Clinical Utility of Wechsler Memory Scale-Revised and Predicted IQ Discrepancies in Closed Head Injury

Transcription

1 @ Pergamon Archives of Clinical Neuropsychology, Vol. 12, No. 8, pp , 1997 Copyright 1997 Nationaf Academy ofneuropsychology Printed inthe USA, All rights reserved /97$ PIIS (97)OO049-8 Clinical Utility of Wechsler Memory Scale-Revised and Predicted IQ Discrepancies in Closed Head Injury Geoffrey Tremont, Richard G. Hoffman, James G. Scott, Russell L. Adams and Marnie J. Nadolne University of Oklahoma Health Sciences Center Comparisons between WechslerMemory Scale-Revised (WMS-R) indexes and WechslerAdult Intelligence Scale-Revised (WAIS-R) IQ scores have beenproposed to identijj severity of memory deficits. Howevec marrynecrologic conditions reduce both intellectual and memory functioning, and thus, examining differences between these scores may be of little value. Closed head injured subjects who completed the WMS-R were divided into either mild injury (MI, n = 41) or moderatelsevere (S1,n = 41) injury groups based on trauma severiry indicators and were matched on age and level of education. The Oklahoma Premorbid Intelligence Estimation (OPIE), a regressionformula that takes into account demographic variables as well as IQ pe~omrarrce, was calculated for each subject. Discrepancy scores were calculated between predicted IQ scores and WAIS-R IQ and WMS-R indexes. SI head-injured subjects displayed significantly larger discrepancies (19 points) between OPIE scores and Delayed Recall Indexes from the WMS-R than the MI subjects (10 points). Significantly larger percentages of subjects in the S1 group displayed significant (> SD) reductions in many of the WMS-R and WAIS-R scoresfrom estimates than subjects in the MI group. Comparing current memory functioning to estimates of premorbid intellectual ability appears to be a sensitive indicator of presence and degree of intellectual and memory dysfunction in head trauma patients. Results also provide evidence that estimates of premorbid intellectual ability can serve as estimates of premorbid memory functioning. O 1997 National Academy of Neuropsyclrology.Published by Elsevier Science Ltd Comparisons between index scores of the Wechsler Memory Scale-Revised (WMS-R) have been made by several authors to distinguish among different clinical groups, differentiate brain-impaired from normals, and evaluate the potential of malingering (Butters et al., 1988; Crossen & Weins, 1988; Mittenberg, Azrin, Millsaps, & Heilbronner, 1993; Wechsler, 1987). However, no study has evaluated how effective these comparisons are in differentiating between brain injury severity levels. Address correspondence to: Geoffrey Tremont, Rhode Island Hospital/Brown University School of Medicine, Department of Psychiatry, Physicians OffIce Building, Suite 430, 593 Eddy Street, Providence, RI gtremont@rihosp.edu. 757

2 758 G. Tremont et al. Comparisons have also been proposed between WMS-R indexes and WAIS-R IQ scores to identify severity of memory deficits and distinguish different necrologic disorders. The rationale for such comparisons is that IQ scores tend to hold better than memory scores in brain-impaired patients (Bomstein, Chelune, & Prifitera, 1989).Bomstein et al. (1989) found that the discrepancy between Full Scale IQ and the Delayed Memory Index significantly differentiated normal subjects from the WMS-R standardization sample and a mixed necrologic group, consisting of mostly seizure patients. One possible problem with comparing memory and obtained IQ is that many necrologic conditions reduce intellectual functioning as well as memory functioning, and thus examining differences between these scores may be of little diagnostic or descriptive value. Additionally, clinicians frequently want to determine patients current level of memory impairment relative to their premorbid functioning. The purpose of the present studyis to evaluate the utility of comparisons between WMS-R index scores and premorbid IQ, using a premorbid index based on both demographic factors and current test performance across two severity levels in patients with closed head injury. A second goal of the study is to identify whether WMS-R index score comparisons are helpful in determining head trauma severity. Subjects and Procedure METHOD Eighty-two (82) head-injured individuals referred as outpatients to a University Neuropsychology Laboratory were divided into two severity groups and equated for age and education. The Mild Head Injury (MI) group consisted of 41 patients whose sustained head trauma resulting in Glasgow Coma Scale (GCS) scores ranging from 13 to 15, loss of consciousness 20 minutes or less, and whose duration of posttraumatic amnesia (PTA) was less than 24 hours. Their average age was years (SD = 10.47), level of education was years (SD = 2,58), and they were examined an average of months postinjury (SD = 22.04). The Moderate/Severe Head Injury (S1) group consisted of 41 patients who sustained head trauma resulting in GCS scores of 12or less, greater than 20 minutes of unconsciousness, and PTA longer than 24 hours. Positive findings on neurodiagnostic procedures were also considered to reflect moderate to severe injury. Their average age was years (SD = 9.61), level of education was years (SD = 1.98), and they were examined an average of months (SD = 55.61) postinjury. Sixty-three percent of MI and 4270of S1subjects were involved in litigation. Subjects completed the WAIS-R and the WMS-R according to standard procedures as part of a comprehensive neuropsychological test battery. Predicted premorbid intelligence was estimated for each subject using the Oklahoma Premorbid Intelligence Estimation, which takes into account demographic variables of age, education, occupation, and race, in addition to performance on the Vocabularyand Picture Completion subtests from the WAIS-R (OPIE, FSIQ Best Formula; Krull, Scott, & Sherer, 1995).The Best FSIQ estimate is calculated with separate regression equations based on a decision rule (i.e., Picture Completion only if non age-corrected scaled score is higher than Vocabulary or use a Vocabulary only if Vocabulary performance is better or the scores are equal). The OPIE formula was developed in response to the criticism that demographic-only estimates have a restricted range of scores that can underestimate ability in individuals with high premorbid functioning and overestimate ability in individuals with low premorbid functioning. It has been validated with 940 subjects from the WAIS-R standardization sample and cross-validated on another set of 940 normal subjects, yielding correlations between predicted and obtained FSIQ, VIQ, and PIQ of.87,.87, and.78, respectively, and producing a wider distribution of scores than

3 Predicted IQ Discrepancies 759 TABLE 1 MeanObtained IQ, Estimated IQ, and Memory Scores for Mild and Moderate Severe Head-Injured Subjects Mild Head-Injured Moderate/Severe Head-Injured (N= 41) (N= 41) M SD M SD Demographic-Only IQ OPIE IQb Full Scale IQ Performance IQ Verbal IQ Verbal memory Visual memory* General memory Attention/Concentration Delayed recall* a Barona, A., Reynolds, C. R., & Chastain, R. (1984). A demographically based index of premorbid intelligence for the WAIS-R. Journal of Consulting and Clinical Psychology, 52, b Vocabulary only FSIQ Estimate = (race) +.85(education).66(occupation +.76(Vocabulary raw score), Picture Completion only FSIQ Estimate = (race) (education).71(occupation (Picture Completion raw score). *p <.05. demographic-only estimations (Krull et al., 1995).The estimate has recently been shown to appropriately estimate theoretical premorbid IQ in a variety of clinical samples, including closed head injury patients (Scott, Krull, Williamson, Adams, & Iverson, 1997). Difference scores were calculated between OPIE scores, Full Scale WAIS-R IQ, and WMS-R indexes. Positive difference scores reflect discrepancies between predicted premorbid IQ and obtained performance on WAIS-R and WMS-R indexes. Within-scale difference scores were also calculated for global index scales. RESULTS WAIS-R, OPIE, and WMS-R scores for the two head-injured groups are presented in Table 1. Mean premorbid IQ estimates based on demographic variables alone are also presented to compare estimates of intellectual functioning (Barona, Reynolds, & Chastain, 1984). Although demographic-only estimates fell slightly below OPIE scores, the difference was not statistical significance and, as seen in previous research, the OPIE yielded a slightly wider distribution of scores (OPIE = vs. Barona = ). No significant differences between litigating and nonlitigating head-injured subjects were found for any of the WMS-R scores or memory-iq comparisons within each severity group, suggesting that litigation status did not impact performance. The moderate/severe group obtained significantly lower Visual memory and Delayed Recall scores than the mildly injured subjects, 497) = 2.31, p <.05 and t(97) = 2.05, p <.05, respectively. No other indexes or IQ scores significantly differentiated the groups. Mean values of the discrepancy scores are presented in Table 2. The S1 group displayed a slightly larger difference between estimated and obtained IQ than the MI group. Both groups showed average reductions of over 8 points, which is consistent with past research with the OPIE (Scott et al., 1996). The only significant between-group memory indexpredicted IQ discrepancy was OPIE-Delayed Recall index, t(97) = 1.96, p <.05, with S1 patients showing over a 17-point difference between DRI and estimated IQ.

4 760 G. Tremont et al. TABLE 2 Means and Standard Deviations of Inter- and Intra-Test Comparisons Mild Head-Injured Moderate/Severe Head-Injured (N= 41) (N= 41) M SD M SD OPIE-FSIQ OPIE-GMI OPIE-DRI* OPIE-ACI ACI-GMI* GMI-DRI ACI-DRI** OPIE = Oklahoma Premorbid Intelligence Estimation, FSIQ = Full Scale IQ, GMI = General Memory Index, DRI = Delayed Recall Index, ACI = Attention Concentration Index. *p <.05. **p <.01. Comparisons among WMS-R indexes revealed that the MI group obtained significantly lower discrepancy scores for the ACI-DRI comparison, indicating that they obtained lower attention scores than the S1subjects. Similarly, MI subjects had lower ACI compared to GMI than S1 subjects. To determine the clinical value of each comparison, the percentage of subjects in each group who obtained difference scores of 15 points or more were calculated and are presented in Table 3. A 15-point difference was selected because of the finding that only 109ZOof normal subjects in the WMS-R standardization sample obtained 15-point discrepancies between FSIQ and DRI (Bornstein et al., 1989). Significantly more S1subjects had 15-point deficits on the OPIE-FSIQ, OPIE-DIU, and ACI-DRI comparisons (X2= 5.89, ~2 = 3.96, Xz= 11.58, respectively). Significant percentages of S1subjects obtained IQ and DRI scores that fell iit least 1 SD below their estimated IQs. Similarly, over one third of S1subjects obtained DRI scores that fell at least 1 SD below their ACI scores. Although the frequency of the differences of other comparisons did not TABLE 3 Percentage of Subjects in Each Group Obtaining 15-point or Greater Deficits on Obtained IQ and Memory Versus Estimated IQ Scores and Compared WMS-R Indexes Mild Head-Injured Moderate/Severe Head-Injured (N= 41) (N= 41) OPIE-FSIQ* OPIE-GMI OPIE-DRI* OPIE-ACI ACI-GMI 7 19 GMI-DRI ACI-DRI* 5 34 OPIE = Oklahoma Premorbid Intelligence Estimation, FSIQ = Full Scale IQ, ACI = Attention/Concentration Index, GMI = General Memory Index, DRI = Delayed Recall Index. *p <.05 (chi-square).

5 Predicted IQ Discrepancies 761 significantly differ between head-injured groups, clinically significant percentages of subjects showed 15-point discrepancies, especially for the WMS-R-OPIE comparisons. DISCUSSION The present study demonstrates that premorbid IQ estimation can serve as a valuable comparison index to assess declines inmemory functioning amonghead-injured patients, and provides initial empirical support for using premorbid intellectual estimates as estimates of premorbid memory functioning. Although obtained WAIS-R and WMS-R scores showed few significant differences between mild and moderate to severely injured patients, within-test and estimated IQ-WMS-R index score comparisons did differentiate the groups, with meaningful numbers of individuals displaying these differences. The most sensitive comparison in the present study was between the OPIE and the DRI, consistent with the notion that delayed memory functioning is the most sensitive WMS-R measure to central nervous system dysfunction (Crossen & Wiens, 1988). Over one half of moderate/severe head-injured patients demonstrated a 15-point orgreater decline when DRI was subtracted from the OPIE. Comparisons between obtained and estimated IQ were less sensitive than the memory and estimated IQ comparisons, with an average loss of intellectual function of 8 and 11points for the mild and moderate/severe head-injured patients, respectively. Slightly larger declines were noted between general memory and predicted premorbid IQ and the difference between ACI and predicted IQ was 15 points or greater in over one third of the total patient sample. Within scale comparisons indicate that more mildly head-injured subjects displayed lower ACIS compared to GMIs than the moderate/severe grc~up.the pattern of lower ACIS compared to GMIs has been observed in malingering research in which normal subjects are instructed to fake performance on the WMS-R (Mittenberg et al., 1993). In the present study, however, litigation had no effect on any of the index score comparisons. It is important to note that these mild head-injured patients were seen over 1 1/2 years postinjury and may represent a rather specific subgroup of patients with continued complaints. Comparing the DRI to the ACI may also be of value clinically, both in evaluating chronicity and motivational factors. On comparison of DRI and ACI, only 5970of mildly injured subjects showed a 15 point or greater discrepancy, whereas over one third of moderate/severe patients demonstrated lower DRI scores. Although the present study identified statistically significant differences of index score comparisons, caution shouldl be used when making clinical determinations on an individual s index score differences because of error variance (Mittenberg, Thompson, & Schwartz, 1991). The present study demonstrates the usefulness of the comparing premorbid IQ estimates to WMS-R index scores as well as current intellectual performance in evaluating both presence and degree of intellectual and memory dysfunction. It should be determined whether other predictors of premorbid ability can also be compared to memory functioning and provide diagnostic or discriminative ability. Additionally, comparisons between IQ or estimated IQ and other memory tests may be helpful in identifying material-specific deficits and literalized processes. Although the present study demonstrates the potential utility of comparisons between predicted premorbid IQ estimates and current memory performance, attention must be paid to factors affecting predicted premorbid IQs (i.e., extremes in age or education continuum), as well as factors affecting Wh4S-R performance, such as age, education, and attentional capacities.

6 762 G. Tremont et al. REFERENCES Bomstein, R. A., Chelune, G. J., & Prifitera,A. (1989). IQ-memorydiscrepanciesin normal clinical samples. PsychologicalAssessrrrent,1, Butters, N., Sahnon, D. E?,Cullum, C. M., Cairns, P., Troster,A. L, Jacobs, D., Moss, M., & Cermrrk,L. S. (1988). Differentiationof anmesticand dementedpatients with the WechslerMemory Scale-Revised.The ClinicalNeurupsychologist,2, Crossen, J. R., & Wiens, A. N. (1988). Wechsler Memory Scale-Revised: Deficits in performance associated with neurotoxicsolventexposure. The ClinicalNeumpsychologist,2, Kmll, K., Scott, J., & Sherer, M. (1995). Estimation of premorbid intelligence from combined performance and demographicvariables. i?re ClinicalNeuropsychologist,9, Mittenherg,W.,Azrin, R., Millsaps,C., & Heilbronner,R. (1993).Identificationof malingeringon the WechslerMemory Scale-Revised.PsychologicalAssessment, 5, Mittenberg,W,,Thompson,G. B., & Schwartz,J.A. (1991).Abnormaland reliabledifferencesamongwechslermemory Scale-Revisedsubtests.PsychologicalAssessment, 3, Scott, J., Kmll, K., Williamson, D., Adams, R., & Iverson, G. (1997). Oklahoma Premorbid Intelligence Estimation (OPIE): Utilization in clinical samples.the ClinicalNeuropsyclwfogirt,11, Wechsler,D. (1987).Manualfor the WechslerMemory Scale-Revised.San Antonio,TX: PsychologicalCorporation.