The study was conducted in the maternity ward and hemodialysis unit at Benjamin Mkapa Hospital. The Benjamin Mkapa Hospital is a tertiary teaching Hospital of the University of Dodoma and is located within the University campus in Dodoma region. The Hospital is a zonal referral serving a population of central Tanzania regions. Daily, 1000 to 1200 patients of different conditions attend attended outpatient clinic which has a bed capacity of 400 beds with bed occupancy of 200 to 250 patients. There is a range of specialized and superspecialized services and Nephrology is among them. Renal replacement therapy in the form of dialysis and kidney transplantation is being offered as the main treatment for kidney failure patients.

The study was a cross-sectional analysis of Hospital-based secondary data from June 2019 to June 2023. The patients’ information was retrieved from the Hospital Health information system and register books in the maternity ward and haemodialysis unit. Pregnant women who had PRAKI were obtained using KDIGO criteria.

The study population comprised pregnant women who were admitted to the maternity ward and sustained acute kidney injury during or post-pregnancy. The kidney injury was confirmed using KIDGO criteria. The treatment of kidney injury was either a haemodialysis service or recovery without dialysis.

All pregnant women were admitted to the maternity ward at BMH within the period of study and confirmed with PRAKI using KDIGO AKIN criteria.

Pregnant women who did not meet the definition of AKI by KDIGO criteria had incomplete/wrong records in the BMH information system or were known to have chronic kidney disease were excluded.

All pregnant women meeting the inclusion criteria and having a complete record of their admission at the BMH from June 2019 to June 2023 were enrolled in the study. The sample size was calculated according to the formula below but the study managed to obtain only 51 pregnant women who had PRAKI and who were enrolled between the year 2019 to 2023.

N = ( Z ) 2 p ( 1 − p ) / d 2

where:

N—was sample size required.

Z—was the desired significance level of 1.96 for a 95% confidence interval.

p—was the proportion of 3.2% obtained from a study in Kenya (Kivai, 2019).

d—was the size of the mean difference of 0.05.

N = ( 1.96 ) 20.032 ( 1 − 0.032 ) / 0.052

N = 48 participants.

48 sample size of participants was determined.

A non-probability purposeful sampling technique was employed whereas directly searching from a Hospital computer database system (medipro-5 and Mapping Tag ERX) women who sustained PRAKI and obtained 51 out of 4007 pregnant women who were admitted in the maternity ward for delivery or referrals women from lower facilities further management due to maternal related complication.

The Questionnaire contained variables such as sociodemographic characteristics, clinical characteristics, mode of treatment and treatment outcome which include complete recovery, dialysis dependency or death were used to extract patient data from the Mapping Tag ERX Computer system of the Hospital. We retrieved information from 4007 women who were admitted for delivery or referred from other facilities as being detected to have kidney injury for further treatment. The diagnosis of PRAKI was made using KIDGO criteria in which patients with an increase in serum creatinine by 26.5 µmo/L within 48 hours of admission or creatinine of more than 50% from the normal creatinine of 84 to 96 micromol/l (µmo/L) within 7 days or there was a failure to pass urine (anuria) in 6-12 hours. The information extracted was entered into Excel, cleaned to ensure its quality then shifted into the SPSS system for data analysis. The participants were classified into three stages according to disease severity using the KDIGO criteria of AKI as follows (Table 1).

Independent Variable included; Social demographic characteristics such as age, sex, place of residence, parity, and Hospital /Place of referral. Social demographic information was checked and extracted from the Medipro-5 and Mapping Tag ERX Tag system which included age, sex, place of residence, and Hospital /Place of referral.

Hemorrhage was determined by haemoglobin and platelet levels as well as findings of blood in urine. Hemorrhage was determined by measuring haemoglobin, platelets, and blood in urine blood; the normal level of haemoglobin was determined to be 13 - 16 g/dl, normal level of platelets was determined to be 150,000 - 450,000/microliter (µL), blood in urine was measured by negative test of blood and no red blood cells. If the woman has a haemorrhage, the amount of haemoglobin was determined to be less than 13 g/d, platelets were determined to be less than 150,000/µL and in urine may have blood-positive and red blood cells

Puerperal sepsis was defined as having a high total leukocyte count and neutrophil absolute count. Puerperal sepsis was defined as high total leukocyte count and absolute neutrophil counts. In our laboratory, the normal total Leukocyte count was 3.5 − 10.8 × 103 per microliter (µL) and the absolute neutrophil count was 1.5 − 7.5 × 103 cells/µL. When the woman had puerperal sepsis the number of total Leukocytes was above 10.8 × 109 cells per litre and the absolute neutrophil count was above 7.5 × 103 cells/µL.

Eclampsia and pre-eclampsia were defined as the patient having high blood pressure and protein in the urine. Pre/Eclampsia was defined as having high blood pressure and the presence of protein in the urine. Normal blood pressure was defined as having blood pressure of 110/60 mmHg to 140/90 mmHg and normal urine protein was a urine dipstick of negative protein. Women with eclampsia were those with a blood pressure of more than 140 mmHg systolic and 90 mmHg diastolic blood pressure and had urine dipstick positive for protein (+1 or more). HELLP syndrome was defined as a syndrome of symptoms of High blood pressure, haemolysis (anaemia), elevated liver enzymes and low platelets HELLP syndrome syndrome with High blood pressure, haemolysis (low haemoglobin), elevated liver enzymes (ALT and AST) and low platelets. A woman who had HELLP syndrome had haemoglobin less than 13 g/dl, alanine transaminase (ALT) was 4 to 36 U/L aspartate transaminase (AST) was 9 to 32 U/L and platelets were less than 150,000/µL

The dependent variable was the patient outcome, defined as either recovery or death. Recovery was counted as the recovery of kidney function to normal functioning either spontaneously or after some sessions of haemodialysis. Normalization of kidney function whereas creatinine and urea return to normal value and good urine output. The normal creatinine was 84 to 96 µmo/L and blood urea nitrogen was 2.1 to 8.5 mmol/L). Therefore, the patient should no longer perform hemodialysis for those who were on haemodialysis or spontaneous recovery of kidney function for those who did not undergo haemodialysis and possibly discharged. Death; Patient counted to have died after death being certified. So death was confirmed by brain death.

The content validity of the research tool was checked by a statistician with good experience in quantitative research methods. The expert was asked to review each question to determine if it addressed the research objectives. The feedback from experts was analyzed and compared to determine the degree of content validity for each question. Any modification suggested was considered before pre-testing and data collection.

A pre-test to check the reliability of the tool was done with respondents with similar criteria as the study sample and appropriate modifications were made. The data collection tool was pre-tested at a nearby facility with a small number of patients (10% of the sample size). The purpose of pre-testing was to verify the adequate collection of desired information as well as ensure the consistency of the questions. In the pre-testing process, minor corrections were made before data collection.

Data was analyzed using STATA version 15 where descriptive analysis to generate proportions for the distribution of participants by the respective variable. Description statistics were conducted for all variables to obtain a clear description of the information; including a distribution of participants by the antecedents of PRAKI, mode of treatment, and treatment outcomes. Data was checked for their completeness and uniformity based on the daily collection.

In this cross-sectional study, 51 pregnant women with PRAKI were studied out of 4007 women who were admitted to the maternity ward in the period between 2019 and 2023. Out of the 51 participants with PRAKI, 74.5% were in the age group between 21 to 25 years. There were 78.4% of women whose parity was less than 5. Half (49.02%) came from the Dodoma Region as shown in Table 2.

At admission, 76.5% of individuals were in stage 3 of KIDGO Creatinine criteria. Reduced urine output was found to be 68% and of these 54.3% had anuria. Of about platelets, 23 (45.1%) were found to have Thrombocytopenia. 92.2% of patients presented with anemia and 63.5% with leukocytosis as shown in Table 3.

The causes of PRAKI were 23.53% PPH, 23.53% Eclampsia and 23.5% pre-eclampsia, as shown in Table 4.

Out of 51 pregnant women with PRAKI, 43.1% of patients received hemodialysis therapy as the renal replacement therapy, as shown in Table 5.

Recovery from renal injury after renal replacement therapy or other therapy was 66.7% and the rest died as shown in Table 6.

Among the individuals who recovered, 19 (86.36%) were treated with hemodialysis therapy, and others recovered spontaneously without renal replacement therapy. The rest did not receive haemodialysis treatment as renal replacement therapy (RRT). Among those who died majority did not get renal replacement therapy as shown in Table 7.

In this study, it was noted that eclampsia/preeclampsia and PPH were the main contributing factors of PRAKI. These findings are comparable to other similar studies such as the one from Muhimbili National Hospital in Tanzania and other studies from other places in Africa and across the globe. The study from Muhimbili National Hospital which is the tertiary Hospital like Benjamin Mkapa Hospital noted that PRAKI was due to HELLP syndrome, severe pre-eclampsia and postpartum haemorrhage [20] . Similarly, the study from Kenya reported that almost half of the patients with PRAKI were due to preeclampsia, eclampsia and HELLP syndrome [19] and that of Morocco indicated PRAKI was due to pre-eclampsia/eclampsia and Haemorrhagic shock were the main causes of PRAKI [4] . Likewise, a study in Malawi demonstrated that pre-eclampsia/eclampsia and antepartum haemorrhage were the main causes of PRAKI [18] . Moreover, a study from Tunisia showed hypertensive disorders (eclampsia/pre-eclampsia and HELLP syndrome) and haemorrhage were the common causes of PRAKI [23] . Furthermore, similar findings were also observed in a study from Somalia where pre-eclampsia/eclampsia and post-partum haemorrhage were reported as the main causes of PRAKI [24] . A systematic review study from the African continent also found pre-eclampsia, post-partum haemorrhage and sepsis as the main causes of PRAKI [25] .

Similarly, haemorrhage and hypertensive disorders as the main causes of PRAKI were also reported other studies from India, Nepal and China. A study from India observed antepartum haemorrhage and pre-eclampsia as the main causes of PRAKI [26] . Likewise, in a study from Nepal, HELLP syndrome and haemorrhages were the major causes of PRAKI [17] . Equally a systematic study from China showed pre-eclampsia/ecmplasia and post-partum haemorrhage as the main causes of PRAKI [6] . In another systematic review, the study demonstrated that preeclampsia is the main cause of PRAKI [27] . Contrasting findings were noted in studies from India and Pakistan which found sepsis as the main cause of PRAKI [16] [28] [29] .

In the study, eclampsia/preeclampsia and PPH might have contributed to the development of PRAKI because these disorders are very common inpregnancy and normally present at third trimester of pregnancy, at delivery or post-delivery periods [1] [16] [30] [31] [32] and are reported to cause acute kidney injury in pregnancy [32] [33] [34] . Pre-eclampsia and eclampsia causes systemic vasoconstriction due to release of antiangiogenic factors. Kidney injury develops due to thrombotic microangiopathy, glomeruloendotheliosis, mesangiolysis and vasoconstriction of peritubular vessels [3] [35] . Likewise, PPH causes kidney injury due to tubular necrosis and cortical necrosis as a result of ischemia [36] . The risk factors of PPH are poor obstetric history, anaemia and prolonged labour [37] . In the study, majority of patients had anaemia which put them in a risk of developing PPH and subsequently PRAKI if no urgent correction of haemorrhage. It is also reported that more PRAKIs occur in low-income countries due to a lack of adequate perinatal care and inappropriate management of pregnancy-related complications [3] [6] . Tanzania is a low-income country the women are at risk of developing PRAKI due to poor management of pregnancy-related complications and lack of adequate perinatal care as shown in the findings of the study.

In this study of 51 patients with PRAKI, Hemodialysis therapy as the mode of treatment was provided to more than half of patients. Of the remaining patients, a quarter recovered spontaneously from renal failure without renal replacement therapy and the rest deserved haemodialysis but missed it. Reasons for missing haemodialysis were ICU admission for mechanical ventilation where there is no set up of renal replacement. Some patients who were admitted in ICU and missed haemodialysis died. This observation was also reported from other studies where patients with PRAKI admitted in ICU due to critical condition the majority were reported to have died [26] .

Similar approach of treatment of pregnant women with PRAKI is demonstrated in similar studies from other places. In a study from Muhimbili National Hospital Tanzania, there were also pregnant women with PRAKI who underwent Haemodialysis [20] . It is also demonstrated in a study from Morocco that a proportional of patients with PRAKI required haemodialysisservices [4] . Moreover, asystematic review study and a study from Pakistan showed that more than quarter of women who had PRAKI needed underwent Haemodialysisservices [16] [34] . Furthermore, a study from Nepal showed more than a half ofwomen who presented with PRAKI required hemodialysis [17] . However, a study from Malawi reported spontaneous recovery with medical management without renal replacement therapy for all women with PRAKI [18] .

More than half of the patients survived at discharge with complete recovery from kidney injury. Of those who survived, the majority recovered following couples of hemodialysis therapy sessions for an average period of 11 days and the rest recovered without renal replacement therapy. Patients who died mostly were those who did not receive hemodialysis services and few deaths occurred to those who underwent haemodialysis services. Patients who were admitted to ICU most of them missed haemodialysis services and died. The explanation of death in these patients could be ICU admission and missing haemodialysis services. Patients admitted to ICU and in need of haemdialysis services are transferred to the haemodialysis unit. Other factors the patients had and might have contributed to high mortality were other co-morbidities such as anaemia, infections and pulmonary congestion.

This is similar to the Study that was done in India which showed patients with PRAKI when admitted to ICU had poor prognoses compared to patients without PRAKI [25] [26] [38] . High mortality was also revealed PRAKI coexist with other medical conditions compared to non-PRAKI of the same underlying disease [26] . Furthermore, patients with eclampsia who later develop PRAKI are reported to have poor outcomes [25] [39] . Black race with PRAKI is also associated with more deaths than other races, a study from the USA [40] . It is also reported that inaccessibility to dialysis services contributes to more deaths of patients with PRAKI as revealed in a study from Cameroon where 31.5% of death rates are due to dialysis inaccessibility [12] .

Death rate due to PRAKI varies in the African continent, it ranges between 0-34.4% as demonstrated in a systematic study from the African continent [25] . A study at Muhimbili National Hospital, Tanzania reported a 4% death rate due to PRAKI [20] . Similarly, a study from Cameroon reported more than half of PRAKI patients required dialysis therapy and at three months of follow-up some of the patients died, others had completely recovered and the rest had partial recovery [12] . Moreover, the Morocco study illustrated that mortality due to PRAKI was 20% [4] . Unlike the study from Malawi where there were no death due to PRAKI and complete recovery from PRAKI [18] .

Whereas a study from India reported complete recovery from PRAKI was three-quarters of the patients and few patients died [2] . A study from Pakistan showed different progression among PRAKI patients; complete resolution, partial resolution and progression to End-stage renal diseases (ESRD). Some of the patients died in three months follow-up [16] . A study from Nepal on patients with PRAKI reported that three-quarters of the patients recovered completely while the rest became dialysis-dependent [17] [41] .