Incidence and predictors of acute kidney injury among women with severe pre-eclampsia at Mbarara Regional Referral Hospital

We prospectively enrolled70 participants on the maternity ward at Mbarara Regional Referral Hospital (MRRH) to determine the incidence and predictors of acute kidney injury in mothers admitted with severe pre-eclampsia. MRRH is a public regional referral hospital that serves as a teaching hospital for Mbarara University of Science and Technology. The hospital serves as a referral for the western region of Uganda but also receives patients from the neighbouring countries like Rwanda, Congo, and Burundi. The hospital offers specialized services, employs eleven obstetricians and 32 midwives together with resident doctors who perform over 10,000 deliveries annually. The hospital also can perform laboratory tests including chemistries, urinalysis, and ultrasonography among others. A resident nephrologist in the hospital can perform acute hemodialysis for patients who require the service.

The study population consisted of all women having severe pre-eclampsia and eclampsia between November 2018 and April 2019. We included all women who were diagnosed with severe pre-eclampsia/eclampsia defined as high blood pressure (≥ 140/90mmHg) plus one or more of the following symptoms of severity (headache, epigastric pain, blurring of vision, convulsions, and proteinuria plus 2 or 3 on urine dipstick). Women who had a history or documentation of chronic kidney disease, chronic hypertension and gestational hypertension were excluded. Eligible women were recruited prior to delivery until the sample size was attained. Consent was obtained and blood samples were taken for baseline renal function tests. Serum creatinine was measured using the kinetic colorimetric method (fixed rare) Jaffe reaction without deproteinization. After 48 h another sample was drawn for a 2nd creatinine. A relative change of serum creatinine value at least 1.5 times the baseline (i.e., at admission) within 48 h was considered AKI and patients were followed up by the study team while in hospital from the time of admission to when they were discharged. In this study, the KDIGO definition of AKI without urine criteria was used. Acute kidney injury was staged as follows: Non-AKI < 1.5 times change of the baseline creatinine, stage 1 as 1.5 to 1.9 times the baseline creatinine and stage 2 as 2.0 to 2.9 times the baseline and stage 3 which is defined as a change > 3.0 times the baseline creatinine. The definition of AKI in pregnancy has not been standardized. Different studies use different parameters (consensus definitions of AKI) which can range from a change in serum creatinine or the need for dialysis [7]. The physiological decrease in serum creatinine in pregnancy compared with similarly healthy, non-pregnant individuals is due to volume expansion and vasodilatation. This in turns leads to increased renal plasma flow and glomerular filtration rate resulting in an increased GFR [8]. The sample size was estimated based on the Fleiss formula (1981) for cohort studies.

The major outcome variable of the study was acute kidney injury.The primary exposure variable was severe pre-eclampsia while the secondary exposure variables were sociodemographic factors, healthcare system factors, obstetric factors which included gravidity/parity, gestation age at diagnosis of pre-eclampsia and admission to the hospital, postpartum hemorrhage, antepartum hemorrhage, eclampsia, HELLP syndrome and previous history of pre-eclampsia. The medical factors were a history of diabetes, and HIV infection.

In this prospective study, our main goal was to establish the incidence and predictors of AKI among mothers admitted with severe pre-eclampsia to the maternity ward of Mbarara Regional Referral Hospital (MRRH). We found a high cumulative incidence of AKI of 42.86% and eclampsia as the main independent predictor of AKI, within 48 h after admission.Our findings were lower than what was reported among severe pre-eclampsia/eclampsia patients in Cameroon among whom 66.7% developed acute kidney injury. This could be because a majority ( 72.5% ) of the enrolled participants in the study done in Cameroon had eclampsia [9] compared to our study which had only 21.4% patients with eclampsia. Furthermore, the Cameroon study was conducted in an intensive care unit and also had 7 years of follow-up. Data from developed countries suggests lower rates of AKI, with estimates as low as 1-2.8% [10]. However, in developing countries, the incidence of AKI is high ranging from 36–66.7% [11, 12], Cameroon (66.7% ) [13], Morocco (66.7%) [14]. The high incidence of AKI in developing countries may be attributed to the late presentation of patients to hospitals for care.

In our study, eclampsia was the main independent predictor of AKI (aOR 5.89 (1.51, 38.88), p-value 0.014. This is in agreement with a prospective hospital-based study done in India which revealed that acute kidney injury was common among patients with eclampsia as high as 25–50% [15]. This may be so because pre-eclampsia and eclampsia are associated with glomerular endotheliosis which decreases the glomerular filtration rate and renal blood flow. An increase in renal vascular resistance predisposes patients to AKI which is characterized by proteinuria and renal failure.

In our study, we noted that antenatal care attendance was associated with a reduction in the incidence of acute kidney injury compared to non-attendance; crude odds ratio 0.14, 95% C.I, 0.027, 0.73, p-value 0.02. However, this was not significant at multivariate analysis and there was no association between the number of times the mothers attended antenatal care and the incidence of AKI. There is evidence to suggest that women with severe pre-eclampsia who present with severe complications are likely to have attended ANC fewer times than those without major complications (Manandhar et al., 2013). This may be so because antenatal attendance is associated with early detection of disease and timely intervention due to the structured routine screening of blood pressure and urine protein during antenatal visits.

However,some studies found no association between the number of times of antenatal care attendance and the severity of pre-eclampsia on presentation [16, 17]. This may be because the disease progression pattern is most times unpredictable and acute complications can occur without warning symptoms.

The inability to use measure more sensitive renal biomarkers for the diagnosis of AKI was one of our study limitations. Serum creatinine takes time to rise and may lead to late detection of AKI. The timing of sample collection in this study was done at admission and 48 h later. This may have prohibited some cases being identified as it is possible that AKI might have been detected even after 48 h and within 7 days from admission. Furthermore, the physiological decrease in serum creatinine in pregnancy compared with similarly healthy, non-pregnant individuals which occurs due to volume expansion and vasodilatation which ultimately leads to increased renal plasma flow and glomerular filtration rate resulting in an increased GFR [8].

Urine output is not routinely monitored quantitatively unless in very sick or catheterized patients.

The incidence of acute kidney injury in patients with pre-eclampsia/eclampsia is high (42.86%).

The presence of eclampsia in patients admitted with severe pre-eclampsia is associated with the development of AKI within 48 h after hospitalization. We recommend a larger prospective study that will include the use of biomarkers for acute kidney injury and closer monitoring of mothers with eclampsia for features of acute kidney injury. It is also important to increase both clinician and patient awareness to minimize the risk of AKI in women with pre-eclampsia.