Address correspondence and reprint requests to Dr. M. Herodes at Department of Neurology and Neurosurgery, 2 Puusepa St., 51014 Tartu, Estonia. E-mail: Marju.Herodes@kliinikum.ee
Summary: Purpose: To study the impact of epilepsy and its treatment on people with epilepsy in Estonia and to analyze how it is affected by the characteristics of epilepsy.
Methods: Clinical and demographic data about patients were obtained from medical notes and mailed self-completed questionnaires (including the RAND 36-Items Health Survey 1.0 (RAND-36)).
Results: Information was collected from 203 patients aged 20–74 years, who all had active epilepsy. A third of the respondents had been seizure free during the last year. Eighty-four percent were receiving monotherapy. More than half of respondents felt stigmatized by epilepsy, 24.7% of them highly so. A third were working full-time, 31.9% were underemployed workers, and 11%, unemployed. Sixty-two percent of these same unemployed or underemployed workers considered their epilepsy to be a significant reason for this situation. Overall, 44% believed they had been treated unfairly at work or when trying to get a job. Study respondents scored lower in all domains on the RAND-36 than did persons from the control group. The biggest differences were found in five domains: Social functioning, Role limitations–physical, Role limitations–emotional, General health, and Vitality.
Conclusions: The clinical characteristics of this study were similar to those of most other series of prevalence cases of epilepsy. The level of employment among persons with epilepsy was not lower than that in the general population. The percentage of stigmatization was high. There were significant differences in the way respondents scored on the stigma scale and on the RAND-36 domains when measuring their health status, depending above all on seizure frequency and type.
Although it is a universal brain disorder, epilepsy is often misunderstood. It is now widely acknowledged that people with epilepsy are as likely to be distressed by social and cultural problems as they are by continuing seizures, and that epilepsy has profound physical, psychological, and social consequences (1). Although current seizure frequency is one of the most important predictors showing the efficacy of treatment, it is not the only measure, especially from the patient‘s viewpoint, commonly used in clinical studies of new antiepileptic drugs (AEDs) (2). The effect of any disease is determined by several factors, including underlying biology, as well as host factors, and available medical interventions, but also by the attitudes and reactions of the surrounding society (3). Several studies have used health-related quality of life in epilepsy as an outcome measure and have also used it to give a broader measure of the burden of the disease (4). Quality of life is difficult to define but might be said to reflect functions in three main areas: physical, social, and psychological (1). Devinsky and Cramer (5) stated that the essence of quality of life is the balance between patients' perceived and desired status. It also is defined by how well one is able to function and how one feels about one's daily life (6), on the assumption that aspects of functional health status have an impact on quality of life. Although no definitive consensus has been reached concerning the essential nature of quality of life, there is some agreement that general health status is one of its main components (7). A variety of instruments are available to evaluate the perception of health in the general population. One of these, which also is among the most widely used questionnaires, is the RAND 36-Item Health Survey 1.0. It is a brief and intensively tested instrument that was derived from longer instruments developed by RAND researchers (Santa Monica, CA, U.S.A.) for the Medical Outcome Study (MOS) and the Health Insurance Experiment (8) to assess health status. The purposes and methods of the RAND study have been fully summarized (9,10). The RAND 36-Item Health Survey 1.0 items are identical to the MOS 36-item short-form health survey (MOS SF-36) described by Ware and Sherbourne (9). They were adapted from longer instruments completed by patients participating in the MOS (11). The conceptual framework is based on the multidimensional World Health Organization definition of health (12). Although the RAND version has a slightly different scoring method, it allows users of the MOS SF-36 and RAND-36 to relate their findings (9).
Because of the emphasis on the phenomenologic experience of the individual, it is necessary that quality of life be determined from the patient's subjective viewpoint, the physician's viewpoint being deliberately excluded, as self-reports are the primary method of assessing it (6) because, with very few exceptions, evaluations conducted by physicians tend to concentrate primarily on seizure management, leaving all else as secondary features (13). It has become relatively common to have patients make a judgment about their medical care (14). This means they must have the courage to express their opinion and show their dissatisfaction. There is a growing awareness of the psychosocial implications of epilepsy. People with epilepsy face social disadvantages not shared by those with other chronic diseases. Psychiatric problems, particularly anxiety, depression, and loss of self-esteem are common among people with epilepsy (15–26). Most patients feel that a prospective employer's knowledge of a diagnosis of epilepsy will make it more difficult for them to get a job (27). Information on these issues has come mainly from developed countries (26,28–33). Very few studies originate from developing countries (34–37), and there is clearly a lack of documented evidence regarding the impact of epilepsy in Eastern Europe (38,39).
Estonia, which is located in Eastern Europe on the coast of the Baltic Sea, regained its independence after the collapse of the Soviet Union. Today 64.6% of the 1.5 million people living in Estonia are ethnic Estonians. Among other nationalities, Russians represent the largest group (∼28.5%). The Russian-speaking population is not evenly distributed throughout the country (40). Southern Estonia revolves around Tartu, the historic university town and the country's second largest city. Tartu, the intellectual and educational center of Estonia, demonstrates relatively typical demographic characteristics for Estonia, with the exception that the percentage of the Russian-speaking population is lower than that in Estonia as a whole. Viljandi County, with a population of 62,336 (41), is considered to be first in the country in terms of the level of development of agriculture, and is located in south-central Estonia. The administrative center of the county is the town of Viljandi, which is situated 81 kms from Tartu.
This report is a comprehensive study of what it is like to have epilepsy in our society. It was conducted to pursue the following objectives: (a) to describe the quality of life for epilepsy patients on the grounds of perceived health status and possible stigma accompanying epilepsy and to analyze how it is affected by the characteristics of epilepsy; and (b) to analyze how quality of life is affected by the sociodemographic characteristics of epilepsy patients, with emphasis on their current employment status.
Design and study sample
The research took place in 1997 through 1998 and followed an epidemiologic survey of epilepsy in the city of Tartu, Estonia. The epidemiologic survey included persons who were residents of Tartu, were aged 20 years and older, and had before or during the period from January 1, 1991, through January 1, 1996, had at least two unprovoked epileptic seizures, at least one of them within the previous 5 years. Data collection for the epidemiologic study consisted of two parts: data registration from a multisource medical register review and data registration from a personal case reexamination. Case records of patients treated at the University Hospital, Outpatients' Clinics, physicians' offices, emergency rooms, or the electroencephalographic laboratory with a diagnosis of epilepsy, convulsions, syncope, amnestic attacks, or abnormal involuntary movements were reviewed, and invitations for reexamination were sent to the relevant persons. During the last 2 years, all the patients were reexamined at least once by one of the authors to specify the types of their seizures.
Our study focused on the analysis of data collected from a sample of 203 patients in the 20- to 74-year age group. The patients were selected at random from the preliminary lists of the epidemiologic study conducted in Tartu, excluding people who were not capable of understanding Estonian (mostly Russian speakers) because no sufficiently well translated and validated questionnaire was available. In Viljandi, primary information about people with epilepsy was gathered through the local epilepsy support group, and clinical information was abstracted from medical notes held in the County Hospital and Outpatients' Clinic register. To evaluate the accuracy of diagnoses, the problematic cases were investigated by one of the authors and reexamined if necessary. All patients gave their consent for participation in the research, and the project was approved by the Ethics Committee of the University of Tartu. In addition, a control group of 200 healthy subjects corresponding in age, sex, and educational level was randomly selected from among the patients receiving treatment from dentists at the University's Dental Clinic. All of the respondents possessed at least a basic education with sufficient ability to read and write, and were capable of understanding and completing the questionnaires.
Clinical information, if needed, was abstracted, once again, from medical notes and during the personal reexamination of subjects. Abstracted information used in the study related to the etiology of epilepsy, classification of seizure type, and current AED therapy. To evaluate the impact of epilepsy on employment status and perceived stigma, the patients were sent a questionnaire by mail. Following the example of other quality of life studies conducted among persons with epilepsy (26,42–44), the questionnaire used a combination of open questions with two previously translated and validated scales (The Stigma of Epilepsy Scale and the RAND 36-Item Health Survey 1.0). In addition, single items that referred specifically to feelings of stigmatization in the area of employment were used. The questionnaire contained a number of scales and questions covering the following issues:
1Demographic characteristics: information was obtained about subjects' sex, age, marital and employment status, and educational level;
2Economic and financial status: patients were asked to state whether they considered it to be “very good,”“good,”“satisfactory,”“moderately bad, or “very bad”;
3Seizure frequency: patients were asked whether they had had seizures once or more in a month, less often than once a month, or not at all in the past year;
4Injuries associated with seizures: subjects who had had at least one seizure in the past year were asked whether they had had a burn or scald, a head injury, milder injuries (including dental injuries), any other injuries (unspecified), or no injuries;
5History of the epilepsy: patients were asked about age at first attack;
6Previous research has shown that patients‘ perception of the severity of their seizure disorder may be more important than seizure frequency in determining their psychological and social well-being (45). Therefore subjects were asked to assess their seizures as “very severe,”“severe,”“medium,” or “light”;
7AED treatment and side effects: patients were asked about the AED they were taking and about the experienced side effects during the past month, as well as about satisfaction with the current treatment and about changes in AED medication in the past year;
8Compliance with medication: patients were asked to state whether they never missed taking their AEDs, missed less often than once a month, missed less often than once a week, or missed more often than once a week. According to other studies, correlations between patient report and objective method have been shown to be high (43);
9Perceived stigma was measured with a three-item scale developed originally for stroke (46), adapted for epilepsy and already used in other quality-of-life studies (26,43). Respondents with epilepsy stated whether they felt that other people (a) were uncomfortable with them, (b) treated them as inferior, or (c) preferred to avoid them. Each of the three items required a yes/no response. An individual's score was the sum of the “yes” responses, and the higher the score, the greater was the perception of stigma. The scale was translated into Estonian by two independent native Estonian speakers with an excellent knowledge of English. The translators then met to discuss and agree on a common version of the questionnaire. Subsequently the common version was evaluated by another native Estonian speaker in terms of conceptual equivalence, linguistic performance, and clarity. The agreed Estonian form was then translated back into English and rated. If modifications were necessary, reformulation was performed in the Estonian version. The internal consistency of the scale was examined using Cronbach's alpha and found to be acceptable (alpha = 0.71) (47). The evidence for the construct validity of the scale was supported by the data received following the hypotheses that patients with frequent seizures and mixed seizure types would score positively on the scale;
10The impact of epilepsy on employment history: those currently un- or underemployed were asked whether this was caused by their epilepsy, whether they had changed jobs in the preceding 2 years because of epilepsy, and whether they had been treated unfairly at work because of epilepsy. Each of the items required a yes/no response. Patients were divided into three groups by seizure type (as having only tonic–clonic, only other types, or both tonic–clonic and other types) and frequency (based on seizure occurrence once or more a month, less often than once a month, or not at all in the past year);
11Health status: respondents were asked to complete a comprehensive generic health status measure, the RAND 36-Item Health Survey 1.0 (RAND-36) (48), which consisted of eight multiitem variables: physical functioning (PF), 10 items; social functioning (SF), two items; role limitations due to physical problems (RP), four items; role limitations due to emotional problems (RE), three items; mental health (MH): five items; energy and vitality (VT), four items; bodily pain (BP), two items; and general perception of health (GH), five items. There is a further unscaled single item on changes in respondents' health over the past year (CHG). The scale was translated into Estonian as described in the Stigma scale. The RAND-36 questionnaire took ∼5 min to complete. As indicated in standard RAND-36 scoring algorithms, for each variable item scores were coded, summed, and transformed onto a scale from 0 (worst possible health state measured by the questionnaire) to 100 (best possible health state). Missing value rates for the items were low and did not exceed 1.5% for any item of the eight scales. Internal-consistency reliabilities (Cronbach's alpha) met the level acceptable for group comparisons (>0.70) across all scales ranging from 0.75 to 0.92. Scaling assumptions were tested in two ways. Corrections, between items and hypothesized scales, were substantial within each scale and reached the level of >0.40 in all instances, supporting the reliability of the RAND-36 scales in both groups. In the epilepsy group, the lowest median item-total correlation was 0.53 for general health, and the highest, 0.84 for bodily pain. Discriminant validity was considered acceptable when these correlations exceeded all correlations between items and other scales. All of the eight scales in both groups passed this test level. The validity of the RAND-36 was assessed using discriminant techniques. The RAND-36 ability to distinguish between a high and a low symptom load was determined through assessment by seizure type and frequency. Discriminative power was examined by comparing the RAND-36 score profiles of the healthy respondents and respondents with epilepsy. A subsequent methodologic article will examine in detail the psychometric properties of the RAND-36 in this sample of people with epilepsy.
The data were analyzed using the statistical analysis package SPSS Professional Statistics 7.5 (49). Tests of significance were χ2 (chi square) and one-way analysis of variance (ANOVA). Attention is drawn to results in which differences were significant at the 5% level or less (p ≤ 0.05). To examine the correlations between different characteristics concerning employment and stigma, we performed a multivariate analysis of the data, but because we found no significant interactions at the 5% level, we were unable to construct corresponding models. Cronbach's alpha was used to test the reliability of the scales. To assess the influence of patient characteristics on quality of life (RAND-36) domains, the effects of the clinical variables (seizure frequency, type of seizures, age at onset, duration of disease), stigmatization and its severity, and sociodemographic variables (age, sex, educational level, employment status, marital status) on each domain total score were studied, using multifactor ANOVA models. Preliminary analyses were carried out to investigate which of the clinical variables predominated. At each stage, factors found to be no longer significant after adjusting for the remainder were excluded. The final model includes only the factors that contributed significantly in predicting the domain score. To determine significant differences between pairs of groups, the Tukey HSD procedure or Bonferroni's method was used (50).
Response to the study
In Tartu, questionnaires to be completed individually were mailed to 110 patients, of whom 78 replied, a response rate of 71%. After sending a reminder, 16 more patients returned their questionnaires. Of all the questionnaires returned, 19 appeared to be unusable: in six cases, >10% of the questionnaire was left unanswered; three were sent back with a note that the person was deceased; five, with a note that the person no longer lived at the address; five, because the person was incapable of understanding the questions because of mental disability. The remaining questionnaires were considered usable and were included in the study. In addition, 15 patients completed the questionnaires while visiting their neurologist at the Outpatients' Clinic. One hundred twenty-two of the questionnaires appeared usable.
In Viljandi, questionnaires were mailed to 120 patients, of whom 85 replied. After being sent a reminder, 10 more patients returned their questionnaires. Eighty-one of the returned questionnaires appeared usable.
Sociodemographic characteristics of sample
The median age of the study population was 41 years (25th and 75th percentiles 29 and 57). The respondents of the study were divided into five age groups: 20–29 years, 26.6%; 30–39 years, 20.7%; 40–49 years, 17.2%; 50–59 years, 13.8%; and 60 years and older, 21.7%. Men accounted for 48.8%. Of the respondents, 40.9% were married or cohabiting, 41.4% were single, and 10.3% were divorced; 44.3% had less than primary (lower than eighth grade) or primary education (eighth or ninth grade), and 55.7%, high school (11th or 12th grade) or university education. Thirty-three percent were working full-time, and 41.9% were un- or underemployed; 0.5% described their economic and financial status as very good; 8.5%, as good; 59%, as satisfactory; 26%, as moderately bad; 6%, as very bad (Table 1).
Table 1. Sociodemographic characteristics of respondents
Retired or receiving disability pension
Less than primary (lower than 8th grade)
Primary (8th or 9th grade)
High school (11th or 12th grade)
The median age of the control group was 40 (25th and 75th percentiles 27 and 56) years. Forty-nine percent were men. Eight percent had less than primary education (lower than eighth grade level), 32% had primary education (eighth or ninth grade level), 49% had a high school education, and 11% had graduated from university.
Disease characteristics of the sample
The median age of the onset of epilepsy was 26.9 years, and the median duration of epilepsy was 11.3 years (25th and 75th percentiles 5.8 and 22.4). Patients were divided into five groups by duration of the disease and into six groups by age at onset of the epilepsy. Of patients, 41.4% reported having only tonic–clonic seizures, 30% reported having both tonic–clonic and other types of seizures, and 28.6% reported having only other types of seizures. Almost a third (34%) had been seizure free in the last year; 39.9 had less than one seizure a month; and 26.1% had one or more seizures a month. Of those who had had at least one seizure in the past year (134 patients), 14% reported having serious injuries (burn, scald); 38%, head injuries; 22%, milder injuries or headache; and 5%, other injuries. More than a fifth (21%) had not experienced any injuries. Those having seizures once or more in a month (χ2 = 11.89; df = 2; p = 0.001) and those having multiple or generalized tonic–clonic seizure types (χ2 = 9.94; df = 2; p = 0.009) were more likely to report a seizure-related injury. Of these, 7.4% described their seizures as very severe; 30%, as severe; 41.4%, as moderate; and 21.2%, as light. There was a significant correlation with the subjective assessment about the severity of the seizures (those assessing their seizures as very severe or severe and those considering them moderate or light were counted together) with reported seizure-related injuries (χ2 = 15.24; df = 4; p = 0.003). Of the 88.7% who were receiving AED treatment, 83.9% were receiving monotherapy. The majority (74.8%) of those receiving monotherapy were receiving carbamazepine (CBZ). The most commonly experienced side effects were memory problems (31%), tiredness (25%), sleepiness (20%), headache (20%), and nervousness (20%). A third (33%) of subjects reported no side effects. The majority of respondents (78%) receiving AED treatment described their epilepsy as very or fairly well controlled by this; 21% stated that the level of control was unsatisfactory. Almost two fifths (41.3%) of those receiving medication had changed it at least once in the past year; 68.4% had changed it once; 22.8%, twice; and 8.8%, three or more times. Of those who had changed their medication once in the past year, 79.3% had changed it because of unsatisfactory control and 20.7% because of side effects. For compliance with medication, 56% of respondents said they never missed taking AEDs, 23% reported missing on average once a month, 14% reported missing once a week, and 7%, more than once a week. Some (17.7%) had some other disease or health problem in addition to epilepsy; 11.8% were receiving medical treatment because of these. The most common additional diseases were diseases of the heart (39%) and joints (33%). The results of the underlying epidemiologic study conducted in Tartu have not yet been published in full. We therefore found it necessary to provide more detailed information about the clinical characteristics of the epilepsy of the study respondents. The main disease characteristics are presented in Table 2.
Table 2. Disease characteristics of respondents
Duration of epilepsy (median)
Age at onset
Tonic–clonic and others
Seizure frequency status in the last year
<1 seizure a month
≥1 seizure a month
Free of medication
Of those receiving AED medication
Receiving 2 AEDs
Receiving ≥3 AEDs
Type of drug
More than half of all respondents (52.4%) felt stigmatized by their epilepsy; 24.7% answered “yes” to all three items, and this shows that they were highly stigmatized. Respondents were more likely to feel stigmatized if they had frequent seizures (χ2 = 23.57; df = 6; p < 0.0001) or mixed seizure types (χ2 = 20.65; df = 6; p < 0.009; Table 3). At the same time, only 37.4% considered their seizures very severe or severe. Those who had experienced seizures during the last year (χ2 = 18.63; df = 1; p < 0.0001) and those who had tonic–clonic type of seizures only or together with other seizure types (χ2 = 7.02; df = 1; p < 0.008) were more likely to score highly (to give two or three “yes” answers) on the stigma scale. Stigmatization was more common among those having university or high school education (χ2 = 12.89; df = 6; p < 0.05). No differences were found in scores on the stigma scale by sex, marital status, or employment status.
Table 3. Reported stigma by seizure type and frequency
Score on stigma scale
Figures in brackets are the numbers on which percentages are calculated.
Tonic–clonic only (n = 84)
Tonic–clonic and other (n = 61)
Other only (n = 58)
χ2 = 20.65, p < 0.009
One or more a month (n = 53)
Less than one a month (n = 81)
None in past year (n = 69)
χ2 = 23.57, p < 0.0001
A third of all respondents were working full-time. Employment status (working either full-time or being underemployed; those retired or receiving disability pension were excluded) was significantly related to age (χ2 = 12.02; df = 4; p = 0.03), seizure frequency (χ2 = 10.81; df = 2; p = 0.004), age at the onset of seizures (χ2 = 15.13; df = 5; p = 0.01) and education (χ2 = 11.38; df = 3; p = 0.01). Sixty-two percent of those who were un- or underemployed named epilepsy as the significant reason for it. Respondents with frequent seizures were more likely to believe it (χ2 = 11.03; df = 2; p = 0.001). During the last 2 years, 29% of respondents had changed jobs (meaning a change of workplace, not change of speciality or loss of job). Men (χ2 = 7.07; df = 1; p < 0.003) and those with frequent seizures (χ2 = 11.79; df = 2; p<0.006) were more likely to do this. Forty-four percent said that they had been treated unfairly at work or when getting a job. There were significant interactions between this opinion and seizure frequency, type, and education: respondents with frequent seizures (χ2 = 16.26; df = 2; p = 0.0001), respondents having tonic–clonic or multiple seizure types (χ2 = 8.94; df = 1; p = 0.002) and respondents who had lower than high school education (χ2 = 7.32; df = 1; p = 0.007) were more likely to report this. We cannot leave unmentioned here that, although it was not asked, several respondents commented on the fact that they had hidden their diagnosis of epilepsy from employers and colleagues because of the fear of discrimination and shame. There also was a significant interaction between full-time work and educational level (those retired or receiving a disablement pension were excluded): the higher the person's education, the more likely he or she was to be working full-time (χ2 = 12.12; df = 6; p = 0.04).
The correlations between the RAND-36 scales and seizure status together with descriptive statistics for the questionnaire are given in Tables 4 and 5. Variance between seizure types was statistically significant in five RAND-36 domains. The comparisons between groups were investigated for each domain using the Tukey HSD test at the 0.05 level. Patients who did not have generalized tonic–clonic seizures or multiple seizure types had significantly higher scores in the Role–physical and Social functioning domains. Those who had multiple seizure types had lower scores than did those with only tonic–clonic seizure types or those with other types of seizures only in the Role–emotional domain. Those who experienced multiple seizure types scored significantly lower in the Emotional well-being and Bodily pain domains compared with those who did not have generalized tonic–clonic seizures (Fig. 1).
Table 4. Descriptive statistics and features of score distributions for the RAND-36 Health Survey by seizure type
CI, 95% confidence interval; SEM, standard error of the mean.
Variance between seizure frequencies. Test of significance was Kruskal–Wallis one-way analysis of variance.
Variance between seizure frequency statuses was statistically significant in seven domains. The differences were significant between all three groups in the Role–emotional domain. Between those who had not had seizures in the past year and those who had had seizures at least once a month or less often than once a month, the differences were significant in the Role–physical, Energy/fatigue, and General health domains. Those experiencing seizures at least once a month scored significantly lower in Emotional well-being and Social functioning compared with those who had been seizure free in the past year. In the Bodily pain domain, the differences were significant between those having seizures once or more in a month compared with those who had had seizures less often than once a month or had not had them in the past year (Fig. 2).
Study respondents scored lower in all domains on the RAND-36 than did persons from the control group, meaning that they were more dysfunctional (Fig. 3). The greatest differences were found in five domains: Social functioning, Role limitations–physical, Role limitations–emotional, General health, and Energy/fatigue (Table 6).
Table 6. Mean scores of dimensions of RAND-36 questionnaire
Variance between groups. Test of significance was Kruskal–Wallis one-way analysis of variance.
Role limitations (physical problems)
Role limitations (emotional problems)
The results of the final models fitted to each RAND-36 domain score, including the factors that remain significant after controlling for the others, are shown in Table 7. Each multifactor model is a main-effect model (no significant interactions were found between the factors). Pairs of groups of significantly different factors were compared using Tukey HSD or Bonferroni's procedures. Scores of the RAND-36 domains were first compared in terms of the clinical variables. Significant differences were found for seizure frequency in all domains, except the Physical functioning domain. Seizure-free patients scored significantly higher than did patients who had experienced seizures during the last year in the Role limitations–physical, Energy/fatigue, Bodily pain, and General health domains. The difference between those who had had seizures once or more in a month compared with those having seizures less often than once a month or not having seizures during the last year was significant in the Role limitations–emotional, Emotional well-being, and Social functioning domains.
Table 7. Results of analysis-of-variance models
Mean square ratio
Age at onset
Role limitations (physical problems)
Age at onset
Role limitations (emotional problems)
Duration of disease
Age at onset
Duration of disease
Age at onset
Duration of disease
Age at onset
In the Role limitations–physical domain, age, stigmatization, stigma severity, and age at onset of epilepsy became significant after controlling for seizure frequency. Younger people were less likely to score low in this domain, and there were significant differences were between the 20–29 and 30–39 age groups compared with people who belonged to the 60 years and older age group. Mean scores for this domain were significantly lower for those who were stigmatized, and for those who expressed very strong feelings of stigma (gave three “yes” answers on the stigma scale) compared with those who expressed less (one “yes” answer). Later age at onset was associated with lower scores; differences were significant between those for whom epilepsy had been diagnosed at the age of 41–50 or older than 50 compared with those for whom it had been diagnosed at younger than 20 years.
In the Role limitations–emotional domain, mean scores were significantly lower for those who were stigmatized and for those who expressed very strong feelings of stigma (gave three “yes” answers on the stigma scale) compared with those who expressed themselves less strongly (one “yes” answer).
In the Energy/fatigue domain employment status, duration of epilepsy, and age at onset of epilepsy were significant. In this domain, mean scores were significantly lower for those currently unemployed, in comparison to those who were in full-time or underemployed work, for those who had had epilepsy for 2 to 5, and 6 to 10 years compared with those who had epilepsy longer and for those whose epilepsy had been diagnosed at the age of 41–50 or older than 50 compared with those for whom it had been diagnosed at younger than 20 years.
In the Emotional well-being domain, those who were stigmatized had significantly lower scores than did those who were not and of those who had had epilepsy 2 to 5 years compared with those who had had epilepsy >20 years.
In the Social functioning domain, stigmatization, stigma severity, employment status, seizure type, and age at onset of seizures became significant. Significantly lower scores were obtained by those who were stigmatized; those who expressed very strong feelings of stigma (gave three “yes” answers) compared with those who did not feel this so strongly (one “yes” answer), those who were currently unemployed compared with those who were in full-time employment or underemployed, those who experienced either tonic–clonic or multiple seizure types compared with those who had only other types of seizures, and those for whom epilepsy had been diagnosed at the age of 41–50 or older than 50 compared with those for whom it had been diagnosed at younger than 20 years.
In the Bodily pain domain, lower scores were related to a shorter (2–5 years) rather than longer (11–20 and >20 years) duration of epilepsy.
In the General health domain, mean scores were significantly lower for those whose epilepsy had been diagnosed at the age of 41–50 or at older than 50 years compared with those for whom it had been diagnosed at younger than 20 years. Those who were stigmatized also scored significantly lower than those who were not, and those who expressed very strong feelings of stigma (gave three “yes” answers) in comparison to those who did not (one “yes” answer).
In the Physical functioning domain, stigmatization, age at onset, and current age were found to be significant. Mean scores in this domain were significantly lower for those who were stigmatized compared with those who were not and for those aged 60 years or older compared with those aged 20–29 years. The overall pattern of variation in terms of age at onset was similar to that in the General health domain.
The importance of measuring quality of life in epilepsy patients has been emphasized (8,13,26,43,44,51–54). In dealing with epilepsy, several authors have drawn attention to the special importance of considering the social aspects. At the same time, recent investigations based on community populations suggest that although significant social difficulties may be experienced, many people with epilepsy cope well in society. However, patients with poor seizure control, multiple seizure types, or associated handicaps have significant social problems (55). Our study focused on adults living in the community. To give a more extensive and accurate survey, the sample for the study was drawn from two Estonian towns differing from each other in several respects. One of them represented the country's urban society, and the other, a mainly provincial and rural population. On January 1, 1996, the estimated crude prevalence ratio of active epilepsy in Tartu was 4.1 per 1,000 (56). When comparing the percentages of sex and age structure of the people with epilepsy in the present study with the same data available about the people with epilepsy of Tartu, there were no significant differences, and thus we consider our study consecutive (Table 8). The clinical characteristics of the present study were similar to most other series of prevalence cases of epilepsy (57–59). Most of the study respondents had generalized seizures with or without other seizure types, the average duration of the disease was 11 years, and patients were predominantly receiving CBZ monotherapy. More than half of those who had experienced seizures during the last year reported having injuries related to them. Findings regarding the rate and severity of seizure-related injuries were slightly higher compared with the results of studies conducted in other countries (43,60). Beran (61) pointed out that the purpose of treating epilepsy may not necessarily be that of seizure eradication but rather the maximal improvement of quality of life for the patient. In comprehensive management, the treating physician must very seriously consider the influence of the therapy on the patients' quality of life (53). Eighty-four percent of our study respondents were receiving monotherapy. This was higher compared with the other studies (26,43,62). The explanation is that all of the patients from Tartu were participating in an epidemiologic survey with consultation by an epileptologist, which often resulted in the correction of medication. The number of untreated cases (11%) was not high and probably reflects insufficient compliance. However, AED prescription patterns had some distinctive features. CBZ was a much more frequently reported drug than in other studies, whereas the percentage of those using valproate (VPA) or phenytoin (PHT) was lower. To our surprise, two patients reported taking bensobarbital, a drug that is no longer officially used in Estonia. The results indicate that treatment strategies in Estonia probably should be modified. Significant numbers of study respondents (67%) reported side effects from the AEDs; the most commonly experienced side effects were nonspecific. In the past year, 41% of respondents had changed their medication; at present, 78% stated that the level of seizure control was satisfactory.
Table 8. Comparison of sex and age structure of Tartu people with epilepsy and those included in the present study
Among people with epilepsy in Tartu
Among people with epilepsy in Tartu in the present study
The problem of stigmatization has been projected as one of the most common social problems faced by persons with epilepsy in a number of studies (23,37,63–66). Stigmatization seems to vary from region to region, and it tends to be more severe outside the developed world (67–72). However, despite its changed manner, it is still a difficult problem in Western countries. As stigmatization is difficult to compare, we collated our results only with the results from the European study (43) in which the same scale was used for measuring stigma. According to this, the highest proportions of stigmatized persons (>60%) were found among the respondents from two highly developed countries (i.e., France and Germany). The study also included respondents from Poland, the Czech Republic, and Hungary, where the percentages of stigma were 32, 55, and 52, respectively. In Estonia, the levels of stigma among people with epilepsy were also high (52%), although 40% of our study's respondents had less than one seizure a month, and 34% had been seizure free in the last year. The majority of patients stated that they were nevertheless satisfied with the current treatment, and the percentage of stigmatization in general and the percentage of severely stigmatized persons was high. The factors influencing the development and maintenance of stigma in different countries are diverse, but we speculate that in general, the higher percentage of stigmatization could be a characteristic of Eastern European countries and could be the result of a general lack of knowledge and of indifference, because the individual's health and well-being was not valued, for a long period, because of the complicated political status. Furthermore, more precise studies from other Eastern European countries could perhaps clarify this topic. Respondents were more likely to feel stigmatized by epilepsy if they had frequent seizures or a combination of seizure types, findings that were in agreement with the results of other studies (26,43,63,65). Stigmatization was more common among educated persons.
Unemployment and part-time employment, being much more frequent in the epilepsy population than in general, have been identified as being among the most serious problems facing people with epilepsy (73,74). The percentage of people working full-time and part-time was 65 in the present study; 11% were unemployed. We do not consider this high because, according to the data of the labor force surveys of the Statistical Office of Estonia, the percentage of employees (both employed and underemployed) residing in Tartu and aged 20 years and older on January 1, 1998, was 63%, the unemployment rate was 9.5%, and 25.5% were pensioners receiving the state pension (75). At the same time, more than half of the study's respondents believed that their employment problems were caused by their disease. A little fewer than half stated that they were being treated unfairly at work. Perceived discrimination may not always correspond to real discrimination (1). Although the findings of this study do not provide evidence of active discrimination against people with epilepsy, this topic must be investigated in greater depth. Of respondents, 55.7% had at least high school education, and problems connected with unemployment or part-time employment were not much expressed among this group. Not surprisingly, seizure frequency was positively related to the unemployed and underemployed workers, but we could not find a relationship with the type of seizure. The finding supports the data of previous research in which lower seizure frequency had been related to the greater likelihood of being employed (1,42,76). The results of our study showed very clearly that there are a variety of reasons for the existence of the stigma. Although it has been found that unemployment and employment problems are on the whole the main source of the stigma (17,42,65), the most educated respondents in our study who had jobs were even more stigmatized.
To assess general health status, a multidimensional instrument, the RAND 36-Item Health Survey 1.0, was used. Although they have proven useful in their countries of origin, such instruments are not directly applicable across nations because of cultural diversity (77). Before using it in our study, we performed a thorough translation and validation process. The construct validity of the scale was supported by the findings that those with frequent seizures did poorly compared with those who experienced infrequent seizures or were currently seizure free. This expected finding was in accordance with other studies (26,41,62,78,79). Although the differences between seizure types were not significant in all the RAND-36 domains, there was a clear tendency toward a greater likelihood of lower scores in the case of patients with generalized tonic–clonic seizures. Patients who experienced both generalized tonic–clonic and other types of seizures did poorly compared with the others, as was to be expected (26,43,78). Discriminant validity was highly acceptable. People with epilepsy had significantly lower scores than did the controls in all domains. Although the mental health of the study respondents was not much worse than that of the control group, their social functioning was significantly lower, and limitations due to emotional problems were more expressed. The results of the European study had previously drawn attention to the fact that it was unclear why respondents with epilepsy scored relatively poorly on the domain concerned with physical function (43). Although current seizure activity remained the most important predictor, there was a concomitant importance of sociodemographic variables (current age and employment status) in quality of life. Older people and people who were currently unemployed were more likely to score lower. The other substantial disease characteristics in explaining the variation in the scores of several domains after controlling for seizure status were age at onset of epilepsy, duration of disease, and seizure type. Age at onset became significant in the case of Physical functioning, Role limitations–physical, Energy/fatigue, Social functioning, and General health. In all those domains, later age at onset was associated with lower scores. Dominian et al. (80) reported an association between depression and older age at onset. Jacoby et al. (26) considered older age at onset to be implicated in feelings of depression and stigma. Duration of disease was significant in the case of Energy/fatigue, Emotional well-being, and Bodily pain. Here, a shorter duration of epilepsy was related to lower scores. Seizure type became significant in relation to Social functioning; those who experienced either tonic–clonic or multiple seizure types scored significantly lower than did those who had only other types of seizures.
To increase the clinical significance of these tests, it is essential to perform repetitive trials. This will be one of the subjects of further investigation. As the RAND-36 was not designed to measure limitations or restrictions specifically associated with epilepsy, a disease-specific instrument may be more sensitive in evaluating variations in patient perception (81,82).
We consider the strength of our study to be that epilepsy diagnosis was based on a clinical assessment. A profound translation and psychometrical testing phase preceded the inclusion of the RAND-36 questionnaire in the research. Although we are aware of the limitations to the generalizability of the study in the interpretation of the results because of a relatively small and somewhat biased sample size, the findings of the study reveal quite clearly that one of the main problems of people with epilepsy in Estonia is their perception of stigmatization. The characteristics describing their disease, its medication, and complications were generally in accordance with the data from other countries, and also marital and educational status (except when assessing the stigmatization) were not statistically significant. Achieving better control of seizures and reducing side effects are essential in improving the quality of life of people with epilepsy, because this reduces the stigma associated with the condition. We emphasize to doctors the importance of psychological support in the care of patients with epilepsy. Unfortunately, the results of our study suggest that, at least in our country, many physicians ignore or do not recognize this actuality, considering it irrelevant.
In conclusion, the findings of this study confirm that psychosocial problems accompany the diagnosis of epilepsy. Although the study demonstrated quality of life decreases in subjects with epilepsy, we consider the results encouraging. No remarkable differences were found in terms of medical problems. A further study is required in this field within our community to help people with epilepsy to better understand their condition, to analyze the reasons for stigma, and if it is not possible to eliminate them completely, then to promote adjustment.
Acknowledgment: This study was supported by grants no. 1869 and 4342 from the Estonian Science Foundation.