development of the quality of life in epilepsy inventory

Epilepsia, 36(11): 1089-1 104, 1995 Lippincott-Raven Publishers, Philadelphia 0 International League Against Epilepsy

Development of the Quality of Life in Epilepsy Inventory

Orrin Devinsky, *$Barbara G. Vickrey, §Joyce Cramer, Kenneth Perrine, ''Bruce Hermann, YKimford Meador, and ?Ron D. Hays

Department of Neurology, New York University School of Medicine, Hospital for Joint Diseases, New York, New York; *Department of Neurology and f U C L A School of Medicine, Los Angeles; $Department of Social Policy, Rand, Santa Monica, California; §Health Services Research, Department of Veterans Affairs Medical Center, West Haven,

and Department of Neurology, Yale University School of Medicine, New Haven, Connecticut; "Departments of Psychiatry and Neurosurgery, University of Tennessee, Memphis, Tennessee; and TlDepartment of Neurology,

Medical College of Georgia, Augusta, Georgia, U.S.A.

Summary: We developed an instrument to measure health-related quality of life (HRQOL) in epilepsy. A 99- item inventory was constructed from the RAND 36-Item Health Survey (generic core), with 9 additional generic items, 48 epilepsy-targeted items, and 6 other items concerning attitudes toward epilepsy and self-esteem. We administered the 99-item inventory to 304 adults with epilepsy at 25 epilepsy centers. Patients and patient- designated proxies completed the inventory and were retested 1-91 days later. A multitrait scaling analysis of these data led to retention of 86 items distributed in 17 multiitem scales (Cronbach's alpha ranged from 0.78 to 0.92). Factor analysis of the 17 multiitem scales yielded four underlying dimensions of health: an epilepsy- targeted dimension, a cognitive factor, mental health, and physical health. Construct validity was supported by significant patient-proxy correlations for all scales and correlations between neuropsychologic tests and self- reported emotional and cognitive function (all p values < 0.05). There were significant negative correlations between the four factor scores derived from the HRQOL

scales and neurotoxicity , systemic toxicity, and health care utilization (except for the correlation between mental health factor and health care utilization; all p values < 0.05). Patients who were seizure-free in the preceding year reported better HRQOL for the overall score, three of the four factor scores, and 8 of the 17 scale scores than did patients with a high frequency of seizures. Relative validity analysis showed that the epilepsy-targeted factor and three of its four component scales were more sensitive to categorization of patients by severity of seizure frequency and type than scales tapping physical health, mental health, o r cognitive function. These cross- sectional data support the reliability and validity of this measure of HRQOL in epilepsy. The addition of an epilepsy-targeted supplement to the generic core improved the sensitivity to severity of epilepsy. The 86 items included in the field testing were supplemented by three additional items to form the Quality of Life in Epilepsy (QOLIE-89) inventory. Key Words: Health-related quality of life-Dimensions of health-Construct validity- Quality of Life in Epilepsy Inventory.

More than 1Y2 million Americans have epilepsy, with - 100,000 new cases identified annually (1). Pa- tients face a diverse range of problems after diagnosis, regardless of their eventual degree of seizure control. It is unfortunate that most adults with symptomatic localization-related (partial) epilepsy continue to have simple or complex partial, or secondarily generalized tonic-clonic seizures (SPS , CPS, GTCS). In the VA studies (2,3), only 22% of untreated or undertreated patients were free of CPS 2 years after a monotherapy antiepileptic drug

Received December 8, 1994; revision accepted May 5 , 1995. Address correspondence and reprint requests to Dr. 0.

Devinsky at Department of Neurology, Hospital for Joint Dis- eases, 301 E. 17 St., New York, NY 10003, U.S.A.

(AED) trial. Careful withdrawal of AEDs in selected patients resulted in 41% relapse at 2 years (4). The possibility of recurrent seizures therefore remains a silent, chronic daily hazard for most epilepsy patients.

Traditional medical care of epilepsy patients fo- cuses on adjustment of medication, based on type, severity, and frequency of seizures and adverse events (AE). However, seizures and AED therapy have impact on patients' lives in many other ways; such effects often linger after long-term remission is achieved. Epilepsy is both a medical diagnosis and a social label (5). The stigma and underlying worry about epilepsy are not usually considered in long- term follow-up of patients with few or no recent seizures. Low self-esteem, lack of independence,

1089

1090 0. DEVINSKY ET AL.

need for AEDs, restrictions on alcohol use and driving, reporting of epilepsy on job and insurance ap- plications, and presence of AEDs in urine tests are chronic problems frequently faced by patients with epilepsy (6-8).

Health-related quality of life (HRQOL) refers to the way in which individuals function and their perceived well-being in physical, mental, and social domains of life. Evaluation of the impact of medications and other treatments for epilepsy in HRQOL is essential because functioning and well-being are the outcomes that are most important to patients. In addition, patients and physicians often differ in their perceptions of traditional medical outcomes such as seizure severity and AE resulting from medications. To date, use of HRQOL measures in epilepsy outcome studies has been limited (9).

HRQOL can be assessed by generic or disease- specific measures or a combination of both. Generic instruments assess a broad range of functioning and well-being and allow comparisons across patients with different diseases as well as patients with the general population. Examples of generic measures include a 36-item health survey developed in the Medical Outcomes Study (MOS) (10,l l), the Sick- ness Impact Profile (12), and the Dartmouth Coop Charts (13). Disease-targeted instruments concen- trate on issues of particular relevance to a specific disease or disorder and therefore may be more sensitive than generic measures to differences within the targeted condition (14).

The Liverpool group has assessed HRQOL in epilepsy by using scales of physical functioning (seizure severity, seizure frequency, activities of daily living, ADL), social functioning, life fulfillment, stigma, impact of epilepsy, and psychological functioning. Although only some of these scales (and subscales) were developed to assess epilepsy patients, all have been validated in epilepsy patients (15-18). In assessing specific clinical issues, the Liverpool group has selected different combina- tions of scales. They have applied QOL measures to assessing lamotrigine (LTG) in patients with refrac- tory epilepsy (19), QOL and quality of services for a community-based population, efficacy of a new AED in children with epilepsy and learning disabil- ities, and early versus delayed treatment of single seizures and early-onset epilepsy (20). Kendrick and Trimble (21) used extensive patient interviews to identify important issues concerning their physical functioning, cognition, emotion, social functioning, and economic/work status. Using a repertory-grid technique with a Construct Importance Scale, patients identified and rated problems during follow-up.

Vickrey et al. (22) incorporated both a generic core instrument and a disease-targeted supplement in developing a HRQOL instrument for patients evaluated for epilepsy surgery, the ESI-55. The ESI-55 includes the RAND 36-Item Health Survey 1 .O as the generic core plus 19 additional items tapping areas regarded as important for persons with epilepsy: cognitive function, role limitations due to memory problems, seizure-specific health perceptions, and overall QOL. The ESI-55 combines strengths of generic and disease-targeted instruments. However, it does not assess some domains such as social isolation and driving limitations that are important to individuals with epilepsy (8). Thus, there was a need to expand on their initial effort and develop a comprehensive HRQOL instrument for use in a general epilepsy population. We describe the development and evaluation of such a measure.

METHODS

Subjects Three hundred four adult, English-speaking epi-

lepsy patients at 25 epilepsy centers in the United Sates were recruited into the study. Patients were recruited based on eligibility criteria and willingness to participate. Inclusion criteria required subjects to be aged 17-65 years, to have a minimum of tenth grade education, and to be able to read English. Patients with a diagnosis of primary generalized (absence, myoclonic, and secondarily generalized tonic-clonic-seizures) or partial (simple and complex partial, and tonic-clonic seizures) epilepsy were included. Patients were excluded who had ac- tive medical or psychiatric illness that caused current symptoms and required treatment, were taking medication that could affect the CNS (other than AEDs), had a change in their AED regimen <4 weeks before study onset, or had undergone a cra- niotomy in the preceding year or neuropsychological testing in the previous 6 months. Initially, only patients with low and moderate seizure frequencies during the preceding year were enrolled. Study eligibility criteria were expanded midway through data collection to include patients who were seizure-free and patients with high seizure frequencies in the preceding 12 months (Table 1). This change in recruitment was designed to permit evaluation of HRQOL as a function of a broader range of epilepsy severity. Patients were required to have a family member or close friend or associate with whom they had contact at least once a week participate in the study as a proxy respondent. Informed consent was obtained from all patients and proxies.

Epilepsia, Vol. 36, No. 11, 1995

DEVELOPMENT OF THE QOLIE INVENTORY 1091

TABLE 1. Criteria for seizure-specific categorization of severity of epilepsy"

No. of seizures in preceding 12 months

Seizure-free > I yr, Low, Moderate, High,

Seizure type n = 21' n = 116 n = 136 n = 31

Simple partial 0 1-20 21-100 101-200 Complex partial 0 1 4 5-12 13-24 Generalized tonic-clonic 0 1 2 4 5-12 Absence 0 1-20 21-100 10 1-200 Myoclonic 0 1-20 21-100 10 1-200

a Subjects with multiple seizure types were classified according to the most severe category of all seizure types. ' All patients in this group had had at least one seizure in the 12- to 24-month period preceding completion of the quality of life questionnaire.

Measures 1. HRQOL study questionnaire (99 items). The

instrument included a widely-used survey of general HRQOL supplemented by items targeted for people with epilepsy. Seventeen multi-item scales were hypothesized using these items, including scales from previous instruments and newly derived scales formed by logical groupings of items into relevant content domains.

Generic core (36 items). The RAND 36-item Health Survey 1 .O (aka SF-36) was adopted as the generic core (10,ll). This measure includes 8 multi-item scales assessing physical functioning (10 items), role limitations due to physical problems (4 items), role limitations due to emotional problems (3 items), social functioning (2 items), bodily pain (2 items), emotional well-being (5 items), energy/fatigue (4 items), and general health perceptions (5 items). It also includes an item on change in health.

Expansion of the generic core (9 items). The two SF-36 role limitations scales were expanded by 3 items to make the scales analo- gous to one another (5 items each). Two items on health perceptions were added to supplement the general health perceptions scale. We added two items that assessed overall QOL using a pictorial QOL chart developed for the Dartmouth COOP (23) and an overall QOL item adapted from a patient preferences study (24) from the Faces Scale (25). This latter item was included in the questionnaire twice to assess reproducibility of a single item in the same administration. An open-ended question that encouraged written responses about additional QOL concerns was also included.

Epilepsy-targeted HRQOL supplement (48 items). The areas that supplemented the generic core were related to issues commonly

reported by epilepsy patients with moderately well-controlled epilepsy (i.e., those not generally considered candidates for epilepsy surgery). These questions were derived from the experience of 6 health professionals in epilepsy and from 30 individual outpatient interviews with epilepsy patients regarding HRQOL issues and a literature review of studies examining seizure severity and psychoso- cia1 aspects of epilepsy.

a. Epilepsy-targeted health perception (2 items). Because the diagnosis of epilepsy and the occurrence of seizures can alter one's perception of health (26), we assessed this issue with 2 epilepsy-targeted items assessing health perception. One item was from the ESI-55 (22) and 1 was created by the study team. These 2 items were added to the generic health perceptions scale in the initial item analysis.

b. Epilepsy-targeted social function (5 items). Epilepsy is associated with problems in social functioning and social isolation (27- 29). Patients with epilepsy often report problems related to self-esteem, stigma, discrimination, social acceptance, and independence (6,8,30,3 1). The actual and perceived stigma associated with seizures and epilepsy causes many individuals to hide their disorder or min- imize social contact (32,33). These 5 items were created by the study team.

c. Working and driving limitations (4 items). Among the psychosocial problems af- fecting adults with epilepsy, driving restrictions and under- and unemployment are commonly reported (8,3636). These 4 items were created by the study team.

d. Cognitive function (21 items). Cogni- tive problems have frequently been reported in patients with epilepsy and may result from underlying brain injury, medications, seizures,



and psychosocial problems (37-39). Further- more, changes in cognitive abilities may occur as a result of changes in AED regimens or after epilepsy surgery (40,41). Items assessed selected memory functions (6 items), attention and concentration (9 items), and language functions (6 items). These items were either developed de novo for the study or taken from longer Medical Outcomes Study (MOS) instruments (4), or from the ESI-55 (22).

e. Seizure worry (8 items). Seizures often occur randomly and can cause injury, embarrassment, and loss of employment and driving privileges. Patients with epilepsy are con- cerned over the lack of control in their lives due to the unpredictable nature of seizures (8,43), as well as direct and indirect effects of seizures (36,4446). These items were either from the ESI-55 (22) or were created by the study team.

f. Medication effects (2 items). AEDs have idiosyncratic and dose-related adverse effects that can affect physical and mental functions (2,3). These 2 items were created by the study team.

g. Social isolation (2 items). In addition to impairing social function, epilepsy can lead to social isolation. These 2 items were adapted from the short-form of the revised UCLA Loneliness Scale (47).

h. Social support (4 items). Epilepsy is often associated with greater needs for social support that may be perceived as due to epilepsy (described in section on epilepsy- targeted social function) or to other or unde- fined factors. These items were drawn from longer instruments in the MOS (48).

Other items (6 items). The questionnaire also included 5 items on attitudes toward seizures and one item on self-esteem, all created specifically for this study.

Proxy HRQOL questionnaire. A proxy questionnaire was constructed by rephrasing the 99 items from the patient questionnaire. Proxies were instructed to answer every question “the way you believe best describes your relative’s or friend’s situation.”

2. Neuropsychological measures. A battery of neuropsychological tests included measures sensitive to a variety of attentionkoncentra- tion, memory, language, learning, sensorimo- tor, and mood problems typically related to epilepsy and AEDs (49). Tests were administered in the following sequence to accommo- date delayed recall for the three memory tests:

Wechsler Memory Scale-Revised, Logical Memory (immediate recall); Rey-Osterreith Complex Figure Test (copy and immediate recall); Rey Auditory Verbal Learning Test; Grooved Pegboard; Controlled Oral Word As- sociation Test; Stroop Color Word Test; Pro- file of Mood States (6 subscales); Logical Memory (delayed recall); Rey-Osterreith Complex Figure (delayed recall); Rey Audi- tory Verbal Learning Test (delayed recall and delayed recognition); Multilingual Aphasia Examination-Visual Naming subtest; and Symbol-Digit Modalities Test (4930).

3. Neurotoxicity and systemic toxicity scales. A directed neurological examination and history, focusing on signs and symptoms of neurotoxicity, was conducted. The neurotoxicity rating scale was a modification of the VA Neurotox- icity Rating Scale (51). This neurotoxicity rating scale assesses the following neurologic symptoms and signs and their severity: diplo- pia, nystagmus, dysarthria, gait, rapid alter- nating movements, tremor, sedation, affect and mood, cognitive function, drug-related dizziness or lightheadedness, drug-related headaches, and other AED-related neurotoxic effects. A total score is derived for neurotoxicity, ranging from 0 (no toxicity) to 465 (maximum toxicity) points. A modification of the VA Systemic Toxicity Rating Scale (51) assessed the following signs or symptoms related to drug use: gastrointestinal problems, impotence (libido or potency), weight gain, changes in hair quality or quantity, and other AED-related systemic effects, providing a total score for systemic toxicity ranging from 0 (no toxicity) to 220 (maximum toxicity) points.

4. Clinical, demographic and health care utilization variables. Patients were interviewed by a physician, nurse, or research assistant to de- termine seizure frequency and type during the past year, age, gender, marital status, level of education, occupational status and level, housing status, seizure history (age of seizure onset, date and type of most recent seizure, current AEDs and dosages), health care utilization (outpatient and inpatient) in the preceding 12 months, and comorbid medical conditions (high blood pressure, diabetes, conges- tive heart failure, heart attack in the last year, emphysema, or other).

Data collection Eligible patients who agreed to participate in the

study were required to come to the medical center


DEVELOPMENT O F THE QOLIE INVENTORY 1093

with their relative or friend who served as a proxy. Patients were interviewed and underwent a baseline neurological examination. After the interview, patients completed the HRQOL questionnaire. The research assistant noted the time the patient started and completed the questionnaire. Patients then performed a battery of neuropsychological tests administered by a trained test examiner who was given a written set of instructions to enhance comparability across sites. Proxies completed both an interview by a physician, nurse, or research assistant and the HRQOL questionnaire. On a second, separate date, 1-91 days after the initial visit (>90% within 30 days), both the patient and the same proxy completed HRQOL questionnaires identical to those completed at the first visit. The target date for follow-up questionnaires was within 3 weeks of the original questionnaire. The follow-up questionnaires were either completed at the time of a return visit to the medical center or at home and returned by mail. Patients were also asked to report any major life events (e.g., death of loved one, medical illness) since the first visit, the number and type of seizures since the first visit, and current AEDs and dosages.

At some sites, patients received monetary com- pensation of ~ $ 3 0 for the first visit and $20 for the second visit to enhance attendance for the two test sessions. Proxies received ~ $ 2 0 for each visit. Pay- ment was made at the discretion of the principal investigator at each site to compensate patients for time and travel expenses.

All questionnaires were reviewed by the site research coordinator for completeness and mailed to a central site in Secaucus, NJ, U.S.A., where they were logged and were then forwarded to RAND (Santa Monica, CA, U.S.A.) for data entry, screen- ing, and analysis.

Analysis plan

Descriptive statistics All scale scores were transformed linearly into

scales of 0-100 points, with higher values representing better functioning and well-being. The RAND scoring algorithm was used for the pain scale (1 1). Mean score, SD, range, and percentage of subjects with maximum and minimum score were calculated for each scale.

Item analysis and scale reliability Multitrait scaling analysis was used to assess item

convergence and discrimination across scales (52). Items that had no item-scale correlations >0.30 were excluded from further consideration. Items

that had higher correlations with a scale other than the hypothesized scale were moved (when substan- tively reasonable), and the multitrait scaling analysis was run again. Cronbach’s alpha (53) was used to estimate scale internal consistency reliability. Test-retest reliability was estimated by calculating Pearson product-moment and intraclass correlations between the first and second set of scores in patients reporting no major life events between test dates.

Content validity Item selection used for this questionnaire maxi-

mized content validity because it was based on in- put from clinicians and patients as well as a review of the literature to identify important QOL issues in epilepsy patients. We also reviewed patient responses on the open-ended question about additional QOL concerns to identify areas not covered by the initial set of items.

Construct validity Several analyses were conducted to evaluate con-

struct validity. After the multitrait scaling analysis was completed, exploratory factor analysis was used to summarize the intercorrelations among the 17 HRQOL scales. Next, we examined correlations between the patient and proxy responses. Correla- tions were also calculated between the HRQOL scales and neuropsychological test scores, the systemic and neurotoxicity scores, educational attainment, employment status, and health care utilization. Finally, we examined the efficiency (54) of the HRQOL measures relative to severity of epilepsy (Table 1). Patients having fewer and less severe seizures were hypothesized to have better HRQOL than patients reporting more frequent and severe seizures. Relative validities were determined by dividing F-ratios from one-way between-group ANOVAs by the F-ratio of a reference scale, with the relative validity of the reference scale set equal to 1 by definition. As a sensitivity analysis, we also examined HRQOL scores by severity category, ad- justing for age, gender, educational attainment, and comorbid medical conditions.

Derivation of overall score We developed an overall score for the HRQOL

measure that is a weighted combination of individual scale scores (55) using a technique previously described (56). First, scores for the four factors underlying the 17 scales were averaged together to produce a criterion score. Scale weights were then derived by regressing this criterion onto each of the scales. Standardized beta coefficients from the re-

Epilepsia, Vol. 36, No. I I , 1995


gression were summed, and each beta coefficient was divided by the sum to derive a relative weight for each scale.

RESULTS

Description of sample The sample consisted of 130 men and 174 women

with a mean age of 36 years (range 17-63 years) and mean age of seizure onset of 17.9 years (SD =

12.1). The number of subjects recruited from each of the 25 participating centers ranged from 1 to 40. There were 116 patients with low, 136 with moderate, and 31 with high seizure frequency based on seizure frequency and type; 21 patients were seizure-free during the preceding year but had at least one seizure in the preceding 12- to 24-month period (Table 1). Forty-eight percent of patients were married or living as married; 52% were single, divorced, or widowed. Seven percent of patients did not complete high school, 37% had a high school diploma or equivalent but no higher education, and 56% had vocational training or college education.

The time that the HRQOL questionnaire was be- gun and the time it was completed were recorded directly on each questionnaire. Based on this data, subjects took an average of 28.4 min (SD 15.6, range, 6-135 min) to complete the questionnaire.

Multitrait scaling Ninety items tapping dimensions of HRQOL and

hypothesized to fall into 17 unique scales were in-

cluded in multitrait scaling analysis. These original scales were the same as those shown in Table 2, except that work/driving limitations and social functions were initially hypothesized to be separate scales and the two items in the health discouragement scale were originally hypothesized to fall into the health perceptions scale. (The five items on attitudes toward seizures and the single item on self- esteem were excluded from the multitrait analysis because they did not directly assess QOL or were exploratory in nature. The single item on change in health, the open-ended question on QOL, and the duplicate overall HRQOL item were also excluded from the multitrait scaling analysis.)

Multitrait scaling analysis of the 17 hypothesized scales indicated that items tapping work and driving limitations were highly correlated with items assessing social function. Therefore, these items were combined into a single scale representing work, driving, and social functioning. The two items on health discouragement that were originally ana- lyzed together with health perceptions items were separated from those items to create a distinct health discouragement scale, based on multitrait scaling results. Four items were dropped from the original questionnaire because all item-scale correlations fell below 0.35 (three items) or had correlations of a similar magnitude across multiple scales (one item), and thus did not discriminate across scales. Therefore, multitrait scaling analysis yielded 86 items placed into 17 multiitem scales as shown in Table 2.

TABLE 2. QOLIE-89 descriptive statistics and reliabilities"

Scoring Proportional

Scale items Mean SD minimum (%) alpha reliabilityb overall score No. of Observed at ceiling Cronbach's Test-retest weight in

Seizure worry 5 58.3 25.8 0 5.9 0.79 0.84 0.06 Medication effects 3 55.3 30.5 0 12.8 0.78 0.64 0.05 Health discouragement 2 69.9 27.7 0 24.4 0.82 0.73 0.07 Work/driving/social function 1 1 66.9 22.9 3.6 3.6 0.86 0.86 0.08 Language 5 74.6 21.0 4.0 7.6 0.88 0.72 0.06 Attentiodconcentration 9 70.0 20.7 1 1 . 1 2.6 0.92 0.86 0.08 Memory 6 54.3 24.2 0 2.0 0.88 0.82 0.07 Overall quality of life 2 67.2 18.4 5.0 4.9 0.79 0.84 0.06 Emotional well-being 5 67.2 19.3 16.0 1 .o 0.83 0.77 0.05 Role limitations: emotional 5 67.8 34.5 0 43.9 0.81 0.67 0.05 Social isolation 2 76.8 25.0 0 33.6 0.88 0.73 0.04 Social support 4 72.5 22.9 0 17.4 0.84 0.78 0.02 Energy/fatigue 4 55.3 21.1 0 0.7 0.84 0.75 0.05 Health perceptions 6 68.3 19.6 20.8 3.0 0.78 0.84 0.06 Physical function 10 85.3 19.8 5.6 34.9 0.89 0.75 0.06 Role limitations: physical 5 69.3 34.5 0 41.2 0.81 0.58 0.07 Pain 2 75.6 24.8 0 31.9 0.87 0.69 0.07 Overall score 86 67.9 15.6 25.6 0 0.97" 0.88 -

a n = 298-304 patients for all data except test-retest reliability. n = 229-232 in the subset of epilepsy patients who were clinically stable (no significant intervening life event) and whose test-retest

Estimated with Mosier's formula (74). interval ranged from 1 to 21 days.



Descriptive statistics Descriptive statistics for the final 17 multiitem

scales are summarized in Table 2. The majority of scales demonstrated good variability. The minimum observed scores (0-100 point scales) in the 17 scales ranged from 0 to 21, and at least some respondents scored 100 (the maximum) on each scale. Mean scores ranged from 54 to 85; 229-232 epilepsy patients were clinically stable (no significant intervening life event) and had a test-retest interval <3 weeks.

Reliability Internal consistency (Cronbach’s alpha) reliabili-

ties for the 17 multiitem scales ranged from 0.78 to 0.92, exceeding the generally accepted criterion of 0.70 for adequate reliability for group comparisons (57). Test-retest reliabilities (product-moment correlations) ranged from r = 0.58 to r = 0.86 for the 17 scales. All scales exceeded the r = 0.70 standard for group comparisons except for the two role limitations scales, pain and medication effects. Intra- class correlation coefficients were also determined and were comparable to the test-retest product- moment correlations, ranging from 0.58 to 0.85.

Factor analysis of the 17 multiitem scales Several factor criteria suggested that four dimen-

sions were needed to account for the intercorrelations among the 17 HRQOL scales (58-60). A four- factor oblique (Promax) rotation was performed (61) yielding (number of scales loading 20.30 in pa- rentheses) an epilepsy-targeted factor (4 scales), a

cognitive factor (3 scales), a mental health factor (6 scales), and a physical health factor (4 scales). Sim- ple stiucture was apparent because no scale loaded 20.30 on multiple factors (Table 3). Estimated correlations between different factors ranged from 0.38 (physical health and epilepsy-targeted) to 0.50 (mental health and epilepsy-targeted).

Content validity A review of patients’ responses to the open-

ended question on QOL showed that some patients had additional concerns regarding finances, specific limitations to participation in athletic activities, pregnancy and possible birth defects, stigma and bother associated with taking pills, and insomnia. However, many other responses to this question were in areas that were covered by the HRQOL measure, such as driving limitations and fatigue.

Patient-proxy agreement Correlations (product-moment correlations) be-

tween patient self-reports and corresponding proxy reports ranged from 0.29 (role limitations: emotional) to 0.57 (work/social function); all correlations were significant at p < 0.0001. A comprehensive comparison of agreement between patient and proxies was reported separately (61a).

Neuropsychological tests-HRQOL correlations All the POMS scales correlated significantly with

the overall QOL, emotional well-being, role functioning-emotional, and energy/fatigue scales (p < 0.0001). Depression, vigor, and confusion from the

TABLE 3 . Promax rotated factor solution for QOLIE-89 scales (n = 299)

Epilepsy- Mental Physical Scale targeted Cognitive health health

Seizure worry 0.91 Medication effects 0.79 Health discouragement 0.65 Work/driving/social 0.62 Function Language Attention/concentration Memory Overall QOL Emotional well-being Role limitation: emotional Social isolation Social support Energy/fatigue Health perceptions Physical function Role limitations: physical Pain

0.95 0.83 0.81

0.70 0.80 0.54 0.69 0.95 0.60

0.64 0.84 0.42 0.78

QOLIE, Quality of Life in Epilepsy. Standardized regression coefficients of the factors on the scales. Factor pattern load-

ings 20.30 are shown.

Epilepsia, Vol. 36, NO. 11. 1995


POMS correlated with multiple scales, including social support, social isolation, and health perceptions (p < 0.0001). Factor analysis of the neuropsychologic test battery yielded six factors: mood, verbal memory, psychomotor, language, visuospatial, and cognitive inhibition domains (61). The mood factor (derived largely from POMS scales) correlated significantly with all the HRQOL scales (r = - 0.20 to r = -0.73; all p-values < 0.0025). Mood related much more strongly than neuropsychologic tests to HRQOL measures, including self-reported cognitive functioning (62). The psychomotor neuropsychology factor (high loadings from Grooved Peg- board and Symbol-Digit Modalities Test) correlated significantly with most of the HRQOL scales (r = -0.20 to r = -0.27; all p-values < 0.001). Al- though some QOLIE-89 scales had significant but modest correlations with the other four neuropsychological factors; these factors had much weaker correlations to the QOLIE-89 scales than the mood and psychomotor factors.

Correlations of toxicity, health care utilization, educational attainment, and employment measures and HRQOL

Correlation of HRQOL factor scores and the overall score with toxicity ratings, health care utilization, education, and employment are summarized in Table 4. Correlations between all four factor scores and overall scores for neurotoxicity (p < 0.0001) and systemic toxicity (p < 0.05 to < 0.0001) were negative. The strongest negative correlations were between these scores and the neurotoxicity scale (r-values ranged from - 0.28 to - 0.44). There were modest negative correlations between the indicators of health care utilization (number of doctor visits and days hospitalized in the preceding year) and the overall score, the epilepsy-targeted factor,

the cognitive factor, and the physical health factor (all p-values < 0.05 except between the physical health factor and the number of days hospitalized). There were modest correlations (range 0.15-0.28) between education or employment and the overall score, the cognitive factor, and the physical health factor (all p-values < 0.05). In contrast, there were no significant correlations between any of these scores and number of AEDs (all p-values > 0.05).

Relationship between epilepsy severity and HRQOL Patients with few or no seizures in the preceding

year showed a consistent pattern in reporting better HRQOL than patients having more frequent seizures (binomial probability, p < 0.00001) (Table 5). The epilepsy-targeted factor and three of four scales that comprise this factor (seizure worry, work/ driving/social function, and health discouragement) had the highest relative validities of all scales and factor scores to the criterion variable of seizure severity. The relative validities ranged from l l .4 for seizure worry to 18.6 for health discouragement. The pain scale was the reference scale (relative validity = 1) and was leas: sensitive to differences in epilepsy severity. Completely seizure-free patients scored higher than patients having the most frequent seizures on overall score, all four factor scores, and eight (seizure worry, health discouragement, work/driving/social function, attention/ concentration, role limitations-emotional, social isolation, energylfatigue, and health perceptions) of the 17 scales (alpha = 0.05, Duncan's multiple- range test).

Additional items and modifications A visual analog item on overall health evaluation

and an item on satisfaction with sexual function

TABLE 4. Correlation coef3cients between QOLIE-89 factor and overall scores and toxicity ratings, health care utilization, education and employmenta

Variable

Neurotoxicity rating Systemic toxicity rating No. of medications' No. of doctor visits in previous year Days hospitalized in previous year Educatiotr' Employ men$

Overall score

-0.44' -0.25" -0.06 - 0.18' - 0. 13d

0.19' 0.28'

Epileps y-targeted

-0.28' -0.25' - 0.05 - 0.17' -0.13d

0.07 0.20'

Cognitive Mental health Physical health

-0.39' -0.31' - 0.19' - 0. 14d -0.01 -0.06 -0.12d -0.04 -0.12d - 0.10

0.15' 0.10 0.22' 0.12

-0.39' - 0.16' -0.06 -0.22' -0.07

0.25' 0.27'

QOLIE, Quality of Life in Epilepsy. a n = 251-299 patients. ' p < 0.0001.

p < 0.001. p < 0.05. p < 0.01. Spearman correlations were calculated for these variables because they are ordinal.



TABLE 5 . Mean QOLIE-89 overall score, factor and individual scale scores, and seizure severity and frequency

Seizure-specific categorization of epilepsy severity"

QOL measure

Overall score Factor scorese

Epileps y-targeted Cognitive Mental health Physical health

Seizure worry Medication effect Health discouragement WorWdrivinglsocial function Language Attentiodconcentration Memory Overall QOL Emotional well-being Role limitations: emotional Social isolation Social support Energylfatigue Health perceptions Physical Role limitations: physical Pain

Scale scores

Seizure - f r e e (>I yr), Low, Moderate, High, n = 21 n = 116 n = 136 n = 31

76.8'

0.56' 0.42" 0.47' 0.32'

74.9' 56.8' 90.0' 77.7' 81.0' 80.5' 61.0' 72.0' 73.4" 90.0' 86.5' 80.3' 63.0' 77.7' 91.5" 73 .O' 76.9'

70.0d

0.23' 0.07'.d 0.02d 0.06'.d

61.4d 59.8' 76.3d 73.1' 75.2' 71.4d 56.7' 68.1' 68.5' 68.8d 77.5c.d 72.5' 53.6'td 69.8c.d 85.0' 72.1' 77.0'

65.2d

- 0.22d -0.03''d -0.07d - 0. 12d

53. Id 52.7" 63.y 62.0d 72.5' 67.3d 5 1.4' 65.8' 64.9' 66.6d 76.0c,d 71.3' 56.4c,d 65.6d 83.8' 63.8' 73.5'

65.Sd

-0.29d -0.13d -0.12d - 0.02c.d

57.4d 49.1' 61.9 58.2d 75.6' 69.4d 52.9' 65.9' 68.1' 68.4d 71.6d 72.2' 50Sd 66.2d 87.9' 66.5' 75.7'

Unadjusted Relative validityx F-ratiob

4.6 9.2

7.6 15.2 1.8 3.6 1.8 3.6 1.5 3.0

5.7 11.4 1.7 3.4 9.3 18.6 8.4 16.8 0.9 1.8 2.2 4.4 1.6 3.2 0.9 1.8 1.9 3.8 2.8 5.6 1.5 3.0 0.6 1.2 1.8 3.6 3 .O 6.0 1 .o 2.0 1.6 3.2 0.5 1 .o

QOLIE, Quality of Life in Epilepsy; HR, health related. Criteria for classification as in Table 1. One-way between-group analysis of variance (unadjusted) of HRQOL scale and epilepsy severity categories.

c,d and Means within a row with different superscripts differ significantly (p < 0.05, Duncan multiple-range test). Largest means andf. If appears above all means in a row, there are no significant differences between the means. Mean denoted by followed by

scores are adjusted for gender, education, age, and present comorbidity. Factor scores were determined from results of factor analysis. Reference scale pain.

were added after field testing. The sexual function item was added because this is often a problem among men and women with epilepsy (63,64) and can be a side effect of some AEDs (2). The overall health evaluation item, adapted from the EuroQol measure (65), was included to have a simple, patient preference summary measure.

Overall score Factor scores, which weight each scale by its fac-

tor loading, were derived for each of the four factor domains. Weights for each scale were derived as described in the analysis plan. The overall score is calculated by summing the product of each scale score times its weight. Overall scores range from 26 to 95 of 100 points, with a mean of 68 points (higher scores indicate better HRQOL). For the overall score, internal consistency reliability was 0.97 and test-retest reliability (product-moment correlations) was 0.88.

The final measure consists of 87 items from the original study questionnaire (86 items distributed into 17 multi-item scales and the single item on change in health) and the 2 additional items on sex-

ual function and overall health (added based on patient responses and expert opinion) (Appendix l).

DISCUSSION

The QOLIE-89 is designed to assess HRQOL in a broad spectrum of adult epilepsy patients. It con- tains 17 scales assessing both general and epilepsy- targeted aspects of physical and mental health, and social and role functioning. Reliability and construct validity were supported by the data collected from 304 patients and their proxies throughout the United States.

Our results demonstrate good internal consistency with reliability coefficients (Cronbach's alpha) >0.70 (the standard for group comparisons) for all scales. Multitrait scaling analysis provided strong support for item discrimination and consistency of scale composition. The factor analysis showed a four-factor solution: mental health, physical health, cognitive, and epilepsy-targeted. The overall QOL scale loaded on mental health, as has been shown with both generic HRQOL instruments and the ESI-55 (22,66).

Epilepsia, Vol. 36, No. 11. 1995


The greatest challenge in developing HRQOL instruments is validation, as there is no gold standard for comparison. To assure content validity (i.e., the extent to which an inventory samples a face-valid representative range of function and perceptions of well-being in the group studied), we supplemented a generic core by adding items relevant to patients with mild to moderate severity epilepsy, based on patient interviews, clinical experience and literature review.

Based on patient responses to open-ended questions, possible areas for revision of the QOLIE-89 include concerns about sleep and pregnancy and reproductive issues. Personal finances were also important issues for people with epilepsy, but these areas are encompassed by general QOL and fall outside the purview of health-related QOL.

The establishment of construct validity of a health status measure is the accumulation of a body of evidence that demonstrates a relationship between the HRQOL measure under study and variables that are hypothesized to be related to HRQOL or also to measure HRQOL (67,68). We used three types of data to assess construct validity for the QOLIE-89 scales: clinical data (seizure data and adverse medication effects), social demographic data (employment and education), proxy data, and neuropsychologic test data.

The significant, but generally modest, correlation between patient and proxy responses in all 17 scales provides additional support for construct validity, consistent with studies of patients having other chronic diseases (69). Although all scales had correlations significant at p < 0.0001, the correlations were as low as 0.29 (role limitations: emotional), suggesting a modest relationship between how patients and their family or close friends perceive their QOL. Careful study of differences as well as agree- ments between patient and proxy measures may provide new insights into the utility of proxy reports of HRQOL and will be the subject of an indepen- dent report.

The neuropsychologic test data, including examiner-administered tests and the self-administered POMS, support the validity of the mood and cognitive self-report constructs of the QOLIE-89. All the POMS scales correlated significantly with the overall QOL, emotional well-being, role functioning- emotional, and energy/fatigue QOLIE-89 scales. QOLIE-89 attention/concentration, memory, and language scales had modest but statistically significant correlations with selected neuropsychological tests in respective areas of attention, memory, and language.

Neurotoxicity, and to a lesser extent systemic

toxicity, of AEDs had highly significant negative correlations with all four QOLIE-89 factor scores and the overall QOLIE-89 score. These findings support the relationship of this measure to measures specific for AED toxicity. Neurotoxicity is associated with greater impairment in self-reported HRQOL than systemic toxicity.

QOLIE-89 measures negatively correlated with health care utilization, including number of doctor visits and days hospitalized in the preceding year. The inverse relation between health care utilization and QOL may result from increased severity of the disorder and the emotional and financial impacts of seeking and receiving health care. HRQOL has also been shown to predict health care utilization (70). Education and employment status both showed positive correlations with QOLIE-89 cognitive and physical health factor scores and with the QOLIE- 89 overall score. These results are consistent with previous studies of socioeconomic determinants of health, demonstrating that individuals with more education and higher income are in better health (71).

For all QOLIE-89 scales, there was a trend for patients with less severe epilepsy to have higher QOL scores as compared with patients with more severe epilepsy. The relative validities (Table 5 ) showed that the epilepsy-targeted factor and the seizure worry, health discouragement, and work/ driving/social function scales were the most power- ful discriminators between groups of different epilepsy severity. The importance of the relative validity analysis is its implications for sample size. Based on this analysis, to discriminate across levels of seizure severity on health discouragement would require a sample size Yidh the size for the pain scale (54). These findings strongly support the addition of the epilepsy-targeted supplement, particularly these three scales, to the generic HRQOL measure.

The less striking relationship between epilepsy severity and physical health, mental health, and cognitive factors (and their component scales) may have several explanations. In many instances, there was a trend for a significant difference that might have been identified had the seizure-free and high seizure frequency groups contained more subjects. For almost every factor and scale, there was a consistent decline in HRQOL scores from seizure-free to low to moderate to high epilepsy severity groups. However, for many of the areas in which significant intergroup differences were not detected, as well as areas in which significant trends were identified, factors other than seizure frequency and type (e.g., AEDs, stress within the family) influence HRQOL. In addition, the particular categories we established



for this epilepsy severity classification system did not have an empirical basis, although they made clinical sense. Therefore, the modest correlations with some aspects HRQOL may reflect the limitations of one’s ability to specify levels of seizure severity that most highly reflect HRQOL of these patients, without such empirical data.

Epilepsy is the paradigm of a disorder in which QOL issues assume paramount importance. Sei- zures are infrequent and brief, while the social and psychological sequelae of seizures and epilepsy and adverse medication effects are persistent. HRQOL assessment can direct research to identify medical and other therapies that have positive impact on QOL (72); e.g., several AEDs may have similar efficacy in terms of seizure control, yet may have differential impact on HRQOL. Such studies

changed the pyramid of antihypertensive therapy, displacing beta blockers to a higher level therapy (73).

The strengths and weaknesses of the QOLIE-89 will be identified through its use. Strengths may include comprehensiveness and the large number of domains included and the ability to derive scale, factor, and overall scores. Overall reliability of the measure is good, and preliminary evidence supports construct validity. Future methodologic research using the QOLIE-89 should address the following: (a) ability to discriminate between treatment groups; (b) responsiveness to change; (c) comparison of the overall score derived in this study and alternative, patient-derived weights for the overall score (e.g., time tradeoff or standard gamble); and (d) simplifying the scoring procedure.

APPENDIX 1

QUALITY OF LIFE IN EPILEPSY QOLIE-89 (Version 1.0) Patient Inventory

Circle one number (between 1 and 5) *l . In general, would you say your health is?

2. Overall, how would you rate your quality of life (IW)?; 10 = Best Possible Quality of L i f e 4 = Worst Possible Quality of Life (as bad as or worse than being dead) *3. Compared to 1 year ago, how would you rate your health in general now (1-5)? 1 = Much better-5 = Much worse *4-13. The following questions are about activities you might do during a typical day. Does your health limit you in these activities? If so, how much? (Circle 1, 2, or 3 on each line.) 1 = Yes, limited a lot, 2 = Yes, limited a little, 3 all

*4. Vigorous activities, such as running, lifting heavy objects, participating in strenuous sports *5. Moderate activities, such as moving a table, pushing a vacuum cleaner, bowling, or playing golf *6. Lifting or carrying groceries *7. Climbing several flights of stairs *8. Climbing one flight of stairs *9. Bending, kneeling, or stooping *lo. Walking more than one mile *I l . Walking several blocks *12. Walking one block *13. Bathing o r dressing yourself 14-18. During the past 4 weeks, have you had any of the following difficulties with your regular daily activities or work as a result of any physical problems? (Please answer YES or NO for each question by circling 1 or 2 on each line.) 1 = Yes, 2 = No *14. Cut down on the amount of time you spend on work or other activities *15. Accomplished less than you would like *16. Were limited in the kind of work or other activities you do *17. Had d#culty performing the work or other activities you do (for example, it took extra effort) 18. Did your work or other activities less carefully than usual 19-23. During the past 4 weeks, have you had any of the following difficulties with your regular daily activities or work as a result of any emotional problems (such as feeling depressed or anxious)? 1 = Yes, 2 = No *19. Cut down on the amount of time you spent on work or other activities *20. Accomplished less than you would like 21. Were limited in the kind of work or other activities you do 22. Had difJiculty performing the work or other activities you do (for example, it took extra effort) *23. Did work or other activities less carefully than usual $24. How much bodily pain have you had during the past 4 weeks? (Circle one number on a scale ranging from I [None] to [Very severe] .) *25. During the past 4 weeks, how much did bodily pain interfere with your normal work (including both work outside the home and housework)? (Circle one number on a scale ranging from 1 [Not at all] to 5 [Extremely].) *26. During the past 4 weeks, to what extent have your physical healt or emotional problems interfered with your normal social activities with family, friends, neighbors, or groups? (Circle one number on a scale ranging from 1 [Not at all] to 5 [Extremely].)

1 = Excellent-2 = Very good-3 = G o o d 4 = Fair-5 = Poor

No, not limited at



*27-35. These questions are about how you FEEL and how things have been for you during the past 4 weeks. How much of the time during the past 4 weeks . . . (Circle one number on a scale ranging from 1 [All of the time] to 6 [None of the time].)

*27. Did you feel full of pep? *28. Have you been a very nervous person? *29. Have you felt so down in the dumps that nothing could cheer you up? *30. Have you felt calm and peaceful? *3 1. Did you have a lot of energy? *32. Have you felt downhearted and blue? *33. Did you feel worn out? *34. Have you been a happy person? *35. Did you feel tired?

3643. How much of the time during the past 4 weeks of the time] .) 36. Has your epilepsy limited your social activities (such as visiting with friends or close relatives)? 37. Have you had difficulty concentrating and thinking? 38. Did you have trouble keeping your attention on an activity for long? 39. Were you discouraged by problems related to your health? 40. Have you worried about having another seizure? 41. Did you have difficulty reasoning and solving problems (such as making plans, making decisions, learning new things)? 42. Were you discouraged by your epilepsy-related problems? *43. Have your physical health or emotional problems interfered with your social activities (like visiting with friends, relatives, etc)? 44-48. Please choose the answer that best describes how TRUE or FALSE each of the following statements is for you . . . on a scale ranging from 1 (Definitely true) or 5 (Definitely false). *44. I seem to get sick (any kind of sickness) a little easier than other people. *45. I am as healthy as anybody I know. *46. I expect my health to get worse. *47. My health is excellent. 48. When there is an illness going around, I usually catch it. 49. How has the QUALITY OF YOUR LIFE been during the past 4 weeks (that is, how have things been going for you)? (Dartmouth

(Circle one number on a scale ranging from 1 [All of the time] to 6 [None

COOP Chart: 1-5) 1 = Very well: could hardly be better-5 = Very bad: could hardly be worse

50. In the past 4 weeks, have you had any trouble with your memory? (Circle one number between 1 and 4.) 1 = Yes, a great d e a l 4 = No, not at all 51-54. Circle one number on a scale ranging from 1 (All of the time) to 6 (None of the time) for how often in the past 4 weeks you have had trouble remembering or how often these memory problems have interfered with your normal work or living.

Choose one for #51-54

1 = All of the time-2 = Most of the time-3 = A good bit of the t i m e 4 = Some of the time-5 = A little of the t i m e 4 = None of the time

51. Names of people 52. Where you put things 53. Things people tell you 54. Things you read hours or days before

55-59. The following questions are about LANGUAGE problems you may have. Circle one number on a scale ranging from 1 (All of the time) to 6 (None of the time) for how often in the past 4 weeks you had trouble speaking or how often these problems interfered with your normal work or living.

Choose one for #55-59

1 = All of the time-2 = Most of the time-3 = A good bit of the t i m e 4 = Some of the time-5 = A little of the t i m e 4 = None of the time

55. Finding the correct work 56. Understanding what others are saying in conversation 57. Understanding directions 58. Understanding what you read 59. Writing

60-64. The following questions are about CONCENTRATION problems you may have. Circle one number on a scale ranging from 1 (All of the time) to 6 (None of the time) for how often in the past 4 weeks you had trouble concentrating or how often these problems interfered with your normal work or living.

I

Choose one for # 6 W 1 I

I 1 = All of the time-2 = Most of the time-3 = A good bit of the t ime-4 = Some of the time-5 = A little of the t i m e 4 = None of the time

60. Concentrating on conversations 61. Concentrating on a task or job 62. Concentrating on reading



63. Concentrating on doing one thing at a time 64. How often do you feel you react slowly to things that are said or done?

65-68. The following questions are about problems you may have with certain ACTIVITIES. Circle one number on a scale ranging from 1 (A great deal) to 5 (Not at all) for how much during the past 4 weeks your epilepsy or antiepileptic medication has caused trouble with . . . Choose one for #65-68 I = A great deal-2 = A lot-3 = Somewhat4 = Only a little-5 = Not at all 65. Working 66. Friendships and relationships (romantic) 67. Leisure time (such as hobbies, going out) 68. Driving

69-73. The following questions relate to the way you FEEL about your seizures. (Circle one number on each line.)

69. How fearful are you of having a seizure during the next month? 1 Very fearful 2 Somewhat fearful 3 Not very fearful 4 Not fearful at all 70. Do you worry about hurting yourself during a seizure? 1 Worry a lot 71. How worried are you about embarrassment or other social problems resulting from having a seizure during the next month? 1 Very worried 2 Somewhat worried 3 Not very worried 4 Not worried at all 72. How worried are you that medications you are taking will be bad for you if taken for a long time? 1 Very worried 2 Somewhat worried 3 Not very worried 4 Not worried at all 73. How well do you do with complicated projects that require organization or planning? Circle one number on a scale ranging from 1 (Very poorly) ro 5 (Very well).

7GO. For each of these PROBLEMS, circle one number for how much they bother you on a scale of 1-5, where

2 Occasionally worry 3 Don’t worry at all

Not at all bothersome .................................................................................................................... Extremely bothersome, 1 2 3 4 5

74. Seizures 75. Memory difficulties 76. Driving limitations 77. Work limitations 78. Social limitations 79. Physical effects of antiepileptic medication 80. Mental effects of antiepileptic medication

81-83. In terms of your satisfaction with your family and social life, circle one number to indicate the following: I Poor 2 Fair 3 Good 4 Very good 5 Excellenmt

81. The amount of togetherness you have with your family and/or friends 82. The support and understanding your family and/or friends give each other 83. The amount you talk things over with your family and/or friends

8488. In terms of your satisfaction with your family and social life, circle one number between I and 5) to indicate the following:

84. Overall, how satisfied were you with your sexual relations during the past 4 weeks? Circle one number on a scale ranging from I to 5. 1 Very satisfied 2 Somewhat satisfied 3 Neither satisfied nor dissatisfied 4 Somewhat dissatisfied 5 Very dissatisfied 85. How limited are your social activities compared with others your age because of your epilepsy or epilepsy-related problems? Circle one number on a scale ranging from 1 to 5. 1 Much more limited 2 Somewhat more limited 3 About the same 4 Somewhat less limited 5 Much less limited 86. During the past 4 weeks, was someone available to help you if you needed and wanted help? Circle one number on a scale ranging from 1 to 5. 1 Yes, as much as I wanted 87. How much of the time during the past 4 weeks did you feel left out? Circle one number on a scale ranging from 1 to 6. 1 All of the time the time 88. During the past 4 weeks, how often did you feel isolated from others? Circle one number on a scale ranging from 1 to 6. 1 Always 2 Very often 3 Fairly often 4 Sometimes 5 Almost never 6 Never

89. How good or bad do you think your health is? On the thermometer scale below, the best imaginable state of health is 100 and the worst imaginable state is 0. Please indicate how you feel about your health by circling one number on the scale. Please consider your epilepsy as part of your health when you answer this question. (Thermometer figure: 0, 10, 20, 30, 40, 50, 60, 70 $0, 90, 100.)

Scale/Item Numbers

2 Yes, quite a bit 3 Yes, some 4 Yes, a little 5 No, not at all

2 Most of the time 3 A good bit of the time 4 Some of the time 5 A little of the time 6 None of

Comments (if any) ..........................................................................................................................................................

Health perceptions: 1, 4 4 4 8 Overall quality of life: 2, 49 Physical function: 4-13 Role limitations-physical: 1618 Role limitations-emotional: 19-23 Pain: 24, 25 Work/driving/social function: 26, 36, 43, 65-68, 76-78, 85 Energy/fatigue: 27, 31, 33, 35



Emotional well-being: 28-30, 32, 34 Attentiodconcentration: 37, 38, 41, 60-64, 73 Health discouragement: 39, 42 Seizure worry: 40, 69-71, 74 Memory: 50-54, 75 Language: 55-59 Medication effects: 72, 79, 80 Social support: 81-83, 86 Social isolation: 87, 88

Change in health: 3 Sexual relations: 84 Overall health: 89

Single Items

* Items derived from the RAND 36-item Health Survey (also termed SF-36).

Acknowledgment: This work was supported by an un- restricted educational grant from Wallace Laboratories.

We thank Karen Spritzer and Eric Wang for assistance in computer programming for data analysis and Debra Weiner and Jackie Gordon for administering the educational grant from Wallace Laboratories. The following sites and individuals (principal investigators in parenthe- ses) participated in this study: University of Alabama, School of Medicine, Birmingham, AL (Dr. R. Edward Faught, Jr.); Department of Neurology/Epilepsy Center, Barrow Neurological Institute, Phoenix, AZ (Dr. Robert S. Fisher); Arizona Comprehensive Epilepsy Program, University of Arizona, Arizona Health Sciences Center, Tucson, AZ (Dr. David M. Labiner); CNI Epilepsy Cen- ter, Colorado Neurological Institute, Englewood, CO (Dr. Ronald E. Kramer); Epilepsy Program, Yale Univer- sity School of Medicine, New Haven, CT (Dr. Susan Spencer); Epilepsy Center, Veterans Affairs Medical Center, West Haven, CT (Dr. Richard H. Mattson); Com- prehensive Epilepsy Center, University of Miami, Miami, FL (Ruth E. Nemire, Pharm.D.); Emory University Clinic, Department of Neurology, Atlanta, GA (Dr. Rob- ert C. Green); Department of Neurology, Medical College of Georgia, Augusta, GA (Dr. Kimford J. Meador); De- partment of Psychology and Social Sciences, Rush- Presbyterian-St. Luke’s Medical Center, Chicago, I L (Dr. Christopher L. Grote); Comprehensive Epilepsy Center, Beth Israel Hospital, Boston, MA, Epilepsy Cen- ter (Dr. Steven C. Schachter); Department of Neurology, New England Medical Center, Boston, MA (Dr. Edward B. Bromfield); Comprehensive Epilepsy Center, Univer- sity of Massachusetts Medical Center, Worcester, MA (Dr . Catherine Phillips); Neuropsychology Program and Comprehensive Epilepsy Program, University of Michi- gan Medical Center, Ann Arbor, MI (Dr. Stanley Berent); The Minnesota Epilepsy Group, P.A., St. Paul, MN (Dr. Frank J. Ritter); Comprehensive Epilepsy Center, NYU Medical Center, Hospital for Joint Diseases, New York, NY (Dr. Orrin Devinsky); Comprehensive Epilepsy Pro- gram, University of Rochester, Rochester, NY (Dr. John Langfitt); Epilepsy Center, State University of New York at Stony Brook, Stony Brook, NY (Dr. Alan Ettinger); Epilepsy Treatment Center, University of Cincinnati Medical Center, Cincinnati, OH (Dr. Michael D. Privit- era); Department of Neurology, Section of Epilepsy and Sleep Disorders, The Cleveland Clinic Foundation, Cleveland, OH (Dr. Susan Stagno); Oregon Comprehen- sive Epilepsy Program, Good Samaritan Hospital and Medical Center, Portland, OR (Dr. Mark S. Yerby); Com- prehensive Epilepsy Center, Graduate Hospital, Philadel-

phia, PA (Dr. Jacqueline French); Neurosciences Re- search Institute, Allegheny-Singer Research Institute, Al- legheny Campus, Medical College of Pennsylvania, Pittsburgh, PA (Dr. Michael E. Mabry); Epi-Care Center, Baptist Memorial Hospital, Semmes-Murphey Clinic and University of Tennessee, Memphis, TN (Dr. Georgia Montouris); and Epilepsy Center, George Washington University Hospital, Washington, DC (Dr. John DeTo- ledo) .

REFERENCES

1 . Annegers JF. Epidemiology and genetics of epilepsy. Neurol Clin 1994;12:15-29.

2. Mattson RH, Cramer JA, Collins JF, et al. Comparison on carbamazepine, phenobarbital, phenytoin and primidone in partial and secondarily generalized tonic clonic seizures. N Engl J Med 1985;313:145-51.

3. Mattson RH, Cramer JA, Collins JF, and the Department of Veterans Affairs Epilepsy Cooperative Study No. 264 Group. A comparison of valproate with carbamazepine for the treatment of partial seizures and secondarily generalized tonic-clonic seizures in adults. N Engl J Med 1992;327:765-

4. Chadwick D, for the Medical Research Council Antiepileptic Drug Withdrawal Study Group. Randomized study of antiepileptic drug withdrawal in patients in remission. Lancer 1991;337:1175-80.

5. Jacoby A. Epilepsy and the quality of everyday life: findings from a study of people with well-controlled epilepsy. Soc Sci Med 1992 ;34: 657-66.

6. Ryan R, Kempner K, Emlen AC. The stigma of epilepsy as a self-concept. Epilepsia 1980;21:433-44.

7. Hermann BP. The relevance of social factors to adjustment in epilepsy. In: Devinsky 0, Theodore WH, eds. Epilepsy and behavior. Wiley-Liss, New York, 1991:23-36.

8. Hayden M, Penna C, Buchanan N. Epilepsy: patient perceptions of their condition. Seizure 1992;1:191-7.

9. National Institutes of Health Consensus Conference, Sur- gery for Epilepsy. JAMA 1990;264:729-33.

10. Ware JE, Sherbourne CD. A 36-item short form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30:473-83.

1 1 . Hays RD, Sherbourne C , Maze1 E. The RAND 36-Item Health Survey 1.0. Health Econ 1993;2:212-27.

12. Bergner M, Babbit R, Pollard W, et al. The Sickness Impact Profile: development and final revision of a health status measure. Med Cave 1981;19:787-805.

13. Nelson EC, Benvick DM. The measurement of health status in clinical practice. Med Care 1989;27:S77-90.

14. Patrick DL, Deyo RA. Generic and disease-specific measures in assessing health status and quality of life. Med Care

15. Baker GA, Smith DF, Dewey M, Morrow J, Crawford PM, 1989;27(~~ppl3):S217-32.



16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29 I

30.

31.

32.

Chadwick DW. The development of a seizure severity scale as an outcome measure in epilepsy. Epilepsy Res 1991;8:

Baker GA, Smith DF, Dewey M, Jacoby A, Chadwick DW. The development of a health-related quality of life measure for patients with intractable epilepsy. Epilepsy Res 1993;16:

Jacoby A. Felt versus enacted stigma: a concept revisited. Evidence from a study of people with epilepsy in remission. SOC Sci Med 1994;38:269-74. Jacoby A, Baker GA, Smith DF, Dewey M, Chadwick DW. Measuring the impact of epilepsy: the development of a novel scale. Epilepsy Res 1994b;16:83-8. Smith DF, Baker GA, Davies G, et al. Outcomes of add-on treatment with lamotrigine in partial epilepsy. Epilepsia 1993 ;34: 3 12-22. Devinsky 0, Baker G, Cramer J. Health-related quality of life scales to assess epilepsy. In: Engel Jr. J, Pedley TA, eds. Epilepsy: a comprehensive textbook. New York: Raven Press (in press). Kendrick AM, Trimble MR. Repertory grid in the assessment of quality of life in patients with epilepsy: the quality of life assessment schedule. In: Trimble MR, Dodson WE, eds. Epilepsy and quality of life. New York: Raven Press, 1994: 151-64. Vickrey BG, Hays RD, Graber J, Rausch R, Engel J, Brook RH. A health-related quality of life instrument for patients evaluated for epilepsy surgery. Med Care 1992;30:299-3 19. Nelson EC, Landgraf JM, Hays RD, Wasson JH, Kirk JW. The functional status of patients: how can it be measured in physicians’ offices? Med Care 1990;28: 11 11-26. Hadorn D, Hays RD. Multitrait-multimethod analysis of health-related quality of life preferences. Med Care 1991 ;29: 82940. Andrews FM, Withey SB. Social indicators of well-being: Americans’ perception of life quality. New York: Plenum Press, 1976. Chaplin JE, Yepez R, Shorvon S, Floyd M. A quantitative approach to measuring the social effects of epilepsy. Neu- roepidemiology 1990;9: 151-8. Thompson P, Oxley J. Socioeconomic accompaniments of severe epilepsy. Epilepsia 1988;29(suppl 1):9-18. Levin R. Psychosocial dimensions of epilepsy: a review of the literature. Epilepsia 1988;29:805-16. Hermann BP. Quality of life in epilepsy. J Epilepsy 19925: 153-65. West P. An investigation into the social construction and consequences of the label epilepsy. Sociol Rev 1979;27:719- 41. Upton D, Thompson PJ. Effectiveness of coping strategies employed by people with chronic epilepsy. J Epilepsy 1992; c.11a 77

245-5 1.

65-8 1.

2. I 17-L I .

, Thompson P, Upton D. Living with intractable epilepsy: impact on family members. Epilepsia 1991 ;32(suppl 1):42.

33. Jacoby A. Felt versus enacted stigma: a concept revisited. SOC Sci Med 1994;38:269-74.

34. Fraser RT, Trejo W, Temkin N, Clemmons D, Dodrill C. Assessing the vocational interests of those with epilepsy. Rehab Psychol 1985;20:29-33.

35. Andermann F, Remillard GM, Zifkin BG, Trottier RG, Drouin P. Epilepsy and driving. Can J Neurol Sci 1988;15:

36. Schachter SC, Shafer PO, Murphy W. The personal impact of seizures: Correlations with seizure frequency, employment, cost of medical care, and satisfaction with physician care. J Epilepsy 1993;6:22+7.

37. Lennox WG. Brain injury, drugs, and environment as causes of mental decay in epilepsy. A m J Psychiatry 1942;99:174- 80.

38. Smith DB, Mattson RH, Cramer JA, et al. Results of a na- tionwide Veterans Administration cooperative study comparing the efficacy and toxicity of carbamazepine, pheno-

371-7.

barbital, phenytoin, and primidone. Epilepsia 1987;28(suppl 3):S50-8.

39. Dodrill CB. Interictal cognitive aspects of epilepsy. Epilep- sia 1992;33(suppl 6):S7-10.

40. Trimble MR. Anticonvulsant drugs and cognitive function: a review of the literature. Epilepsia 1987;28(suppl 3):S3745.

41. Novelly RA, Augustine EA, Mattson RH, et al. Selective memory improvement or impairment in temporal lobectomy for epilepsy. Ann Neurol 1984;15:64-7.

42. Stewart AL, Sherbourne CD, Hays RD, et al. Summary and discussion of MOS measures. In: Stewart AL, Ware JE, eds. Measuring functioning and well-being: the medical outcomes study approach. Durham, NC: Duke University Press, 1992:345-71.

43. DeVellis RF, DeVellis BM, Wallston BS, Wallston KA. Ep- ilepsy and learned helplessness. Basic Appl SOC Psychol 19803 :241-53.

44. Mittan R. Fear of seizures. In: Whitman S, Herman B, eds. Psychopathology in epilepsy. New York: Oxford University Press, 1986:9&121.

45. Baker GA, Smith DF, Dewey M, Morrow J, Crawford PM, Chadwick DW. The development of a seizure severity scale as an outcome measure in epilepsy. Epilepsy Res 1991;8: 245-5 1.

46. Duncan JS, Sander JWAS. The Chalfont Seizure Severity Scale. J Neurol Neurosurg Psychiatry 1991 ;54:873-6.

47. Hays RD, DiMatteo MR. A short-form measure of loneliness. J Personality Assess 1987;51:69-81.

48. Sherbourne CD, Stewart AL. The MOS social support survey. SOC Sci Med 1991;32:705-14.

49. Perrine K. A new quality of life inventory for epilepsy patients: interim results. Epilepsia 1993;34(suppl 4):S28-33.

50. Lezak MD. Neuropsychological assessment, 2nd ed. New York, Oxford Press, 1983:239-622.

51. Cramer JA, Smith DB, Mattson RH, Delgado-Escueta AV, and the VA Epilepsy Cooperative Study No. 118 Group. A method of quantification for the evaluation of antiepileptic drug therapy. Neurology 1983;33(suppl I):26-37.

52. Hays RD, Hayashi T. Beyond internal consistency reliability: rationale and user’s guide for the Multitrait Analysis Program on the microcomputer. Behav Res Method Znstr Comput 1990;22: 167-75.

53. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika 1951;16:297-334.

54. Liang MH, Larson MG, Cullen KE, Schwartz JA. Compar- ative measurement and efficiency and sensitivity of five health status instruments for arthritis research. Arthritis Rheum 1985;28:542-7.

55. Vickrey BG, Perrine KR, Hays RD, et al. QOLIE-89 (version 1.0) scoring manual. Santa Monica, CA: RAND, 1993.

56. Vickrey BG, Hays RD, Spritzer K. Methodologic issues in quality of life assessment for epilepsy surgery. In: Chadwick DW, Baker GA, Jacoby A, eds. Quality of life and quality of care in epilepsy: update 1993. London: Royal Society of Medicine, Round Table Series, 1993;3 1 :27-37.

57. Nunnally J. Psychometric theory, 2nd ed. New York: Mc- Graw Hill, 1978.

58. Cattell RB. The scree test for number of factors. Multivari- ate Behav Res 1966;1:245-76.

59. Guttman LA. Some necessary conditions for common factor analysis. Psychometrika 1954;19: 149-61.

60. Montanelli RD, Humphreys LG. Latent roots of random data correlation matrices with squared multiple correlations on the diagonal: a Monte Carlo study. Psychometrika 1976; 41:341-8.

61. Hendrickson AE, White PO. PROMAX: a quick method for rotation to oblique simple structure. Br J Stat Psychol 1964; 17 : 65-70.

61a. Hays RD, Vickrey B, Hermann BP, et al. Agreement between self reports and proxy reports of quality of life in epilepsy patients. Quality Life Res 1995;4:159-68.

Epilepsia, Vol. 36, NO. 11, 1995


62. Perrine K, Hermann BP, Meador KJ, et al. The relationship of neuropsychological functioning to quality of life in epilepsy. Arch Neurol (in press).

63. Cramer JA, Jones EE. Reproductive function in epilepsy. Epilepsia 1991 ;32(suppl 6):S19-26.

64. Morrell MJ. Sexual dysfunction in epilepsy. Epilepsia 1991; 32(suppl 6): S38-45.

65. Brazier J, Jones J, Kind P. Testing the validity of the Euro- qol and comparing it with the SF-36 health survey questionnaire. Qua1 Life Res 1993;2:169-80.

66. Hays RD, Wasson JH, Nelson E. Measuring the health of older patients: use of the COOP charts. Presented at the meeting of the American Public Health Association. Chi- cago, IL, October 1989.

67. Jaeschke R, Guyatt GH. How to develop and validate a new quality of life instrument. In: Spilker B, ed. Quality of life assessments in clinical trials. New York: Raven Press, 1990: 47-57.

68. Shumaker SA, Anderson RT, Czajkowski SM. Psychologi- cal tests and scales. In: Spilker B, ed. Quality oflife assess-

ments in clinical trials. New York: Raven Press, 1990:95- 113.

69. Sprangers MAG, Aaronson NK. The role of health care pro- viders and significant others in evaluating the quality of life of patients with chronic disease: a review. J Clin Epidemiol 1992;45: 743-60.

70. Manning WG, Wells KB. The effects of psychological dis- tress and psychological well-being on use of medical services. Med Care 1992;30:541-53.

71. Mutcher JE, Burr JA. Racial differences in health and health care service utilization in later life: the effect of socioeconomic status. J Health SOC Behav 1991;32:342-56.

72. Baker GA, Smith DF, Dewey M, Jacoby, Chadwick DW. The initial development of a health-related quality of life model as an outcome measure in epilepsy. Epilepsy Res

73. Croog SH, Levine S, Testa M, et al. The effects of antihypertensive therapy on the quality of life. N Engl J Med 1986; 3 14: 1657-64.

74. Mosier CI. On the reliability of a weighted composite. Psy- chometrika 1943;8: 161-8.

1993 ; 16~65-8 1 .


development of the quality of life in epilepsy inventory

Documents