Quality of life pre and post epilepsy surgery

Abstract

The aim of this work was to assess the health-related quality of life (HRQL) of patients pre and post surgical treatment for epilepsy. A total of 145 patients were interviewed during their pre-surgical assessment on the telemetry unit, Queen Square. The HRQL assessment comprised the quality of life assessment schedule (QOLAS), the EuroQol EQ-5D and the epilepsy surgery inventory (ESI-55). A total of 40 patients were followed up, of which 22 had undergone surgery and achieved 75% or greater reduction in seizures. The QOLAS scores for the patients who achieved 75% or greater seizure reduction post-op were significantly lower (i.e. improved HRQL) compared to baseline. The descriptive data suggest that the EQ-5D may not be capturing all of the QOL issues of relevance to patients with chronic, intractable epilepsy and the EQ-5D may not be valid for this group. Most patients queried the visual analogue scale (VAS) which asks for an overall rating of the respondent’s self-perceived health. The most frequent comments, from 42% of patients, was that ‘health’ did not include their epilepsy. Despite this, the group whose seizures were reduced had significantly higher VAS scores at follow-up. We can conclude that the VAS is sensitive to clinical change. The baseline EQ-5D utility and follow-up scores were compared. There were no significant changes in QOL scores for either group. The patients who achieved 75% or greater reduction in seizures post-op scored significantly higher (i.e. better QOL) on 2/3 composite scores of the ESI-55 at follow-up. The QOLAS, the EQ-5D VAS and the ESI-55 were sensitive to clinically defined outcome. The results for the EQ-5D profile and the EQ-5D utility suggest that the EQ-5D is not a valid and responsive instrument for use in patients with intractable epilepsy.

Introduction

A number of recent studies have assessed health-related quality of life (HRQL) pre and post definitive surgical treatment for intractable epilepsy. The findings are complex and our current knowledge is limited by a lack of long-term studies, absence of standardised patient populations and paucity of pre- and post-operative comparisons using standardised QOL and seizure assessment instruments (Spencer, 1996). Other unresolved methodological issues include what percentage of seizure reduction is the most appropriate outcome measure and what is the most appropriate follow-up period. Whilst it has been demonstrated that post-operative seizure freedom is associated with significant improvements in QOL (Hermann et al., 1992, Kim and Kim, 1995, Kellett et al., 1997), a number of other seizure based outcomes have been used and the picture for different degrees of seizure reduction is less clear. Some researchers have chosen a 75% (Bladin, 1992, Hermann et al., 1992, Malgrem et al., 1997) and some a 90% reduction in seizures (Rose et al., 1996, McLachlan et al., 1997).

There is no agreed follow-up period for assessing QOL post-surgery. Researchers have chosen periods as diverse as 3 months (Kim and Kim, 1995), 6–8 months (Hermann et al., 1992), 1 year (Rose et al., 1996), 2 years (McLachlan et al., 1997) and 4 years (Malgrem et al., 1997).

Also, a number of QOL measures have been used and this diversity of instruments makes intra-study comparisons difficult. QOL measures used have included the ESI-55 (Vickrey et al., 1995a, Rose et al., 1996, McLachlan et al., 1997) the SF-36 (Malgrem et al., 1997) QOLIE-89 (Kim and Kim, 1995) the Liverpool battery (Kellett et al., 1997) the WPSI and the GHQ (Hermann et al., 1992).

Although a number of epilepsy-specific measures are now available (Hays, 1995, Cramer, 1996), recent papers have raised the question of the sensitivity and the face validity of the instruments (Gilliam et al., 1997, Leidy et al., 1998). It is argued that more data are needed on the instruments’ sensitivity to change and that many of the more established measures do not tap issues of concern to patients such as driving, independence and pregnancy/birth defects. Moreover, existing instruments used in epilepsy yield a profile score which cannot be aggregated into a single, overall score. The single index score, however, is required for cost-utility evaluations (Hays et al., 1996), a research area of growing importance as the costs of health care come under increasing scrutiny (Spilker, 1996). No measure of HRQL has emerged as ideal for QOL surgery and further psychometric testing of all currently available instruments is needed.

Given the issues of the validity and the responsiveness of QOL scales in epilepsy, we assessed the HRQL of patients pre and post epilepsy surgery using (i) the QOLAS (Kendrick and Trimble, 1994, Selai and Trimble, 1998), a measure shown to have validity for the patient and to be sensitive to change; (ii) the ESI-55 (Vickrey et al., 1992), an established measure for use in epilepsy surgery; and (iii) the EQ-5D (EuroQol Group, 1990, Brooks, 1996), a measure specifically designed to yield a single, overall score for use in economic analyses.