Epilepsia, 52(5):965–974, 2011 doi: 10.1111/j.1528-1167.2010.02973.x
FULL-LENGTH ORIGINAL RESEARCH
Relationship of clinical and quality of life trajectories following the onset of seizures: Findings from the UK MESS Study *Ann Jacoby, ySteven Lane, zAnthony Marson, and zGus A. Baker, on behalf of the MESS Study Group *Division of Public Health, yCentre for Medical Statistics and Health Evaluation and zDivision of Neurosciences, University of Liverpool, Liverpool, United Kingdom
SUMMARY Purpose: We defined a series of clinical trajectories represented among adult patients with new-onset seizures across a 4-year follow-up period; and linked these clinical trajectories to the quality of life (QOL) profiles and trajectories of those experiencing them. We examined both between- and within-group differences. Methods: Analyses were based on 253 individuals completing QOL questionnaires at baseline and 2 and 4 years subsequently. Based on patient self-report, we defined five ‘‘clinical trajectory’’ groups: individuals experiencing a single seizure only; individuals entering early remission; individuals experiencing late remission; individuals initially becoming seizure-free but subsequently relapsing; individuals with seizures persisting throughout follow-up. QOL profiles at each time point were compared using a validated QOL battery, NEWQOL. Key Findings: Even at baseline, there were significant between-group differences, with patients experiencing a
Approximately 6% of individuals in the general population will experience a seizure during the course of their lifetime (Hauser & Kurland, 1975). The risk of further seizures has been shown to decrease with increasing time from the initial event (Sander, 1993), and many people experiencing a first seizure will never have a second (Sander & Sillanpa, 1997). However, once they do, they then fulfil the commonly accepted (Hart et al., 1990) criterion for a diagnosis of epilepsy, and treatment with antiepileptic drugs (AEDs) becomes the norm. Recent evidence suggests that in persons experiencing a single or few seizures, there is little to be gained clinically in the longer term from a policy of immediate AED treatment, even though the short-term risk of Accepted December 15, 2010; Early View publication February 14, 2011. Address correspondence to Professor Ann Jacoby, Division of Public Health, Whelan Building, The Quadrangle, Brownlow Hill, Liverpool L69 3GB, U.K. E-mail:
[email protected] Wiley Periodicals, Inc. ª 2011 International League Against Epilepsy
single seizure only reporting the best QOL profile and those with seizures subsequently persisting across all time points reporting the worst. By 2 years, the QOL profiles of individuals experiencing early remission were similar to those of single seizure patients, as were those for late remission and relapse patients. Significance: A consistent pattern was seen, with ‘‘single seizure’’ individuals doing best and individuals with persistent seizures doing worst. Of particular concern is that even at baseline, individuals whose seizures persisted were doing poorly for QOL, suggesting the possibility that underlying neurobiologic mechanisms were operating. In contrast, our findings support previous reports of only short-lived and small QOL decrements for individuals experiencing a single or few seizures. KEY WORDS: Single seizure, Persistent seizures, Newonset, Psychosocial, Stigma, Resilience.
further seizures is reduced (Kim et al., 2006). Previous research also indicates that the likelihood of seizure remission is high in treated patients (Sander & Sillanpa, 1997). Where remission is prolonged, withdrawal of AEDs can then be considered, although the largest study to consider the outcomes of AED withdrawal estimated the risk of relapse as twice that of remaining on treatment, with negative effects for quality of life (AEDWS Group, 1991; Jacoby et al., 1992). Approximately 20–30% of patients will fail to respond to treatment and continue having seizures, with only about 20% experiencing even short-lived seizure-free periods (Sander & Sillanpa, 1997). Chronic illness has been proposed as a fundamental form of human suffering, with the affected individual experiencing many losses (Charmaz, 1983). Given the threatened daily life losses the onset of seizures poses, it would be surprising if these were not reflected in the QOL profiles of those individuals who experience them. Velissaris et al. (2007) conducted an in-depth phenomenologic study to explore the immediate psychological impact of a first seizure, and noted
965
966 A. Jacoby et al. that persons thus affected typically reported shock and fear at what had happened, anxieties about the possibility of seizures recurring, an increased sense of vulnerability and social activity limitations—reactions collectively representing loss of control. In some patients this sense of lost control was pervasive and in others it was fairly limited, but by 3 months patients in both groups were reporting largely normal psychological functioning. In a subsequent analysis (Velissaris et al., 2009), these authors noted that cognitive complaints were common, more so in those experiencing pervasive loss of control, but also a dramatic decline in such complaints, such that their 3-month scores were similar to those experiencing only limited loss of control. Dworetzky et al. (2000) likewise concluded that having a single seizure has only a modest impact on QOL. They compared domain scores on the SF-36 health status measure for persons experiencing a single seizure (PSS) with age-adjusted population norms. Scores were comparable for six domains, although significantly lower on the energy/vitality and physical role functioning domains. Nonetheless, more than one third (38%) of PSS, when asked at 1-year follow-up, considered that their seizure had had a moderate to extreme impact on their QOL, and the authors suggest that anxiety about the seizure episode may contribute to this. Lindsten et al. (2002, 2003) examined 10-year QOL outcomes in a cohort of patients 17 years and older who were experiencing newly diagnosed unprovoked seizures and their sex- and age-matched controls. Among those reporting them, 27% had experienced only a single seizure. These authors reported that across the 10-year period, most leisure-time activities, and marital and driving status were unaffected by the onset of epilepsy; the negative effects that were noted (for physical activity, traveling abroad, and overall activity) occurred later rather than earlier in followup, suggesting causal factors other than seizures. The authors suggested a possible link between reduced activity and the finding that patients also reported reduced income compared to controls, and higher morbidity. However, they also noted that patients had significantly lower income levels than controls before the index seizure. Employment rates were similar for patients and controls across the 10-year period and did not evolve negatively among patients after seizure onset. Therefore, these authors too present a generally optimistic picture of the likely QOL trajectory of patients experiencing a single or few seizures, concluding that important QOL effects will be present only for those whose seizures do not remit. In this article, we define a series of clinical trajectories represented among adult patients taking part in the Multicentre Study of Early Epilepsy and Single Seizures (MESS; ISRCTN 98767960) across the 4-year period of their follow-up; and link these clinical trajectories to the QOL profiles of those experiencing them. We examine QOL differences between the clinical trajectory groups at each Epilepsia, 52(5):965–974, 2011 doi: 10.1111/j.1528-1167.2010.02973.x
time point and QOL trajectories over time for each group. Our a priori hypotheses were: 1. at baseline, QOL profiles of all groups would be similar; 2. QOL profiles would diverge over time, with patients experiencing only a single seizure presenting the best profiles by end of follow-up and patients whose seizures persisted throughout follow-up the worst; 3. individuals experiencing a few seizures only then entering early remission would have QOL profiles similar to those experiencing only a single seizure; 4. individuals seizure-free in the early follow-up phase who subsequently relapsed would have poorer 4-year QOL profiles than those whose seizures persisted initially but subsequently remitted, since their hopes of achieving long-term seizure remission were not realized.
Methods MESS was a pragmatic randomized trial that aimed to assess the role of antiepileptic drug (AED) treatment for the management of a single or few seizures. Study methods have been described in detail elsewhere (Marson et al., 2005) but, in brief: patients were cued for entry if they had an adequately documented recent history of one or more clinically definite, spontaneous, unprovoked, epileptic seizures, previously untreated, and both clinician and patient were uncertain about whether to commence treatment. Recruited patients were randomized to either immediate or deferred AED treatment and followed for a maximum of 4 years. Demographic and clinical information was collected for all randomized patients, as well as for patients eligible but not consenting to randomization. MESS was approved by the U.K. Multi-Region Ethics Committee (and by the relevant ethics committees for participating non-U.K. centers). QOL study participants Eligibility criteria for the MESS QOL study were: adult (defined as aged 16 years and above), living in the United Kingdom, and without any major learning disability. Patients were tracked prospectively for QOL at entry to MESS (‘‘baseline’’: respondents recruited between 1994 and 2001) and at 2-year (between 1996 and 2003) and 4-year follow-up (between 1998 and 2005). Follow-up at 2-year intervals reflected that 2-year remission was a primary clinical outcome, at which point patients would be eligible for AED withdrawal and, at the time the study commenced, for restoration of a U.K. driving licence. QOL questionnaires were mailed to patients immediately following study entry and at 2 and 4 years after the initial data sweep. At each data-collection phase, a single mailed reminder was sent to nonresponders 3 weeks after the initial mail-out; and those failing to respond were contacted again by telephone after an additional 3-week period. The impact of randomization policy on QOL has been published elsewhere (Jacoby et al., 2007). In this article, we
967 Clinical and Quality of Life Trajectories examine patients’ QOL outcomes for specified clinical trajectories, regardless of randomization status. Based on patient self-report at each time point, five clinical trajectories were defined (Figure S1): 1. experiencing a single seizure only prior to baseline, no further seizures during follow-up (the ‘‘single seizure’’ group); 2. experiencing more than one seizure prior to baseline, none subsequently (the ‘‘early remission’’ group); 3. experiencing a single or multiple seizures prerandomization, further seizures up to 2-year follow-up, none subsequently (the ‘‘late remission’’ group); 4. experiencing single or multiple seizures prerandomization, seizure-free between baseline and 2-year follow-up, further seizures between 2- and 4-year follow-up (the ‘‘relapse’’ group); 5. experiencing a single or multiple seizures prerandomization, seizures persisting throughout the entire follow-up period (the ‘‘persistent’’ group) At baseline, the patient’s consulting physician completed clinical proforma detailing seizure number, date of occurrence, and classification. Comparison of patient-based and clinician-based information identified a small number of discrepancies in relation to number of seizures: therefore, seven individuals describing themselves as having had a single seizure only were recorded in the clinical notes as having had more than one.1 Because the primary focus for the present paper is patient-reported QOL, and since the timing of clinical follow-up did not always align precisely with that of the QOL assessments, the trajectories defined in the preceding paragraph are based on patient report in the QOL assessments. Description of QOL battery QOL was assessed using an initial version of a QOL battery, NEWQOL, designed to examine physical, psychological, and social functioning in persons with new-onset seizures (Table S1) and shown to be valid, reliable, and relevant to patients (Abetz et al., 2000). In addition, measures of subjective cognitive function (Epilepsy Surgery Inventory subscale, Vickrey et al., 1992), anxiety/depression (Hospital Anxiety & Depression Scale, Zigmond & Snaith, 1983), self-esteem (Rosenberg, 1965), mastery (Pearlin & Schooler, 1978), and overall QOL (Terrible-Delighted Faces Scale, Andrews & Withey, 1976), and items relating to education, employment, and driving status were included. Statistical methods All analyses were conducted using SPSS for Windows, version 17.0 (SPSS Inc, Chicago, IL, U.S.A.). Differences between the defined clinical trajectory groups were exam1 Six individuals classified from the QOL questionnaire data as ‘‘single seizure’’ were recorded in the clinical record as having had two seizures; and one was recorded as having had three seizures.
ined using analysis of variance (ANOVA) and KruskalWallis tests for continuous data, depending on whether they were normally or nonnormally distributed. Categorical variables were examined using the chi-square test; if the expected cell frequencies were
