Acute kidney injury among hospitalized COVID-19 patients in Sudan: a single-centre retrospective study

Abstract

Background

Acute kidney injury (AKI) is a common complication of COVID-19 and may lead to persistent kidney dysfunction. Little is known about Sudan. This study aimed to assess the incidence, clinical characteristics, and outcomes of AKI among COVID-19 patients in tertiary hospital settings.

Methods

We retrospectively evaluated 868 of 990 adults with confirmed COVID-19 admitted to a Khartoum Teaching Hospital’s COVID-19 isolation center, Khartoum, Sudan in 2021. Patients on maintenance dialysis or with incomplete core data were excluded. AKI was defined was defined using Kidney Disease Improving Global Outcomes criteria and staged according to Risk, Injury, Failure, Loss, and End-stage renal disease criteria. Outcomes included in-hospital mortality and renal status at discharge. Comparisons used chi-square tests, and multivariable logistic regression identified independent predictors of survival.

Results

AKI occurred in 46% of hospitalized patients, mainly older males with hypertension or diabetes. The severity ranged widely. In-hospital mortality was 69%, and renal recovery among survivors was frequently incomplete. Multivariable analysis showed that increasing age (adjusted OR 0.970 per year), elevated D-dimer (adjusted OR 0.834 per µg/mL), and requirement for mechanical ventilation (adjusted OR 0.101) were independently associated with lower survival, whereas admission creatinine, urea, and nephrotoxin exposure were not independent predictors.

Conclusion

COVID-19–associated AKI was common and linked with high mortality and incomplete renal recovery. Early risk stratification using age and D-dimer, close renal monitoring, and strengthened critical care and dialysis capacity are essential in resource-limited settings. Prospective multicenter studies are needed to refine risk prediction and optimize management strategies.

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of coronavirus disease 2019 (COVID-19), was first identified in Wuhan, China, in December 2019 [1]. Clinical manifestations range from asymptomatic infection to respiratory and gastrointestinal symptoms and, in severe cases, multiorgan systemic involvement [2, 3]. Emerging evidence indicates that the kidneys are an important target of SARS-CoV-2 infection, likely related to the abundant expression of angiotensin-converting enzyme 2 (ACE2) receptors on renal tubular epithelial cells and podocytes, as well as their presence in pulmonary alveolar epithelium and vascular endothelium, which may contribute to the multisystem manifestations of COVID-19 [4].

During the previous SARS epidemic, acute kidney injury (AKI) occurred in approximately 6.7% of patients and was associated with very high mortality (> 90%) [5]. Although SARS-CoV-2 differs biologically from SARS-CoV, its impact on renal function remains incompletely understood [6–8]. Reports during the COVID-19 pandemic indicate substantial variability in AKI incidence, ranging from 5% to 46%, and a marked increase in demand for nephrology services and renal replacement therapy [6, 9–12]. While emerging studies suggest a potential direct or indirect effect of SARS-CoV-2 on kidney function, the magnitude of AKI risk and its clinical implications remain uncertain [13]. Data on COVID-19–associated AKI from Sudan remain scarce, making it difficult to fully understand its epidemiology and clinical outcomes in this setting and underscoring the need for locally generated evidence to inform clinical management and resource planning in resource-constrained healthcare systems. Accordingly, this study aimed to assess the incidence, clinical characteristics, and short-term outcomes of AKI among hospitalized patients with COVID-19 in Sudan.

Methods

Study design and setting

We conducted a retrospective cohort study at Khartoum Teaching Hospital’s COVID-19 isolation center, Khartoum, Sudan. All adult patients admitted between January 1 and December 31, 2021, were identified from the hospital’s admission registry.

Participants and eligibility criteria

All adult patients (≥ 18 years) with real-time polymerase chain reaction (RT‒PCR)-confirmed SARS-CoV-2 infection and a hospital stay of ≥ 24 h were eligible. Patients on maintenance hemodialysis and those with incomplete key data were excluded from the study.

Data collection and definitions of variables

Data was extracted from patients’ files by two independent, trained researchers using a standardized checklist to ensure complete and consistent capture of all relevant information. The checklist comprises three parts: (1) demographic data (age, sex); (2) disease characteristics (comorbidities, classification of AKI, and clinical outcomes); and (3) and nephrotoxin exposure and initial AKI management. Nephrotoxin exposure (NSAIDs, contrast media, RAAS inhibitors, aminoglycosides) was retrieved from medication lists and clinical notes and coded as a binary variable (any documented exposure vs. none); detailed information on timing, route, dose, and duration was inconsistently recorded and therefore not captured.

AKI was diagnosed using KDIGO thresholds levels, where an increase of serum creatinine (SCr) 0.3 mg/dL or 1.5 times from the baseline indicated the presence of AKI. The SCr level at admission represents the baseline for monitoring renal function changes throughout the hospital stay. Baseline renal function was defined by admission SCr as pre-admission values were not documented for the majority of patients; the number of patients with documented pre-admission SCr is reported in the Results to permit assessment of potential misclassification. To determine the severity of AKI, the RIFLE criteria were applied [14, 15], categorizing patients into stages of Risk (SCr increase by 1.5 times), Injury (SCr increase by two times), Failure (SCr increase by three times), Loss (complete loss of kidney function for more than four weeks), and End-Stage Renal Disease (complete loss of function for more than three months). Medication data during hospitalization were abstracted as structured categories (hemodynamic optimization, fluid–electrolyte therapy, diuretic use, and renal replacement therapy); route and agent were recorded when documented, but dosing and timing were inconsistently available and therefore are not reported quantitatively.

Outcomes included in-hospital mortality and renal status at discharge among patients with AKI. Renal outcomes were categorized based on serum creatinine (SCr) change relative to admission as:1) complete recovery (SCr returned to baseline or ≤ 1.1× baseline), 2) partial recovery (≥ 25% fall from peak but > 1.1× baseline), 3) no recovery (persistent elevation ≥ 1.5× baseline), 4) persistent dysfunction (abnormal SCr patient not dialysis at discharge), and 5) dialysis dependence at discharge. Mortality was identified from hospital records, and renal data were taken from the last available labs before discharge or death.

To minimize selection bias, we included all consecutive admissions meeting the criteria. We addressed information bias by using standardized definitions and two independent data collectors. No formal sample size calculation was performed; the cohort reflects all eligible cases during the study period.

Statistical analysis

Statistical analyses were conducted via the Statistical Package for Social Sciences (SPSS) version 30 (Armonk, NY: IBM Corp.). AKI onset ≤ 48 h after admission was classified as early AKI, whereas onset > 48 h was classified as late AKI [16]. Normality was checked using the Kolmogorov–Smirnov test. Continuous variables are presented as the means ± SDs or median ± IQR. Categorical variables are reported as frequencies and percentages and were compared via the chi-square test. Binary logistic regression was used to test the association between survival, demographics, baseline characteristics and laboratory markers. The model identified independent predictors of survival and reported odds ratios (ORs) with 95% confidence intervals. A P value below 0.05 indicated statistical significance.

Results

Among the 990 patients admitted with COVID-19 to Khartoum Teaching Hospital’s isolation unit, 50 were on maintenance dialysis, and 72 had incomplete core data (e.g. missing admission SCr, RIFLE staging, or unavailable discharge status), leaving 868 patients available for analysis. The incidence of AKI in this cohort was 46% (400 of 868), corresponding to 40.4% of all admissions.

Among 400 patients with COVID-19–associated AKI, the mean age was 62.4 ± 18 years; 61% were male, 62.7% were ≥ 60 years, and 7.5% were active smokers. Hypertension or diabetes was present in 94.5% of the patients; respiratory disease, heart failure, and multiple comorbidities occurred in 8%, 4.8%, and 42.5%, respectively (Table 1).

Table 1.

Patient demographics and comorbidities (n = 400)

Characteristic	n (%)
Age mean 62.4 ± 18.1 (years) Age Group ≥ 60 < 60	251 (62.7) 149 (37.3)
Gender Male Female	244 (61) 156 (39)
Smokers	30 (7.5)
Comorbidities Hypertension Diabetes mellitus Respiratory (Asthma/COPD/TB) Heart failure Combined comorbidities	214 (53.5) 164 (41) 32 (8) 19 (4.8) 170 (42.5)

According to the RIFLE criteria, 53.3% of AKI patients were at Risk, 12.8% were at Injury, 8.3% were at Failure, 7.5% were at Loss, and 18.3% were at End-stage (Table 2). AKI was diagnosed within 48 h in 73.5% of patients (mean time to diagnosis, 2.3 ± 2.2 days). Initial management included hemodynamic optimization in 83%, fluid–electrolyte therapy in 48.5%, and loop diuretics in 46% of patients. Nephrotoxic exposures were common (79.5%), most frequently NSAIDs (30%), contrast media (27.8%), and RAAS inhibitors (26%).

Table 2.

AKI Severity and initial management of the study cohort (n = 400)

Variables	n (%)
AKI severity Risk Injury Failure Loss of kidney function End-stage renal failure (ESRF)	213(53.3) 51(12.8) 33(8.3) 30(7.5) 73(18.3)
Initial Management Interventions Correction of primary hemodynamic disturbances Fluid and electrolyte therapy Loop diuretic use	332 (83) 194 (48.5) 184 (46)

Baseline laboratory tests demonstrated marked renal impairment and systemic inflammation (mean SCr 3.07 ± 3.6 mg/dL; D-dimer 3.5 ± 4 µg/mL) (Table 3). Urinalysis showed hematuria in 41.3% and proteinuria in 42.5%. Overall, 39.3% (157/400) of AKI patients required hemodialysis with the primary indications are detailed in summarized in Table 4.

Table 3.

Baseline clinical and laboratory characteristics within 24 h of admission among patients with COVID-19–associated AKI (n = 400)

Variables	Mean (± SD)/Median (± IQR)
D-dimer (µg/mL)	3.5 ± 4 a
BUN (mg/dL)	60 ± 70 a
Urea (mmol/L)	109.6 ± 68.5 b
SCr (mg/dL)	3.07 ± 3.6 b
WBC (x 10^9/L)	17.2 ± 7.8 b
• Lymphocytes count (%)	9 ± 13 a
• Neutrophil count (%)	71.75 ± 23.4
Hemoglobin (g/dL)	10.45 ± 3 a

^a Median (± IQR)

^b Mean (± SD)

Table 4.

Indications for hemodialysis among patients with COVID-19–associated AKI (n = 157)

Hemodialysis Indications	n (%)
BUN > 100 mg/dL	68 (17)
Encephalopathy and/or uremia	23 (5.8)
Life threatening electrolyte imbalances	17 (4.3)
Refractory acidosis	17 (4.3)
Volume overload unresponsive to diuretics	32 (8)

*Indications were not mutually exclusive

Despite critical care support, 61.8% were admitted to ICU, 41% received vasopressors, and 79.8% needed mechanical ventilation, and in-hospital mortality remained 69%. Among 124 survivors (31%), 11.5% regained baseline renal function, 19% developed CKD, and 9.5% progressed to maintenance dialysis.

Chi-Square test showed no significant association between timing of AKI onset or nephrotoxin exposure and mortality, renal recovery, or progression to chronic kidney disease (Table 5). By contrast, survivors were significantly younger and had lower D-dimer levels than non-survivors (Table 6), whereas admission serum creatinine, BUN and urea levels, comorbidity burden, and nephrotoxin exposure did not differ significantly.

Table 5.

Association between the timing of AKI diagnosis, exposure to nephrotoxic drugs and clinical outcomes

Variable	Clinical Outcomes			P value
	In hospital mortality	Progressed to CKD	Recovered normal kidney function
AKI diagnosis Early (≤ 48 h) Late (> 48 h)	197 (67.0%) 81 (76.4%)	59 (20.1%) 17 (16.0%)	38 (12.9%) 8 (7.5%)	0.164
Nephrotoxic drugs a Exposed Not exposed	224 (70.4%) 54 (65.9%)	59 (18.6%) 17 (20.7%)	35 (11.0%) 11 (13.4%)	0.7

* Chi-Square Test

^a Exposure = any documented exposure to the listed agents; timing, route, dose and duration were not recorded

Table 6.

Predictors of survival among patients with COVID-19–Associated AKI

Variables	Survived (n = 124)	Not survived (n = 276)	Adjusted OR	95% CI	P-value
Age (year) a	53.9 ± 20.2	66.1 ± 15.7	0.970	0.953–0.989	0.002
Female b	55/156 (35.3%)	101/156 (64.7%)	0.962	0.512–1.809	0.904
No. of comorbidities a	1.66 ± 0.94	1.79 ± 1.1	1.072	0.776–1.480	0.673
Mechanical ventilation c	66/319 (20.7%)	253/319 (79.3%)	0.101	0.048–0.212	< 0.001
Severity of AKI d - Stage 2: Injury - Stage 3: Failure - Stage 4: Loss of RF - Stage 5: ESRF	19 / 51 (37.3%) 12 / 33 (36.4%) 7 / 30 (23.3%) 32 / 73 (43.8%)	32 / 51 (62.7%) 21 / 33 (63.6%) 23 / 30 (76.7%) 41 / 73 (56.2%)	0.783 1.201 0.864 1.224	0.323–1.897 0.456–3.162 0.291–2.570 0.484–3.094	0.588 0.711 0.793 0.669
D-dimer level (µg/mL) e	3.3 (2.754)	5.46 (4.77)	0.838	0.76–0.92	< 0.001
SCr on admission (mg/dL) a	4.16 (4.46)	3.51 (3.51)	0.977	0.93–1.13	0.655
Urea (mmol/L) a	103.14 (75.2)	114.87 (63.25)	0.995	0.990–1.01	0.127
BUN (mg/dL) e	54.5 (31)	95.16 (100)	1.003	0.990–1.01	0.140
Nephrotoxic drug’s Exposure f	94 /122 (77%)	224/278 (80.6%)	0.613	0.31–1.23	0.167

Binary logistic regression

^a Mean (± SD)

^b Compared with male

^c Compared with patients who did not require mechanical ventilation

^d Compared with stage 1: Risk

^e Median (± IQR)

^f Compared to those who exposed to Nephrotoxic drug

Multivariable logistic regression adjusted for 10 demographic, baseline characteristics and laboratory markers such as age, sex, number of comorbidities, need for mechanical ventilation, admission serum creatinine, urea, BUN, D dimer level, nephrotoxin exposure and AKI stage was used to identify independent predictors of survival. Higher age, higher D dimer, and need for mechanical ventilation showed significant independent associations with lower survival. Sex, comorbidity count, admission creatinine, urea, BUN, nephrotoxin exposure, and AKI stage showed no significant association with mortality (Table 6). The model showed good fit and discrimination, with χ² 87.4 and p below 0.001, a Nagelkerke pseudo-R² ≈ 0.26, and overall classification accuracy of 80.1%.

Discussion

In this first study of COVID-19–associated AKI in Sudan, we observed an incidence of 46% among patients, a rate exceeding those reported in Italy (22.4%), Jordan (25.3%), and Tunisia (28%) yet parallelling cohorts in China (46%) and New York (45%) [17–19]. These findings underscore a critical interplay between delayed hospital presentation, a high burden of comorbidities (94.5% with hypertension or diabetes), and constrained diagnostic resources that likely amplified AKI detection and severity. Because diabetes, hypertension, and heart failure independently increase AKI risk, renal dysfunction in this retrospective cohort should be interpreted cautiously and not attributed solely to COVID-19.

In our cohort, in-hospital mortality among COVID-19–associated AKI patients was 69%, which was markedly higher than the 50% mortality reported in a recent Egyptian study of COVID-19–AKI patients [20] and exceeded the rates of 37% and 50% reported in multicenter cohorts from New York and Europe [18]. These differences likely reflect study-specific factors, including later presentation with advanced respiratory failure and multiorgan involvement in our cohort, together with systemic constraints such as strained critical-care capacity in Sudan, including limited ICU availability, ventilatory support, and renal replacement therapy resources. Similar challenges have been documented across African COVID-19 critical-care settings, where resource limitations and delayed presentation were associated with poorer outcomes [21]. Quantitatively, multivariable analysis demonstrated that mortality was primarily driven by markers of overall disease severity, older age (adjusted OR 0.970 per year), elevated D-dimer (adjusted OR 0.834 per µg/mL), and the need for mechanical ventilation (adjusted OR 0.101) rather than by AKI stage itself. These findings suggest that, in severe COVID-19, AKI frequently reflects the burden of systemic illness, particularly respiratory failure and coagulopathy, rather than acting as an independent determinant of death. This interpretation is supported by our multivariable model, in which markers of systemic severity, particularly mechanical ventilation requirement and D-dimer elevation, remain independently associated with mortality after adjustment for renal parameters and comorbidity burden. This highlights the importance of integrated cardiopulmonary–renal assessment when managing COVID-19 patients with AKI. In the same line previous study done in USA showing the mortality was predominantly associated with critical respiratory involvement and coagulation abnormalities, and AKI severity alone was not consistently an independent predictor of outcome [22].

AKI covered the full RIFLE spectrum; over half of the patients were at “Risk,” and nearly one-fifth reached “End-stage”. Rapid onset (mean 2.3 ± 2.2 days) and persistent renal dysfunction, with only 11.5% of survivors regaining baseline function and 30.6% being discharged with residual impairment, highlight the pandemic’s long-term effect on kidney health [23]. Although hematuria and proteinuria rates (41.3% and 42.5%, respectively) aligned with prior data [7], our reliance on estimated baseline creatinine may have modestly inflated incidence estimates.

Although nephrotoxin exposure was frequently documented (79.5%), the absence of an observed association with outcomes should be interpreted in light of the retrospective chart-based exposure assessment, which did not consistently capture timing or dose. Accordingly, AKI risk in this cohort likely reflects the broader severity of COVID-19 rather than isolated nephrotoxic exposure alone [24].

Further analyses revealed that only advanced age and elevated D-dimer were robust predictors of in-hospital mortality, whereas admission creatinine and urea levels lacked prognostic value. Notably, admission SCr was higher among survivors but was not independently associated with mortality after adjustment, suggesting that overall disease severity rather than initial renal dysfunction primarily influenced outcomes.

Our cohort’s exceptionally high need for ICU admission (61.8%), mechanical ventilation (79.8%), and dialysis (39.3%) compared with approximately 50% and 30%, respectively [25]. Three strategic imperatives have emerged for resource-limited settings: [1] early risk stratification leveraging age and D-dimer; [2] real-time renal monitoring with predefined intervention thresholds; and [3] scaling critical care and dialysis capacity in anticipation of surges.

This study has limitations inherent to its single-center, retrospective design. Pre-admission serum creatinine was unavailable for most patients; therefore, admission SCr was used as the pragmatic baseline for AKI ascertainment, an approach commonly adopted in acute-care cohorts when historical data are lacking [14, 25]. Urine-output criteria could not be applied because reliable hourly documentation was not consistently available, and AKI classification was thus based on serum-creatinine criteria alone. Finally, the absence of a non-AKI comparator group limits causal inference regarding AKI-specific risk factors. Despite these limitations, our study provides robust, real-world evidence from a large consecutive cohort with standardized data abstraction and multivariable adjustment, and it offers a practical framework for future prospective longitudinal, multicenter investigations to refine AKI risk models and optimize renal-support strategies in low-resource settings.

Conclusion

AKI affects nearly half of COVID-19 admissions and was associated with a high in-hospital mortality rate and incomplete renal recovery among survivors. Advanced age, elevated D-dimer levels, and markers of critical illness were the strongest predictors of death. In Sudan and other resource-limited settings, early risk identification, close renal monitoring, and strict nephrotoxin stewardship underpinned by expanded critical care and dialysis capacity are essential to improve outcomes. Prospective, multicenter validation of risk stratification models and therapeutic approaches is urgently needed.

Author contributions

Maab, Emmanuel, and Kannan contributed to the conceptualization of the study. Maab, Emmanuel, and Yaman were involved in methodology development. Funding acquisition and supervision were carried out by Maab. Emmanuel, Bashir, and Yaman conducted the investigation. Saad and Moataz performed the formal analysis and contributed to software and visualization along with Abdul Salam. Data curation was handled by Moataz, Bashir, and Abdul Salam. Resources were provided by Bashir, Abdul Salam, and Kannan. Project administration was led by Kannan. The original draft was prepared by Maab and Emmanuel, while Moataz, Saad, Bashir, Yaman, Abdul Salam, and Kannan were involved in review and editing. All authors read and approved the final manuscript.

Funding statement

The authors of the study received no funding during the duration of the study.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study adhered to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Medical Ethical Committee – Ministry of Health – Khartoum State (Serial Number: KMOH-REC-2022-NO.2.3). Written informed consent was obtained from all eligible participants, along with assurances of data privacy and anonymity.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.