- Email: [email protected]
Factors associated with quality of life in a low-income population with epilepsy
Epilepsy Research 127 (2016) 168–174
Contents lists available at www.sciencedirect.com
Epilepsy Research journal homepage: www.elsevier.com/locate/epilepsyres
Factors associated with quality of life in a low-income population with epilepsy Camilo Alfonso Espinosa Jovel (MD) a,c,∗ , Sergio Ramírez Salazar (MD) b , Carlos Rincón Rodríguez (MSc.) b , Fidel Ernesto Sobrino Mejía (MD) a,c a
University of La Sabana, Medical School, Neurology Postgraduate Department, Bogotá, Colombia University of La Sabana, Medical School, Bogotá, Colombia c Kennedy Western Hospital, Neurology Department, Bogotá, Colombia b
a r t i c l e
i n f o
Article history: Received 30 December 2015 Received in revised form 17 August 2016 Accepted 31 August 2016 Available online 1 September 2016 Keywords: Quality of life Epilepsy Low income Depression Daytime sleepiness
a b s t r a c t Objective: Currently few studies describe the variables that impact quality of life (QoL) in patients with epilepsy in low-income populations. The study aimed to establish relationships between QoL scores obtained through the QOLIE-10 inventory and clinical variables in patients older than 18 years diagnosed with epilepsy. Methods: We conducted an observational, descriptive, and cross-sectional study. We conducted consecutive recruitment of the data for all patients with an epilepsy diagnosis who were treated in the neurology department of Kennedy Western Hospital located in Bogota, Colombia. The variables that were statistically significant in the bivariate analysis were included in a multiple linear regression model. Results: 220 patients were evaluated. The 50th percentile of the total score of the QOLIE-10 scale was 70 (95% CI: 67,5–75). The demographic profile was characterized by low level of education, unemployment, and single marital status. The variables included in the regression model that significantly affected QoL were depression (p < 0.001), severe daytime sleepiness (p = 0.030), structural/metabolic etiology of epilepsy (p = 0.021), drug resistant epilepsy (p = 0.015), and epilepsy with undetermined antiepileptic drug response (p = 0.007). Conclusions: The QoL in patients with epilepsy from a low-economic population is determined primarily by depression, severe daytime sleepiness, etiology of epilepsy (structural/metabolic etiology), and the type of therapeutic response to antiepileptic drugs (drug resistant epilepsy and undetermined antiepileptic drug response). These data suggest the need to promote early diagnosis and treatment of psychiatric comorbidities and sleep disorders, as well as effective and timely therapeutic interventions to prevent drug resistance in epilepsy. © 2016 Published by Elsevier B.V.
1. Introduction Epilepsy is a chronic disease defined by the International League against Epilepsy (ILAE) as a brain disorder, characterized by a predisposition to present seizures, generating neurobiological, cognitive, psychological, and social consequences (Berg et al., 2010). Recent studies have shown that epilepsy is a prevalent disease with high social and economic impact, and is more frequent in lowincome countries (Ngugi et al., 2010). The estimated prevalence for active epilepsy (with seizures in the past 5 years) in developed
∗ Corresponding author. Permanant address: Campus del Puente del Común, Km. 7, Autopista Norte, Bogotá, Colombia. E-mail address: camilo [email protected] (C.A. Espinosa Jovel). http://dx.doi.org/10.1016/j.eplepsyres.2016.08.031 0920-1211/© 2016 Published by Elsevier B.V.
countries is 4.9 per 1.000, compared with 12.7 per 1.000 for rural areas in developing countries (Ngugi et al., 2010). In Colombia, the latest study showed that the overall prevalence is 11.3 per 1000, with small regional variations, except in the eastern region where prevalence is 23 per 1000 (Velez and Eslava-Cobos, 2006). Epilepsy is not only a common disease, but also a disabling one; these patients have a higher risk of premature death, psychosocial dysfunction, and poor quality of life (QoL) (Fazel et al., 2013). QoL is a broad concept that is somewhat subjective; it is defined by the World Health Organization as the perception that an individual has of his place in existence, in the context of culture and system securities in which they live, and in relation to their goals, expectations, standards, and concerns (Jacoby et al., 2009). The QoL for epilepsy patients has been evaluated for more than 20 years with the application of different questionnaires that allow
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
standardization of this concept, and in the recent years, QoL has become one of the main outcomes of therapeutic interventions for epilepsy patients (Vergara Palma et al., 2015). The QOLIE-10 inventory, which is a reduced version of QOLIE-31, is an effective questionnaire for measuring the QoL in patients with epilepsy, and recently Viteri et al. validated the Spanish version of the QOLIE-10 inventory, proving it to be valid and reliable (Viteri et al., 2008). Recent studies have shown that the main determinants of QoL in patients with epilepsy are seizure frequency, depression, and sleep disorders, while the evidence for age, marital status, education level, employment status, and antiepileptic drugs as predictors of QoL is inconsistent (Taylor et al., 2011). Since depression and sleep disorders represent frequent comorbidities in patients with epilepsy, and have significant impact on the QoL of these people, some questionnaires have been recently validated in the Spanish language in order to objectify these diagnoses. Examples include the NDDI-E inventory for depression in epilepsy which was validated in the Spanish language by Di Capua et al. (2012), and the Epworth scale for daytime sleepiness, which was validated in the Colombian population by Chica-Urzola et al. (2007). Currently there are few studies that describe the variables that affect the QoL in patients with epilepsy in low-income populations. For this reason, and because Kennedy Western Hospital is a public hospital in Bogota, Colombia, where most patients with epilepsy have a low income and a high degree of social vulnerability (Espinosa Jovel et al., 2014), we decided to conduct this study in order to describe the main clinical and socio-demographic variables that affect the QoL in this population. 2. Materials and methods 2.1. Study population The Kennedy Western Hospital is a public hospital which functions as a tertiary care center for people from Kennedy (Bogotá, Colombia) and nearby areas, representing a population of approximately 2,741,000 people according to national statistics. The majority of people have a low socioeconomic status, with the highest unemployment rate (16.3%) of all districts in Bogota (http://www.culturarecreacionydeporte.gov.co/localidades/ kennedy). A total of 53% of the inhabitants of Kennedy live in poverty and 13.3% live in extreme poverty (http://www. culturarecreacionydeporte.gov.co/localidades/kennedy). In 2014, the neurology department of the Kennedy Western Hospital, evaluated a total of 2649 patients diagnosed with epilepsy, both in outpatient and emergency consultation. We also have recently published a demographic and clinical description of patients with epilepsy attending the Kennedy Western Hospital, showing that 86.8% have a low socioeconomic status, 76.7% are unemployed, and only 10.2% have completed college (Espinosa Jovel et al., 2014). 2.2. Study design We conducted an observational, descriptive, cross-sectional study, in which we used a consecutive sampling technique, taking information from patients older than 18 years who had an epilepsy diagnosis for at least 1 year and were receiving antiepileptic drugs at the time of the evaluation. We included patients who were treated at the neurology department of Kennedy Western Hospital between September 2014 and March 2015. We excluded patients with physical and/or mental limitations that did not allow proper registration of data collection, as well as patients with a history of epilepsy surgery. Epilepsy surgery is widely associated with a better QoL, and we have previously shown that in our population surgical intervention could change the QOLIE-10 score by almost
169
21 points (Vergara Palma et al., 2015). For this reason, we decided to exclude patients with a history of epilepsy surgery, thus avoiding false-positive results in the QOLIE-10 inventory. All patients gave informed consent. This study was approved by the research ethics committee of the University of La Sabana, by Act No. 46 of September 2014. The information was obtained during a specialized consultation, where the instrument of data collection was applied by the neurologist, including sociodemographic variables (age, sex, education, marital status, and occupational activity) and clinical variables (risk factors for epilepsy, age of diagnosis, type of seizures, frequency of seizures, treatment with antiepileptic drugs, and therapeutic response). The etiology and the classification of epilepsy was based on the age of presentation, semiology, evolution of the disease, and familial history, and complemented with electroencephalogram, video EEG, brain MRI, and neuropsychological assessment (when available in the medical record). We did not have access to genetic confirmation in cases of probable genetic epilepsy. During the consultation, the QOLIE-10 inventory, the NDDI-E, and the Epworth scale were applied by the neurologist. Based on the validation studies of each of these questionnaires (Di Capua et al., 2012; Chica-Urzola et al., 2007), major depression was defined as a score ≥13 on NDDI-E, and mild to severe daytime sleepiness was defined as scores of 10–16 and 17–24 respectively on the Epworth scale. The definition of epilepsy used in this study was based on the recommendations of the ILAE report in 2014 (at least 2 unprovoked (or reflex) seizures occurring >24 h apart) (Fisher et al., 2014). The definition of therapeutic response to antiepileptic drugs was based on the proposal of ILAE in 2010, where 3 types are defined: controlled, resistant, and undetermined epilepsy (Kwan et al., 2010). We also consider (subjectively) that some patients with 1 or 2 mild focal seizures in the last year were controlled, since they had no impact on daily functioning and it was their basal situation during many years. 2.3. Statistical analysis A description of each variable, based on the median and interquartile range for quantitative variables and absolute and relative frequency for categorical variables, was made. A bivariate analysis was made to determine the association between sociodemographic and clinical variables (including the Epworth and NDDI-E score scales) and the total score of the QOLIE-10 scale. The non-parametric Wilcoxon test and Kruskal-Wallis test were used on categorical independent variables, and the Spearman’s rank correlation coefficient on quantitative independent variables. The variables that were statistically significant in the bivariate analysis were included in a multiple linear regression model. The outcome variable used in the regression model was the total score of the QOLIE-10 inventory, which was defined as a quantitative continuous variable. Model assumptions were evaluated according to the Shapiro-Wilk test. The scatter plot between residuals and model predictions and the coefficient of determination r2 was calculated. The presence of collinearity was evaluated. The data were analyzed with the STATA 12 program. 3. Results 3.1. Sociodemographic and clinical data A total of 220 patients were included, of whom 51.8% (n = 114) were men. The 50th percentile (P50) of age was 35 years (interquartile range: 32–37), with a range of 18–79 years. The P50 for the age of diagnosis of epilepsy was 14 years (interquartile range: 12.9–15). A total of 6.8% (n = 15) of patients were illiterate, and only 14.5% (n = 32) had technical careers and/or a college education. A total
170
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
Table 1 Main sociodemographic and clinical aspects. Age (P50, interquartile range, range)
35 (32–37), 18–79
Age at diagnosis of epilepsy (P50, interquartile range, range)
14 (12.9–15), 1–69
QOLIE- 10 score (P50, interquartile range, range)
70 (67.5–75), 32.5–100
NDDI-E score Frequency, (percentage)
≥13 <13
86 (39) 134(60.9)
Epworth score Frequency, (percentage)
0–9 10–16 17–24
138 (62.7) 61 (27.7) 21 (9.5)
Sex Frequency, (percentage)
Female Male
106 (48.1) 114 (51.8)
Level of education Frequency, (percentage)
Illiterate Incomplete primary school Complete primary school Incomplete secondary school Complete secondary school Technical/college career
15 (6.82) 1 (0.47) 70(31.82) 45 (20.4) 57 (25.9) 32 (14.55)
Married
32 (14.5)
Separated Single Free Union Widower
23 (10.4) 126 (57.2) 36 (16.3) 3 (1.3)
Labor activity Frequency, (percentage)
No Yes
147 (66.8) 73 (33.1)
Family history of epilepsy Frequency, (percentage)
No Yes
153 (69.5) 67 (30.4)
Etiology of epilepsy Frequency, (percentage)
Unknown cause Structural/metabolic Genetic
82 (37.2) 99 (45) 39 (17.7)
Type of seizure Frequency, (percentage)
Focal with impaired consciousness Focal with secondary generalization Focal without impaired consciousness Generalized
48 (21.8) 86 (39) 8 (3.6) 78 (35.4)
Seizure frequency Frequency, (percentage)
No seizures in the last year Seizures in the last year
79 (35.9) 141 (64)
Antiepileptic treatment Frequency, (percentage)
Monotherapy Polytherapy
123 (55.9) 97 (44)
Response to antiepileptic drugs Frequency, (percentage)
Controlled Undetermined Resistant
103 (46.8) 57 (25.9) 60 (27.2)
– Marital Status Frequency, (percentage)
of 57.2% (n = 126) of patients were single, and 66.8% (n = 147) were unemployed at the time of evaluation. Of the 147 patients who were unemployed, 55.7% (n = 82) considered their epilepsy the cause of unemployment. A total of 35.4% (n = 78) of patients presented generalized seizures, and 55.9% (n = 123) were in monotherapy with antiepileptic drugs at the time of evaluation. A total of 27.2% (n = 60) of patients were classified as having drug-resistant epilepsy. A total of 39% (n = 86) of patients had major depression, as measured by a score of ≥13 on the NDDI-E scale. A total of 9.5% (n = 21) of patients had severe daytime sleepiness as measured by a score of ≥17 on the Epworth scale. The P50 of the total score of the QOLIE-10 scale was 70 (interquartile range: 67.5–75) with a range of 32.5–100. Table 1. 3.2. Bivariate analysis After performing the bivariate analysis, the following variables had statistical significance with the total score of the QOLIE-10 inventory: marital status (p = 0.016), unemployment secondary to epilepsy (p < 0.001), type of seizures (p = 0.007), seizures
in the last year (p < 0.001), polytherapy (p < 0.001), therapeutic response (p < 0.001), etiology of epilepsy (p = 0.049), and NDDI-E and Epworth scores (p < 0.001). Table 2. Fig. 1 shows the relationship between the total score of QOLIE-10 and major depression (≥13 NDDI-E). The higher the score on the NDDI-E, the lower the perceived QoL in the QOLIE-10 scale. The age, sex, educational level, and age of diagnosis did not show statistical significance within the total score of the QOLIE-10 inventory. 3.3. Linear regression analysis Within the variables with statistical significance in the bivariate analysis, we selected those with the lower P50 in the total score of QOLIE-10 for the regression model. Due to the low number of patients in the variable “Unemployment secondary to epilepsy” (n = 82), we decided not to include this variable in the regression analysis, as it could negatively impact the efficacy of the model. A first linear regression model was carried out with the variables mentioned before, documenting an r2 of 0.5215. Model
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
171
Table 2 Variables associated with the total score of the QOLIE-10 in the bivariate analysis. 50th percentile of total score of QOLIE-10
P25-P75 of total score of QOLIE-10
P value
Marital Status
Separated Married Single Widowed Free Union
60 76.2 70 75 75
55–75 65–85 57.5–80 60–82.5 72.5–80
0.016*
Unemployment secondary to epilepsy
Yes No
60 80
55–72.5 67.5–85
<0.001**
Type of seizure
Focal with impaired consciousness Focal with secondary generalization Focal without impaired consciousness Generalized
67.5 68.7 81.2 75
53–75 57.5–80 66.8–86.8 60–83.1
0.007*
Presence of seizure in the last year
Yes No
62.5 80
60–67.5 77.5–80
<0.001*
Antiepileptic treatment
Polytherapy Monotherapy
60 77.5
55–75 65–85
<0.001**
Therapeutic response to antiepileptic drugs
Resistant Undetermined Controlled
60 65 80
52.5–64.3 57.5–75 72.5–85
<0.001*
Etiology of epilepsy
Structural/metabolic Unknown Genetic
67.5 72.5 75
62.5–72.5 59.3–82.5 62.5–85
0.049*
Major depression NDDI-E
≥13 <13
60 80
52.5–67.5 70–85
<0.001**
Daytime sleepiness Epworth
≤9 10–16 17–24
75 62.5 57.5
62.5–83.1 52.5–77.5 53.7–66.2
<0.001*
* **
Kruskal Wallis test. Wilcoxon test.
Fig. 1. Scatter plot showing the relationship between the total score of QOLIE-10 and major depression (≥13 NDDI-E). The higher the score on the NDDIE, the lower the perceived quality of life in the QOLIE-10 scale. Data were analyzed with the correlation coefficient of Spearman.
assumptions from the Shapiro-Wilk test (prob > z = 0.04449) and the scatter plot of the residuals vs. predictions were verified. In this model, 8 atypical cases were detected in which the QoL
was significantly different from the rest of the study population, so we decided to exclude these patients from the model since the residuals were below −24.7. In the second model, the
172
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
Table 3 Linear regression model.
Marital status: Separated Focal seizure with impaired consciousness Focal seizure with secondary generalization Presence of seizures in the last year Polytherapy with antiepileptic drugs Undetermined therapeutic response Drug-resistant epilepsy Structural/metabolic etiology Major depression (score ≥13 NDDI-E) Severe daytime sleepiness (Epworth: 17–24) Mild daytime sleepiness (Epworth 10–16) a
Coefficient
IC 95%
P value
−3.475 −2.012 0.496 −2.446 −2.653 −6.104 −6.209 −3.254 −13.835 −4.955 −1.674
−7.561–0.609 −5.748–1.722 −2.965–3.064 −6.612–1.719 −5.638–0.330 −10.525–(−)1.684 −11.193–(−)1.224 −6.011–(−)0.496 −16.586–(−)11.083 −9.417–(−)0.494 −4.644–1.294
0.095 0.289 0.974 0.248 0.081 0.007a 0.015a 0.021a <0.001a 0.030a 0.627
Variables with significant association.
8 cases previously mentioned were excluded, so 212 patients were analyzed. The variables included in the second regression model, which showed a significant correlation with the total score of the QOLIE-10 scale were: major depression (NDDI-E ≥13) (p < 0.001), severe daytime sleepiness (Epworth 17–24) (p = 0.030), structural/metabolic etiology of epilepsy (p = 0.021), drug resistant epilepsy (p = 0.015), and epilepsy with undetermined antiepileptic drug response (p = 0.007). This model explains 60% of the variability in QoL (r2 = 0.5968). Table 3. The model assumptions were verified by a Shapiro-Wilk test (prob > z = 0.62013) and the scatter plot of the residuals vs. predictions. 4. Discussion In this study, through a multiple linear regression model, we demonstrated that the variables that affect the QoL in patients with epilepsy who have a low-income and are in a socially vulnerable condition are: depression (NDDI-E ≥13), severe daytime sleepiness (Epworth 17–24), structural/metabolic etiology of epilepsy, drug resistant epilepsy and therapeutic response to antiepileptic drugs, specifically, the undetermined antiepileptic drug response. 4.1. Sociodemographic profile The demographic profile of the population evaluated in this study was characterized by low education, unemployment, and single status. These findings are not exclusive to people with low incomes (The RESt-1 Group, 2000). The REST-1 group showed that in some European countries (Italy, Germany, Spain, Holland, England, Portugal, and Russia) patients with epilepsy studied less than the general population, had higher unemployment rates, and were more likely to be single or divorced (The RESt-1 Group, 2000). One of the main differences was based on educational profile, as the illiteracy rate of patients with epilepsy in our study was much higher than the illiteracy rate of epilepsy patients documented in European populations with strong economic resources (6.8% vs. 2%) (The RESt-1 Group, 2000). Although the sample size and methodology used in our study did not reflect the demographic profile of all patients with epilepsy in our population, it allowed us to make an initial approach. Although in the bivariate analysis a significant relation between marital status and labor activity (unemployment secondary to epilepsy) within the total score of the QOLIE-10 scale was documented, in the linear regression model these demographic variables were found not significant for perception of QoL. Regarding the labor activity, although there is a high percent of unemployment in the patients evaluated (66.8%), this variable was found not significant for perception of QoL. The evidence of labor activity as a predictor of QoL is controversial, and some studies of other populations with similar demographic characteristics
showed that low socioeconomic status is an important predictor of QoL in patients with epilepsy (Alanis-Guevara et al., 2005).
4.2. Clinical profile The clinical profile of patients with epilepsy assessed in this study was similar to that observed in high-income populations. Regarding the type of seizures, most of our patients had focal seizures (64.6%), and although the epidemiology of the types of seizures is highly variable and depends on many conditions, some studies in high-income populations have found similar data (Forsgren, 1992; Luengo et al., 2001). As for the etiology of epilepsy, and taking into account the classification proposed by the ILAE in 2010, 45% of our patients had a structural or metabolic cause of epilepsy, followed by etiology of unknown cause in 45.7% of patients. Similar etiological profiles have been found in highincome population studies. The clinical response pattern of our population toward antiepileptic drugs was similar to reported patterns in high-income populations (Torres-Ferrús et al., 2013). In our study 46.8% of patients were controlled and 27.2% were classified as having drug resistant epilepsy. These data are similar to those reported by Brodie et al., in Glasgow, Scotland, where it was reported that about 59% of patients will remain without seizures and be considered controlled, 25% will be resistant to antiepileptic drugs, and the remaining 16% will fluctuate between relapses and periods of freedom from seizures over 1 year (Brodie et al., 2012). In the linear regression model, the clinical variables that impacted most significantly on perception of QoL were: structural/metabolic etiology of epilepsy, drug resistant epilepsy and therapeutic response to antiepileptic drugs, specifically, the undetermined antiepileptic drug response. This data is consistent with other published data in similar populations (Lavados et al., 1992; Milovanovic´ et al., 2014). The frequency of seizures represents one of the main predictors of QoL in most of the populations studied (Alanis-Guevara et al., 2005; Milovanovic´ et al., 2014; Loring et al., 2004); however, although our study documented a significant association in the bivariate analysis, this variable did not show a significant impact on the perception of QoL in the linear regression model. Regarding antiepileptic drugs, we did not specify the type of medications; we only mentioned the number of medications (polytherapy or monotherapy). Although this variable was statistically significant in the bivariate analysis, in the linear regression model it did not show a strong association within the QOLIE-10 total score. It is well known that some medications can influence the mood (e.g. valproate and lamotrigine, among others) (Chiu et al., 2013) and thus indirectly affect the QoL. Some others medications can be associated with daytime sleepiness (e.g. benzodiazepines, among others) and also indirectly affect the QoL in patients with epilepsy (Perucca and Gilliam, 2012). As we did not specify the type of med-
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
ication, it could represent a selection bias that must be considered when analyzing the data. 4.3. Depression Depression is a major psychiatric comorbidity in patients with epilepsy, and may occur in up to 30–35% of cases, more frequently in patients with focal onset epilepsy and drug resistant epilepsy (Tellez-Zenteno et al., 2007). In our study, 39% of patients had depressive symptoms with a score ≥13 on the NDDI-E scale, posing a probable diagnosis of major depressive disorder, and showing a similar prevalence documented in multiple populations, which is independent of the socioeconomic profile (Alanis-Guevara et al., 2005; Milovanovic´ et al., 2014; Loring et al., 2004; Tellez-Zenteno et al., 2007). Some studies have shown that depression is underdiagnosed, undertreated, and representative of one of the main predictors of QoL, with even more impact than the seizure frequency (Boylan et al., 2004). Not only was there a high prevalence of depression documented in our population, this was the variable with the greatest impact on the perception of QoL, even more than other important variables such as seizure frequency, excessive daytime sleepiness, etiology of epilepsy, and therapeutic response to antiepileptic drugs. Some studies in populations with similar sociodemographic characteristics found data comparable to ours. In Brazil, Tedrus et al. documented that about 20.5% of patients with epilepsy have depression, representing the main predictor of the perception of QoL measured by QOLIE-31 inventory (Tedrus et al., 2013). 4.4. Sleep disorders Although sleep disorders in patients with epilepsy are a prevalent comorbidity with multiple pathophysiologic consequences, they are widely underdiagnosed (Manni and Terzaghi, 2010). In the clinical manifestations of sleep disorders, daytime sleepiness is one of the most common symptoms, and in patients with epilepsy it is a clinical marker of sleep apnea (Manni and Terzaghi, 2010). Sleep disorders and excessive daytime sleepiness are one of the main predictors of QoL in patients with epilepsy, and some studies in similar populations have found data comparable to ours (de Weerd et al., 2004). A total of 27.7% of the patients in our study had mild daytime sleepiness, and 9.5% had severe daytime sleepiness, which may reflect an underlying sleep disorder. It is important to mention that the daytime sleepiness is just one symptom of a specific disturbance, and in patients with epilepsy it can be influenced by many factors (Manni and Terzaghi, 2010). We did not study specific sleep disturbances, as this was not the aim for this study; however, it is important to take into account that the adverse effects of drugs, the decline in physical activity, and the presence of metabolic comorbidities (e.g. hypothyroidism, among others) are conditions that can also promote daytime sleepiness in patients with epilepsy (Manni and Terzaghi, 2010; de Weerd et al., 2004). Regardless of the cause of daytime sleepiness, we documented that this variable (Severe daytime sleepiness with an Epworth score of 17–24) negatively impacted the QoL, which increases the need to promote early diagnosis and prompt treatment of sleep disorders in patients with epilepsy. 5. Study limitations This study was based on a medical record, not in a population register, so the sample size and the sampling strategy, which were based on consecutive recruitment, do not allow for evaluation of the whole population. The regression model used in this study explains 60% of the variability in QoL, and although the concept of QoL can be very broad and be influenced by multiple variables unrelated to
173
the health and the wealth of the person, some clinical and sociodemographic conditions were not measured in the study (severity of seizures, anxiety disorders, specific structural/metabolic cause of epilepsy, type of antiepileptic drug, and family support, among others) and could also have affected the perception of QoL in this population. In addition, the “seizure frequency” variable was defined as a qualitative nominal variable, and not as a quantitative continuous variable, representing a probable selection bias. In addition, when the general QoL questionnaires such as SF-36, are compared to specific questionnaires about epilepsy such as QOLIE-31 or QOLIE-10, the latter could be less like to reveal the effect of demographic, sociocultural, and economic variables on QoL (Stavem et al., 2000). 6. Conclusions The QoL in epilepsy patients from a low-income population is determined primarily by depression, severe daytime sleepiness, structural/metabolic etiology of epilepsy, drug resistant epilepsy and therapeutic response to antiepileptic drugs, specifically, the undetermined antiepileptic drug response. These data suggest the need to promote early diagnosis and treatment of psychiatric comorbidities and sleep disorders, as well as effective and timely therapeutic interventions to prevent drug resistance in epilepsy, in order to improve the perception of QoL. Epilepsy must be considered a syndrome, in which seizures are the main symptom, but not the only one, and there are multiple associated conditions such as psychiatric comorbidity, sleep disorders, and psychosocial dysfunction, which are common, disabling, and possibly much more influential in these patients’ perceptions of QoL. Conflict of interest The authors have no conflict of interest to disclose. Acknowledgments The authors acknowledge the members of the Neurology department and the Neurology postgraduate program of the University of La Sabana: Dr. Erik Sánchez, Dr. Juan Vergara, Dr. Daniel Hedmont, Dr. Gustavo Barrios, Dr. Javier Vicini, and Dr. Maria Claudia Angulo; the Medical school of the University of La Sabana: Dr. Fernando Ríos, Dr. Maria José Maldonado, and Dr. Diana Diaz; the patients treated in the Neurology department of the Kennedy Western Hospital; the directors of the Kennedy Western Hospital, the nursing personnel and doctors in training at Kennedy Western Hospital. References ˜ E., Corona, T., et al., 2005. Sleep disturbances, Alanis-Guevara, I., Pena, socioeconomic status, and seizure control as main predictors of quality of life in epilepsy. Epilepsy Behav. 7, 481–485. Anon, http://www.culturarecreacionydeporte.gov.co/localidades/kennedy, 2016. Berg, A.T., Berkovic, S.F., Brodie, M.J., et al., 2010. Revised terminology and concepts for organization of seizures and epilepsies: report of the ILAE Commission on Classification and Terminology, 2005–2009. Epilepsia 51, 676–685. Boylan, L.S., Flint, L.A., Labovitz, D.L., et al., 2004. Depression but not seizure frequency predicts quality of life in treatment-resistant epilepsy. Neurology 62, 258–261. Brodie, M.J., Barry, S.J., Bamagous, G.A., et al., 2012. Patterns of treatment response in newly diagnosed epilepsy. Neurology 78, 1548–1554. Chica-Urzola, H.L., Escobar-Córdoba, F., Eslava-Schmalbach, J., 2007. Validación de la escala de somnolencia de epworth. Rev. Salud Pública 9, 558–567. Chiu, C.T., Wang, Z., Hunsberger, J.G., et al., 2013. Therapeutic potential of mood stabilizers lithium and valproic acid: beyond bipolar disorder. Pharmacol. Rev. 65, 105–142. Di Capua, D., Garcia-Garcia, M.E., Reig-Ferrer, A., et al., 2012. Validation of the Spanish version of the neurological disorders depression inventory for epilepsy (NDDI-E). Epilepsy Behav. 24, 493–496.
174
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
Espinosa Jovel, C.A., Pardo, C.M., Moreno, C.M., et al., 2014. Demographic and social profile of epilepsy in a vulnerable low-income population in Bogotá, Colombia. Neurología, pii: S0213-4853(14)00242-4. Fazel, S., Wolf, A., Långström, N., et al., 2013. Premature mortality in epilepsy and the role of psychiatric comorbidity: a total population study. Lancet 382, 1646–1654. Fisher, R.S., Acevedo, C., Arzimanoglou, A., et al., 2014. ILAE official report: a practical clinical definition of epilepsy. Epilepsia 55, 475–482. Forsgren, L., 1992. Prevalence of epilepsy in adults in northern Sweden. Epilepsia 33, 450–458. Jacoby, A., Snape, D., Baker, G.A., 2009. Determinants of quality of life in people with epilepsy. Neurol. Clin. 27, 843–863. Kwan, P., Arzimanoglou, A., Berg, A.T., et al., 2010. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 51, 1069–1077. Lavados, J., Germain, L., Morales, A., et al., 1992. A descriptive study of epilepsy in the district of El Salvador, Chile, 1984–1988. Acta Neurol. Scand. 85, 249–256. Loring, D.W., Meador, K.J., Lee, G.P., 2004. Determinants of quality of life in epilepsy. Epilepsy Behav. 5, 976–980. Luengo, A., Parra, J., Colás, J., et al., 2001. Prevalence of epilepsy in northeast Madrid. J. Neurol. 248, 762–767. Manni, R., Terzaghi, M., 2010. Comorbidity between epilepsy and sleep disorders. Epilepsy Res. 90, 171–177. ´ M., Martinovic, ´ Zˇ , Toˇskovic, ´ O., 2014. Determinants of quality of life in Milovanovic, people with epilepsy in Serbia. Epilepsy Behav. 31, 160–166. Ngugi, A.K., Bottomley, C., Kleinschmidt, I., et al., 2010. Estimation of the burden of active and life-time epilepsy: a meta-analytic approach. Epilepsia 51, 883–890.
Perucca, P., Gilliam, F.G., 2012. Adverse effects of antiepileptic drugs. Lancet Neurol. 11, 792–802. Stavem, K., Loge, J.H., Kaasa, S., 2000. Health status of people with epilepsy compared with general reference population. Epilepsia 41, 85–90. Taylor, R.S., Sander, J.W., Taylor, R.J., et al., 2011. Predictors of health-related quality of life and costs in adults with epilepsy: a systematic review. Epilepsia 52, 2168–2180. Tedrus, G.M., Fonseca, L.C., Carvalho, R.M., 2013. Epilepsy and quality of life: socio-demographic and clinical aspects, and psychiatric co-morbidity. Arq. Neuropsiquiatr. 71, 385–391. Tellez-Zenteno, J.F., Patten, S.B., Jetté, N., et al., 2007. Psychiatric comorbidity in epilepsy: a population-based analysis. Epilepsia 48, 2336–2344. Social aspects of epilepsy in the adult in seven European countries., 2000. Epilepsia 41, 998–1004. Torres-Ferrús, M., Toledo, M., González-Cuevas, M., et al., 2013. Aetiology and treatment of epilepsy in a series of 1.557 patients. Rev. Neurol. 57, 306–312. Velez, A., Eslava-Cobos, J., 2006. Epilepsy in Colombia: epidemiologic profile and classification of epileptic seizures and syndromes. Epilepsia 47, 193–201. Vergara Palma, J., Espinosa Jovel, C.A., Vergara, T., et al., 2015. Impact of epilepsy surgery on the quality of life of a low-income population through the application of the Qolie-10 scale. Epilepsy Res. 110, 183–188. Viteri, C., Codina, M., Cobaleda, S., et al., 2008. Validation of the Spanish version of the QOLIE-10 quality of life in epilepsy questionnaire. Neurologia 23, 157–167. de Weerd, A., de Haas, S., Otte, A., et al., 2004. Subjective sleep disturbances in patients with partial epilepsy: a questionnaire based study on prevalence and impact on quality of life. Epilepsia 45, 1397–1404.
Contents lists available at www.sciencedirect.com
Epilepsy Research journal homepage: www.elsevier.com/locate/epilepsyres
Factors associated with quality of life in a low-income population with epilepsy Camilo Alfonso Espinosa Jovel (MD) a,c,∗ , Sergio Ramírez Salazar (MD) b , Carlos Rincón Rodríguez (MSc.) b , Fidel Ernesto Sobrino Mejía (MD) a,c a
University of La Sabana, Medical School, Neurology Postgraduate Department, Bogotá, Colombia University of La Sabana, Medical School, Bogotá, Colombia c Kennedy Western Hospital, Neurology Department, Bogotá, Colombia b
a r t i c l e
i n f o
Article history: Received 30 December 2015 Received in revised form 17 August 2016 Accepted 31 August 2016 Available online 1 September 2016 Keywords: Quality of life Epilepsy Low income Depression Daytime sleepiness
a b s t r a c t Objective: Currently few studies describe the variables that impact quality of life (QoL) in patients with epilepsy in low-income populations. The study aimed to establish relationships between QoL scores obtained through the QOLIE-10 inventory and clinical variables in patients older than 18 years diagnosed with epilepsy. Methods: We conducted an observational, descriptive, and cross-sectional study. We conducted consecutive recruitment of the data for all patients with an epilepsy diagnosis who were treated in the neurology department of Kennedy Western Hospital located in Bogota, Colombia. The variables that were statistically significant in the bivariate analysis were included in a multiple linear regression model. Results: 220 patients were evaluated. The 50th percentile of the total score of the QOLIE-10 scale was 70 (95% CI: 67,5–75). The demographic profile was characterized by low level of education, unemployment, and single marital status. The variables included in the regression model that significantly affected QoL were depression (p < 0.001), severe daytime sleepiness (p = 0.030), structural/metabolic etiology of epilepsy (p = 0.021), drug resistant epilepsy (p = 0.015), and epilepsy with undetermined antiepileptic drug response (p = 0.007). Conclusions: The QoL in patients with epilepsy from a low-economic population is determined primarily by depression, severe daytime sleepiness, etiology of epilepsy (structural/metabolic etiology), and the type of therapeutic response to antiepileptic drugs (drug resistant epilepsy and undetermined antiepileptic drug response). These data suggest the need to promote early diagnosis and treatment of psychiatric comorbidities and sleep disorders, as well as effective and timely therapeutic interventions to prevent drug resistance in epilepsy. © 2016 Published by Elsevier B.V.
1. Introduction Epilepsy is a chronic disease defined by the International League against Epilepsy (ILAE) as a brain disorder, characterized by a predisposition to present seizures, generating neurobiological, cognitive, psychological, and social consequences (Berg et al., 2010). Recent studies have shown that epilepsy is a prevalent disease with high social and economic impact, and is more frequent in lowincome countries (Ngugi et al., 2010). The estimated prevalence for active epilepsy (with seizures in the past 5 years) in developed
∗ Corresponding author. Permanant address: Campus del Puente del Común, Km. 7, Autopista Norte, Bogotá, Colombia. E-mail address: camilo [email protected] (C.A. Espinosa Jovel). http://dx.doi.org/10.1016/j.eplepsyres.2016.08.031 0920-1211/© 2016 Published by Elsevier B.V.
countries is 4.9 per 1.000, compared with 12.7 per 1.000 for rural areas in developing countries (Ngugi et al., 2010). In Colombia, the latest study showed that the overall prevalence is 11.3 per 1000, with small regional variations, except in the eastern region where prevalence is 23 per 1000 (Velez and Eslava-Cobos, 2006). Epilepsy is not only a common disease, but also a disabling one; these patients have a higher risk of premature death, psychosocial dysfunction, and poor quality of life (QoL) (Fazel et al., 2013). QoL is a broad concept that is somewhat subjective; it is defined by the World Health Organization as the perception that an individual has of his place in existence, in the context of culture and system securities in which they live, and in relation to their goals, expectations, standards, and concerns (Jacoby et al., 2009). The QoL for epilepsy patients has been evaluated for more than 20 years with the application of different questionnaires that allow
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
standardization of this concept, and in the recent years, QoL has become one of the main outcomes of therapeutic interventions for epilepsy patients (Vergara Palma et al., 2015). The QOLIE-10 inventory, which is a reduced version of QOLIE-31, is an effective questionnaire for measuring the QoL in patients with epilepsy, and recently Viteri et al. validated the Spanish version of the QOLIE-10 inventory, proving it to be valid and reliable (Viteri et al., 2008). Recent studies have shown that the main determinants of QoL in patients with epilepsy are seizure frequency, depression, and sleep disorders, while the evidence for age, marital status, education level, employment status, and antiepileptic drugs as predictors of QoL is inconsistent (Taylor et al., 2011). Since depression and sleep disorders represent frequent comorbidities in patients with epilepsy, and have significant impact on the QoL of these people, some questionnaires have been recently validated in the Spanish language in order to objectify these diagnoses. Examples include the NDDI-E inventory for depression in epilepsy which was validated in the Spanish language by Di Capua et al. (2012), and the Epworth scale for daytime sleepiness, which was validated in the Colombian population by Chica-Urzola et al. (2007). Currently there are few studies that describe the variables that affect the QoL in patients with epilepsy in low-income populations. For this reason, and because Kennedy Western Hospital is a public hospital in Bogota, Colombia, where most patients with epilepsy have a low income and a high degree of social vulnerability (Espinosa Jovel et al., 2014), we decided to conduct this study in order to describe the main clinical and socio-demographic variables that affect the QoL in this population. 2. Materials and methods 2.1. Study population The Kennedy Western Hospital is a public hospital which functions as a tertiary care center for people from Kennedy (Bogotá, Colombia) and nearby areas, representing a population of approximately 2,741,000 people according to national statistics. The majority of people have a low socioeconomic status, with the highest unemployment rate (16.3%) of all districts in Bogota (http://www.culturarecreacionydeporte.gov.co/localidades/ kennedy). A total of 53% of the inhabitants of Kennedy live in poverty and 13.3% live in extreme poverty (http://www. culturarecreacionydeporte.gov.co/localidades/kennedy). In 2014, the neurology department of the Kennedy Western Hospital, evaluated a total of 2649 patients diagnosed with epilepsy, both in outpatient and emergency consultation. We also have recently published a demographic and clinical description of patients with epilepsy attending the Kennedy Western Hospital, showing that 86.8% have a low socioeconomic status, 76.7% are unemployed, and only 10.2% have completed college (Espinosa Jovel et al., 2014). 2.2. Study design We conducted an observational, descriptive, cross-sectional study, in which we used a consecutive sampling technique, taking information from patients older than 18 years who had an epilepsy diagnosis for at least 1 year and were receiving antiepileptic drugs at the time of the evaluation. We included patients who were treated at the neurology department of Kennedy Western Hospital between September 2014 and March 2015. We excluded patients with physical and/or mental limitations that did not allow proper registration of data collection, as well as patients with a history of epilepsy surgery. Epilepsy surgery is widely associated with a better QoL, and we have previously shown that in our population surgical intervention could change the QOLIE-10 score by almost
169
21 points (Vergara Palma et al., 2015). For this reason, we decided to exclude patients with a history of epilepsy surgery, thus avoiding false-positive results in the QOLIE-10 inventory. All patients gave informed consent. This study was approved by the research ethics committee of the University of La Sabana, by Act No. 46 of September 2014. The information was obtained during a specialized consultation, where the instrument of data collection was applied by the neurologist, including sociodemographic variables (age, sex, education, marital status, and occupational activity) and clinical variables (risk factors for epilepsy, age of diagnosis, type of seizures, frequency of seizures, treatment with antiepileptic drugs, and therapeutic response). The etiology and the classification of epilepsy was based on the age of presentation, semiology, evolution of the disease, and familial history, and complemented with electroencephalogram, video EEG, brain MRI, and neuropsychological assessment (when available in the medical record). We did not have access to genetic confirmation in cases of probable genetic epilepsy. During the consultation, the QOLIE-10 inventory, the NDDI-E, and the Epworth scale were applied by the neurologist. Based on the validation studies of each of these questionnaires (Di Capua et al., 2012; Chica-Urzola et al., 2007), major depression was defined as a score ≥13 on NDDI-E, and mild to severe daytime sleepiness was defined as scores of 10–16 and 17–24 respectively on the Epworth scale. The definition of epilepsy used in this study was based on the recommendations of the ILAE report in 2014 (at least 2 unprovoked (or reflex) seizures occurring >24 h apart) (Fisher et al., 2014). The definition of therapeutic response to antiepileptic drugs was based on the proposal of ILAE in 2010, where 3 types are defined: controlled, resistant, and undetermined epilepsy (Kwan et al., 2010). We also consider (subjectively) that some patients with 1 or 2 mild focal seizures in the last year were controlled, since they had no impact on daily functioning and it was their basal situation during many years. 2.3. Statistical analysis A description of each variable, based on the median and interquartile range for quantitative variables and absolute and relative frequency for categorical variables, was made. A bivariate analysis was made to determine the association between sociodemographic and clinical variables (including the Epworth and NDDI-E score scales) and the total score of the QOLIE-10 scale. The non-parametric Wilcoxon test and Kruskal-Wallis test were used on categorical independent variables, and the Spearman’s rank correlation coefficient on quantitative independent variables. The variables that were statistically significant in the bivariate analysis were included in a multiple linear regression model. The outcome variable used in the regression model was the total score of the QOLIE-10 inventory, which was defined as a quantitative continuous variable. Model assumptions were evaluated according to the Shapiro-Wilk test. The scatter plot between residuals and model predictions and the coefficient of determination r2 was calculated. The presence of collinearity was evaluated. The data were analyzed with the STATA 12 program. 3. Results 3.1. Sociodemographic and clinical data A total of 220 patients were included, of whom 51.8% (n = 114) were men. The 50th percentile (P50) of age was 35 years (interquartile range: 32–37), with a range of 18–79 years. The P50 for the age of diagnosis of epilepsy was 14 years (interquartile range: 12.9–15). A total of 6.8% (n = 15) of patients were illiterate, and only 14.5% (n = 32) had technical careers and/or a college education. A total
170
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
Table 1 Main sociodemographic and clinical aspects. Age (P50, interquartile range, range)
35 (32–37), 18–79
Age at diagnosis of epilepsy (P50, interquartile range, range)
14 (12.9–15), 1–69
QOLIE- 10 score (P50, interquartile range, range)
70 (67.5–75), 32.5–100
NDDI-E score Frequency, (percentage)
≥13 <13
86 (39) 134(60.9)
Epworth score Frequency, (percentage)
0–9 10–16 17–24
138 (62.7) 61 (27.7) 21 (9.5)
Sex Frequency, (percentage)
Female Male
106 (48.1) 114 (51.8)
Level of education Frequency, (percentage)
Illiterate Incomplete primary school Complete primary school Incomplete secondary school Complete secondary school Technical/college career
15 (6.82) 1 (0.47) 70(31.82) 45 (20.4) 57 (25.9) 32 (14.55)
Married
32 (14.5)
Separated Single Free Union Widower
23 (10.4) 126 (57.2) 36 (16.3) 3 (1.3)
Labor activity Frequency, (percentage)
No Yes
147 (66.8) 73 (33.1)
Family history of epilepsy Frequency, (percentage)
No Yes
153 (69.5) 67 (30.4)
Etiology of epilepsy Frequency, (percentage)
Unknown cause Structural/metabolic Genetic
82 (37.2) 99 (45) 39 (17.7)
Type of seizure Frequency, (percentage)
Focal with impaired consciousness Focal with secondary generalization Focal without impaired consciousness Generalized
48 (21.8) 86 (39) 8 (3.6) 78 (35.4)
Seizure frequency Frequency, (percentage)
No seizures in the last year Seizures in the last year
79 (35.9) 141 (64)
Antiepileptic treatment Frequency, (percentage)
Monotherapy Polytherapy
123 (55.9) 97 (44)
Response to antiepileptic drugs Frequency, (percentage)
Controlled Undetermined Resistant
103 (46.8) 57 (25.9) 60 (27.2)
– Marital Status Frequency, (percentage)
of 57.2% (n = 126) of patients were single, and 66.8% (n = 147) were unemployed at the time of evaluation. Of the 147 patients who were unemployed, 55.7% (n = 82) considered their epilepsy the cause of unemployment. A total of 35.4% (n = 78) of patients presented generalized seizures, and 55.9% (n = 123) were in monotherapy with antiepileptic drugs at the time of evaluation. A total of 27.2% (n = 60) of patients were classified as having drug-resistant epilepsy. A total of 39% (n = 86) of patients had major depression, as measured by a score of ≥13 on the NDDI-E scale. A total of 9.5% (n = 21) of patients had severe daytime sleepiness as measured by a score of ≥17 on the Epworth scale. The P50 of the total score of the QOLIE-10 scale was 70 (interquartile range: 67.5–75) with a range of 32.5–100. Table 1. 3.2. Bivariate analysis After performing the bivariate analysis, the following variables had statistical significance with the total score of the QOLIE-10 inventory: marital status (p = 0.016), unemployment secondary to epilepsy (p < 0.001), type of seizures (p = 0.007), seizures
in the last year (p < 0.001), polytherapy (p < 0.001), therapeutic response (p < 0.001), etiology of epilepsy (p = 0.049), and NDDI-E and Epworth scores (p < 0.001). Table 2. Fig. 1 shows the relationship between the total score of QOLIE-10 and major depression (≥13 NDDI-E). The higher the score on the NDDI-E, the lower the perceived QoL in the QOLIE-10 scale. The age, sex, educational level, and age of diagnosis did not show statistical significance within the total score of the QOLIE-10 inventory. 3.3. Linear regression analysis Within the variables with statistical significance in the bivariate analysis, we selected those with the lower P50 in the total score of QOLIE-10 for the regression model. Due to the low number of patients in the variable “Unemployment secondary to epilepsy” (n = 82), we decided not to include this variable in the regression analysis, as it could negatively impact the efficacy of the model. A first linear regression model was carried out with the variables mentioned before, documenting an r2 of 0.5215. Model
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
171
Table 2 Variables associated with the total score of the QOLIE-10 in the bivariate analysis. 50th percentile of total score of QOLIE-10
P25-P75 of total score of QOLIE-10
P value
Marital Status
Separated Married Single Widowed Free Union
60 76.2 70 75 75
55–75 65–85 57.5–80 60–82.5 72.5–80
0.016*
Unemployment secondary to epilepsy
Yes No
60 80
55–72.5 67.5–85
<0.001**
Type of seizure
Focal with impaired consciousness Focal with secondary generalization Focal without impaired consciousness Generalized
67.5 68.7 81.2 75
53–75 57.5–80 66.8–86.8 60–83.1
0.007*
Presence of seizure in the last year
Yes No
62.5 80
60–67.5 77.5–80
<0.001*
Antiepileptic treatment
Polytherapy Monotherapy
60 77.5
55–75 65–85
<0.001**
Therapeutic response to antiepileptic drugs
Resistant Undetermined Controlled
60 65 80
52.5–64.3 57.5–75 72.5–85
<0.001*
Etiology of epilepsy
Structural/metabolic Unknown Genetic
67.5 72.5 75
62.5–72.5 59.3–82.5 62.5–85
0.049*
Major depression NDDI-E
≥13 <13
60 80
52.5–67.5 70–85
<0.001**
Daytime sleepiness Epworth
≤9 10–16 17–24
75 62.5 57.5
62.5–83.1 52.5–77.5 53.7–66.2
<0.001*
* **
Kruskal Wallis test. Wilcoxon test.
Fig. 1. Scatter plot showing the relationship between the total score of QOLIE-10 and major depression (≥13 NDDI-E). The higher the score on the NDDIE, the lower the perceived quality of life in the QOLIE-10 scale. Data were analyzed with the correlation coefficient of Spearman.
assumptions from the Shapiro-Wilk test (prob > z = 0.04449) and the scatter plot of the residuals vs. predictions were verified. In this model, 8 atypical cases were detected in which the QoL
was significantly different from the rest of the study population, so we decided to exclude these patients from the model since the residuals were below −24.7. In the second model, the
172
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
Table 3 Linear regression model.
Marital status: Separated Focal seizure with impaired consciousness Focal seizure with secondary generalization Presence of seizures in the last year Polytherapy with antiepileptic drugs Undetermined therapeutic response Drug-resistant epilepsy Structural/metabolic etiology Major depression (score ≥13 NDDI-E) Severe daytime sleepiness (Epworth: 17–24) Mild daytime sleepiness (Epworth 10–16) a
Coefficient
IC 95%
P value
−3.475 −2.012 0.496 −2.446 −2.653 −6.104 −6.209 −3.254 −13.835 −4.955 −1.674
−7.561–0.609 −5.748–1.722 −2.965–3.064 −6.612–1.719 −5.638–0.330 −10.525–(−)1.684 −11.193–(−)1.224 −6.011–(−)0.496 −16.586–(−)11.083 −9.417–(−)0.494 −4.644–1.294
0.095 0.289 0.974 0.248 0.081 0.007a 0.015a 0.021a <0.001a 0.030a 0.627
Variables with significant association.
8 cases previously mentioned were excluded, so 212 patients were analyzed. The variables included in the second regression model, which showed a significant correlation with the total score of the QOLIE-10 scale were: major depression (NDDI-E ≥13) (p < 0.001), severe daytime sleepiness (Epworth 17–24) (p = 0.030), structural/metabolic etiology of epilepsy (p = 0.021), drug resistant epilepsy (p = 0.015), and epilepsy with undetermined antiepileptic drug response (p = 0.007). This model explains 60% of the variability in QoL (r2 = 0.5968). Table 3. The model assumptions were verified by a Shapiro-Wilk test (prob > z = 0.62013) and the scatter plot of the residuals vs. predictions. 4. Discussion In this study, through a multiple linear regression model, we demonstrated that the variables that affect the QoL in patients with epilepsy who have a low-income and are in a socially vulnerable condition are: depression (NDDI-E ≥13), severe daytime sleepiness (Epworth 17–24), structural/metabolic etiology of epilepsy, drug resistant epilepsy and therapeutic response to antiepileptic drugs, specifically, the undetermined antiepileptic drug response. 4.1. Sociodemographic profile The demographic profile of the population evaluated in this study was characterized by low education, unemployment, and single status. These findings are not exclusive to people with low incomes (The RESt-1 Group, 2000). The REST-1 group showed that in some European countries (Italy, Germany, Spain, Holland, England, Portugal, and Russia) patients with epilepsy studied less than the general population, had higher unemployment rates, and were more likely to be single or divorced (The RESt-1 Group, 2000). One of the main differences was based on educational profile, as the illiteracy rate of patients with epilepsy in our study was much higher than the illiteracy rate of epilepsy patients documented in European populations with strong economic resources (6.8% vs. 2%) (The RESt-1 Group, 2000). Although the sample size and methodology used in our study did not reflect the demographic profile of all patients with epilepsy in our population, it allowed us to make an initial approach. Although in the bivariate analysis a significant relation between marital status and labor activity (unemployment secondary to epilepsy) within the total score of the QOLIE-10 scale was documented, in the linear regression model these demographic variables were found not significant for perception of QoL. Regarding the labor activity, although there is a high percent of unemployment in the patients evaluated (66.8%), this variable was found not significant for perception of QoL. The evidence of labor activity as a predictor of QoL is controversial, and some studies of other populations with similar demographic characteristics
showed that low socioeconomic status is an important predictor of QoL in patients with epilepsy (Alanis-Guevara et al., 2005).
4.2. Clinical profile The clinical profile of patients with epilepsy assessed in this study was similar to that observed in high-income populations. Regarding the type of seizures, most of our patients had focal seizures (64.6%), and although the epidemiology of the types of seizures is highly variable and depends on many conditions, some studies in high-income populations have found similar data (Forsgren, 1992; Luengo et al., 2001). As for the etiology of epilepsy, and taking into account the classification proposed by the ILAE in 2010, 45% of our patients had a structural or metabolic cause of epilepsy, followed by etiology of unknown cause in 45.7% of patients. Similar etiological profiles have been found in highincome population studies. The clinical response pattern of our population toward antiepileptic drugs was similar to reported patterns in high-income populations (Torres-Ferrús et al., 2013). In our study 46.8% of patients were controlled and 27.2% were classified as having drug resistant epilepsy. These data are similar to those reported by Brodie et al., in Glasgow, Scotland, where it was reported that about 59% of patients will remain without seizures and be considered controlled, 25% will be resistant to antiepileptic drugs, and the remaining 16% will fluctuate between relapses and periods of freedom from seizures over 1 year (Brodie et al., 2012). In the linear regression model, the clinical variables that impacted most significantly on perception of QoL were: structural/metabolic etiology of epilepsy, drug resistant epilepsy and therapeutic response to antiepileptic drugs, specifically, the undetermined antiepileptic drug response. This data is consistent with other published data in similar populations (Lavados et al., 1992; Milovanovic´ et al., 2014). The frequency of seizures represents one of the main predictors of QoL in most of the populations studied (Alanis-Guevara et al., 2005; Milovanovic´ et al., 2014; Loring et al., 2004); however, although our study documented a significant association in the bivariate analysis, this variable did not show a significant impact on the perception of QoL in the linear regression model. Regarding antiepileptic drugs, we did not specify the type of medications; we only mentioned the number of medications (polytherapy or monotherapy). Although this variable was statistically significant in the bivariate analysis, in the linear regression model it did not show a strong association within the QOLIE-10 total score. It is well known that some medications can influence the mood (e.g. valproate and lamotrigine, among others) (Chiu et al., 2013) and thus indirectly affect the QoL. Some others medications can be associated with daytime sleepiness (e.g. benzodiazepines, among others) and also indirectly affect the QoL in patients with epilepsy (Perucca and Gilliam, 2012). As we did not specify the type of med-
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
ication, it could represent a selection bias that must be considered when analyzing the data. 4.3. Depression Depression is a major psychiatric comorbidity in patients with epilepsy, and may occur in up to 30–35% of cases, more frequently in patients with focal onset epilepsy and drug resistant epilepsy (Tellez-Zenteno et al., 2007). In our study, 39% of patients had depressive symptoms with a score ≥13 on the NDDI-E scale, posing a probable diagnosis of major depressive disorder, and showing a similar prevalence documented in multiple populations, which is independent of the socioeconomic profile (Alanis-Guevara et al., 2005; Milovanovic´ et al., 2014; Loring et al., 2004; Tellez-Zenteno et al., 2007). Some studies have shown that depression is underdiagnosed, undertreated, and representative of one of the main predictors of QoL, with even more impact than the seizure frequency (Boylan et al., 2004). Not only was there a high prevalence of depression documented in our population, this was the variable with the greatest impact on the perception of QoL, even more than other important variables such as seizure frequency, excessive daytime sleepiness, etiology of epilepsy, and therapeutic response to antiepileptic drugs. Some studies in populations with similar sociodemographic characteristics found data comparable to ours. In Brazil, Tedrus et al. documented that about 20.5% of patients with epilepsy have depression, representing the main predictor of the perception of QoL measured by QOLIE-31 inventory (Tedrus et al., 2013). 4.4. Sleep disorders Although sleep disorders in patients with epilepsy are a prevalent comorbidity with multiple pathophysiologic consequences, they are widely underdiagnosed (Manni and Terzaghi, 2010). In the clinical manifestations of sleep disorders, daytime sleepiness is one of the most common symptoms, and in patients with epilepsy it is a clinical marker of sleep apnea (Manni and Terzaghi, 2010). Sleep disorders and excessive daytime sleepiness are one of the main predictors of QoL in patients with epilepsy, and some studies in similar populations have found data comparable to ours (de Weerd et al., 2004). A total of 27.7% of the patients in our study had mild daytime sleepiness, and 9.5% had severe daytime sleepiness, which may reflect an underlying sleep disorder. It is important to mention that the daytime sleepiness is just one symptom of a specific disturbance, and in patients with epilepsy it can be influenced by many factors (Manni and Terzaghi, 2010). We did not study specific sleep disturbances, as this was not the aim for this study; however, it is important to take into account that the adverse effects of drugs, the decline in physical activity, and the presence of metabolic comorbidities (e.g. hypothyroidism, among others) are conditions that can also promote daytime sleepiness in patients with epilepsy (Manni and Terzaghi, 2010; de Weerd et al., 2004). Regardless of the cause of daytime sleepiness, we documented that this variable (Severe daytime sleepiness with an Epworth score of 17–24) negatively impacted the QoL, which increases the need to promote early diagnosis and prompt treatment of sleep disorders in patients with epilepsy. 5. Study limitations This study was based on a medical record, not in a population register, so the sample size and the sampling strategy, which were based on consecutive recruitment, do not allow for evaluation of the whole population. The regression model used in this study explains 60% of the variability in QoL, and although the concept of QoL can be very broad and be influenced by multiple variables unrelated to
173
the health and the wealth of the person, some clinical and sociodemographic conditions were not measured in the study (severity of seizures, anxiety disorders, specific structural/metabolic cause of epilepsy, type of antiepileptic drug, and family support, among others) and could also have affected the perception of QoL in this population. In addition, the “seizure frequency” variable was defined as a qualitative nominal variable, and not as a quantitative continuous variable, representing a probable selection bias. In addition, when the general QoL questionnaires such as SF-36, are compared to specific questionnaires about epilepsy such as QOLIE-31 or QOLIE-10, the latter could be less like to reveal the effect of demographic, sociocultural, and economic variables on QoL (Stavem et al., 2000). 6. Conclusions The QoL in epilepsy patients from a low-income population is determined primarily by depression, severe daytime sleepiness, structural/metabolic etiology of epilepsy, drug resistant epilepsy and therapeutic response to antiepileptic drugs, specifically, the undetermined antiepileptic drug response. These data suggest the need to promote early diagnosis and treatment of psychiatric comorbidities and sleep disorders, as well as effective and timely therapeutic interventions to prevent drug resistance in epilepsy, in order to improve the perception of QoL. Epilepsy must be considered a syndrome, in which seizures are the main symptom, but not the only one, and there are multiple associated conditions such as psychiatric comorbidity, sleep disorders, and psychosocial dysfunction, which are common, disabling, and possibly much more influential in these patients’ perceptions of QoL. Conflict of interest The authors have no conflict of interest to disclose. Acknowledgments The authors acknowledge the members of the Neurology department and the Neurology postgraduate program of the University of La Sabana: Dr. Erik Sánchez, Dr. Juan Vergara, Dr. Daniel Hedmont, Dr. Gustavo Barrios, Dr. Javier Vicini, and Dr. Maria Claudia Angulo; the Medical school of the University of La Sabana: Dr. Fernando Ríos, Dr. Maria José Maldonado, and Dr. Diana Diaz; the patients treated in the Neurology department of the Kennedy Western Hospital; the directors of the Kennedy Western Hospital, the nursing personnel and doctors in training at Kennedy Western Hospital. References ˜ E., Corona, T., et al., 2005. Sleep disturbances, Alanis-Guevara, I., Pena, socioeconomic status, and seizure control as main predictors of quality of life in epilepsy. Epilepsy Behav. 7, 481–485. Anon, http://www.culturarecreacionydeporte.gov.co/localidades/kennedy, 2016. Berg, A.T., Berkovic, S.F., Brodie, M.J., et al., 2010. Revised terminology and concepts for organization of seizures and epilepsies: report of the ILAE Commission on Classification and Terminology, 2005–2009. Epilepsia 51, 676–685. Boylan, L.S., Flint, L.A., Labovitz, D.L., et al., 2004. Depression but not seizure frequency predicts quality of life in treatment-resistant epilepsy. Neurology 62, 258–261. Brodie, M.J., Barry, S.J., Bamagous, G.A., et al., 2012. Patterns of treatment response in newly diagnosed epilepsy. Neurology 78, 1548–1554. Chica-Urzola, H.L., Escobar-Córdoba, F., Eslava-Schmalbach, J., 2007. Validación de la escala de somnolencia de epworth. Rev. Salud Pública 9, 558–567. Chiu, C.T., Wang, Z., Hunsberger, J.G., et al., 2013. Therapeutic potential of mood stabilizers lithium and valproic acid: beyond bipolar disorder. Pharmacol. Rev. 65, 105–142. Di Capua, D., Garcia-Garcia, M.E., Reig-Ferrer, A., et al., 2012. Validation of the Spanish version of the neurological disorders depression inventory for epilepsy (NDDI-E). Epilepsy Behav. 24, 493–496.
174
C.A. Espinosa Jovel et al. / Epilepsy Research 127 (2016) 168–174
Espinosa Jovel, C.A., Pardo, C.M., Moreno, C.M., et al., 2014. Demographic and social profile of epilepsy in a vulnerable low-income population in Bogotá, Colombia. Neurología, pii: S0213-4853(14)00242-4. Fazel, S., Wolf, A., Långström, N., et al., 2013. Premature mortality in epilepsy and the role of psychiatric comorbidity: a total population study. Lancet 382, 1646–1654. Fisher, R.S., Acevedo, C., Arzimanoglou, A., et al., 2014. ILAE official report: a practical clinical definition of epilepsy. Epilepsia 55, 475–482. Forsgren, L., 1992. Prevalence of epilepsy in adults in northern Sweden. Epilepsia 33, 450–458. Jacoby, A., Snape, D., Baker, G.A., 2009. Determinants of quality of life in people with epilepsy. Neurol. Clin. 27, 843–863. Kwan, P., Arzimanoglou, A., Berg, A.T., et al., 2010. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 51, 1069–1077. Lavados, J., Germain, L., Morales, A., et al., 1992. A descriptive study of epilepsy in the district of El Salvador, Chile, 1984–1988. Acta Neurol. Scand. 85, 249–256. Loring, D.W., Meador, K.J., Lee, G.P., 2004. Determinants of quality of life in epilepsy. Epilepsy Behav. 5, 976–980. Luengo, A., Parra, J., Colás, J., et al., 2001. Prevalence of epilepsy in northeast Madrid. J. Neurol. 248, 762–767. Manni, R., Terzaghi, M., 2010. Comorbidity between epilepsy and sleep disorders. Epilepsy Res. 90, 171–177. ´ M., Martinovic, ´ Zˇ , Toˇskovic, ´ O., 2014. Determinants of quality of life in Milovanovic, people with epilepsy in Serbia. Epilepsy Behav. 31, 160–166. Ngugi, A.K., Bottomley, C., Kleinschmidt, I., et al., 2010. Estimation of the burden of active and life-time epilepsy: a meta-analytic approach. Epilepsia 51, 883–890.
Perucca, P., Gilliam, F.G., 2012. Adverse effects of antiepileptic drugs. Lancet Neurol. 11, 792–802. Stavem, K., Loge, J.H., Kaasa, S., 2000. Health status of people with epilepsy compared with general reference population. Epilepsia 41, 85–90. Taylor, R.S., Sander, J.W., Taylor, R.J., et al., 2011. Predictors of health-related quality of life and costs in adults with epilepsy: a systematic review. Epilepsia 52, 2168–2180. Tedrus, G.M., Fonseca, L.C., Carvalho, R.M., 2013. Epilepsy and quality of life: socio-demographic and clinical aspects, and psychiatric co-morbidity. Arq. Neuropsiquiatr. 71, 385–391. Tellez-Zenteno, J.F., Patten, S.B., Jetté, N., et al., 2007. Psychiatric comorbidity in epilepsy: a population-based analysis. Epilepsia 48, 2336–2344. Social aspects of epilepsy in the adult in seven European countries., 2000. Epilepsia 41, 998–1004. Torres-Ferrús, M., Toledo, M., González-Cuevas, M., et al., 2013. Aetiology and treatment of epilepsy in a series of 1.557 patients. Rev. Neurol. 57, 306–312. Velez, A., Eslava-Cobos, J., 2006. Epilepsy in Colombia: epidemiologic profile and classification of epileptic seizures and syndromes. Epilepsia 47, 193–201. Vergara Palma, J., Espinosa Jovel, C.A., Vergara, T., et al., 2015. Impact of epilepsy surgery on the quality of life of a low-income population through the application of the Qolie-10 scale. Epilepsy Res. 110, 183–188. Viteri, C., Codina, M., Cobaleda, S., et al., 2008. Validation of the Spanish version of the QOLIE-10 quality of life in epilepsy questionnaire. Neurologia 23, 157–167. de Weerd, A., de Haas, S., Otte, A., et al., 2004. Subjective sleep disturbances in patients with partial epilepsy: a questionnaire based study on prevalence and impact on quality of life. Epilepsia 45, 1397–1404.