INTRODUCTION

Status epilepticus (SE) is a neurological emergency which may occur in people with epilepsy (PWE) as well as in those with- out epilepsy. It is associated with high morbidity, mortality and low quality of life thus in PWE episodes of SE have added clinical relevance (1,2). In 2015, the International League Against Epilepsy defined SE as a bilateral tonic–clonic activity lasting longer than 5 minutes, and absence SE and as focal SE as exceeding 10 minutes (3). Because of the time-locked defi- nition of SE ascertaining the incidence of SE in the community is wrought with many challenges (4) thus most available stud- ies are hospital based. In a review of population-based studies, Sanchez et all (4) reported an overall incidence of SE 9.9 to 41 per 100,000/year ranging from 3.5 to 41 per 100 000 per year. Recent studies from Europe have reported incidence rate of

36.1to 81.1 per 100 000 per year based on the new ILAE 2015 definition of SE(5,6).

Few studies have reported on SE in Africa(7,8). Bhalla et al (7) in 2014 reported an incidence of 10.8 per 100 000 population and Kariuki et al (8) reported a prevalence rate of 2.3 per 1000. Apart from using different definitions for SE, these studies included people without epilepsy and children. Based on few available studies in SSA, SE in Africa is reported to be high in Children, hospital-based studies and PWE who are not on anti- epileptic drugs (9,10). Most cases of documented SE are likely

to be seizures with predominantly motor features. In Austria, the age and sex adjusted incidence of a first episode of non- convulsive SE and SE with prominent motor phenomena was 12.1 and 24 per 100 000 adults per year, respectively (5).

In Sub-Saharan Africa (SSA), risk factors for SE include infec- tions and non-adherence to medication. In Kenya, for example, documented risk factors for SE were neurologic impairments, acute encephalopathy, previous hospitalization, and presence of antibody titers to falciparum malaria and HIV (8). Risk factors for SE and seizure clusters seizure (SC) have been reported to be similar (11,12). SE is a strong determinant of quality of life in PWE/their caregivers and the cost of treating epilepsy (1,13,14) . In SSA, SE may be treated at home and possibly in unorthodox way which may result in high morbidity and mor- tality.

Despite the potential for poor outcomes of SE in PWE, there is little, or no research related to the subject in Nigeria. Recogniz- ing incidence and patterns of SE is helpful in preventing unto- ward consequences of SE. The aim of this study was to describe the caregiver reported frequency and pattern of SE in PWE attending neurology outpatient clinic in Enugu.

METHODS

This was a cross-sectional descriptive study carried out in 3 major hospitals in the city of Enugu, Enugu State Nigeria. The Hospitals were the University of Nigeria Teaching Hospital, Enugu State University Teaching Hospital, and the Memfys Hospital for Neurosurgery. These three tertiary hospitals offer both secondary and tertiary epilepsy care in the city and receive patients from all parts of southeast Nigeria and beyond.

Study participants :The plan was to collect the data of all con- secutive consenting epilepsy patients who attend the neurology clinic of these hospitals. All consecutive consenting patients accompanied by caregivers in which case the caregiver must have been an eyewitness of the seizure were included in the study. We excluded patients with possible psychogenic non- epileptic seizures and single epileptic seizures. Cases where seizure duration could not be truly determined with some level of certainty were excluded from the present analysis. Cases of incomplete data or illegible lasting less than 5 minutes were also excluded. The study period was 9 months.

Data collection tools and procedures: Data was collected using

asemi-structured questionnaire. The questionnaire contained open ended and as well as multiple choice questions and con- sisted of three principal sections. The first section was on gen- eral information about the respondents such as age and sex and related characteristics. The second section contained questions on the age on onset of epilepsy, age of first treatment and fami- ly history. The third section was focused on clinical characteris- tics of seizures such as prediction of seizures, seizure triggers, duration of seizures in minutes or recurrent seizures during which the patients did not recover consciousness. The question- naire was constructed in simple English to match the expected reading level of most persons that completed primary school (6 years of basic education). Cues to remember seizure duration were given to informants. Such cues include questions like (1) do you think that seizures lasted as long as the time you have been in this clinic? (2) did seizures last as long as it took you to get to a hospital or call for help? The study questionnaire was designed by the principal investigator and reviewed by the oth- er authors and senior colleagues in neurology both in Enugu and outside. The mean time for filling out the questionnaire was 5 minutes. All questionnaires were filled in the English lan- guage either by the caregiver (self-administered) and some- times with help from the investigators. Data were collected by neurology registrars, senior registrars and consultants.

Completed questionnaires were retrieved the same day. In cases where the respondent did not understand English, a translation in the local language was used. In such cases, the items on the questionnaire were read out to the respondents, and their en- dorsed options were ticked by the investigator. The study proto- col was reviewed by the ethics committee of the Teaching Hos- pitals. All participants gave their informed consent after reading or having the consent form read for them.

34

Operational definitions: SE was defined as seizures lasting more than 5 minutes for generalized seizures and 10 minutes for focal seizures3. CS was defined as series of seizures close- ly grouped in time with shorter than normal inter ictal periods or as an increase over the patient's typical seizure frequency in a day or week (15). Epilepsy was defined based on the In- ternational League Against Epilepsy criteria.

Data analysis: The SPSS version 22 (IBM Corporation, New York, USA) was used for data management and statistical analysis. Data were presented in tables. The statistical methods included Mann-Whitney U test for unpaired observations and Chi-squared test for comparison of categorical data. Distribu- tion of types of seizures was calculated as the percentage of participants. Mean and median were calculated, and values were presented as tables where applicable. In all, p < 0.05 was regarded as statistically significant. Conclusions were drawn at this level of significance at 95% confidence level.

RESULTS

Description of participants: Data of 154 PWE were reviewed and analysed. Males were 80(51.9%) and 74(48.1%) were females. The male to female ratio was 0.9:1. Most participants were aged 20-29 years (42.9%) with a mean age of 32.4(17.3) years. Males were older than females by almost 7 years (p<0.0). Other characteristics of the patients are shown in Table 1. Before presenting to the hospital 40(26%) used tradi- tional herbal drugs while 36(23.4%) resorted to prayer as a sole means of treatment.

Seizure characteristics: Reported seizures characteristics are shown in Table 2. The mean age of onset of epilepsy in the cohort was 20.9 years (with a median of 17 years); earlier in females (14.4 years) than males 26.3 years) (p <0.01). The peak age of onset of epilepsy in SE was 0 to 9 years. The mean time taken from the age of onset to the age of going to the hospital was similar in both males and females. P=0.2. The 6-month seizure freedom was 9.7%. More males reported a family history of epilepsy. P=0.03. About 45.5% reported a clinical history of generalized seizures. Among those that had focal seizures clinically, 38(45.2%) could always predict the onset of seizures while the rest did so sometimes. Clinically seizures were reported to be similar every time by 107 (69.5%).

A total of 56(36.4%) confirmed that they had experienced pro- longed seizures that lasted more than 5 minutes for generalized seizures and 10 minutes for focal seizures in the past. Sex distri- bution SE showed that males (36.3%) and females (36.5%) re- ported a history of status in the past. P=0.98. Seizure related admissions were reported in 14(9.1%) of PWE more in females 11(14.9%) than males 3(3.8%). See Table 3. Figure 1, showed that history of SE appears to be bimodal; 40% below the age of 20 years and 50% after the after of 50 years.

Table 3 shows the proportion of PWE with various risk factors who had SE. About 54.5% of those with stroke, 47.9% of pa- tients who had CS reported a history of SE. A large proportion of dementia cases also had CS and SE although the overall number was small. Factors that correlated with past history of SE were: history of CS, family history of epilepsy and having no past history of seizure related admissions.

Table 3. Distribution of status epilepticus by various documented risk factors

*Multiple risk factor was recorded. #Percentage of risk factors.

DISCUSSION

Seizure frequency and pattern are strong determinants of quali- ty of life in epilepsy as well as the burden of epilepsy. Alt- hough in PWE, seizures are generally sporadic or even infre- quent, however, they may experience prolonged seizures (status epilepticus) and even cluster seizures. Identification of SE in the community is very important because of the associat- ed high morbidity, mortality as well as the associated high di- rect and indirect health costs in epilepsy(13,14).

In the index study, the male to female ratio of PWE with a his- tory of SE was 0.9:1. SE was reported by 36.4%: males (36.3%) and females 36.5%). P=0.98. About 54.5% of PWE who had stroke experienced SE. A large proportion of demen- tia cases also had SE although the numbers were small. Fur- thermore, the age distribution of SE was bimodal (before 20 years and after 50 years), and a large proportion of PWE with SE also had experienced SC. Factors that correlated with SE were history of SC, family history of epilepsy and no history of seizure related admissions.

The seizure related characteristics in the index study were sim- ilar to other published works from Nigeria (16,17). The age distribution of the patients in the index study may suggest a changing pattern of epilepsy risk factors or increasing aware- ness of epilepsy in the country. Younger mean-age-of-onset in females may be attributed to a better health seeking behaviour among females. Another factor responsible for this may be relatively large proportion of individuals with hypertension, diabetes and stroke. These disorders are generally commoner in older males. Clinically, 54.5% had focal seizures, a finding which may be explained by the high rates of risk factors for focal seizures in the study. This is similar to previous studies (18). Six-month seizure freedom in the index study was a mere 9.7% while 24.7% reported within 24 hours of seizures. These

findings support previous reports on seizure control in PWE in Nigeria(17). These may be related to several factors including non-adherence, uses of unorthodox medicine, seizure severity and alcohol (18).

The sex and age distribution of PWE who reported a history of SE showed same sex distribution and a bimodal age distribu- tion. In the US, Dham et al(19) reported a bimodal distribution of SE with the first peak in the first decade of life and the sec- ond after 60 years. Double peak in the incidence of SE was also reported in a review by Sanchez et al(4). Current demo- graphic in Nigeria and SSA have shown a rise in the older age group; therefore, SE is likely to become a common problem and an important health issue in years to come. Similar to some previous studies, the gender distribution of SE has been report- ed to be similar in males and females. Male to female ratio varied also from one study to the other with some reporting more males and others more females (4). In Ethiopia, Amare et al(20) reported a male-to-female ratio of SE of 1.5:1 which is different from the index study. In a study by Kariuki et al (8) there was an equal gender distribution.

There are no community-based studies on SE in Nigeria. Com- munity-based studies are frequently limited by recall bias and the ability of onlookers to recognize seizure-types and record their duration appropriately. Even in hospital settings in Nige- ria, EEG monitory is not frequently carried out hence there are likely to be low rates of detection.

The definition of SE has evolved over the past decades, how- ever, the incidence of SE has not differed much using different definitions. Leitinger et al (5) reported that reducing the diag- nostic time of SE increased the incidence only moderately by 10%. The overall incidence of SE range from 5.2 to 41 per 100,000/year (7) with an average of 9.9 per 100 000. Kantanen, et al (6) reported an annual age-adjusted incidence of 81.1/100,000 based on the new ILAE 2015 definition of SE.

The age and sex adjusted incidence of a first episode of SE, Non-convulsive statue epilepticus and SE with prominent mo- tor phenomena (including Convulsive SE) was 36.1 per 100 000 adults per year in Austria (5). In SSA, Kiruiki and his col- leagues reported an overall prevalence of 2.3 per 1,000 from three sites in Kenya(8). Their study included children and were limited to predominantly motor seizures and people without epilepsy. The reported incidence of SE in the index study (36.3%) is similar in males and females. This finding is within the rates reported by previous studies. The prevalence of SE in this study may be affected by several factors listed in table 4 which have been linked to SE in previous studies. Further- more, our cohort may represent patients with severe forms of epilepsy.

The incidence of SE is affected by age, geographical location, comorbidities and possibly family history (6,21). Similar to the index study in most adult studies there is a spike after the age of 50 (5,22). Geographical factors affect socio cultural charac- teristics of the population as well as disease pattern(4). For an example, whereas in Kenya (8), Malaria and HIV were associ- ated with SE, in Finland (6) alcohol withdrawal was the single most common acute symptomatic etiology in the study by Kantenen et al (6,23). The relationship between SE and non- adherence and no previous hospital visit have also been docu- mented in PWE(8). These two factors are important in SSA because large treatment gap and poverty.

In the index study, SE was reported in 54.5% of those with a history of stroke and 35.3% of those with Traumatic Brain Injury (TBI). Other cases with small but significant proportion of SE were those with meningitis and alcohol abuse. These findings are in support of previous studies that reported stokes, TBI and infections as common causes of SE a pattern which tends to vary between countries (4). SE has also been reported to be common in neurodegenerative disorders (4). African studies have reported infections and non-adherence to be high on the list of risk factors(8-10,20). Sadarangani et al(10), in Kenyan children, found that 71% of SE cases had an infectious cause, 53% attributed to malaria. Likewise, Amare et al (20) described CNS infections as the primary source of SE in Ethi- opia. However, these studies were not limited to PWE. SE may also be related to the premorbid state of the patients (4,24,25). Metabolic disorders such as hyperglycemia, uremia and acidosis of other etiologies are common causes of SE in non-epilepsy patients and may trigger status in PWE. Another risk factor for SE reported in the literature is family history.

38

In a population-based twin study reported a high pattern of SE concordance between monozygotic twins compared to dizygot- ic twins, linking familial predisposition and possible genetics factors to the risk of developing SE(25). Family history of epi- lepsy was reported in 14% of our cohort and correlated to a history of SE in the index study.

Factors that may precipitate/cause SE can also precipitate SC. These risk factors include TBI, longer duration of epilepsy and poor seizure control(26). SE has a direct effect on mortality, quality of life and increased health cost. It leads to repeated admissions in the emergency room or even in the intensive care unit. In the index study seizure related admissions negatively correlated to a history of SE. The reason for this is not clear. Large prospective studies are needed to shed more light on this finding. SE is a condition for which data on incidence, etiolo- gy, risk factors and outcomes are required for proper decision- making and for the allocation of resources by policy makers. These resources need to be used in the development of strate- gies that help improve prevention, diagnosis and reduce mor- bidity and mortality.

Limitations: This study has some limitations. Firstly, data used in this might have been affected by recall bias which may affect the true incidence of SE. Secondly, the timing of seizures may not be very accurate and subjective. Thirdly, only predom- inately motor seizures are observable, and subtle form of sei- zure are likely to be overlooked. Our study addressed only sur- vivors. Mortality rates of SE are important in assessing the true burden of this complication of epilepsy.

Finally, questionnaires were administered in English which may introduce some language bias because some medical terms in English do not have direct local equivalents. These limitations notwithstanding, this study has provided data for comparison for future studies. Large multi-center and commu- nity-based studies are needed to accurately document the preva- lence of SE in Nigeria.

Conclusion:The reported lifetime history of SE among PWE attending a tertiary hospital clinic in Enugu is high. This may suggest both poor seizure control and/or seizure severity. Care- ful patient education will improve both adherence and emergen- cy management of epilepsy to reduce the morbidity of epilepsy in the community.