Outcomes of acute kidney injury in critically ill adults with tropical acute febrile illness: Prospective observational study

Abstract

BACKGROUND

Tropical acute febrile illness (TAFI) is a major illness with a large number of intensive care unit (ICU) admissions in tropical regions. It is frequently complicated by acute kidney injury (AKI) and associated poor outcomes. The data on the predictors of outcomes, in terms of in-hospital mortality and renal outcomes, amongst critically-ill TAFI patients from our region remain limited.

AIM

To evaluate the clinical profile and outcomes of AKI amongst critically-ill TAFI patients and to identify the factors associated with in-hospital mortality.

METHODS

This prospective observational cohort study was conducted amongst the patients admitted in ICU at a tertiary care teaching hospital in the state of Rajasthan (India) between June and September 2025. Adult patients (aged ≥ 18 years), with laboratory-confirmed TAFI and AKI diagnosed using AKI network criteria, were included in the study. Demographic, clinical, laboratory, and organ-support variables were recorded. The primary outcome was all-cause in-hospital mortality while the secondary outcomes included renal recovery at the time of discharge and a composite net negative outcome (NNO) (defined as in-hospital mortality or persistent renal dysfunction during discharge). Univariate and exploratory multivariable logistic regression analyses were conducted to identify the predictors of the outcomes.

RESULTS

The data collected from a total of 68 critically-ill TAFI patients with AKI was analyzed. Their median age was 52 years while 37 (54.4%) were females. Scrub typhus 43 (63.24%) was found to be the most common etiology, followed by dengue 9 (13.24%). AKI stage I was found to be the most frequent (37; 54.4%) at ICU admission, while 19 (27.94%) required renal replacement therapy. Overall, the in-hospital mortality was 8 (11.8%) while NNO was observed in 23 (33.8%) patients. In terms of multivariable analysis, higher APACHE II scores and lower arterial oxygen tension at admission independently predicted the mortality. Dialysis requirement was found to be a strong predictor of NNO (odds ratio 21.95), along with hypoxemia and greater illness severity.

CONCLUSION

AKI in critically-ill TAFI patients has the modest mortality yet it results in substantial morbidity and incomplete renal recovery outcomes. Early hypoxemia, illness severity, and the need for dialysis are the key predictors of adverse outcomes. The study findings emphasize the importance of a comprehensive severity assessment beyond the existing creatinine-based AKI staging method.

Key Words: Tropical acute febrile illness; Tropical fever; Acute kidney injury; In-hospital mortality; Persistent renal injury; Observational cohort study; Outcome predictors; Scrub typhus; Dengue; Multiple organ dysfunction syndrome

Core Tip: Acute kidney injury (AKI) with tropical acute febrile illness is associated with significant morbidity and low mortality. Early hypoxemia, higher illness severity scores, and the requirement for dialysis independently predict adverse outcomes. This study highlights the need for early risk stratification in this patient subgroup beyond creatinine-based AKI staging.

INTRODUCTION

Tropical acute febrile illness (TAFI) is a common clinical syndrome in tropical and subtropical regions and a major factor behind hospital and intensive care unit (ICU) admissions in developing countries. TAFI is defined as an acute febrile illness with less than two weeks duration, documented fever (≥ 37.5 °C) and non-specific systemic symptoms without an obvious localizing focus of infection[1]. A spectrum of etiologies-malaria, dengue, leptospirosis, scrub typhus, enteric fever, and other tropical infections-accounts for considerable morbidity and mortality rates across India. Seasonal surges during monsoon and post-monsoon months, coupled with delay in seeking appropriate healthcare and limited early diagnostic access, contribute to severe disease presentations and high complication rates[2].

Acute kidney injury (AKI) is one among the most frequent and serious complications of TAFI and is associated with higher mortality, prolonged hospitalization, and heavy utilization of the healthcare resources. According to the AKI network (AKIN) criteria, AKI is defined by a sudden (within 48 hours) increase in serum creatinine level, up to ≥ 0.3 mg/dL or an increase up to ≥ 1.5 times of the baseline value, or a reduction in urine output to < 0.5 mL/kg/hour for at least six hours[3]. The AKI mechanisms in tropical infections are multifactorial and may include prerenal hypoperfusion from volume depletion, direct pathogen-related renal injury, immune-mediated inflammation, rhabdomyolysis, disseminated intravascular coagulation, and hemodynamic instability[4].

The incidence of AKI varies widely across the tropical infections. For instance, 5%-15% in uncomplicated dengue to over 40%-80% in severe malaria or leptospirosis cases. Further, the severity of AKI is heavily influenced by multiple factors such as age, comorbidities, severity of the tropical infections, and delayed presentation[5,6]. The development of AKI in critically-ill patients significantly worsens the outcomes, with steeply increasing chances of mortality amongst those requiring renal replacement therapy (RRT)[7]. Despite its clinical importance, the prospective data on TAFI-associated AKI in Indian ICUs remain limited.

Given this gap, the aim of the present study is to evaluate the clinical profile and the outcomes of AKI amongst critically-ill TAFI patients and to identify the factors associated with in-hospital mortality.

MATERIALS AND METHODS

This observational cohort study was conducted amongst the patients admitted in ICU between June and September 2025 at a tertiary care teaching hospital in the state of Rajasthan (India). The study was conducted after obtaining the approval from the Institutional Ethics Committee and was registered with Clinical Trials Registry of India with ref No. REF/2025/08/112774.

The study population included all the adult patients (aged ≥ 18 years), admitted to the ICU, with a confirmed diagnosis of tropical fever (based on positive microbiological or serological tests, and included dengue, malaria, leptospirosis, scrub typhus, enteric fever, or mixed infections) and were diagnosed with AKI, fulfilling the AKIN criteria, at or during the ICU admission[3].

For this study, patients with known or newly-diagnosed chronic kidney diseases, or without complete essential inclusion data were excluded. Informed written consent was obtained from the patients prior to their participation. Demographic, clinical and laboratory data were collected from the patient’s medical records, such as their age, gender, comorbidities, microbiological etiology of the tropical fever, duration of the fever before admission, hemodynamic status, need for vasopressor, mechanical ventilation, presence of organ dysfunction, need for organ support, detailed initial laboratory parameters (e.g., complete blood count, liver and renal function tests, electrolytes, coagulation profile, biomarkers) and the worst values of the laboratory parameters. Apart from these variables, the information regarding the primary outcome variable i.e., all-cause in-hospital mortality, was collected. The data regarding secondary outcome variables such as the renal function status at discharge, composite of net negative outcome (NNO) [in-hospital mortality or, persistent renal dysfunction (need for RRT, or raised creatinine > 1.5 times) at discharge], length of ICU stay, length of hospital stay and the duration of RRT at hospital was also collected.

Statistical analysis

The sample size was calculated using the single-proportion formula. Assuming 95% confidence level (Z_1-α/2 = 1.96), an expected mortality incidence of 20% amongst the TAFI-associated AKI patients, and an absolute precision of 10%, a minimum of 62 patients was yielded[8]. To account for an anticipated 10% attrition due to dropout from the study or early patient discharge against medical advice before outcome assessment, six patients were included, thus arriving at the final sample size to be 68 patients.

The collected data was processed and tabulated using the Microsoft Excel (Microsoft Corporation, Redmond, WA, United States). The tabulated data was analysed further using IBM SPSS Statistics for Windows, Version 25.0 (IBM Corp., Armonk, NY, United States).

The collected data was first tested for normality using the Shapiro-Wilk test. Then, the continuous variables were presented as mean ± SD or median with interquartile range (IQR) as found suitable in normality testing. The categorical variables were expressed as n (%). Patients were stratified into two groups based on “the in-hospital mortality” criterion. The independent groups were compared using student t-test or Mann-Whitney U test for continuous variables, as found appropriate whereas χ² test or Fisher’s exact test was used for categorical variables. Further, exploratory binomial regression analysis was conducted for in-hospital mortality and NNO at discharge separately by including the variables with P < 0.1 in univariate analysis or variable, known from the literature to influence outcomes. Thus, the independent predictors of the poor outcomes were identified. Adjusted odds ratio (OR) with 95% confidence interval (CI) was calculated and presented. All the statistical tests used in the study were two-tailed and P < 0.05 was considered to be statistically significant.

RESULTS

During the study period, 2639 patients visited various departments of the study institute for fever and suspected TAFI. Amongst them, 294 (11.14%) patients tested positive for TAFI, either microbiologically or serologically. Out of this, 159 (6.02%) patients were admitted to the ICU while a cohort of 68 (2.58%) patients with TAFI in ICU and positive AKI criteria was included in the study (Figure 1).

Figure 1 STROBE flow diagram of the study participants. ICU: Intensive care unit; AKI: Acute kidney injury.

The median (IQR) age of the study cohort was 52 (37.5-60.75) years, with predominantly 37 (54.4%) female patients. These patients were presented to the hospital with various symptoms among which nausea/vomiting 52 (76.5%), headache 40 (58.8%), myalgia 40 (58.8%), and altered sensorium 33 (48.5%) were commonly found. Amongst all the prevalent tropical fever aetiologies at the study ICU, Scrub typhus was the most commonly diagnosed aetiology (43; 63.24%) followed by dengue (either alone or with co-infection) (18; 26.5%) and typhoid (either alone or with co-infection (7; 10.3%) in TAFI with AKI patients. The median duration of the fever before admission was 7 days (IQR 5-10.5), and the median hospital stay was 9 days (IQR 5-12.75). At the time of ICU admission, the median serum creatinine was 1.35 mg/dL (IQR 0.9-2.9), with AKI stage I being the most common stage i.e., 37 (54.4%). While 19 (27.94%) patients required RRT, 9 (13.24%) required intermittent hemodialysis. Vasopressors and invasive ventilation were required in 28 (41.2%) and 27 (39.7%) patients, respectively. The in-hospital mortality was 8 (11.8%), while the NNO at discharge (mortality or persistent renal failure) occurred in 23 (33.8%) patients (Table 1 and Supplementary Table 1).

Table 1 Baseline characteristics of the study population.

Variable	Value
Age (years)	52 (37.5-60.8)
Hospital stay (days)	9 (5-12.8)
ICU stay (days)	7 (4-9.8)
Fever duration (days)	7 (5-10.5)
Temperature (°F)	100.5 (99.4-101.6)
Heart rate (minute)	105.3 ± 18.0
SBP (mmHg)	113.9 ± 18.8
DBP (mmHg)	74 (66-82)
Respiratory rate (minute)	22 (20-26)
Oxygen saturation (%)	96 (94-98)
Hemoglobin (g/dL)	10.8 ± 2.2
Platelet count (mm3)	80000 (51000-124750)
Serum creatinine (mg/dL)	1.35 (0.90-2.90)
Serum urea (mg/dL)	68.2 (43.5-127.0)
Bilirubin (mg/dL)	1.4 (0.8-3.6)
AST (U/L)	126 (67-212)
ALT (U/L)	76 (47-114)
APACHE II (admission)	15.6 ± 7.8
SOFA score (admission)	8.6 ± 3.9
Gender
Female	37 (54.4)
Male	31 (45.6)
Common symptoms
Fever	68 (100)
Nausea/vomiting	52 (76.5)
Headache/myalgia	40 (58.8)
Shortness of breath	36 (52.9)
Altered sensorium	33 (48.5)
Comorbidities
Diabetes mellitus	15 (22.1)
Hypertension	17 (25.0)
Chronic liver disease	6 (8.8)
Oxygen/ventilation
Invasive ventilation (any time during ICU stay)	27 (39.7)
NIV (any time during ICU stay)	9 (13.2)
Vasopressor use	28 (41.2)
AKI stage on admission
Stage 1	37 (54.4)
Stage 2	27 (39.71)
Stage 3	2 (2.9)
Renal replacement therapy
Dialysis required	19 (27.9)
Tropical fever etiology
Scrub typhus	43 (63.2)
Dengue (± coinfection)	18 (26.5)
Typhoid (± coinfection)	7 (10.3)
Outcomes
In-hospital mortality	8 (11.8)
Net negative outcome	23 (33.8)

Presented as mean ± SD, n (%) or median (IQR). AKI: Acute kidney injury; ALT: Alanine aminotransferase; APACHE-II: Acute Physiology and Chronic Health Evaluation II; AST: Aspartate aminotransferase; DBP: Diastolic blood pressure; ICU: Intensive care unit; IQR: Interquartile range; NIV: Non-invasive ventilation; SBP: Systolic blood pressure; SOFA: Sequential organ failure assessment.

There was no attrition (due to dropout from the study or early patient discharge against medical advice before outcome assessment) in the study population. In univariate analysis (in terms of in-hospital mortality) (Table 2 and Supplementary Table 2), no statistically significant difference was found in demographic or baseline clinical parameters such as the age, gender, comorbidities, hemodynamic variables, or biochemical parameters between the survivors and non-survivors. The median age of the non-survivors was 45 years compared to the 54 years among survivors (P = 0.319). Both duration of hospital stay and fever duration were also comparable between the groups. In terms of infectious etiology, dengue positivity (P = 0.014) showed a significant association with in-hospital mortality, whereas scrub typhus, and typhoid were found to be not significantly related to the mortality outcomes. The renal outcomes at discharge were found to be strongly associated with mortality (P < 0.001), as all the patients who died had either persistent renal failure or incomplete recovery. Other factors, including AKI stage at admission, dialysis requirement, vasopressor support, and the disease severity scores (SOFA and APACHE-II), were not found to be statistically significant predictors of mortality on univariate analysis.

Table 2 Univariate analysis for in-hospital mortality.

Variable	Survivors (n = 60)	Non-survivors (n = 8)	P value
Age (years)	54 (39-61)	45 (26-60)	0.319
Hospital stay (days)	9 (6-13)	6 (4-9)	0.192
FiO2 on admission (%)	32 (21-53)	55 (25-67)	0.099
Hemoglobin on admission (g/dL)	10.9 ± 1.8	9.6 ± 4.1	0.115
Serum creatinine on admission (mg/dL)	1.35 (0.9-2.9)	1.28 (0.6-5.7)	0.165
APACHE-II on admission	15.1 ± 7.4	19.1 ± 10.6	0.176
SOFA score on admission	8.7 ± 3.8	8.5 ± 4.3	0.919
Bleeding	5 (7.4)	3 (37.5)	0.016
NIV use during ICU stay	6 (10.0)	3 (37.5)	0.031
Blood product transfusion	6 (10.0)	4 (50.0)	0.003
Dengue infection	13 (21.7)	5 (62.5)	0.014
AKI stage ≥ 2 on admission	25 (41.7)	4 (50.0)	0.848
Dialysis requirement	16 (26.7)	3 (37.5)	0.521
Renal non-recovery at discharge	2 (3.3)	8 (100)	< 0.001
Invasive ventilation	24 (40.0)	3 (37.5)	0.892
Vasopressor use	26 (43.3)	2 (25.0)	0.322

Presented as mean ± SD, n (%) or median (IQR). Variables with P < 0.10 were considered for multivariable analysis. AKI: Acute kidney injury; APACHE-II: Acute Physiology and Chronic Health Evaluation II; ICU: Intensive care unit; IQR: Interquartile range; NIV: Non-invasive ventilation; SOFA: Sequential organ failure assessment.

Considering the limitation of event-per-variable ratio in this study, an exploratory binary logistic regression model was constructed to identify the independent predictors of in-hospital mortality (Table 3). The overall model was found to be statistically significant (χ² = 21.83, P = 0.016) and it demonstrated good calibration (Hosmer-Lemeshow χ² = 1.59, P = 0.991). The model explained 53.3% of the variance in mortality (Nagelkerke R² = 0.533). Younger age (OR = 0.889, 95%CI: 0.800-0.989; P = 0.030) and higher FiO₂ (OR = 1.095, 95%CI: 1.008-1.190; P = 0.032) requirement on admission were found to be the independent predictors of mortality. Further, shorter duration of fever prior to the ICU admission (OR = 0.764, 95%CI: 0.576-1.014; P = 0.062), and higher APACHE-II score on admission (OR = 1.525, 95%CI: 0.997-2.333; P = 0.052) demonstrated a borderline association and a trend towards the increased mortality. Other variables of the model, including serum creatinine levels (admission and worst values), APACHE-II score at 48 hours, SOFA scores (admission and 48 hours), and dialysis requirement were found to be not independently associated with in-hospital mortality in the adjusted model.

Table 3 Exploratory multivariable logistic regression analyses for in-hospital mortality.

Variable	β (SE)	Odds ratio, exp (β)	95%CI for OR	P value
Age (years)	-0.117 (0.054)	0.889	0.800-0.989	0.030
Fever duration (days)	-0.269 (0.144)	0.764	0.576-1.014	0.062
FiO2 on admission	0.091 (0.042)	1.095	1.008-1.190	0.032
Serum creatinine on admission (mg/dL)	0.982 (0.750)	2.671	0.614-11.619	0.190
Worst serum creatinine (mg/dL)	-0.307 (0.839)	0.736	0.142-3.808	0.714
APACHE-II on admission	0.422 (0.217)	1.525	0.997-2.333	0.052
APACHE-II at 48 hours	-0.395 (0.261)	0.673	0.403-1.124	0.130
SOFA score on admission	-0.877 (0.470)	0.416	0.165-1.046	0.062
SOFA score at 48 hours	0.684 (0.543)	1.981	0.684-5.743	0.208
Dialysis requirement	-0.867 (1.407)	0.42	0.027-6.618	0.538

Model statistics: Omnibus χ² = 21.83, df = 10, P = 0.016; -2 Log likelihood = 27.43; Cox & Snell R² = 0.275; Nagelkerke R² = 0.533; Hosmer-Lemeshow χ² = 1.59, df = 8, P = 0.991. Disclaimer: Binary logistic regression models were used. Odds ratios (OR) < 1 indicate reduced odds of adverse outcome with increasing predictor values. Variables with P < 0.05 were considered statistically significant. Results should be interpreted cautiously given the limited number of outcome events. APACHE-II: Acute Physiology and Chronic Health Evaluation II; SOFA: Sequential organ failure assessment; OR: Odds ratio; CI: Confidence interval.

Another exploratory binary logistic regression model was developed to identify the independent predictors of NNO, defined as either in-hospital mortality or persistent renal failure at discharge (Table 4). The overall model was found to be highly significant (χ² = 38.57, P < 0.001), demonstrated a good calibration (Hosmer-Lemeshow P = 0.690), and explained 60.0% of the variance in outcomes (Nagelkerke R² = 0.600). Younger age (OR = 0.917, 95%CI: 0.849-0.990; P = 0.027), shorter duration of fever (OR = 0.838, 95%CI: 0.728-0.965; P = 0.014) prior to ICU admission, higher FiO₂ requirement on admission (OR = 1.079, 95%CI: 1.025-1.137; P = 0.004), greater illness severity on admission, reflected by higher APACHE-II scores (OR = 1.410, 95%CI: 1.071-1.856; P = 0.014), and the requirement for dialysis (OR = 0.042, 95%CI: 0.006-0.294; P = 0.001) were found to be strongly associated with adverse outcomes. Lower SOFA scores on admission were found to be associated with higher likelihood of NNO (OR = 0.582, 95%CI: 0.361-0.940; P = 0.027). This unexpected direction of association can likely be explained by an overlap with other severity markers in the model and should be cautiously interpreted. Other variables, including serum creatinine levels, APACHE-II and SOFA scores at 48 hours, were found to be not independently associated with NNO after multivariable adjustment.

Table 4 Exploratory multivariable logistic regression analyses for net negative outcome at hospital discharge.

Variable	β (SE)	Odds ratio, exp (β)	95%CI for OR	P value
Age (years)	-0.086 (0.039)	0.917	0.849-0.990	0.027
Fever duration (days)	-0.177 (0.072)	0.838	0.728-0.965	0.014
FiO2 on admission	0.076 (0.026)	1.079	1.025-1.137	0.004
Serum creatinine on admission (mg/dL)	-0.051 (0.441)	0.951	0.401-2.256	0.909
Worst serum creatinine (mg/dL)	0.182 (0.533)	1.2	0.422-3.414	0.733
APACHE-II on admission	0.343 (0.140)	1.41	1.071-1.856	0.014
APACHE-II at 48 hours	-0.222 (0.142)	0.801	0.606-1.059	0.119
SOFA score on admission	-0.540 (0.244)	0.582	0.361-0.940	0.027
SOFA score at 48 hours	0.404 (0.252)	1.497	0.914-2.454	0.109
Dialysis requirement	-3.176 (0.996)	0.042	0.006-0.294	0.001

Model statistics: Omnibus χ² = 38.57, df = 10, P < 0.001; -2 Log likelihood = 48.45; Cox & Snell R² = 0.433; Nagelkerke R² = 0.600; Hosmer-Lemeshow χ² = 5.62, df = 8, P = 0.690. Disclaimer: Binary logistic regression models were used. Odds ratios (OR) < 1 indicate reduced odds of adverse outcome with increasing predictor values. Variables with P < 0.05 were considered statistically significant. Results should be interpreted cautiously given the limited number of outcome events. APACHE-II: Acute Physiology and Chronic Health Evaluation II; SOFA: Sequential organ failure assessment; OR: Odds ratio; CI: Confidence interval.

DISCUSSION

A total of 68 critically-ill adults with TAFI, complicated by AKI, was investigated in this study. Scrub typhus was found to be the most common etiology, followed by dengue and other tropical infections. Most of the patients were presented with early-stage AKI at ICU admission; however, 27.9% required RRT during hospitalization. Both vasopressor and invasive ventilator support were required in case of 41.2% and 39.7% of the patients, respectively. Overall, the in-hospital mortality rate was found to be 11.8%. The results from the univariate analysis revealed that dengue infection, bleeding manifestations, and blood product transfusion were significantly associated with mortality. A composite NNO (mortality or persistent renal dysfunction at discharge) occurred in 33.8% of the patients and was associated with greater respiratory compromise, higher APACHE II scores at admission, and the requirement for dialysis. AKI stage and serum creatinine levels at admission did not predict mortality or NNOs. From the exploratory multivariable analysis, both dialysis requirement and lower arterial oxygen tension at admission were found to be the strongest predictors of adverse outcomes.

In this single-centre observational cohort study of critically-ill adults diagnosed with TAFI, AKI was found to be associated with considerable morbidity, including a high prevalence of persistent renal dysfunction at discharge, despite relatively modest in-hospital mortality. The recent ICU-based studies from the tropical regions similarly demonstrate that AKI, complicating TAFI, is linked to incomplete renal recovery and adverse composite outcomes, underscoring the fact that survival alone underestimates the overall disease burden[9,10].

Scrub typhus was the most common etiological diagnosis found in this study cohort, which reflects the evolving epidemiology of tropical infections in India, where rickettsial diseases are increasingly recognized as significant causes of severe illnesses requiring intensive care[11]. Simple etiological patterns have been reported in other Indian and Southeast Asian ICUs, managing mixed tropical infections[12]. Although scrub typhus accounted for the largest proportion of cases in this study, the dengue infection was found to be disproportionately associated with higher mortality and adverse outcomes. This finding suggests that though scrub typhus heavily burdens the ICU numerically, the dengue-associated AKI may follow a more severe clinical course, once the critical illness develops[13].

The strong association between dengue infection and adverse outcomes is clinically significant. Contemporary cohort studies and systematic reviews indicate that the dengue-associated AKI results from multifactorial mechanisms, including plasma leakage, immune-mediated endothelial injury, shock, rhabdomyolysis, and systemic inflammation, which results in multi-organ dysfunction[13]. In the current study cohort, both bleeding manifestations and blood product transfusions were found to be prevalent among non-survivors, a pattern consistent with literature linking severe dengue, coagulopathy, and hemorrhagic complications to poorer outcomes amongst critically-ill patients[14].

Although most of the patients were presented with early-stage AKI upon ICU admission, nearly one-third required RRT during hospitalization. This pattern highlights the dynamic and progressive nature of renal injury in tropical infections[15]. Importantly, neither the AKI stage at admission nor the absolute serum creatinine level independently predicted the mortality or adverse outcomes. Instead, the need for dialysis emerged as the strongest predictor of poor prognosis. Recent ICU studies consistently demonstrate that the requirement for RRT reflects cumulative illness severity and multi-organ failure rather than isolated renal dysfunction, thereby outperforming the creatinine-based AKI classifications in prognostication[16]. These findings reinforce the recognized limitation of creatinine-based staging in critically-ill patients, particularly in the context of systemic inflammatory and infectious syndromes[17].

The overall in-hospital mortality, observed in this study cohort, was comparable to that of the data reported in recent ICU studies of tropical infection-associated AKI[10,11]. However, when composite outcomes were considered, approximately one-third of the patients experienced a NNO at discharge, emphasizing the fact that AKI-related morbidity extends beyond mortality alone[9]. Persistent renal dysfunction at discharge showed a significant association with adverse outcomes, a pattern consistent with growing evidence that the survivors of AKI are at high risk of progression to chronic kidney disease and other long-term renal sequelae[18,19]. These findings highlight the importance of structured post-discharge renal follow-up in patients, recovering from critical illnesses due to tropical infections.

Respiratory compromise at ICU admission, reflected by lower PaO₂ levels and higher FiO₂ requirements, was found to be consistently associated with mortality and NNOs in this study. Recent literature on severe dengue and other tropical infections identifies early hypoxemia and respiratory failure as the markers of severe systemic disease and the predictors of poor outcomes[20]. This finding supports the concept of organ cross-talk in critical illness, wherein concurrent respiratory and renal dysfunction synergistically amplify the risk of adverse outcomes[21].

Illness severity, assessed based on APACHE II score at admission, independently predicted the adverse outcomes in this study, whereas the SOFA scores did not consistently correlate with mortality. The findings from the recent comparative analyses suggest that APACHE II score may better capture the baseline physiological derangement and an overall disease burden in heterogeneous ICU populations with severe infections, potentially explaining its superior prognostic performance in this cohort[22,23].

From a critical care perspective, the current study findings suggest that risk stratification in TAFI-associated AKI should extend beyond creatinine-based staging to incorporate the markers of global illness severity (APACHE-II), respiratory dysfunction, and the need for organ support[17]. Particular vigilance is warranted in patients with dengue-associated AKI, given its strong association with mortality and poor renal recovery. Early ICU admission, close monitoring, and timely initiation of organ support may help in preventing the progression to irreversible organ dysfunction[24].

Strengths and limitations

This study is a prospective observational cohort design that used real world data for TAFI and AKI in a tertiary care setting. This study evaluated the cohort after clinical- and lab-confirmed diagnoses, which shows a comprehensive inclusion strategy. The study included a clinically important composite outcome in the form of mortality or persistent renal dysfunction at the end of hospitalization to make a complete assessment beyond mortality. However, a single-centre design, restricted ICU population and a modest sample size limit the generalizability of the study. With limited number of events, this study had a very low event-per-variable ratio and so, the results of the regression have to be cautiously interpreted. Long-term renal outcomes beyond hospital discharge were not assessed, which may potentially underestimate the true burden of chronic kidney disease following AKI in tropical infections[18,19]. As an observational study with possibilities for residual confounding, the findings are interpreted as exploratory.

CONCLUSION

In this observational study, AKI with TAFI was found to be associated with considerable morbidity and modest in-hospital mortality. Age, hypoxia on admission and overall disease severity (APACHE-II) on admission were found to be associated with poor mortality outcomes. Creatinine-based AKI staging, as a single approach, has been found to have a limited prognostic utility for both mortality and morbidity outcomes. Amongst all the tropical infections, dengue fever-related AKI has been found to have poor clinical outcomes.

ACKNOWLEDGEMENTS

The authors extend their gratitude to the Department of Microbiology, Central Research Facility (CRF), the Department of Internal Medicine, and the management of MGUMST and MGMC & H, Jaipur, for their cooperation and support in facilitating access to relevant data, thereby enabling the successful completion of this study.