A total of fifty-five (55) cases of eclampsia were included out of four hundred and forty-four (444) patients hospitalized in the intensive care units in the two hospitals, a rate of 12.39%.

The average age of eclamptics was 24.67 ± 1.41 years. The median was 23 years with the extremes of 15 and 38 years.

In most cases (85.45%), eclamptics had been referred from another hospitals for a proportion of 14.55% of patients who came by themselves. Primigravidae represented 52.73%. The prenatal consultation was carried out at least four times in thirty-four eclamptics, or 61.82% of cases.

Table 1 represents the sociodemographic data of eclamptics.

Medical history

No medical history was identified in forty-nine eclamptics or 89.09%. Pregnancy-related hypertension and eclampsia were found as medical histories in three (5.45%) and five (9.09%) patients respectively.

Clinical condition of eclamptics

Eclampsia occurred antepartum in forty patients (72.73%). Admission in intensive cares occurred within twenty-four hours following eclampsia in forty-eight patients (87.27%).

Neurologically, it was eclamptic status in twenty patients (36.37%) and four (9.09%) of them had a Glasgow score less than 9.

On the cardiovascular level, twenty-seven patients (49.09%) had a systolic blood pressure greater than 160 mm Hg and nine (16.37%) a diuresis less than

0.5 ml/kg/hour. Arterial oxygen saturation was less than 95% in four eclamptics (7.27%).

We present in Table 2 the clinical condition of eclamptics upon admission.

Severe anemia < 7 g/dl and severe thrombocytopenia < 50.000/mm³ were found respectively in four (7.27%) and five (9.09%) eclamptics. Cytolysis with AST > three times normal was found in eight eclamptics (14.54%). Acute kidney injury with serum creatinine > 1.5 times normal was also noted in eight (14.54%) other eclamptics.

Haptoglobin, schistocytes and LDH had not been performed.

The Table 3 represents the biological disturbances found upon admission of patients in intensive care.

Caesarean section was the main mode of delivery in forty-seven eclamptics (85.45%). General anesthesia with rapid sequence induction was the anesthesia technique in 95.75% of cesarean patients.

All eclamptics were placed on the magnesium sulfate protocol. Noradrenaline was used in four of the eclamptics who were shocked (7.27%). Mechanical ventilation was indicated in thirty-two eclamptics or 58.18%. Labile blood products (concentrated red blood cells and fresh frozen plasma) were administered to twenty-two (40%) eclamptics of whom four (7.41%) had been transfused with platelets.

Renal function was severely degraded in three patients (5.56%) who had received hemodialysis.

The length of stay was on average 4.27 ± 1.27 days.

Nine eclamptics died during hospitalization, giving a mortality rate of 16.36%.

Complications that occurred in patients during their stay in intensive care are shown in Table 4.

To identify the factors associated with eclamptic mortality, we carried out a univariate analysis, increasing the independent variables with mortality. Table 5 shows the associations of variables with mortality. History of pregnancy-induced hypertension or preeclampsia, need for norepinephrine, severity of initial neurological condition, and occurrence of complications during intensive care stay were associated with mortality. Table 4 represents the factors associated with mortality after crossing the variables.

Eclamptics admitted in intensive care units were those who had presented vital distress. The incidence was 12.39%, a relatively high frequency which could be linked to a very high rate (96.36%) of referral of eclamptics from peripheral health centers to the two hospitals, to an insufficient monitoring of risky pregnancies (38.18% of eclamptics had less than four prenatal consultations). Our results were comparable to those of work carried out in Dakar by Bah AO et al. [5] in 2000, Békoin-Abhé et al. [6] in 2018 in Ivory Coast, who reported incidences of 10.37% and 12.1% respectively. The incidence in our study was higher than those of the studies carried out in Benin by Adisso et al. in 2006 at CHU Parakou (0.76%) [3] , Ditrinou et al. in 2007 at CNHU-HKM (2.98%) [4] , Ahounou et al. in 2017 at CHUD-OP (5.5%) [5] . It was much lower than the incidence in the work of Moukoro P et al. (21.52%) [7] .

The study population was young as confirmed by the average age which was 24.67 ± 1.41 years. Extreme ages, under eighteen and over thirty-five, are known to be risk factors for preeclampsia [8] . Five eclamptics in our population were found in these age intervals. Our results were comparable to those of several African authors, such as Thiam L et al., Diouf et al., and Buambo-Bamanga et al., who found an average age of eclamptics, respectively 25 years, 24 years and 22 years [8] [9] [10] .

The clinical profile of eclamptics in our study was that of a young woman in a moderate to severe coma (63.64%), who presented repeated convulsive attacks (32.73%) or eclamptic status epilepticus (36.37%), who was placed on mechanical ventilation (40%), in a precarious hemodynamic state, treated with blood transfusion (40%) and noradrenaline (9.09%). In the work of Baye S et al., the frequency of comas (92.1%) [11] was significantly higher than ours. Elombila M et al., reported in their work on severe eclampsia, a need for mechanical ventilation of 4.2% [12] , a rate ten times lower than our results. This disparity is linked to the fact that our study population consists only of eclamptics admitted to intensive care, which is not the case in the work of Elombila M et al.

A history of pre-eclampsia and/or pregnancy-related hypertension increases the risk of severe pre-eclampsia sevenfold [13] [14] . These two risk factors were associated with mortality in our study (p = 0.01). Yancey et al., Sibaï et al., obtained the same results in their work [15] [16] . It is therefore important to implement a monitoring policy in level 3 maternity wards for women with a history of pre-eclampsia or pregnancy-related hypertension in order to reduce maternal mortality. Furthermore, the severity of eclampsia marked by a severe alteration of the state of consciousness (p = 0.02), hemodynamic instability (p = 0.02), the occurrence of complications (p = 0.0002), was associated with mortality and constituted a poor prognosis factor for eclamptics in the intensive care units of both hospitals. Reducing maternal mortality linked to eclampsia will therefore require the implementation of at least four prenatal consultations recommended by the WHO, the establishment of a compulsory anesthetic consultation for all pregnant women, the recommendation to refer early, towards a level 3 maternity ward, any pregnant woman diagnosed with pre-eclampsia, the recruitment of anesthetists and the opening of a position of anesthetist in any level 3 maternity unit, the improvement of the technical platform for intensive care in hospitals in Benin and the free healthcare or the establishment of universal health insurance for all pregnant women.