Abstracts

RATIONALE: Studies have been published which assessed health-related quality of life (HRQOL) in epilepsy, using different instruments. The information conveyed by these studies must be assessed in the light of the validity and comparability of the questionnaires used. Objective: To evaluate the instruments used for the assessment of HRQOL and to assess whether and to what extent they have been validated for use with epilepsy patients.
METHODS: We searched the Medline data-base for the years 1985-1999 for the key words epilepsy, quality of life and health status indicators. Articles were accepted if they were written in English and the patients[scquote] age was [gte]15 years. Excluded were letters, reviews, editorials, and articles dealing with costs, life satisfaction, nursing or clinical management of epilepsy. For each eligible paper we extracted the following items: source of the article, instrument(s) used, population surveyed, type of epilepsy, and study design. For validation studies we collected information on the reliability, validity, and responsiveness.
RESULTS: The literature search yielded 308 articles, 48 of which were eligible for review: 38 published in neuroscience, 4 in general medicine, and 6 in QOL journals. A total of 21 HRQOL instruments, generic or specific for epilepsy, were used. Three instruments (ESI-55, QOLIE-89, SF-36) were used in more than five studies and nine only once; 38 studies used one instrument, 5 two instruments, and 5 three or more instruments. Populations studied were: North-american (25 studies), British (9), Scandinavian (4), Japanese (3), and other (3); four studies investigated mixed populations. Most studies (24) included different types of epilepsy; 14 studies evaluated patients included in surgical programs. Out of 48 studies, 24 examined HRQOL in different target populations, 9 described the development and the validation of an instrument, 8 simply validated the instrument in the original language or in its translation, and 7 assessed HRQOL in the context of a clinical trial. Eleven instruments (both generic and specific) were validated for use in epileptic patients: while all instruments included measures of reliability (test-retest analysis, Chronbach[scquote]s [alpha], inter-rater agreement), and validity (content validity, construct or criterion validity), responsiveness (differences between means, t-tests, effect size) was tested only in five. Validated instruments were used in 68% of observational and in all clinical trials, but one.
CONCLUSIONS: Fairly valid and reliable questionnaires are generally used in clinical trials, but they may not be used in observational studies.