Acute Kidney Failure during Preeclampsia at Parakou University Hospital in Benin in 2019: Diagnostic and Therapeutic Aspects

Background: Acute renal injury during preeclampsia is a relatively common complication in intensive care and grafted with significant morbidity and mortality. Objective: Describe the diagnostic and therapeutic aspects of acute kidney injury during preeclampsia at Parakou University Hospital center in Benin in 2019. Patients and Method: Cross-sectional descriptive and analytical study with prospective data collection carried out from March 18 th -August 18 th , 2019 (5 months) in the intensive care unit and the maternity ward. Women with preeclampsia were included. Acute renal failure, as dependent variable was diagnosed and classified according to the criteria of Kidney Disease Improving Global Outcomes (KDIGO). Results: A total of 88 patients were included in the study among whom, 39 developed an acute kidney failure (44.32%). The average age was 26.42 ± 6.23 years. The majority of them (44.32%) were at their first pregnancy. The pregnancy had not been followed in 61.54% of cases. The majority of patients with an acute kidney failure were classified as stage 1(46.15%). The patients admitted to the intensive care unit were at stages 2 and 3 for acute kidney failure. The clinical picture was complicated by: eclampsia in 38.46% of case, a HELLP syndrome in10.26% of cases and an acute pulmonary edema in 12.82% of cases. Medical treatment was based on hydro-electrolytic rebalancing, magnesium sulphate, antihypertensive and the transfusion of red blood cells and fresh frozen plasma. Cesarean section was performed for 76.92% of cases. The outcome was positive for 71.79% of patients. Conclusion: Preeclampsia is a risk factor for acute kidney injury and is a poor maternal prognosis factor.


Introduction
According to the World Health Organization (WHO), Pregnancy is defined as a period of about 9 months during which a woman carries the embryo, and then the fetus which grows in her womb and, for most women, is a time of great happiness and fullness [1]. Unfortunately, during pregnancy, the woman and the fetus face various health risks and many of these conditions expose them to morbid risks during pregnancy, childbirth and in the days after the postpartum period. Among them, we can mention the toxemia of pregnancy called preeclampsia [2]. Preeclampsia is an expression of kidney damage. It involves hypertension blood pressure, edema and proteinuria [3]. In developing countries, however, the incidence remains high. It's the case from India in 2010 where 20% of AKF cases were pregnancyrelated [13]. In Morocco the incidence was 66 cases per 10,000 pregnancies in 2013 [14]. This bearing is still highest in sub-Saharan Africa where in 2006 in Senegal, 50% of AKF cases among adults were due to pregnancy [7]. In the literature, preeclampsia appears to be the first cause of these AKF during pregnancy. In Benin, it is a frequent complication evaluated at 4.8% in Porto-Novo in 2017 by Tshabu-aguemon C, et al. but the AKF only represented 0.61% of complications [15]. In Parakou, approximately 11.73% of pregnant women developed an AKF in 2016 according to Ahoui S, et al. and proteinuria was significantly associated with occurrence of this AKF [7]. Faced with this high risk of occurrence of an AKF during the preeclampsia, we initiated this work which objective was to describe the diagnostic and therapeutic aspects of AKF in preeclampsia in Parakou University Hospital center of Benin.

Study framework
The study took place in the intensive care unit and the Maternity Department of Parakou University Hospital center. It is a multipurpose resuscitation unit working 24 hours a day with 11 beds. Maternity ward is a multi-purpose care unit (gynecological and prenatal consultations, childbirth, various surgeries, vaccination unit, family planning and various advices units).

Study method
It was a Cross-sectional descriptive and analytical study with prospective data collection, which took place over a period of five months from March 18 th -August 18 th , 2019.
The study population consisted of all admitted pregnant women or delivered women for preeclampsia in the maternity ward and for all delivered women admitted for preeclampsia in the maternity ward and transferred to intensive care unit.

Inclusion criteria:
It was included in the study any patient admitted for preeclampsia (pregnant ≥ 20 WA or postpartum < 42 days) with PAS (systolic blood pressure) ≥ 140 mmHg and / or PAD (diastolic blood pressure) ≥ 90 mmHg; with a proteinuria ≥ 2 crosses; with or without edema of the lower limbs and having given their free and informed consent or that of the family to participate in the study.
Non-inclusion criteria: It were not included the patients who came with added preeclampsia and other vasculo-renal syndromes.
Exclusion criteria: It was excluded patients with a history of kidney diseases, diabetes and patients who did not have a kidney exam during their hospitalization.

Sampling
We carried out an exhaustive recruitment of patients admitted for preeclampsia and meeting our inclusion criteria. The variables studied were classified in two groups.

Dependent variable:
Represented by acute renal failure which has been diagnosed and classified according to the Kidney Disease Improving Global Outcomes (KDIGO) criteria of 2012 revised in 2017 [16,17]. A single criterion is sufficient to establish the stage of severity. The Classification of acute renal injury according to KDIGO is presented in table 1. Renal injury in preeclampsia is defined by level of serum creatinine exceeding 10 mg.L 1 [18].

Socio demographic data
Age, level of education, profession, married or not, place of residence, way of life (supplementation of salt at the table, fruits and vegetables consumption (at least 5 fruits and vegetables per day)).

Clinical data
Mode of admission (referred or direct admission) and reason for reference, the previous treatment; obstetrician gynecologist history (preeclampsia/eclampsia, pregnancy, parity, period between two pregnancies called inter reproductive space (<2 years or ≥ 2 years), number of prenatal consultations (ANC), place childbirth, term of pregnancy; general signs (general condition (good or bad), arterial hypertension, edema of the lower limbs, proteinuria, jaundice, hemoglobinuria; physical signs: Glasgow Coma Score (GCS), diuresis (oliguria, anuria or normal), haemorrhages, convulsive fits, crackling rales, hydrops.

Obstetric treatment
Caesarean section or vaginal birth.

Data collection method and tool
It was done through an individual interview with each patient during which an individual file was completed with in addition the referral sheets, patient diaries, medical records and operative reports. An informed consent form has been completed and signed by patients or by a representative of the family. The study was conducted after obtaining authorization from local ethics and biomedical research committee of the University of Parakou.

1.
Those with serum creatinine < 10 mg/dL were considered to have no AKF and classified in Non-Cases.

2.
Those with a serum creatinine ≥ 10 mg/dL benefited from a second renal assessment on D2 of the first assessment with the results found: • An increase in serum creatinine meeting the definition of KDIGO and were included as Case; • An increase in creatinemia but which did not meet the definition of KDIGO or a decrease in the latter and were considered to have no AKF and classified in Non-Cases.
The AKF was monitored by measuring the serum creatinine on the 7 th day, 14 th day and at one (1) month and the criterion of good recovery was based on a return of lower serum creatinine to the patient's baseline. Transaminases and CBC were repeated according to initial disruption.

Data processing and analysis
The data were coded and analyzed using EPI INFO software. The data have been captured and analyzed using the Epi Info's software versions 3.5.1 and 7.1.1.14 from Center of Disease Control (CDC) respectively. The data analysis was carried out in two parts: 1) A descriptive section which consisted of calculating the percentages for the qualitative variables and means and standard deviation for quantitative variables.
2) An analytical component which consisted in measuring the association between the dependent variable and the independent variables. The measure of association used was the Odd ratio (OR).
The tests Chi-square, Fischer, or Yates-corrected Chi-square statistics were used as appropriate to determine the degree of significance of the association (p-value). The materiality threshold was set at 0.05.

Sociodemographic data of patients admitted for preeclampsia
During the study period, 1204 patients were admitted to the Gyneco-obstetrics department of Parakou University Hospital center. Among them, 104 (8.64%) developed preeclampsia and according to our inclusion criteria we collected 88 cases (84.62%). The average age was 26.42 ± 6.23 years with extremes of 16 and 45 years. The slice of age most represented is that of 20 to 30 years or 51.14% (n=45). Fifty-seven female patients were out of school (64.77%) compared to 31 female patients enrolled (35.23%). The self-employed women were 42 (47.73%), housewives 32 (36.36%), pupils/students 8 (9.09%) and 6 patients (6.82%) were civil servants. The women married represented 81.82% (n=72) of our population and unmarried 18.18% (n =16). Sixty-eight patients (77.27%) lived in urban areas compared to 22.73% (n=20) in rural area. Eighty-one patients (92.05%) were taking salt supplementation at the table and 69 patients (78.41%) regularly consumed fruits and vegetables.
Paraclinical data: Table 3 shows the distribution of patients admitted for preeclampsia according to paraclinical data.    Table 5 shows the averages and extremes of the laboratory tests of the patients with acute kidney injury

2) Laboratory data of patients with AKI
The distribution of patients according to the stage of ARI (KDIGO 2012) is recorded in table 6 (n=39).
Therapeutic data for patients with acute renal failure: Among the patients with an AKF, 25 (64.10%) were treated in the maternity ward against 14 (35.90) transferred to intensive care unit. The medical treatment consisted on a hydro-electrolytic rebalancing or even a vascular filling based on crystalloids (Salted serum 0.9%, Serum glucose 5%, and Ringer lactate) associated with bicarbonate serum in the case of metabolic acidosis and hyperkalemia. All patients received oxygen therapy using glasses. Magnesium sulfate was administered to 24 patients (74.36%). The antihypertensive drugs were used in 79.49% of cases (n=31). The drugs used were: Nicardipine; 29 patients (93.54%), Clonidine; one patient (3.23%) and combination Clonidine + Nicardipine; one patient (3.23%). Furosemide has been used in 6 cases (15.38%). The other therapies consisted of transfusions (Red blood cell; 7 cases and Fresh frozen plasma 2 cases).
In our study, 30 patients (76.92%) gave birth by caesarean compared to 9 (23.08%) by vaginal birth. The indication for dialysis was given in the 8 patients with anuria and in one patient with an acute pulmonary edema refractory to diuretics but it was not performed in any case for financial reasons.

Clinical and paraclinical data
The study allowed us to note that most patients were referred from centers peripheral health (76.92%). This finding is identical to that made by Tondi ZMM, et al. who reported that 71.9% of the patients in their series had been referred [19]. This high rate references found both in our study and in the international literature shows the inadequacy of the implementation of a decision tree in our maternity hospitals which explains the references of pregnant women at the stage of complications, a situation that could be avoided.
High blood pressure in pregnancy was the main reason for referral (53.33%) in our series followed by eclampsia (20%), altered state of consciousness (20%), oliguria (10.00%) and anuria (6.67%). On the other hand, in the study by Missamou A, et al. the alteration of consciousness was the main reason for admission (62%) [20]. Nulliparous (46.15%) were the most affected in our study unlike the studies by Arora [21][22][23]. Tondi ZMM, et al. reported in their study that 53.1% of patients were first-time mothers [19]. This percentage diversity could be explained by the fact that our study concerned only cases of preeclampsia and according to data from the literature, nulliparity is a risk factor [24] while other studies all involved obstetric AKF cases.
According to WHO recommendations, at least 4 antenatal consultations are recommended for the proper follow-up of a normal pregnancy or at best one per month [25]. In our study, the pregnancy was not monitored (absence of prenatal consultations) or had been poorly monitored (less than 4 antenatal consultations) in more than half of the patients (61.54%). This observation has also been made by others including Hachim [21,23]. This finding could be explained by the fact that not only the majority of patients were not educated (no culture of pregnancy monitoring) but also they were self-employed with low income monthly (47.73%). Yet the antenatal/ prenatal consultation (ANC) offers the opportunity for patients to do check-ups to detect preeclampsia early as well as the AKF. However it did not exist in our study a statistically significant association between the number of ANC and the occurrence of an AKF.  [27,19]. These different findings confirm the inconsistency of edemas during preeclampsia.
On admission, the average SBP was 183.95 ± 26.1 mmHg and the average DBP 113.46 ± 23.03 mmHg. High blood pressure was severe in 79.49% of patients, which is corroborated by the data in the literature according to which it is the patients with severe preeclampsia which have more AKF [11,28].  [19,7]. For Kabbali N, et al. more than half of the patients were in oliguria (57%) [29]. In the series by Hachim K, et al., 71% of patients were in oliguria [26]. That in pregnant and postpartum women, our results are similar to those data in the literature that oliguria is the main functional sign of AKI.  [29,26,19]. Creatinine, substance nitrogen comes from the breakdown of creatine, especially present in the muscles. It's an indicator of kidney condition. The higher its value, the more it reflects the degree of impairment of kidney function. Thus in our study, the high value of the average creatinemia reflects the significant degree of renal impairment in our patients.
In our series we have classified the patients according to the practical recommendations of KDIGO 2012 unlike the other authors who mostly used Risk Injury Failure Loss End-stage (RIFLE). Nevertheless there is a correspondence which can be made. KDIGO stages 1, 2 and 3 will correspond respectively to R, I and F of RIFLE. In our series, the patients presented more AKF at stage 1 (46.15%) and this finding is similar to that reported by Ngomas [20]. The classification of KDIGO was set up because it synthesizes the previous classification which is part of the RIFLE. She avoids considering the most severe form of an Acute Kidney Failure (AKF) which is fundamental since it is now accepted that minor modifications on renal function have a significant impact on the outcome of patients with short and long term.

Therapeutic data
According to data in the literature, the mechanism underlying hypertension is dominated by vasoconstriction. In our series, 79.49% of our patients received antihypertensive treatment including a    [19,29]. This treatment is justified by the fact that although the woman is edematous during PE, it is vascular empty which is responsible for the occurrence of functional AKF notified above. Thus the management of an AKF during pregnancy is going well and well by vascular filling in order to perfuse the kidney, as in the population in general. In our series, diuretics (Furosemide) were used in only 6 patients (15.38%). Among these 6 patients, 5 had presented acute pulmonary edema, hence the need to use diuretics to ensure water depletion in the lung parenchyma. The only remaining one who received secondary diuretic treatment is a case of AKF with anuria that benefited from an initial vascular filling during 48 hours without success and given the installation of a hypervolemia  [19]. The use of magnesium sulphate for the most part in our study is based on data from the literature. In effect during the preeclampsia, there is cerebral vasoconstriction in the brain posterior which causes (depending on the degree) oxygen weaning resulting in the onset of seizures. Magnesium sulfate by its mode of action (blocking of voltage-dependent calcium channels and sodium channels at the level of excitable cells) causes a vasodilator effect on small blood vessels allowing intracerebral reperfusion. Therefore, it is the drug of choice in the treatment and prevention of eclampsia. But during the AKF its use is limited or even prohibited by some authors. Before the time required for making the diagnosis of an AKF according to KDIGO, it was used in the vast majority of patients in our sample without knowledge of their kidney condition.
In our series, 17.95% of patients received a blood transfusion. This rate is very low compared to that reported by Keita [29,19]. In contrast, in Ansari MR, et al. the majority of patients gave birth vaginally (71%) [22]. The relatively high rate of caesarean section in our study could be explained by several settings. First, the patients were referred (76.92%) and in an emergency table that required immediate intake in the operating theatre for a cesarean. The free of charge has contributed to faster and more effective action on the part of the nursing staff; secondly in case of AKF, the expulsion of the fetus must be done quickly to limit progression of kidney injury [9].
The indication for dialysis was made in 23.08% of cases (n=9). For some financial reasons and lack of a universal health insurance program, none of them was able to benefit from this dialysis. This observation was made by Tondi ZMM, et al. who for the same reasons only one patient was able to be dialysed out of the 8 in her series [19]. With a percentage higher but still low, in their series, Missamou A, et al. and Ngomas J, et al. could dialyze respectively 4% and 4.9% of patients who were in need [20,23]. This low rate of dialysis is the reflection of a health system in our countries where social security is absent and health care costs are largely covered by the households themselves. The high cost of dialysis therefore becomes an obstacle to its implementation. On the other hand, for Zelmat S, et al. 25% of patients could be dialyzed [30]. In Benin, a health insurance project could be set up and its application would help to reduce the mortality of these patients and give better results.

Conclusion
At the end of this work, which aimed to describe the diagnostic aspects and treatments for acute renal failure during preeclampsia in Hospital University center of Parakou in 2019, we can remember that the AKF is very present with a frequency of 44.32%. Those most affected are young, first-time pregnant and none parity women. The typical profile of a patient at risk of developing an AKF is a patient with a history of preeclampsia, primiparous, between 32-36 WA and with a short intergenesic space. The anomaly of the most common urine output was oliguria. Treatment is essentially based on hydro-electrolytic rebalancing or even vascular filling, the use of antihypertensive drugs, of magnesium sulfate, the transfusion (red blood cell and fresh plasma frozen). The death rate from AKF in the present study is quite significant and could have been low if patients had been on dialysis.