significant differences were obtained on the MMS, with right CVA patients receiving higher mean MMS scores, t(55) = 2.618, p <.05. However, this difference is believed to reflect the heavy language emphasis of the MMS (Grace etal., 1995), rather than a significant overall cognitive difference between the groups. For both CVA groups, mean HVOT total scores were below the standard cut off for impaired performance. However, the mean age for the entire sample was above the age range for which the standards were derived (Hooper, 1983), so caution should be used in interpretation of level of impairment. A recent publication (Libon et al., 1994) did reveal higher mean HVOT performance in their sample of normal elderly subjects than that seen in the present study, suggesting that performance of these subjects was impaired. Further analysis of quantitative HVOT performance revealed significant differences between the CVA groups, t(65) = 4.421, p <.01, with fight CVA patients demonstrating the poorest performance. Analysis of qualitative performance was conducted using the scoring procedure developed by the authors. The procedure demonstrated adequate interscorer reliability (mean r =.93, range =.64 to.99). Examination of qualitative errors revealed the highest error rates for part responses, followed by don't know/no response errors. Other error types were committed less often, and varied in frequency between the CVA groups. Comparison of error types between groups revealed significant differences for part responses U = 686.5, p <.05, language-based responses U = 291.0, p <.01, and unformed/unassociated responses U = 690.0, p <.05. Part and unformed/unassociated errors were higher in fight CVA patients,

5 m Laterality Differences Right D Left 2 Me 1 0 Part* Lang Pala Unfo" Parsev Dknr Note: Part = Part responses Lang = Language-based responses, Pala = Part/language-based responses, Unfo = Unformed/unassociated responses, Per = Perseverative responses, Dk = Don't know/no responses. * indicates a statistically significant difference (1~<.05) between CVA groups FIGURE 2. ItVOT errors for right and left CVA patients. while language-based errors were greater in left CVA patients. Figure 2 presents the HVOT qualitative error rates for the CVA groups. DISCUSSION This study demonstrated differences in HVOT performance between right and left CVA groups. Differences were found in quantitative total score, with fight CVA patients performing more poorly. This finding is contradictory to previous studies of HVOT performance (Boyd, 1981; Fitz et al., 1992; Wang, 1977), which failed to show laterality differences. However, the finding of poorer performance in fight CVA patients is consistent with literature on laterality of visuospatial functioning. A possible reason for the disparity in fmdings is subject differences across studies. The present study employed geriatric patients with a restricted age range, whereas other studies have also included younger patients (Boyd, 1981; Fitz et al., 1992; Wang, 1977). In addition, the current sample was limited to CVA patients. Previous research has included heterogeneous brain-injured patients in whom more diffuse impairment may have been present (Boyd, 1981; Wang, 1977). The current findings also reflect a dissociation between overall cognitive status (as measured by the MMS) and HVOT performance. The right CVA group achieved significantly higher scores on the MMS than the left CVA group, though scored more poorly on the HVOT. In fact, the fight hemisphere group scored above the standard MMS cut off of 24 for general cognitive impairment. This finding is not surprising given the emphasis on language-based performance of the MMS, and strongly suggests that poorer performance of the fight CVA patients is not merely an artifact of general cognitive impairment.

6 228 J. D. Nadler et al. Analysis of qualitative errors revealed highest rates for part responses and don't know/no response errors for both groups. These responses reflect disordered visuospatial processing, synthesis, and integration, suggesting that HVOT performance is affected by functions subserved by the fight hemisphere (Benton & Tranel, 1993; Kolb & Whishaw, 1990; Lezak, 1983). This hypothesis is further supported by the significantly higher rates of part and unformed/unassociated responses seen in fight CVA patients in comparison to left CVA patients. Language-based errors, which are related to left hemisphere dysfunction (Benson, 1993; Kolb & Whishaw, 1990; Lezak, 1983), were, indeed, significantly greater in left CVA patients. The current findings are consistent with theories of lateralization of function in the brain and suggest that while HVOT performance largely involves visuospatial processing (i.e., fight hemisphere functions), other cognitive abilities, such as language skills, are also relevant to performance on this test. Thus, poor performance on the HVOT may be indicative of poor visuospatial processing, but it may also suggest other forms of cognitive impairment (e.g., anomia). Further study of HVOT performance in other populations is needed so that the current results can be generalized beyond a geriatric stroke population. The study also illustrates the clinical utility of qualitative analysis of neuropsychological test performance. Given the multifactorial nature of many neuropsychological tests, relying exclusively on quantitative evaluation of test performance (total score) with no consideration of qualitative responses (errors) may significantly limit the understanding of a patient's brain impairment and may result in a misinterpretation of their cognitive abilities (Kaplan, 1988). As suggested in the current study, a patient with an anomia could be misinterpreted as having a visuospatial processing disturbance due to poor performance on the HVOT. Similarly, a patient with a visuospatial processing disturbance could be misidentified as having an a_r.omia as a result of subnormal performance on a confrontation (visual) naming test. Hence, it is essential that clinicians attend to the qualitative aspects of test performance in interpreting neuropsychological test results. REFERENCES Benson, D. E (1993). Aphasia. In K. M. Heilman & E. Valenstein (Eds.), Clinical neuropsychology (pp ). New York: Oxford University Press. Benton, A., & Tranel, D. (1993). Visuoperceptual, visuospatial, and visuoconstructive disorders. In K. M. Heilman & E. Valenstein (Eds.), Clinical neuropsychology (pp ). New York: Oxford University Press. Boyd, J. L. (I 981 ). A validity study of the Hooper Visual Organization Test. Journal of Consulting and Clinical Psychology, 49, Boyd, J. L. (1982). Reply to Rathbun and Smith: Who made the Hooper blooper? Journal of Consulting and Clinical Psychology, 50, Fitz, A. G., Conrad, P. M., Hom, D. L., & Sarff, P. L. (1992). Hooper Visual Organization Test performance in lateralized brain injury. Archives of Clinical Neuropsychology, 7, Grace, J., Nadler, J. D., White, D. A., Guilmette, T. J., Giuliano, A. J., Monseh, A. U., & Snow, M. G. (1995). Folstein versus Modified Mini-Mental State Exam in geriatric stroke: Stability, validity, and screening utility. Archives of Neurology, 52, Folstein, M. E, Folstein, S. E., & McHugh, P. R. (1975). Mini-Mental State: A practical method for grading the cognitive status of patients for the clinician. Journal of Psychiatric Research, 12, Hooper, H. E. (1958). The Hooper Visual Organization Test: Manual. Beverly Hills: Western Psychological Services. Hooper, H. E. (1983). The Hooper ~rtsual Organization Test 1983 edition: Manual. Los Angeles: Western Psychological Services. Kaplan, E. (1988). A process approach to neuropsychological assessment. In T. Boll & B. K. Bryant (Eds.), Clinical neuropsychology and brain function: Research, measurement, and practice. Washington, DC: American Psychological Association. Kolb, B., & Whishaw, I. Q. (1990). Fundamentals of human neuropsychology (3rd ed.). New York: W. H. Freeman and Company.

7 Laterality Differences 229 Lezak, M. D. (1983). Neuropsychological assessment (2nd ed.). New York: Oxford University Press. Libon, D. J., Glosser, G., Malamut, B. L., Kaplan, E., Goldberg, E., Swenson, R., & Sands, L. P. 0994). Age, executive functions, and visuospatial functioning in healthy older adults. Neuropsychology, 8(1 ), Rathbun, L, & Smith, A. (1982). Comment on the validity of Boyd's validation study of the Hooper Visual Organization Test. Journal of Counseling and Clinical Psychology, 50, Wang, P. L. (1977). Visual organization ability in brain-damaged adults. Perceptual and Motor Skills, 45,