Epidemiological, Diagnostic and Evolutionary Profile of Seizures in Young Infants at Albert Royer (Dakar/Senegal) ()
1. Introduction
The neonatal period is a high-risk period for neurological morbidity and mortality, one of the most common manifestations of which is seizures [1] [2]. The exact prevalence of neonatal seizures is difficult to estimate, due to a wide variability in its frequency in most studies worldwide (1.5‰ - 5.5‰ in newborns). This could be explained by the disparity in inclusion criteria, the diversity in the methodology used to identify seizures and the lack of description of the clinical presentations of seizures [3]-[5]. Investigating the etiology of an infant’s seizures is a crucial step in ensuring appropriate therapeutic management. Thanks to early and accurate detection of the underlying cause of the seizures, recent advances in neuroimaging techniques, particularly electroencephalographic monitoring, have significantly improved the prognosis, as well as new neuroprotective strategies such as hypothermia. The management of seizures is often problematic in developing countries, as resuscitation equipment and anticonvulsants are sometimes lacking due to their unavailability or unsuitable galenic presentations. In sub-Saharan Africa, particularly Senegal, the main studies concern perinatal asphyxia. The aim of our study was to describe the epidemiological, clinical, therapeutic and prognostic characteristics of seizures in young infants in the neonatology department of the Albert Royer National Children’s Hospital (Senegal).
2. Materials and Methods
2.1. Study Setting
The study took place in the neonatology unit of the Albert Royer Children’s Hospital, a national referral service, which receives and admits infants aged 0 - 2 months, all from other health facilities in the regions of Senegal.
2.2. Type, Period and Study Population
This was a retrospective, descriptive and analytical study, conducted from 1 January 2012 to 30 September 2018, of infants admitted who presented with seizures.
2.3. Inclusion Criteria
All infants aged 0 days to 2 months with seizures diagnosed at the clinic examination during the study period were included.
2.4. Data Collected and Parameters Studied
Data were collected on a pre-established survey form after consultation with hospital records, the health record and the liaison form for referred children. The parameters studied were epidemiological (age, sex, geographical origin), diagnostic (clinical and paraclinical signs, etiologies), therapeutic (ventilatory support, etiological treatment) and evolutionary (mortality, complications), relating to the characteristics of the mother and the infant.
2.5. Data Entry and Statistical Analysis
Data were entered and analyzed using SPSS software version 20.0. Qualitative variables were expressed as percentages and quantitative variables as means or medians with their ranges. For statistical comparisons, we used Pearson’s Chi-squared and the p-value (p), with a 95% confidence interval when the p-value was significant (p < 0.05), depending on the conditions of application.
3. Results
3.1. Epidemiological and Socio-Demographic Data
A total of 351 infants were included, representing a hospital frequency of 8.5%. The most common maternal age group was between 18 and 35 years. The majority of mothers (66%) were from the suburbs of Dakar. The socio-economic level was low in half the cases (50.0%).
3.2. Maternal Medical and Gynecological-Obstetric Data
The average gestational age was 2.5 and the average parity was 1.9. Primiparous women accounted for 46.1% of the total, compared with 53.9% of multiparous women. Almost all the mothers (99.1%) had made at least 3 antenatal visits. Urogenital infections (6.3%), gravid hypertension (2.8%) and funicular abnormalities (3.1%) were the main medical and gynecological-obstetric pathologies associated with pregnancy. Prolonged labor lasting more than 12 hours occurred in 10.5% of cases. Delivery was by non-instrumental vaginal way in the majority of cases (80.9%).
3.3. Clinical Data
Most of the infants (43.6%) had not cried at birth and had been resuscitated in the delivery room. The Apgar score at 5 minutes was less than 7 in 12.8% of cases. The majority of infants (63%) were born at term, compared with 3.1% of premature babies and 4% of post-mature babies (Figure 1).
Trophicity was normal in 68% of cases. The sex ratio (M/F) was 1.4. The mean age of the infants was 6.7 days [0 days - 2 months]. Half of the cases (52.7%) were admitted before 72 hours of life. The clinical features of the seizures were dominated by focal tonic seizures (47.3%), multifocal clonic ѕеizսres (41.9%), subtle seizures as oral-buccal-lingual movements (19.9%) and abnormal eye movements (15.9%). A status epilepticus was noted in 8% of cases. The total duration of seizures was less than 72 hours in the majority of cases (70.3%). Other neurological abnormalities were dominated by abnormal primitive reflexes (74.6%) and lethargy (35.8%). Other clinical signs associated with seizures were dominated by disturbances in thermal regulation (66.2%) and respiratory distress (50.2%). The main causes of seizures were hypoxic-ischemic encephalopathy (48.7%), metabolic disturbances (48.1%), particularly dysnatremia, central ոеrvοսѕ system infections (15.6%), and bilirubin еոϲерhаlоpathу (4.5%) (as shown in Table 1).
![]()
Figure 1. Distribution of cases according to gestational age.
Table 1. Distribution of cases by causes.
Causes |
Number (n) |
Percentage (%) |
Hypoxic-ischemic encephalopathy |
171 |
48.7 |
Metabolic disturbances |
169 |
48.1 |
Central nervous system infections |
55 |
15.6 |
Bilirubin enceрhаlоpathу |
16 |
4.5 |
Intracranial hemorrhage |
10 |
2.7 |
Ischemic stroke |
8 |
2.2 |
Neonatal epilepsy |
6 |
1.7 |
Congenital hydrocephalus |
3 |
0.7 |
Congenital malaria |
2 |
0.5 |
Viral encephalitis |
2 |
0.5 |
Aminoacidopathy |
1 |
0.2 |
Unknown |
20 |
5.7 |
3.4. Paraclinical Data
Laboratory tests abnormalities included hyponatremia (17.1%), hypernatremia (9.4%), renal failure (9.9%), hepatic cytolysis (6.8%), hypomagnesaemia (6.5%), hypocalcaemia (4.5%), hyperkaliemia (1.4%) and hepatocellular failure (0.2%). The most common Transfontanellar ultrasound abnormalities were hemorrhagic lesions (6.1%) and hypoxic-ischemic white matter lesions (3.1%).
3.5. Therapeutic Data
Phenobarbital was administered as the first line anticonvulsant in the majority of cases (84%), followed by diazepam (19.9%) (Figure 2).
Figure 2. Distribution of cases according to anticonvulsant treatment.
An oral relay was instituted in 16.5% of cases, and sodium valproate was used most frequently. The associated therapeutic measures are shown in the following table (Table 2).
Table 2. Associated therapeutic measures.
Other therapeutic measures |
Number (n) |
Percentage (%) |
Invasive ventilation |
86 |
24.5 |
Magnesium sulphate |
9 |
2.5 |
Correction of hypoglycemia |
20 |
5.7 |
Correction of hyponatremia |
45 |
12.8 |
Correction of hypocalcaemia |
8 |
2.28 |
Rehydration |
49 |
13.9 |
Phototherapy |
14 |
3.9 |
Antibiotic therapy |
38 |
10.8 |
3.6. Evolutionary Data
The case fatality rate was 39.5%. The majority of deaths (59.2%) occurred within 72 hours of hospitalization. The etiologies and clinical signs associated with the highest mortality were stroke (62.5%), dehydration (48.1%), hypernatremia (45.5%) and EAI (35.1%). The main sequela observed were delayed psychomotor development (20.6%), cerebral palsy (6.0%), epilepsy (4.8%) and neurological sequela (1.2%). Cerebral palsy occurred more frequently in infants who had convulsed for more than 48 hours (28.6% cases).
3.7. Factors Associated with Mortality
Mortality was significantly associated with prematurity (p = 0.002), macrosomia (p = 0.006), intrauterine growth restriction (p = 0.006), resuscitation at birth (p = 0.003), dehydration (p = 0.005), shock (p = 0.008), postnatal age greater than 48 hours (p < 0.003) and status epilepticus (p = 0.01) (Table 3).
Table 3. Factors associated with mortality.
Parameters |
Survival |
Death |
OR |
p-value |
Number |
% |
Number |
% |
Term |
|
|
|
|
|
|
< 37 SA |
2 |
18 |
9 |
82 |
6.99 |
0.002* |
> 37 SA |
141 |
61 |
90 |
39 |
- |
- |
Trophicity |
|
|
|
|
|
|
Eutrophy |
146 |
61 |
92 |
39 |
- |
|
Macrosomia |
23 |
82 |
5 |
18 |
0.34 |
0.006* |
Hypotrophy |
24 |
51 |
23 |
49 |
1.52 |
0.006* |
Type of seizures |
|
|
|
|
|
|
Generalized |
46 |
71 |
19 |
29 |
- |
- |
Localized |
87 |
63 |
52 |
37 |
1.45 |
0.05 |
Resuscitation at birth |
|
|
|
|
|
|
Yes |
99 |
66 |
52 |
34 |
0.63 |
0.003* |
No |
213 |
58 |
154 |
42 |
- |
- |
Dehydration |
|
|
|
|
|
|
Yes |
32 |
62 |
20 |
38 |
0.75 |
0.005* |
No |
250 |
54 |
209 |
46 |
- |
- |
Shock |
|
|
|
|
|
|
Yes |
7 |
27 |
19 |
73 |
3.55 |
0.008* |
No |
275 |
57 |
210 |
43 |
- |
- |
Postnatal age |
|
|
|
|
|
|
≤ 48h |
111 |
63 |
65 |
37 |
- |
- |
> 48h |
99 |
58 |
72 |
42 |
1.24 |
0.003* |
Status epilepticus |
|
|
|
|
|
|
Yes |
14 |
50 |
14 |
50 |
0.63 |
0.01* |
No |
196 |
61 |
123 |
39 |
- |
- |
*: significant p-value (< 0.05); SA = week of amenorrhea
4. Discussion
In our study, the frequency of hospitalization was high at around 8.5%, demonstrating its relatively common occurrence in the infant period in our regions. A lower result was found in Morocco in 2017 (4.5%) [6]. However, this difference can be explained by our longer study period. The most represented maternal age group was between 18 and 35 years, in contrast to a study conducted in California where maternal age over 40 years was an independent risk factor associated with the occurrence of infant seizures [7]. Throughout the literature, primiparity is associated with a higher risk of childbirth difficulties [8]. In our study, most of the mothers were primiparous. Almost all the mothers had undergone at least 3 antenatal visits. However, these pregnancies were insufficiently monitored and this situation was correlated with the socio-economic level, which was considered low in the majority of cases. In our series, we noted medical and gyneco-obstetric complications such as urogenital infection, gravid arterial hypertension, funicular anomalies and prolonged labor, thus demonstrating their role in the genesis of neonatal complications. These results are comparable to those of other studies on the subject [7] [9]. In our study, vaginal delivery predominated (80.9%), unlike in Morocco, where the caesarean section rate was higher (45%). This would reduce the risk of childbirth complications, particularly perinatal asphyxia [6]. Most of the infants (43.6%) had required resuscitation in the delivery room, a result consistent with the presence of hypoxic-ischemic encephalopathy as the main cause of seizures. This situation would appear to be linked to poor resuscitation, which worsens the neurological prognosis and the incidence of seizures.
In our study, seizures mainly concerned full-term infants. Indeed, term and post-term infants are more exposed to the consequences of anoxo-ischemia, such as seizures [6] [7]. With regard to trophicity, hypotrophy and macrosomia may expose the baby to complications that can lead to seizures, such as metabolic disorders (hypoglycemia, hypomagnesaemia), obstetric trauma or ischemic lesions (leucomalacia) [10]. As reported in the literature, we noted a slight male predominance, reflecting the excess male morbidity in the neonatal period [6] [11] [12]. The majority of infants (64.1%) were less than 7 days old. Similar results have been reported by some authors, with greater susceptibility during the first week of life, particularly the first 48 hours [3] [6] [7] [13]. Neonatal seizures have unique clinical features compared with those of older infants and children. In addition to their discreet nature, which may escape the scrutiny of medical staff, they were focal in two-thirds of cases, in line with the literature [14]. In our study, they were focal tonic, multifocal clonic or subtle seizures as in the majority of cases, or of the chewing or eye-revulsion type, a result similar to that noted in Morocco [6]. The frequency of atypical seizures is also a feature of the neonatal period [15]. Archaic reflexes were blunted in most cases and hypotonia was often present. The main other associated signs were disturbances in thermal regulation and respiratory distress. These clinical signs were related to the etiology and were sometimes part of a multivisceral failure. Perinatal asphyxia, the cause of hypoxic-ischemic encephalopathy, is a worldwide public health problem. Metabolic disturbances are the 2nd leading cause of seizures, namely dysnatremia, hypoglycemia, hypomagnesaemia and hypocalcaemia, and central ոеrvοսѕ system infections are in 3rd place [16]. In our context, the main causes of infant seizures were anoxic-ischemic encephalopathy, metabolic disturbances and meningitis. These results are similar to those of other studies on the subject [6] [17]-[19]. Diagnosis and investigation in infants are made difficult by the non-accessibility of certain complementary examinations. The main finding was that the essential test for confirming seizures, the electroencephalogram, was not available, and none of the children were able to benefit from it at the initial stage. Electroencephalogram plays an essential role in detecting subclinical seizures, confirming the epileptic nature of the seizures, assessing the prognosis, guiding treatment and sometimes contributing to the diagnosis [20]. Its unavailability is therefore a real obstacle to the monitoring and management of infant seizures. In addition, magnetic resonance imaging is rarely performed in the neonatal period for reasons of availability, accessibility and organization. It is also an essential test for determining the prognosis in the event of seizures and neurological damage in the newborn [21] [22]. In our context, the most available examination is the Transfontanellar ultrasound, which has been able to demonstrate a certain number of lesions related to seizures. It is available on the ward at the patient’s bedside. The standard biochemical work-up is also available, enabling the detection of a number of metabolic disorders associated with seizures.
Phenobarbital was the main anticonvulsant used in our study and in the study conducted in Morocco in 2017 [6]. It remains the treatment of choice for neonatal seizures, as shown by recent European and American surveys [17] [23]-[27]. Unfortunately, its shortage during the study period resulted in the use of diazepam, which is not recommended in the neonatal period. The drugs most commonly used in cases of infant seizures refractory to phenobarbital are phenytoin, levetiracetam, lidocaine and midazolam, but the majority remain unavailable. In addition to anticonvulsant treatment, the following therapeutic means have been used: correction of metabolic disturbances, administration of antibiotics and various means of resuscitation. Therapeutic hypothermia is not used, even though two large studies have demonstrated its efficacy in neuroprotection [28] [29]. In our study, neonatal convulsions are marked by a very high mortality rate of around 39.5%. A lower result (9%) was reported in Morocco in 2017 [6]. This situation could be explained by late referrals with a serious clinical picture, poor quality of initial management, inadequate technical facilities and unavailable drugs. Most deaths (59.2%) occurred within 72 hours of hospitalization. There was a high incidence of premature death associated with neonatal seizures [17] [30]. In survivors, the most common sequela reported across studies are neurological disorders, developmental delays and postnatal epilepsy [31]-[38]. In our work, the evolution was marked by the occurrence of epilepsy and neurological sequela such as cerebral palsy, language disorders, blindness and delayed psychomotor development. These sequela are relatively frequent and pose the problem of long-term follow-up. In our context, the socio-cultural realities and low socio-economic level make this care difficult. The following risk factors for death were identified: prematurity (p = 0.02), macrosomia (p = 0.006), intrauterine growth retardation (p = 0.006), the presence of dehydration (p = 0.005) or shock (p = 0.008), age greater than 48 hours and status epilepticus (p = 0.01). The association between the presence of status epilepticus and mortality supports the hypothesis that optimal seizure control in terms of both detection and treatment improves the neurological prognosis in these infants. Furthermore, improving short- and long-term neurological outcomes requires correct management of the risk factors of prematurity and intrauterine growth retardation, as well as fluid and electrolyte disturbances [17] [30] [39] [40].
5. Conclusion
Neonatal seizures have unique clinical features and are often subclinical. This requires immediate assessment to determine the cause, institute emergency treatment specific to the etiology and limit cerebral dysfunction. Hence there is a need to equip neonatology departments with continuous video-EEG equipment in order to refine the detection of electrographic seizures, but also to optimize the quality of perinatal care for better management of risk factors for neonatal seizures.