Abstract
Objective
Epilepsy is the most common neurological condition in children in Low and Middle Income Countries (LMIC). Although the diagnosis remains majorly clinical, EEG is fundamental in assessment and management of epilepsy. However, its uptake remains low in LMIC. We aimed to determine the proportion of children with epilepsy who have had an EEG, the proportion of epileptiform EEGs, and the factors associated with epileptiform EEGs among the children with epilepsy at a neurology clinic in South Western Uganda.
Methods
We conducted a retrospective record review of 427 children less than 18 years old attending the Epilepsy Clinic at Mbarara Regional Referral Hospital. Records were selected by simple random sampling. A waiver of consent was obtained to study the medical records. Data was collected using REDCAP and analysed with STATA version 17. Proportions were reported as percentages and factors associated with epileptiform EEGs were analysed using the modified Poisson regression analysis.
Results
Out of 427 participants in this study, 149 (34.9%) had done EEG. Epileptiform EEGs were observed among 102 (68.5%) children. The factors independently associated with epileptiform EEGs among the children with epilepsy included participants’ sex and type of seizure. Having an epileptiform EEG was 1.98 times more likely among females than males (cPR:1.98, P -value 0.005) while it was 1.87 times more likely in children with generalized seizures compared to those with focal seizures (cPR:1.87, P-value=0.019).
Significance
There is need for more EEGs to be done especially among children with generalized seizures. More health workers also need to be trained in electro-encephalography.
Keywords: epilepsy, electroencephalogram, seizures, epileptiform EEG, neurology
Introduction
In children, epilepsy is the most common neurological condition. It affects all children regardless of geography, age, race, or social class.1 The incidence of epilepsy peaks in children and is highest throughout the first year of life.1 The prevalence of active epilepsy in High-income countries is about 4.5 per 1000 inhabitants.2 Africa has 20% of the global burden alone.3 Epilepsy is normally a secondary diagnosis in sub-Saharan Africa because of the elevated risks that are routinely present from birth and the negative neurological effects of infectious diseases present during and after infancy.4 In Uganda, 2% of children under the age of fifteen in south-western Uganda are reported to have epilepsy.4 This is a significant socioeconomic burden on the healthcare system and families of the children with epilepsy.5
Epilepsy is a clinical diagnosis, highly dependent on collateral history from parents, caretakers, or witnesses.6 Such history may not be very reliable and informative because of the variable interpretations of the seizure events by the witnesses.7 In such instances, the Electro-Encephalogram (EEG) plays a critical role in justifying the clinical suspicion. For example the EEG will help to distinguish between an absence-type seizure with a widespread inter-ictal epileptiform discharge and a focal seizure with loss of consciousness and focal inter-ictal epileptiform discharge.7–9 This offers clinicians confidence and goal-centered care. Hence, it has an invaluable role in the proper management of a child with epilepsy.10 In already clinically diagnosed epilepsy, EEG has a role in treatment cessation and monitoring.7,8 This is a very important aspect to determine the duration of follow-up following seizure freedom and for prognostic purposes.1,10
Access delays for EEG are documented in high-income nations, even though this is an important and quite inexpensive investigation.11 In developing countries, the situation is even worse with prolonged delays of more than 6 months.11,12 These delays are due to high cost of acquiring and running an EEG service, and limited access to facilities with the capacity to do an EEG.13 In our Paediatric epilepsy clinic in Mbarara Regional Referral Hospital, South Western Uganda, there are more than 20 children per clinic day with the majority presenting with epileptiform disorders, yet EEGs are not performed for all of them and when done, it is mostly after initiation of treatment. This adds to the delay in diagnosis, which in certain children with epilepsy can have long-lasting effects on their behavioral and cognitive abilities.14,15
Studies done in LMIC have shown that EEG evaluation changed epilepsy classification in about 55% of the cases.12,16 Given the limited availability of EEG services in many resource-constrained settings, understanding the extent to which EEG is utilized is important for assessing access to diagnostic services and informing resource allocation. In addition, determining the yield of epileptiform EEG findings is clinically relevant because it provides insight into the diagnostic value of EEG in routine practice and may help optimize its use among children most likely to benefit from the investigation.12 Identifying factors associated with epileptiform EEG findings is also important, as such information may support evidence-based prioritization of EEG referrals, improve diagnostic pathways, and enhance the efficiency of pediatric epilepsy care in settings with limited resources.10,12,14–16
Therefore, this study set out to determine the proportion of children who had undergone an EEG test, the proportion of epileptiform EEGs among them, and the factors associated with epileptiform EEGs.
Methods
Study Design and Setting
We retrospectively studied medical records of 427 children less than 18 years attending the epilepsy clinic at Mbarara Regional Referral Hospital (MRRH). MRRH is a government-owned referral hospital situated about 300km Southwest of the capital city, Kampala. The Paediatrics Epilepsy Clinic at MRRH runs every Wednesday from 8:00 am to 6:00 pm as a specialized outpatient clinic for children aged 1 month to 16 years. This clinic is run by a Pediatrician, trained in both paediatric electro encephalography and childhood epilepsy. He is assisted by Pediatric residents, a medical officer, nurses, a data clerk, and an EEG technician. The clinic receives outpatient referrals from all over South-western Uganda and the neighbouring countries of Rwanda and the Democratic Republic of Congo. Children aged 16 to 18 who have been under long-term treatment, continue to get care from the Paediatric clinic. In addition to epilepsy, children development delay, autism, cerebral palsy, neuro-cutaneous disorders, and post-infectious CNS sequelae are cared for in this clinic. About 20–30 children are seen each clinic day.
The epilepsy clinic provides clinical assessment, treatment, and referrals to other specialty clinics. EEG is mostly recommended for children with non-refractory seizures on AEDs. Magnetic resonance imaging (MRI) and CT scans are provided at the hospital, but at a cost. Many children are unable to do the CT and MRI due to the prohibitive costs. Every Thursday, the children who need routine EEGs are scheduled and get their EEGs completed on time. However, unscheduled EEGs can be performed in the event of an emergency. The same team members who oversee the Paediatrics Epilepsy Clinic often provide clinical assessments for children who require emergency care for epilepsy in the pediatric ward. The attendants of the patients can contact the pediatric epilepsy clinic through emergency telephone contacts at any time if they have any questions. Every child who attends the Epilepsy Clinic has a file archived in the records office and whose contents are also digitized in the computerized data system.
The EEG Procedure
The routine EEG done in the clinic takes 30 minutes for a cooperative patient. It often takes longer in children who require sedation. The clinic uses a 25 electrode Cadwell Arc Essentia EEG, 2018 19029002E3A0618005 machine. These electrodes are placed on different locations on the scalp as per the standard procedure. A special gel is applied on the electrodes for easy conduction and 4 major phases are required; a wakeful state for 10 minutes, a hyperventilation state for 5 minutes, photic stimulation for 5 minutes, and finally the sleep stage for 10 minutes. Potential differences between locations on the scalp are then recorded through the different phases. These measured potentials are the result of electrical currents generated in the brain and propagated to the scalp. The recordings are stored in a computer for interpretation.
Study Population
We reviewed all medical records of all children with epilepsy attending the epilepsy clinic of MRRH for the 5 years staring from January 2019 to December 2024.
We excluded files with incomplete records. These included files without participant demographics, and having less than 3 variables of interest. Also, records that had EEGs with excessive muscle artifacts and inconclusive EEG readings were excluded.
Sample Size Calculation
For the study to be powered to obtain the right proportion of children who had undergone an EEG test, we used the Modified Kish and Leslie Formula, n=(Z2p(1-p))/d2. We set our p at 0.5 since there are no similar studies done in a setting similar to ours. Adjusting for a 10% non-response rate, the estimated sample size was 427 children with epilepsy.
The sample size required to study the proportion of children with epileptiform EEG was computed using OpenEpi software. The expected proportion of epileptiform EEG was assumed to be 53.1% obtained from a single-center cross-sectional study done at a public tertiary hospital in Pakistan to assess the predictors of abnormal electroencephalogram and neuroimaging in children presenting to the emergency department with new-onset afebrile seizure.17 Using a finite population of 182 EEGs available in the clinic data base, and assuming a 95% confidence interval gave a sample size of 124 EEGs was sufficient to answer this objective. However, we found 149 records from our data and we decided to take a higher sample size to further increase the power of our study.
The sample size for factors associated with epileptiform EEGs was also calculated using OpenEpi online software. A study conducted among children with epilepsy in a Saudi was considered.14 In this study, children who had no family history of epilepsy were less likely to have an abnormal EEG compared to those children with family history of epilepsy (aOR 0.20, p-value <0.001). Assuming a 95% confidence interval and a statistical power of 80%, this gave a sample size of 80 participants.
We also decided to consider a higher sample size of 149 instead of 80 to increase the power of our study to 98.3%.
Sampling Technique
We used simple random sampling to select records of children with epilepsy who met the inclusion criteria. Eligible records were assigned random numbers using the RAND() function in Microsoft Excel, sorted in ascending order, and the required sample size was selected from the resulting list.
Study Procedure
We obtained the records of all children in the epilepsy clinic from the registry and identified the files for all children who had done an EEG. One of the investigators, trained in electroencephalography interpreted the EEGs and also screened out all EEGs with excessive muscle artefacts that could cause inconclusive interpretation. We then extracted the data from these files. These included social and demographic information: patient age, date of birth, sex, physical address, level of education. Clinical characteristics included family history of epilepsy, history of antenatal and birth events, CNS infections, age at seizure onset, seizure description, baseline frequency, duration on prescribed treatment. The data was collected using REDCap software and exported into Stata software version 17 for cleaning and analysis.
Data Analysis
We obtained the proportion of children who had done EEG by dividing the number of children who had EEG reports in their medical records by the total number of children with epilepsy. The proportion of children who had an epileptiform EEG was also obtained by dividing the number of children who had an epileptiform EEG by the total number of children who had done an EEG. These prevalences and the 95% confidence interval were calculated.
To obtain the factors associated with epileptiform EEGs, we used bivariable modified Poisson regression analysis. We reported crude prevalence ratios, 95% confidence intervals, and their P -values in tabular form. Variables with P-values <0.20 and all those that were biologically plausible were then entered in a multivariable modified Poisson regression analysis. Adjusted odds ratios, their P -values, and 95% confidence intervals were reported and presented in a table The backward elimination method was used to get a parsimonious model. Factors were considered significant if they had a P -value less than 0.05.
Quality Assurance and Control
We pretested the data extraction tool and collection process to ensure clarity and the ability to collect the required information to answer the study objectives. All the collected data was checked for completeness and any inconsistencies were corrected using information from the physical registers.
Ethical Considerations
We obtained ethical approval from the Mbarara University of Science and Technology Research Ethics Committee (Reference number: MUST-2024-1539). Administrative approval, to access the records and conduct the study was obtained from the Hospital Director of Mbarara Regional Referral Hospital. Before collecting data, a waiver of consent was obtained from MUST-REC to allow for the use of patient records. Names and other means of identity were not used during the data collection. The researcher ensured that all information obtained was kept in strict confidence and only used for the study. Electronic data collected using REDCap from the study was protected in a password-locked database. Passwords to the database were accessible to the principal investigator only. All files obtained from the records office for purposes of this study were returned. Special attention was taken to ensure no files were moved out of hospital premises. Our study complies with all the regulations under the Declaration of Helsinki.
Results
Study Flow Chart
Proportion of Children with Epileptiform EEGS Among Those Who Had Done an EEG
Out of 427 participants enrolled in this study, 34.9% (149/427, 95% CI: 30.5–39.6%) children had done EEG and 68.5% (102/149, 95% CI: 60.5–75.5%) children had an epileptiform EEG. This is shown in Figure 1.
Figure 1.
Figure showing the flow of patients in the study.
Baseline Characteristics of Participants Who Had Done Electroencephalogram
Most of the participants were females 80.5% (120/149) with a mean age of 6.95 and had no family history of epilepsy 77.9% (116/149). Among the participants, 14.1% (21/149) had a history of birth asphyxia, and only 2.7% (4/148) had a prior history of head injury before the first episode of convulsions. This is shown in Table 1.
Table 1.
Sociodemographic and Medical Characteristics of Participants with EEG Results
| Variable | Total | Non-Epileptiform EEG | Epileptiform EEG | P-value |
|---|---|---|---|---|
| N=149 | N=47 | N=102 | ||
| Age category | 0.577 | |||
| <5 years | 26 (17.4) | 7 (14.9) | 19 (18.6) | |
| ≥5 years | 123 (82.6) | 40 (85.1) | 83 (81.4) | |
| Participants sex | <0.001* | |||
| Male | 29 (19.5) | 18 (38.3) | 11 (10.8) | |
| Female | 120 (80.5) | 29 (61.7) | 91 (89.2) | |
| History of febrile seizures | 0.251 | |||
| No | 124 (83.2) | 38 (80.9) | 86 (84.3) | |
| Yes | 16 (10.7) | 4 (8.5) | 12 (11.8) | |
| Unknown | 9 (6.0) | 5 (10.6) | 4 (3.9) | |
| History of CNS infections prior to first episode of epilepsy | 0.610 | |||
| No | 123 (82.6) | 37 (78.7) | 86 (84.3) | |
| Yes | 22 (14.8) | 8 (17.0) | 14 (13.7) | |
| Unknown | 4 (2.7) | 2 (4.3) | 2 (2.0) | |
| Family history of Epilepsy | 0.802 | |||
| No | 116 (77.9) | 36 (76.6) | 80 (78.4) | |
| Yes | 33 (22.1) | 11 (23.4) | 22 (21.6) | |
| History of Neonatal jaundice | 0.199 | |||
| No | 131 (87.9) | 38 (80.9) | 93 (91.2) | |
| Yes | 10 (6.7) | 5 (10.6) | 5 (4.9) | |
| Unknown | 8 (5.4) | 4 (8.5) | 4 (3.9) | |
| History of Birth asphyxia | 0.537 | |||
| No | 124 (83.2) | 40 (85.1) | 84 (82.4) | |
| Yes | 21 (14.1) | 5 (10.6) | 16 (15.7) | |
| Unknown | 4 (2.7) | 2 (4.3) | 2 (2.0) | |
| History of status epilepticus | 0.224 | |||
| No | 116 (77.9) | 37 (78.7) | 79 (77.5) | |
| Yes | 26 (17.4) | 6 (12.8) | 20 (19.6) | |
| Unknown | 7 (4.7) | 4 (8.5) | 3 (2.9) | |
| History of head injury prior to first episode of epilepsy | 0.668 | |||
| No | 117 (78.5) | 39 (83.0) | 78 (76.5) | |
| Yes | 4 (2.7) | 1 (2.1) | 3 (2.9) | |
| Unknown | 28 (18.8) | 7 (14.9) | 21 (20.6) |
Note: *Chi-square p-value <0.05.
Abbreviations: EEG, Electro-encephalogram; CNS, Central nervous system.
Most children with epilepsy had generalized seizures 79.2% (118/149) and reported adherence to anti-epileptic drugs 79.9% (19/149). In more than half of the participants 53.0% (69/149), no treatment adjustment was made following an EEG. This is shown in Table 2.
Table 2.
Epilepsy-Related Characteristics of Participants Who Had Done Electroencephalogram
| Variable | Total | Non-Epileptiform EEG | Epileptiform EEG | P-value |
|---|---|---|---|---|
| N=149 | (N=47) | (N=102) | ||
| Type of seizure | ||||
| Focal | 19 (12.8) | 11 (23.4) | 8 (7.8) | <0.001* |
| Generalized | 118 (79.2) | 28 (59.6) | 90 (88.2) | |
| Others | 12 (8.0) | 8 (17.0) | 4 (3.9) | |
| Type of AEDs | 0.082 | |||
| Phenobarbitone | 44 (29.5) | 9 (19.1) | 35 (34.3) | |
| Sodium valproate | 52 (34.9) | 22 (46.8) | 30 (29.4) | |
| Carbamazepine | 45 (30.2) | 15 (31.9) | 30 (29.4) | |
| Others (specify) | 8 (5.4) | 1 (2.1) | 7 (6.9) | |
| Frequency of seizures | 0.514 | |||
| At least Monthly | 64 (43.0) | 18 (38.3) | 46 (45.1) | |
| Weekly | 66 (44.3) | 21 (44.7) | 45 (44.1) | |
| Daily | 19 (12.8) | 8 (17.0) | 11 (10.8) | |
| Time after seizure to EEG | 0.739 | |||
| ≥1 week | 95 (63.8) | 67 (65.7) | 28 (59.6) | |
| <1 week | 27 (18.1) | 17 (16.7) | 10 (21.3) | |
| Unknown | 27 (18.1) | 18 (17.7) | 9 (19.1) | |
| Treatment adjustment after EEG | 0.770 | |||
| No | 79 (53.0) | 27 (57.4) | 52 (51.0) | |
| Yes | 53 (35.6) | 14 (29.8) | 39 (38.2) | |
| Unknown | 17 (11.4) | 6 (12.8) | 11 (10.8) |
Note: *Chi-square p-value<0.05.
Abbreviations: AED, Anti-epileptic drug; EEG, Electro-encephalogram.
Factors Associated with Epileptiform EEGs Among the Children with Epilepsy at MRRH Using Modified Poisson Regression Analysis
The factors independently associated with epileptiform EEGs among the children with epilepsy at MRRH included participants’ sex and type of seizure. Having an epileptiform EEG was 1.98 times more likely among females than males (cPR:1.98, p-value 0.005) while it was 1.87 times more likely in children with generalized seizures compared to those with focal seizures (cPR:1.98, p-value=0.019). This is shown in Table 3.
Table 3.
Factors Associated with Epileptiform EEGs Among the Children with Epilepsy
| Variable | Non-Epileptiform EEG | Epileptiform EEG | Bivariate Analysis | p-value | Multivariate Analysis | P -value |
|---|---|---|---|---|---|---|
| (N=47) | No (N=102) | cPR (95% CI) | aPR (95% CI) | |||
| Age category | ||||||
| < 5 years | 7 (14.9) | 19 (18.6) | Ref | Ref | ||
| ≥5 years | 40 (85.1) | 83 (81.4) | 0.92 (0.71–1.20) | 0.555 | 0.87 (0.69–1.09) | 0.224 |
| Sex | ||||||
| Male | 18 (38.3) | 11 (10.8) | Ref | Ref | ||
| Female | 29 (61.7) | 91 (89.2) | 2.00 (1.23–3.22) | 0.005* | 1.98 (1.24–3.18) | 0.005* |
| History of Neonatal jaundice | ||||||
| No | 38 (80.9) | 93 (91.2) | Ref | Ref | ||
| Yes | 5 (10.6) | 5 (4.9) | 0.70 (0.37–1.32) | 0.277 | 0.73 (0.46–1.14) | 0.163 |
| Unknown | 4 (8.5) | 4 (3.9) | 0.71 (0.35–1.42) | 0.329 | 0.80 (0.50–1.27) | 0.342 |
| Type of seizure | ||||||
| Focal | 11 (23.4) | 8 (7.8) | Ref | Ref | ||
| Generalized | 28 (59.6) | 90 (88.2) | 1.81 (1.06–3.10) | 0.031* | 1.87 (1.11–3.14) | 0.019* |
| Others | 8 (17.0) | 4 (3.9) | 0.79 (0.30–2.07) | 0.634 | 0.93 (0.34–2.52) | 0.880 |
| Type of AEDs | ||||||
| Phenobarbitone | 9 (19.1) | 35 (34.3) | Ref | Ref | ||
| Sodium valproate | 22 (46.8) | 30 (29.4) | 0.73 (0.55–0.96) | 0.023* | 0.80 (0.62–1.03) | 0.087 |
| Carbamazepine | 15 (31.9) | 30 (29.4) | 0.84 (0.65–1.08) | 0.176 | 1.01 (0.81–1.25) | 0.955 |
| Others | 1 (2.1) | 7 (6.9) | 1.10 (0.81–1.49) | 0.537 | 1.17 (0.78–1.76) | 0.436 |
| Frequency of seizures | ||||||
| At least monthly | 18 (38.3) | 46 (45.1) | Ref | Ref | ||
| Weekly | 21 (44.7) | 45 (44.1) | 0.95 (0.76–1.19) | 0.647 | 0.90 (0.74–1.11) | 0.343 |
| Daily | 8 (17.0) | 11 (10.8) | 1.81 (0.53–1.22) | 0.306 | 0.76 (0.53–1.10) | 0.150 |
Note: *P-value<0.05.
Abbreviations: cPR, Crude Prevalence Ratio; aPR, Adjusted Prevalence Ratio.
Discussion
Proportion of Children with Epilepsy Who Have Had an EEG Test
This study found that 34.9% (30.5–39.6) of children attending the clinic had an EEG test done as part of their clinical care. Whereas there is no universally accepted proportion throughout the literature, ideally every child with epilepsy should have a baseline EEG within 48–72 hours following a seizure, even those who have done it, less than 18% are performed within the recommended time. This proportion is generally low because of the constrained resources evidenced by a single EEG machine, one trained technician, and one trained pediatrician who interprets the EEGs. This limits the number of children with epilepsy to have a baseline EEG. This low proportion puts some children at risk of receiving a misdiagnosis and incorrect treatment consequently.18
This was found to be low when compared to a study done in Pakistan from July 2019 to June 2021 among children with epilepsy aged one month to 18 years which was 62.7%.17 This low proportion could be explained due to differences in the study site. The epilepsy clinic is an outpatient setting in a government-owned hospital compared to an emergency setting in a private hospital in the Pakistan study. We however found no other studies that documented the proportion of children with epilepsy who have had an EEG test done.
Proportion of Epileptiform EEGs
Our study showed that two-thirds of the children attending the epilepsy clinic had epileptiform EEGs with a proportion of 68.5%. This proportion is high because of the selection criteria for EEGs in our clinic. EEGs are done for children with non-refractory seizures while on AEDs. Additionally, The patients who attend the clinic are a selected population where most of them are referred with a clinical suspicion of epilepsy.19 This selection makes it more likely to select a child whose EEG would otherwise be epileptiform. Additionally, routine scalp EEGS have comparatively a low and wide sensitivity ranging from 25% to 56% in the diagnosis of epilepsy.20 This wide range can be explained by the diverse patient selection and clinical details needed for the diagnosis of epilepsy in different studied populations. This makes comparisons very difficult due to wide variations in methodology and multiple confounding variables impacting the results.
The proportion of epileptiform EEGs was similar to that reported in Zambia. A retrospective observational study in Lusaka found the proportion of epileptiform EEGs at 70.6%. This could be because both studies used selected populations. The study in Zambia considered only referrals (referral bias) rather than a representative sample from the population.21 Both studies also had similarities in the study settings where all facilities are tertiary centers located in low- and middle-income countries (LMIC).
However, the proportion of epileptiform EEGs in our study is higher than that found in a study conducted in Saudi Arabia and Colombia that found a proportion of 49.7% and 40% epileptiform EEGs respectively.14,22 The proportion in our study is higher than that found in these studies because of patient selection. Our study had a smaller percentage of neonates and infants compared to the study in Saudi Arabia. Epileptiform patterns in neonates and infants are physiologically harder to diagnose. This is due to the incomplete CNS myelination and the limbic system that is resistant to synchronization in an immature state.18 In the study in Colombia, only patients with autism were included which is a small proportion.22
Factors Associated with Epileptiform EEGs Among the Children with Epilepsy Who Have Had an EEG
In this study, the two factors that were associated with having an epileptiform EEG were being female and having generalized seizures.
Females with epilepsy were 1.98 times more likely to have an epileptiform EEG than the males (cPR:1.98, p-value 0.005). This is similar to what was found in a study conducted in Saudi Arabia that agrees with the relationship that the female gender is associated with epileptiform EEGs.14 This difference in sex observed in the current study can be explained by the mean age of the participants being 7 years. This is about the time when puberty begins which causes hormonal fluctuations and eventual increased seizure episodes.23 The menstrual cycle has been linked to seizure exacerbation due to fluctuations in oestrogen and progesterone. These seizures are also known to be drug refractory.24 This is however different from other studies which have documented male predominance of epileptiform EEGs. A study done in Medellin, Colombia had a predominance of male gender. This could be due to the differences in the baseline characteristics of the children. The Colombia study was done among patients with autism spectrum disorder which is more prevalent in males.22 However, this finding should be interpreted with caution due to the selection criteria and possible residual confounding. More research needs to be done to further confirm if indeed females are more likely to have epileptiform EEGs.
Children who had generalized seizures were 1.87 times more likely to have epileptiform EEGs compared to those with focal seizures (cPR: 1.87, p-value=0.019). Other studies have shown conflicting results with no association between seizure types and interictal discharges.14,15,17 This conflict can be explained by the difficulty in clinical classification for clinicians. In our setting, clinicians are more likely to rely on a narrative of an eyewitness. Even with a good eyewitness accounts, seizures in the paediatric population tend to rapidly generalize due to dysfunctional cortical-subcortical interaction.18 Evolution of seizures is common in children where they start as focal and end up generalized.25 This finding therefore needs to be studied more.
Strengths and Limitations of the Study
We reviewed records over a 5-year period which allowed us to get an adequate sample size for our study population which increases its internal and external validity. Records for five years are representative of all the patients that attended the clinic since its debut in 2019. MRRH is a tertiary referral hospital, so we received children with epilepsy from a diverse population that allows our data to be generalized.
Our study had some limitations including; a retrospective design leading to missing data; selection of participants from an epilepsy clinic rather than a general paediatric pool and absence of inter-rater reliability for EEG interpretation.
Conclusion
The proportion of children with epilepsy who get a baseline EEG is low, and among these children with confirmed epilepsy, the majority of the EEGs are epileptiform.
Having generalized seizures and being female are independently associated with having an epileptiform EEG pattern.
Clinical Relevance and Future Directions
We recommend that clinicians prescribe EEG more routinely in the care of Children with Epilepsy. Females and children with generalized seizures should be considered a priority in getting an EEG done.
We recommend more staff to be trained in electro physiology and also for the government to procure more EEG machines.
Prospective studies should be carried out to determine more factors associated with epileptiform EEGs.
Acknowledgments
We acknowledge all the staff in the Epilepsy Clinic at Mbarara Regional Referral Hospital.
Funding Statement
This study did not receive any funding from any organization.
Ethical Publication Statement
We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
References
- 1.Beghi E. The epidemiology of epilepsy. Neuroepidemiology. 2020;54(2):185–10. doi: 10.1159/000503831 [DOI] [PubMed] [Google Scholar]
- 2.Coste J, Mandereau-Bruno L, Carcaillon-Bentata L, Mikaeloff Y, Bouilleret V. Prevalence, demographic and spatial distribution of treated epilepsy in France in 2020: a study based on the French national health data system. J Neurol. 2024;271(1):519–525. doi: 10.1007/s00415-023-11953-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kariuki SM, Ngugi AK, Kombe MZ, et al. Prevalence and mortality of epilepsies with convulsive and non-convulsive seizures in Kilifi, Kenya. Seizure. 2021;89:51–55. doi: 10.1016/j.seizure.2021.04.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Duggan M. Epilepsy in rural Ugandan children: seizure pattern, age of onset and associated findings. Afr Health Sci. 2010;10(3):218–225. [PMC free article] [PubMed] [Google Scholar]
- 5.Atugonza R, Kakooza-Mwesige A, Lhatoo S. Multiple anti-epileptic drug use in children with epilepsy in Mulago hospital, Uganda: a cross sectional study. BMC Pediatric. 2016;16:34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rana KS. Rational use of EEG in childhood epilepsy. Indian J Pediatr. 2000;67(1 Suppl):S22–31. [PubMed] [Google Scholar]
- 7.Nariai H. Frontiers in EEG as a tool for the management of pediatric epilepsy: past, present, and future. Epilepsia Open. 2025. doi: 10.1002/epi4.70116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.The role of EEG in patients with suspected epilepsy - PubMed. Available from: https://pubmed.ncbi.nlm.nih.gov/32364504/. Accessed June 14, 2026.
- 9.Mikhaeil-Demo Y, Gonzalez Otarula KA, Bachman EM, Schuele SU. Indications and yield of ambulatory EEG recordings. Epileptic Disord. 2021;23(1):94–103. doi: 10.1684/epd.2021.1249 [DOI] [PubMed] [Google Scholar]
- 10.Das JC. Electroencephalogram (EEG) in the management of epilepsy in children. Mymensingh Med J. 2014;23(2):406–411. [PubMed] [Google Scholar]
- 11.McLane HC, Berkowitz AL, Patenaude BN, et al. Availability, accessibility, and affordability of neurodiagnostic tests in 37 countries. Neurology. 2015;85(18):1614–1622. doi: 10.1212/WNL.0000000000002090 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Acheampong M. Utilization of Paediatric Electroencephalogram at Komfo Anokye Teaching Hospital. African J Curr Med Res. 2022. doi: 10.31191/AFRIJCMR.V5I1.97 [DOI] [Google Scholar]
- 13.Romagnoli S, Franchi F, Ricci Z. Processed EEG monitoring for anesthesia and intensive care practice. Minerva Anestesiologica. 2019;85(11):1219–1230. doi: 10.23736/S0375-9393.19.13478-5 [DOI] [PubMed] [Google Scholar]
- 14.Owolabi LF, Reda AA, Ahmed RE, Enwere OO, Adamu B, AlGhamdi M. Electroencephalography findings in childhood epilepsy in a Saudi population: yield, pattern and determinants of abnormality. J Taibah Univ Med Sci. 2021;16(1):86–92. doi: 10.1016/j.jtumed.2020.10.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Owolabi LF, Sale S, Owolabi SD, Nalado A, Umar M, Taura AA. Electroencephalography abnormalities in generalized epilepsy and their predictors: a multicenter experience. Ann Afr Med. 2018;17(2):64–69. doi: 10.4103/aam.aam_2_17 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lagunju IOA, Oyinlade AO, Atalabi OM, et al. Electroencephalography as a tool for evidence-based diagnosis and improved outcomes in children with epilepsy in a resource-poor setting. Pan Afr Med J. 2015;22:328. doi: 10.11604/pamj.2015.22.328.7065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ali N, Haider S, Mustahsan S, et al. Predictors of abnormal electroencephalogram and neuroimaging in children presenting to the emergency department with new-onset afebrile seizures. BMC Pediatr. 2022;22(1):619. doi: 10.1186/s12887-022-03668-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Fontana E, Negrini F, Francione S, et al. Temporal lobe epilepsy in children: electroclinical study of 77 cases. Epilepsia. 2006;47(s5):26–30. doi: 10.1111/j.1528-1167.2006.00873.x [DOI] [PubMed] [Google Scholar]
- 19.Namusisi J, Kyoyagala S, Nantongo J, et al. Poor seizure control among children attending a tertiary hospital in South Western Uganda – a retrospective study. Ijgm. 2023;16:895–904. doi: 10.2147/IJGM.S398318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Smith SJM. EEG in the diagnosis, classification, and management of patients with epilepsy. J Neurol Neurosurg Psychiatry. 2005;76 Suppl 2(Suppl 2):ii2–7. doi: 10.1136/jnnp.2005.069245 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ciccone O, Mathews M, Birbeck GL. Management of acute seizures in children: a review with special consideration of care in resource-limited settings. Afr J Emergency Med. 2017;7:S3–S9. doi: 10.1016/j.afjem.2017.09.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Arteaga A, Vélez E, Cornejo W, et al. Epilepsy and electroencephalographic abnormalities in patients with diagnosis of idiopathic autism spectrum disorder in Medellín. Int J Psychol Res. 2021;14(1):115–120. doi: 10.21500/20112084.5335 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Li Q, Zhang Z, Fang J. Hormonal changes in women with epilepsy. Neuropsychiatr Dis Treat. 2024;20:373–388. doi: 10.2147/NDT.S453532 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Shiono S, Williamson J, Kapur J, Joshi S. Progesterone receptor activation regulates seizure susceptibility. Ann Clin Transl Neurol. 2019;6(7):1302–1310. doi: 10.1002/acn3.50830 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ighodaro ET, Maini K, Arya K, Sharma S. Focal Onset Seizure. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2025. Available from: http://www.ncbi.nlm.nih.gov/books/NBK500005/. Accessed September 17, 2025. [PubMed] [Google Scholar]