Mortality and predictors of acute kidney injury in adults: a hospital-based prospective observational study

Acute kidney injury (AKI) is a major global public health problem. It is expensive to manage and associated with a high rate of prolonged hospitalization and in-hospital mortality. Little is known about the burden of acute kidney injury in moderate to low-income countries. We aim to assess predictors of in-hospital mortality among AKI patients admitted to the medical ward. We prospectively identified patients meeting kidney disease improving global outcomes (KIDGO) AKI definitions from April to August 2019. Patients with underlying CKD and patients hospitalized for less than 48 h were excluded. The Cox regression model was fitted to identify predictors of mortality and statistical significance was considered at the p-value of less than 0.05. A total of 203 patients were enrolled over 5 months. Out of this, 121(59.6%) were males, 58(28.6%) were aged greater than 60 years, and 141(69.5%) had community-acquired acute kidney injury. The most common causes of AKI were Hypovolemia 99(48.77%), Glomerulonephritis 51(25.11%), and sepsis 32(15.79%). The overall in-hospital mortality rate was 12.8%. Stage 3 AKI (AHR = 9.61, 95% CI 1.17–28.52, p = 0.035), duration of AKI (AHR = 7.04, 95% CI 1.37–36.08, p = 0.019), length of hospital stay (AHR = 0.19, 95% CI 0.05–0.73, p = 0.012), and hyperkalemia (AHR = 3.61, 95% CI 1.12–11.71, p = 0.032) were significantly associated with in-hospital mortality. There is a high rate of acute kidney injury-related in-hospital mortality in adult patients admitted to the medical ward. The severity of AKI, hyperkalemia duration of AKI, and a short length of hospital stay were predictors of 30-days in-hospital mortality. Most of the causes of AKI are preventable and patients may benefit from early identification and treatment of these reversible causes.

www.nature.com/scientificreports/ services, the high morbidity and mortality associated with acute kidney injury in Ethiopia are challenging to the community. Government-run dialysis services in Ethiopia and access to dialysis are limited and some patients are referred to private dialysis facilities where treatment is limited by the inability to cover the costs of treatment 10,11 . Limited data is available regarding renal function tests in hospitalized patients, risk factor recognition, outcomes, and associated factors of AKI in Ethiopia 10,12 . Therefore, this study was aimed to assess predictors of in-hospital mortality among AKI patients admitted to the medical ward of Jimma Medical Center, Southwest Ethiopia.

Methods and materials
Study design and population. A hospital-based prospective observational study was conducted from April to August 2019 at the medical ward of Jimma Medical Center, Southwest Ethiopia. The internal medicine department of the hospital has over 80 beds for inpatients and has two renal units 13 . All adult patients with the diagnosis of acute kidney injury were consecutively enrolled in the study and followed till discharge from the hospital or death. Single population proportion formula was used to determine the sample size. Patients with age ≥ 18 years, patients with confirmed AKI, and patients who were willing to sign written informed consent were included in the study. On the other hand; patients who cannot be able to give the required information, hospitalized for less than 48 h, and those patients with underlying chronic kidney disease (CKD) were excluded from the study.
Sample size determination and sampling technique. The proportion of patients with AKI-related mortality was 31.2%, as reported from an earlier study done in Sudan among adult acute kidney injury patients 14 .
The required sample size was determined by using single population proportion formula: where n = the desirable sample size, Za/2 = 1.96 (the critical value at 95% level of significance), p = 0.312 (proportion of patients with AKI related mortality), d = 0.05 (acceptable marginal error) and 1 − p = proportion of patients that do not possess the character of interest. Replacing with values results in: The total number of adults patients with AKI admitted to the medical ward of the hospital during the last 6 months were 212. The number of the source population (N) was 212, which is below 10,000. Correction formula issued to determine final minimum sample size (nf).
By adding 10% drop out, the final minimum sample size was 142.
The consecutive sampling technique was used to collect data from patients who met the inclusion criteria.
Data collection and procedures. Data abstraction checklist and structured questionnaire were developed from a review of the literature [14][15][16] . We collected information about patient demographics, diagnosis, comorbidities (hypertension, diabetes, heart failure, chronic liver disease), laboratory and other investigation results (repeated measures of creatinine, urea, electrolytes, complete blood count, and liver function tests). The data collection tool was first pretested on 10 (5%) patients at Mizan-Tepi University teaching hospital to validate and check the consistency, applicability, and understandability of the questionnaire and data abstraction format before the actual data collection.
AKI and outcomes. AKI [17][18][19] . Comorbidities were defined as any chronic medical conditions that coexist with AKI. Renal recovery was defined as a return to baseline kidney function at hospital discharge. Duration of AKI was defined as the number of days from the first day the patient met the AKI criteria until they no longer did. Prolonged length of hospital stay was defined as hospital stays greater than 7 days 20,21 .
Outcomes The main outcome variable was all-cause mortality. Mortality was defined as death before hospital discharge. Other outcomes included recovery of kidney function at discharge (improved), non-recovery AKI, referral, and self-discharge. Patients were followed starting from diagnosis of AKI to occurrence of the event. was used for analysis. Descriptive statistics, such as frequency, percentage, mean, median, and standard deviation were used to summarize patients' baseline clinical characteristics. Frequency and proportion were used for categorical data, while the mean and standard deviation for continuous variables. Crosstabs were used for the comparison of proportions of categorical variables. The Cox regression model was fitted to determine independent predictors of mortality. Statistical significance was considered at the p-value of less than 0.05 on multivariate analysis.
Ethical approval and consent to participate. The institutional review board of Jimma University, institute of health, research ethics unit approved the study (Reference Number IHRPGA-574-2019), and a letter of permission was forwarded to the administration of Jimma Medical Center. Written informed consent was secured from all participants and the collected data was kept confidential. All methods were carried out in accordance with the relevant guidelines and regulations.

Results
Baseline characteristics of patients.  Table 2.
Based on KIDGO criteria, eighty-six patients (42.4%) had stage 3 AKI. Most patients (87.2%) had prolonged lengths of hospital stay. The median length of hospital stay was 11 days. More than half (57.2%) of patients had persistent AKI (AKI duration ≥ 7 days). The mean AKI duration was 6.59 days. From patients with urine output assessment at diagnosis of AKI, 29.4% had oliguric and 70.6% had non-oliguric AKI. The duration of oliguria was 6 ± 2 days. At discharge, only 64.5% had urine output assessment. From this, 3.53% were anuric and 7.07% were oliguric. Clinical characteristics of AKI patients are outlined in Table 3. www.nature.com/scientificreports/

Outcomes of AKI.
Mortality. The overall in-hospital rate of mortality was 12.8% (n = 26). Glomerulonephritis and hypovolemia were the most common causes of AKI which accounted for more than 65% of deaths (  Table 5. Other outcomes One hundred eight (53.2%) patients were discharged improved (Return to baseline kidney function). Of those patients discharged with improved renal function, 13% had sustained reversal and the other 40.2% had a late reversal. The average duration of late reversal was 11.30 days. The minimum and maximum duration of late reversal were 8 and 21 days respectively. Patients discharged with non-recovery AKI were 62 (30.5%). The remaining 4(2%) patients were referred for further investigation and treatment and 3(1.5%) patients were self-discharged.
The probability of survival among the proportion of patients with stage 3 AKI decreases with time compared to patients with stage 1and stage 2 AKI (Fig. 1).

Discussion
The pattern of acute kidney injury (AKI) in this study is similar to that reported in developing countries. In the present study; community-acquired AKI accounted for 69.5% of cases, almost within a range of results reported from a study done in Sudan 14 . AKI in developing countries is believed to be community-acquired, occur between 70 and 90% of cases 9,22 . Lack of resources for early detection and treatment, patient's late arrival to health facilities with advanced-stage disease, and delay in diagnosis can contribute to the increase in the burden of AKI in developing countries.
The overall in-hospital mortality of AKI in our study (12.8%) is much lower compared to findings from other studies, 22.6% in Egypt, 36.9% in Cameroon, 44.4% in Malawi and, 20.3% in Singapore 6,16,23,24 . The variation with these studies might be due to; the difference in patient's site of admission (i.e. ICU vs general medical ward), study design, and underlying disease of patients. Our study excludes patients with underlying CKD, which is a known predictor of in-hospital mortality in patients with AKI. The result of this study was comparable with other studies, 13% in Egypt 25 and 10.8% in Sanford-USA 26 . Table 5. Bivariate and multivariate cox regression analysis of predictors of mortality (n = 26). AKI, acute kidney injury; HR, hazard ratio; AHR, adjusted hazard ratio; CI, confidence interval; LOHS, length of hospital stay; Dur-AKI, duration of AKI; RRT, renal replacement therapy. *Statistically significant. www.nature.com/scientificreports/ Among those who died, the severity of AKI as per KDIGO staging, presence of hyperkalemia, duration of AKI, and length of hospital stay were significant predictors of in-hospital mortality. Patients with stage 3 AKI had a higher risk of 30-days in-hospital mortality when compared with patients with a less severe stage of the disease ( Table 5). The reason might be due to the patient's' delayed admission to the hospital with an advanced stage of the disease and missed attempts to prevent the progression of the disease. This is consistent with several other studies done in Egypt, Brazil, and India respectively 25,27,28 . Another study also showed that even patients with a 25-50% increase in peak serum creatinine levels from baseline have an increased risk of mortality. Stage of AKI is reported to be a predictor for mortality and decreased kidney function, correlate with short term and even longer-term adverse outcomes 1,29 . In this study; patients with a short length of hospital stay (≤ 7 days) were shown to have a high risk of inhospital mortality. This is similar to a result reported from a study done in Nigeria 30 . This may be due to limited resources to detect AKI early in health centers and primary hospitals, which leads to late referral and admission of patients with a severe stage of AKI, usually with complications, and died early on hospital admission. Persistent AKI can be associated with worse clinical outcomes include increased length of stay, time on the ventilator, and days with cardiovascular failure 31 . We found that patients with persistent AKI had an increased risk of in-hospital mortality. This is similar to findings from other studies where persistent AKI is shown to be associated with increased risk of mortality 25,32 . Studies suggest AKI duration along with the stage of AKI should be used as a parameter for the prediction of worse outcomes especially. This may help physicians in preventing in-hospital and post-discharge mortality especially in critically ill, high risk patients 33 .
The in-hospital mortality rate of patients with hyperkalemia was shown to be 30.7% 34 . In this study, hyperkalemia patients had higher mortality rates than patients with normokalemia. Other studies also showed that, compared to patients with normokalemia, hyperkalemia was associate with a higher percentage of death 28,35 . Therefore, early identification of high-risk patients and maintaining serum potassium level within a safe range may help to avoid the risk of hyperkalemia related-mortality.
Our study has some limitations. First, urine output assessment was done in around 70% of patients. Second, patients with underlying CKD were excluded from the study. Furthermore, no outcome monitoring was performed beyond the period of hospitalization. Therefore, important information, including the incidence of CKD after AKI, mortality rate, and ESRD, and other non-renal outcomes are unknown.
In conclusion, the pattern of AKI is consistent with other studies from the developing world. There is a high rate of acute kidney injury-related in-hospital mortality in adult patients admitted to the medical ward. The severity of AKI, hyperkalemia, duration of AKI, and short length of hospital stay were predictors of mortality. Most of the causes of AKI are preventable and patients may benefit from early identification and treatment of these reversible causes.

Availability of data and materials
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.