Dr. Panopalis is recipient of a Canadian Institutes of Health Research fellowship.
Systemic Lupus Erythematosus
Impact of memory impairment on employment status in persons with systemic lupus erythematosus
Article first published online: 29 NOV 2007
Copyright © 2007 by the American College of Rheumatology
Arthritis Care & Research
Volume 57, Issue 8, pages 1453–1460, 15 December 2007
How to Cite
Panopalis, P., Julian, L., Yazdany, J., Gillis, J. Z., Trupin, L., Hersh, A., Criswell, L. A., Katz, P. and Yelin, E. (2007), Impact of memory impairment on employment status in persons with systemic lupus erythematosus. Arthritis & Rheumatism, 57: 1453–1460. doi: 10.1002/art.23090
- Issue published online: 29 NOV 2007
- Article first published online: 29 NOV 2007
- Manuscript Accepted: 3 MAY 2007
- Manuscript Received: 22 JAN 2007
- General Clinical Research Center, Moffitt Hospital, University of California, San Francisco
- National Center for Research Resources. Grant Number: 5-M01-RR-00079
- US Public Health Service
- Rosalind Russell Medical Research Center for Arthritis
- NIH. Grant Numbers: K24-AR-02175, R01-AR-44804
- State of California Lupus Fund
- Arthritis Foundation
- Agency for Healthcare Research and Quality
- National Institute of Arthritis and Musculoskeletal and Skin Diseases. Grant Numbers: 1-R01-HS-013893, P60-AR-053308
- Systemic lupus erythematosus;
- Memory impairment;
To assess the specific contribution of memory impairment to employment status in persons with systemic lupus erythematosus (SLE).
A total of 832 patients with SLE were surveyed and data collected on demographics, SLE symptoms and activity, health status, depression, medications, health resource utilization, and current employment status. Participants underwent screening for memory impairment and based on their scores were categorized to 3 levels of memory function: intact, mild-moderate impairment, and severe impairment. Employment status was compared across impairment levels using multivariate logistic regression, adjusting for sociodemographic characteristics (i.e., age, sex, race, education, and marital status), employment status at year of diagnosis, disease activity, disease duration, and depression.
In the intact memory function group, 54.2% were employed, versus 40.6% in the mild-moderate impairment group and 31.0% in the severe impairment group. In the intact memory function group, 29.2% were unable to work, versus 40.6% in the mild-moderate impairment group and 56.3% in the severe impairment group. After multivariate adjustment, increasing levels of memory impairment predicted a decreased likelihood of being employed: odds ratio (OR) 0.70, 95% confidence interval (95% CI) 0.48–1.02 for the mild-moderate impairment group and OR 0.57, 95% CI 0.32–1.00 for the severe impairment group. Participants with memory impairment were more likely to report being unable to work: OR 1.36, 95% CI 0.90–2.04 for the mild-moderate impairment group, and OR 1.99, 95% CI 1.12–3.55 for the severe impairment group. These findings were statistically significant only in the severe impairment groups.
The findings suggest that severe memory impairment is an important factor associated with employment status in persons with SLE.
Neuropsychiatric systemic lupus erythematosus (SLE) is an important and common manifestation of SLE and may present with major and often dramatic clinical events such as seizures, psychosis, or stroke (1). Other presentations of neuropsychiatric SLE, such as mild cognitive impairment or subtle behavioral changes, are less conspicuous and are often overlooked. Cognitive impairment in particular, appears to be under-recognized by both physicians and patients, even though it has been shown to occur commonly in patients with SLE, with prevalence estimates ranging from 26–81% (2–5). Although neuropsychological studies have not revealed a consistent profile of cognitive deficits, most have shown a pattern consistent with subcortical brain involvement, with deficits involving memory and recall, fluency, and attention (6, 7). Memory appears to be one of the most consistently affected domains of cognitive functioning in patients with SLE (4); however, the effects of memory impairment on important outcomes, such as employment and quality of life, remain understudied.
Employment is one of the most important aspects of psychosocial functioning. It not only provides earnings and social status, but also imparts a sense of self-worth and the opportunity to socialize with peers and to create social networks (8). Several studies have shown that persons with SLE experience a considerable amount of work disability. Partridge et al (9) reported that 3.4 years following diagnosis, 40% of patients who had been employed at some time since diagnosis quit work completely. Mau et al (10), using data from a German national database, estimated standardized employment ratios (SER) of observed versus expected cases to compare employment in persons with rheumatic diseases with the general population. In patients with SLE, the SERs for women with a disease duration of 6–10 years and for >10 years were 0.80 and 0.68, respectively. In a prior study, Yelin et al (11) using data from the University of California, San Francisco (UCSF) Lupus Outcomes Study (LOS) cohort (which is also used in the present study) demonstrated that employment among persons with SLE declined from 74% at the time of diagnosis to 48% at the time of the study (2004). Among those working at the time of diagnosis, the proportion employed declined by 15%, 36%, and 63% after 5, 10, and 20 years, respectively. The principal determinants of work loss included demographic factors such as increasing age, female sex, and low levels of education, and work characteristics, such as a physically and psychologically demanding job in combination with low levels of control. Only 1 study has attempted to demonstrate the specific influence of cognitive impairment. Utset et al (12) evaluated work status in patients with SLE with respect to disease characteristics, fatigue, anxiety, depressive symptoms, quality of life, and neuropsychological test performance. Subjects who reported formal work disability were considerably more likely to be neurocognitively impaired.
In the present study we used data derived from the UCSF LOS, a large and heterogeneous cohort of persons with SLE from the US, to examine the impact of memory impairment on employment status. Study participants were classified to 3 levels of memory function (intact, mild-moderate impairment, and severe impairment) and the employment status of patients in the 3 groups was then compared. Because employment status may be influenced by various sociodemographic and disease factors, multivariate analysis was used to determine the specific contribution of memory impairment to employment status after taking these factors into account.
SUBJECTS AND METHODS
The study cohort consisted of 832 subjects participating in the second annual interview of the UCSF LOS, a large cohort of persons with SLE from the US enrolled in an ongoing longitudinal survey. Details regarding eligibility and enrollment of participants have been described elsewhere (11). Briefly, subjects previously enrolled in a UCSF SLE genetics study were invited to reenroll in the LOS. All participants met at least 4 of the 11 American College of Rheumatology (ACR) revised criteria for the classification of SLE (13) after chart review by a rheumatologist or a registered nurse working under the supervision of a rheumatologist. The original 982 participants enrolled in the first wave of interviews were recruited from both clinical and community based sources: 22% from UCSF-affiliated clinics, 11% from non-UCSF rheumatology offices, and 67% from various community-based sources (e.g., lupus support groups, conferences, newsletters, websites). Of the 900 LOS participants initially enrolled prior to the end of 2003, 832 (92%) were available for the second annual interviews, which were conducted between January 2004 and April 2005. In order to focus our attention on patients of working age, the study was limited to subjects ages <65 years at the time of their second interview, the normal age of retirement in the US. The study protocol was approved by the UCSF Committee on Human Research.
Data were derived from structured, 1-hour telephone interviews conducted by trained interviewers. The survey included validated items pertaining to demographic and socioeconomic characteristics, SLE disease activity and manifestations, medications, general health, mental health, memory function, employment, health care utilization, and health insurance coverage.
The primary outcome of this study was current employment status. Participants were queried about their work situation for both the year of diagnosis and the current year (i.e., in the last week) by reporting on whether they 1) had a job and were at work, 2) had a job but were not at work, 3) were looking for work, 4) were keeping house, 5) were in school, 6) were unable to work, 7) were retired, or 8) other. If they reported that they did not have a job, they were asked if they had done any work for pay or profit, including unpaid work in a family business or a farm. Participants were considered employed if they reported that they had a job (whether or not they were at work in the last week) or they had done any work for pay or profit. In the US, this is the method used by the Bureau of Labor Statistics to derive labor force data from the Current Population Survey (14). Participants were also queried on the specific type of occupation in which they were employed at the year of diagnosis and the current year. Based on their responses, participants' occupations were categorized into 1 of 4 groups: unskilled labor, semi-skilled labor, managerial or technical, or professional.
Patients were screened for memory impairment by telephone interview using the Hopkins Verbal Learning Test-Revised (HVLT-R), a measure of episodic verbal learning and memory. The HVLT-R has been shown to be a valid and reliable measure of impairment in the domains of verbal learning and memory (15, 16). The test consists of a 12-item word list that is presented to patients on 3 successive learning trials yielding 3 primary learning and memory indices: a total recall score, a delayed recall score, and a percent retention score. The 3 available raw scores (total recall, delayed recall, percent retention) were transformed into Z scores using age-matched, normative values for the HVLT-R (17). For the purposes of classifying subjects, we calculated a composite Z score, which represented the mean of the 3 individual Z scores. This is a technique that is commonly used when tests yield multiple scores; it allows for a reduction in the amount of data in the analysis and facilitates interpretation of the results. The composite HVLT-R score was used to classify subjects into 1 of 3 groups: intact memory group if they scored −0.5 SD or higher than the age-matched normative value; mild-moderate impairment group if they scored −0.5 to −2.0 SD, and severe impairment if they scored less than −2.0 SD below age-matched population means.
Disease and general health status.
Disease activity was assessed using the Systemic Lupus Activity Questionnaire (SLAQ) (18), a validated, patient-reported assessment of disease activity in SLE. This measure has been found to correlate strongly with the revised Systemic Lupus Activity Measure (18). In addition, disease activity was assessed by patient self-report of the presence or absence of disease flare over the previous 3 months. Disease duration was calculated as the number of years since self-reported diagnosis of SLE. General health status was assessed with the Medical Outcomes Study Short Form 12 (19) and depressive symptoms by the Center for Epidemiologic Studies Depression Scale (CES-D) (20).
Information regarding various demographic characteristics was collected, including age, sex, race/ethnicity, marital status, and education. Race/ethnicity was dichotomized into white or nonwhite (includes African American, Hispanic/Latino, Asian/Pacific Islander, or other). Marital status was categorized as being currently married or living with a partner versus other. Education was reported as less than high school, high school graduate, some college, trade or vocational school, college graduate, or postgraduate degree.
After classifying subjects based on memory function, we compared the demographic, disease, and employment characteristics of the 3 groups as defined by intact memory, mild-moderate impairment, and severe impairment. Demographic, disease, and employment characteristics were expressed using means, SD, and proportions, as appropriate, and statistical tests (chi-square and analysis of variance) were used to identify differences among the 3 groups. We used univariate and multivariate logistic regression to assess the impact of memory function on the probability of being employed, and of not being able to work. Covariates in the multivariate logistic regression models included age, sex, ethnicity (white versus nonwhite), marital status (currently married or living with a partner), education (some college or higher versus no college), employment status at diagnosis (employed versus unable to work), disease activity (SLAQ score), disease duration, and depressive symptoms (CES-D score <16 versus ≥16). All statistical analyses were performed using SPSS software, version 14.0 (SPSS, Chicago, IL).
Ninety-one subjects ≥65 years of age were excluded from the analysis, leaving a final study sample of 741 subjects with SLE. Based on HVLT-R scores, 452 (61%) were classified as having intact memory function, 202 (27%) as having mild-moderate impairment, and 87 (12%) as having severe impairment. The demographic and disease characteristics of the 3 memory function groups are shown in Table 1. The 3 groups were similar in terms of age, sex, and marital status but varied in several other demographic characteristics. Participants with mild-moderate or severe memory impairment were more likely to be nonwhite, to have a lower annual household income, and to have a lower level of educational attainment. Participants in all 3 groups had a similar disease duration; however, those with mild-moderate and severe memory impairment had greater disease activity as expressed by the SLAQ score and were more likely to have high CES-D scores, suggesting more depressive symptoms.
|Intact (n = 452)||Mild to moderate impairment (n = 202)||Severe impairment (n = 87)|
|Age, mean ± SD years||45.5 ± 10.6||46.1 ± 11.1||46.3 ± 11.2|
|White, %||77.0||69.3||60.9||< 0.05|
|Married or living with partner, %||58.4||61.9||51.7|
|High school||98.2||96.0||93.0||< 0.05|
|Some college||91.4||82.2||76.7||< 0.05|
|College graduate||48.2||29.2||22.1||< 0.05|
|Household annual income, %|
|$20,000 to <$60,000||33.2||35.6||37.9|
|$60,000 to <$100,000||30.3||28.2||10.3||< 0.05|
|Disease duration, mean ± SD years||13.4 ± 8.1||12.9 ± 8.1||13.6 ± 9.2|
|Disease activity (SLAQ), mean ± SD score||11.5 ± 7.4||13.1 ± 7.9||15.5 ± 10.5||< 0.05|
|Flare in last month, %||46.6||46.9||57.8|
|SF-12, mean ± SD score|
|MCS-12||41.6 ± 6.2||41.9 ± 6.2||39.8 ± 5.8|
|PCS-12||40.0 ± 6.5||40.1 ± 6.5||38.7 ± 6.2|
|CES-D score ± SD||13.3 ± 11.6||17.8 ± 13.4||22.5 ± 16.3|
|Depressive symptoms, %†||35.2||48.0||58.6||< 0.05|
The employment status of subjects in each group for the current year and the year of diagnosis is shown in Table 2. In the intact memory group, 54.2% of participants were employed (i.e., had a job or did any work for pay or profit) in the current year, compared with 40.6% in the mild-moderate impairment group, and 31.0% in the severe impairment group. In the group with intact memory, 29.2% reported that they were unable to work in the current year, compared with 40.6% in the group with mild-moderate impairment, and 56.3% in the severe impairment group. There was no significant difference among the 3 groups in the proportion that were employed in the year of diagnosis (Table 2), and these figures did not differ considerably from reported rates of employment in an age and sex-matched normal population (73.6%) (21). There was a significant difference in the proportion of subjects reporting that they were unable to work at year of diagnosis, although the absolute numbers were small. Among participants not employed in the current year, there was no significant difference among the 3 groups in the duration of time since their last regular job. The majority had not had regular work in >3 years (71.1% in the intact group, 70.4% in the mild-moderate impairment group, and 76.7% in the severe impairment group) but this difference was not statistically significant (P = 0.65; data not shown).
|Employment status||Memory function||P|
|Intact (n = 452)||Mild to moderate impairment (n = 202)||Severe impairment (n = 87)|
|Current year, %|
|Unable to work||29.2||40.6||56.3||< 0.001|
|Year of diagnosis, %|
|Unable to work||6.6||10.9||16.1||0.009|
The frequency of employment in 4 occupational categories (unskilled, semiskilled, managerial or technical, and professional) at year of diagnosis and at current year is shown in Figure 1. In the year of diagnosis, the 3 memory function groups displayed a similar distribution among the various occupational categories. However, in the current year, only the intact memory function group showed the expected increase in the proportion of subjects in the professional occupations, which is typical of persons in this age group. Subjects in the mild-moderate impairment and severe impairment groups did not demonstrate a similar increase, and those in the severe impairment group demonstrated substantial decreases in the proportion of subjects in managerial and professional occupations.
The unadjusted and regression-adjusted odds ratios (ORs) for being employed and of being unable to work are shown in Tables 3 and 4. Prior to adjustment, increasing levels of memory impairment predicted a decreased likelihood of being employed, OR 0.58 (95% confidence interval [95% CI] 0.41–0.81) for the mild-moderate impairment group, and OR 0.38 (95% CI 0.23–0.62) for the severe impairment group (Table 3). After adjustment for important covariates, a similar pattern was observed in the mild-moderate impairment group (OR 0.70, 95% CI 0.48–1.02), and in the severe impairment group (OR 0.57, 95% CI 0.32–1.00). Correspondingly, the OR for being unable to work increased with greater severity of memory impairment (Table 4). Prior to adjustment, participants with memory impairment were more likely to report being unable to work: OR 1.66 (95% CI 1.17–2.34) for the mild-moderate impairment group and OR 3.13 (95% CI 1.95–5.00) for the group with severe impairment. After adjustment, participants with memory impairment were still more likely to report being unable to work: OR 1.36 (95% CI 0.90–2.04) for the mild-moderate impairment group and OR 1.99 (95% CI 1.12–3.55) for the severe impairment group. After multivariate adjustment, only the results for the severe impairment group, with respect to either employment status or an inability to work, remained statistically significant. The most important predictors of employment and of being unable to work were education, employment status at diagnosis, disease activity and duration, depressive symptoms, and memory function.
|Variable||Unadjusted OR (95% CI)||Regression-adjusted OR (95% CI)†|
|Age, years‡||0.97 (0.96, 0.98)||0.97 (0.96, 0.99)|
|Female||0.68 (0.38, 1.20)||0.81 (0.42, 1.56)|
|White||1.20 (0.86, 1.66)||1.39 (0.96, 2.03)|
|Married or living with partner||1.29 (0.96, 1.72)||1.18 (0.84, 1.65)|
|Education§||2.70 (1.69, 4.30)||1.84 (1.09, 3.11)|
|Employed at year of diagnosis||2.28 (1.62, 3.19)||2.42 (1.65, 3.55)|
|SLAQ‡||0.91 (0.89, 0.93)||0.92 (0.90, 0.95)|
|Disease duration, years‡||0.98 (0.96, 1.00)||0.97 (0.95, 0.99)|
|Depressive symptoms¶||0.36 (0.25, 0.50)||0.64 (0.44, 0.94)|
|Mild to moderate||0.58 (0.41, 0.81)||0.70 (0.48, 1.02)|
|Severe||0.38 (0.23, 0.62)||0.57 (0.32, 1.00)|
|Variable||Unadjusted OR (95% CI)||Regression-adjusted OR (95% CI)†|
|Age, years‡||1.04 (1.03, 1.06)||1.04 (1.02, 1.05)|
|Female||1.20 (0.66, 2.16)||0.93 (0.44, 1.94)|
|White||1.12 (0.80, 1.57)||0.98 (0.65, 1.49)|
|Married or living with partner||0.59 (0.43, 0.79)||0.60 (0.41, 0.87)|
|Education§||0.36 (0.23, 0.55)||0.53 (0.32, 0.89)|
|Unable to work at year of diagnosis||4.93 (2.85, 8.52)||2.64 (1.37, 5.07)|
|SLAQ‡||1.13 (1.11, 1.16)||1.11 (1.08, 1.14)|
|Disease duration, years‡||1.01 (1.00, 1.03)||1.02 (1.00, 1.05)|
|Depression¶||4.29 (3.12, 5.90)||1.71 (1.15, 2.54)|
|Mild to moderate||1.66 (1.17, 2.34)||1.36 (0.90, 2.04)|
|Severe||3.13 (1.95, 5.00)||1.99 (1.12, 3.55)|
The findings of this study suggest that memory-impaired persons with SLE, as assessed by a brief screening measure, experience greater work disability than those with intact memory function. These findings were observed even after adjusting for important covariates such as disease duration, disease activity, education level, depressive symptoms, and demographic variables. Employment status at the time of diagnosis, level of educational attainment, greater disease activity and duration, and depressive symptoms were also predictive of greater work disability.
Although severe cognitive impairment and frank dementia may occur in patients with SLE, in the vast majority of cases cognitive impairment will be mild and often so subtle it is unrecognized. In 1 study examining the prevalence of neuropsychiatric syndromes, 80% of the subjects with SLE were found to have cognitive impairment (memory was the most frequently affected domain) and of these, 70% were classified as mild (4). The results of our study demonstrate that impairment in even 1 domain of cognitive functioning, memory, may have a considerable impact on employment. Studies of elderly populations (22, 23), patients with multiple sclerosis (24), and patients with the human immunodeficiency virus (25) have also found that even mild impairment in cognitive functioning and memory is associated with an increase in work disability and difficulties performing activities of daily living.
Our findings are consistent with those of Utset et al (12), who sought to determine the clinical, serologic, demographic, psychological, and neurocognitive correlates of work disability in SLE. In their study of 50 subjects with SLE, 50% were identified as having neurocognitive impairment; additionally 32% of the subjects had obtained formal work disability, 16% had self-reported work disability, and 52% denied having work disability. Subjects who reported formal work disability were more likely to be neurocognitively impaired OR 14.4 (95% CI 3.0–68.2). After adjustment for various demographic and clinical variables, only neurocognitive dysfunction and scores on a disease damage index were significantly associated with formal work disability.
We observed a strong correlation between depressive symptoms and memory impairment in our study, in which 59% of patients with severe memory impairment were possibly depressed, compared with 48% in the mild-moderate group, and 35% in the intact memory group. In addition to being strongly correlated with memory function, the presence of depressive symptoms was also found to be an important predictor of both employment and of being unable to work. Reports of depression are common in patients with SLE and have been associated with poorer performance on neurocognitive tests (26, 27). Some authors have argued that cognitive deficits in SLE may result largely from psychological disorders such as depression, and the stress of living with a chronic disease. Alternatively, organic central nervous system involvement may be the cause of both depression and cognitive dysfunction. In fact, some magnetic resonance imaging findings in patients with cognitive dysfunction are often similar to those of patients with primary depression, specifically ischemic and nonspecific white matter changes (28, 29). Furthermore, depression may result as a direct consequence of cognitive impairment and the associated difficulties that it may cause in daily life. Despite the complexity of the relationship between cognitive impairment and depression, and despite the high prevalence of depressive symptoms in our study, we found that impairment in memory function was an important independent predictor even after adjusting for depressive symptoms. Defining the precise role of depression, a highly treatable disorder, should be the aim of future studies.
The association between memory function and employment status in patients with SLE underscores the need for judicious assessment of cognitive function and for the development of appropriate strategies that will either reverse cognitive impairment or help patients overcome obstacles they may face in daily life. Pertaining to employment, persons with cognitive impairment may benefit from training programs that have been used in other conditions, such as severe mental illness and traumatic brain injury (30). These programs have been designed to either help patients compensate for impaired cognitive function or to enhance or restore defective function. In a study of persons with mental illness and a history of job failure (31), half the subjects were randomized to receive cognitive training. Over a period of 1 year, subjects receiving cognitive training were significantly more likely to be employed (69.6% versus 4.8%), worked more hours, and earned higher wages than those who did not receive the training. Strategies to get people with SLE back to work would not only help restore their self-esteem and improve their quality of life, but would also help offset the large indirect costs associated with this condition (32–34).
For our current study, we chose the HVLT-R as a screening measure to detect possible memory impairment. The HVLT-R provides a brief measure of episodic verbal memory that has been found to be correlated with the California Verbal Learning Test, a more extensive measure of verbal memory (35). Furthermore, the HVLT-R was selected because it can better capture deficits resulting from subcortical white matter brain dysfunction (i.e., verbal learning and memory), which are considered to be the primary cognitive deficits in patients with SLE (36). Although the HVLT-R has not been validated for use via telephone, we chose this measure because other telephone-validated measures (e.g., the mini-mental status exam) would be less sensitive to the kinds of deficits observed in patients with SLE. Nonetheless, we acknowledge that this telephone-administered assessment of memory function is an important limitation of our study. Subjects identified as memory-impaired using these methods would ideally require more extensive evaluation using a conventional battery of tests such as those that have been proposed by the ACR (1). Another limitation of our study was its cross-sectional design. Due to this limitation, we do not have an assessment of premorbid cognitive functioning prior to the onset of the SLE disease processes. In addition we did not evaluate change in memory functioning prospectively. Although few studies of chronic medical conditions can adequately capture premorbid functioning, future studies evaluating change in cognition in relation to employment outcomes are warranted. A future longitudinal study will provide important information about change in memory functioning in relation to other important variables, such as depression, disease factors, and socioeconomic status.
Evaluation of cognitive function may need to become part of the routine assessment of patients with SLE, even in the absence of overt dysfunction, because even mild impairment appears to interfere with employment and quality of life in general. One challenge that remains is finding a means to identify patients at risk for poor outcomes, such as inability to maintain employment, who would benefit most from potential treatment strategies. Assessments that take 1 hour or longer are too long and too expensive to be used routinely. One group has evaluated the use of a shorter cognitive test that appears to be promising and may be easily administered in a busy clinical setting (37). Future research should also help define the specific neurocognitive deficits in patients with SLE that lead to work disability. This will allow for the development of treatment strategies that will be best suited for helping maintain employment and preserving quality of life.
Dr. Panopalis had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Panopalis, Julian, Yazdany, Criswell, Katz, Yelin.
Acquisition of data. Criswell, Yelin.
Analysis and interpretation of data. Panopalis, Julian, Yazdany, Gillis, Trupin, Yelin.
Manuscript preparation. Panopalis, Julian, Yazdany, Gillis, Trupin, Hersh, Criswell, Katz, Yelin.
Statistical analysis. Panopalis, Julian, Yazdany.
- 14US Department of Labor: Bureau of Labor Statistics. How the government measures unemployment. URL: http://www.bls.gov/cps/cps_htgm.htm. 2001.
- 17Hopkins Verbal Learning Test-Revised: professional manual. Odessa (FL): Psychological Assessment Resources; 2002., .