INTRODUCTION

Acute kidney injury (AKI) is the sudden loss of renal function that causes retention of nitrogenous wastes and alterations in fluid, electrolyte, and acid-base balance. It can be differentiated into prerenal, intrarenal, and postrenal etiologies, which affect patients’ outcomes. (1) Acute kidney injury is a predictor of poor prognosis in hospitalized patients, especially those in critical condition, where its association with hemodynamic instability increases the risk of multiple organ dysfunction and death. (2)

Hemodynamic instability in AKI is related to a decrease in effective intravascular volume, endothelial dysfunction, release of mediators of inflammation, and neurohormonal activation. These factors lead to tissue hypoperfusion, which perpetuates renal damage and affects other vital organs, producing a state of hemodynamic dysfunction condition is characterized by hypotension, tachycardia, and alterations in systemic vascular resistance. (3)

There are various tools for assessing the prognosis and mortality of hospitalized patients, such as the SOFA (Sequential Organ Failure Assessment) score and the Elixhauser Comorbidity Index. The SOFA score is a widely used metric in intensive care units to predict mortality based on multiple organ dysfunction. Its calculation is complex and requires the combination of numerous clinical and laboratory parameters. (4) Conversely, the Elixhauser Comorbidity Index enables the stratification of mortality risk based on patients’ comorbidities. (5) However, these systems have not been specifically designed to assess hemodynamic instability in patients with AKI, which may limit their applicability in this population.

The assessment of hemodynamic instability in patients with AKI is essential to guide therapeutic interventions and stratify the risk of complications. However, there is currently no standardized index that quantifies hemodynamic instability and its association with in-hospital mortality, allowing for an objective assessment of the hemodynamic status of these patients.

The aim of the present study was to develop the Hemodynamic Instability Index in Acute Kidney Injury (HII-AKI) and analyze its association with in-hospital mortality in patients hospitalized for AKI. This novel index could allow for the early identification of patients at higher risk of death using less extensive clinical data than, for example, the SOFA score. Consequently, this will facilitate the optimization of clinical decisions and management strategies for this vulnerable population.

METHODS

RESULTS

A total of 5060 patients hospitalized for AKI were analyzed; mean age was 68.25 ± 15.50 years and the distribution by sex was balanced (51.5% were men and 48.50% were women). The patients exhibited the following metabolic and hemodynamic abnormalities: anion gap 12.01 ± 3.87 mEq/L, serum bicarbonate 23.42 ± 4.92 mEq/L, BUN 33.24 ± 19.39 mg/dL, GFR 63.49 ± 30.21 mL/min/1.73 m², and creatinine levels on admission 1.34 ± 0.72 mg/dL. Mean hemoglobin levels were 10.77 ± 2.43 g/dL, platelet count 227.33 ± 115.40 ×10³/μL, and WBC 11.32 ± 17.23 ×10³/μL. Mean systolic and diastolic blood pressure were 120.57 ± 22.25 mmHg and 67.74 ± 13.04 mmHg, respectively. The comorbidities recorded included chronic renal disease (25.4%), heart failure (31.8%), lung disease (30.4%), diabetes mellitus (37.5%), and hypertension (67.9%). The median SOFA score on admission was 2 points (IQR: 1-3) and the median length of hospital stay was 7 days (IQR: 4-13), reflecting a population with a high burden of comorbidities and physiological instability (Table 1)

Principal component analysis (PCA) identified seven pathophysiological components with eigenvalues > 1, which explained 62% of the total variance, with acceptable sample adequacy (KMO = 0.591; Bartlett p < 0.001). The first component encompassed renal function markers (creatinine, BUN, and GFR). The second component captured respiratory and acid-base parameters (respiratory rate, oxygen saturation, and bicarbonate). The third component represented hemodynamic measures (systolic and diastolic blood pressure, and pulse pressure). The remaining components reflected metabolic and hematological axes. Based on the factor loadings of the hemodynamic axis, the five variables with the greatest direct contribution to cardiovascular instability were selected—respiratory rate, oxygen saturation, pulse, systolic blood pressure, and diastolic blood pressure—and used to construct the Hemodynamic Instability Index in Acute Kidney Injury (HII-AKI) (Table 2)

PCA identified five key variables of hemodynamic instability in acute kidney injury: respiratory rate, oxygen saturation, pulse, systolic blood pressure, and diastolic blood pressure. These variables were weighted according to their factor loadings to construct the Acute Renal Failure Hemodynamic Instability Index (AKI-HII) formula. Pulse and respiratory rate had a positive contribution, while oxygen saturation and systolic and diastolic blood pressures had a negative influence. Before adjustment, HII-AKI values ranged from -129.21 to 22.49 (mean -60.81, standard deviation 19.22). To avoid negative values and facilitate clinical interpretation, 129.21 was added to all values. This adjustment resulted in an adjusted index with a range of 0 to 151.70, maintaining the same standard deviation, and with an adjusted mean of 68.39 (Table 3).

The HII-AKI demonstrated effective performance in discriminating in-hospital mortality, with an AUCROC of 0.742 (95% CI 0.722 - 0.762; p < 0.001), surpassing the performance of other scales and biomarkers evaluated. The SOFA score exhibited an AUC-ROC of 0.723 (95% CI 0.702-0.744; p < 0.001), with a similar though slightly lower performance while the AUCROC of the Elixhauser Comorbidity Index was 0.465 (95% CI 0.442-0.489; p = 0.003), suggesting a limited discriminatory ability. Other clinical parameters, such as anion gap (AUC-ROC = 0.658), WBC count (AUCROC = 0.672), and age (AUC-ROC = 0.527) showed moderate to low performance. Oxygen saturation exhibited the lowest performance, with an AUC-ROC of 0.348 (95% CI 0.323-0.374; p < 0.001), indicating an inverse association with in-hospital mortality.

Patients with elevated HII-AKI (≥72.20) had significant differences in multiple clinical parameters. They were younger, and median length of hospital stay was longer [10 (IQR 5-19) days vs. 6 (IQR 4-12) days; p < 0.001]. These patients demonstrated elevated anion gap levels, reduced bicarbonate levels, and elevated BUN, findings that reflect greater severity of metabolic acidosis. In addition, hemoglobin concentration and blood pressure were significantly lower. Heart rate and respiratory rate were also higher, as well as white blood cell count. Oxygen saturation was reduced, reflecting significant cardiovascular and respiratory instability (Table 4)

Kaplan-Meier survival analysis showed significant differences in survival between patients with high HII-AKI (≥72.2) and those with a lower index. Median survival was significantly lower in the group with high HII-AKI (46.36 vs. 81.56 days). The log rank (MantelCox) test demostrated a statistically significant difference between the two groups (p < 0.001). In the Cox regression analysis, HII-AKI emerged as the strongest predictor of in-hospital mortality, with a hazard ratio (HR) of 2.394 (95% CI 2.008 - 2.855). Compared to other severity scales, the SOFA score also showed a significant association with mortality (HR = 1.559; 95% CI 1.308 - 1.858), although the impact was lower than that of the HII-AKI. The Elixhauser Comorbidity Index did not demonstrate a significant association with mortality (Table 5).

DISCUSSION

This study employed principal component analysis (PCA) to identify the pathophysiological axes of AKI, with seven components explaining 62% of the total variance. These components reflect the dynamic interaction between renal function, acid-base balance, electrolytes, hemodynamics, metabolism, and age. The results, supported by an acceptable sample adequacy (KMO = 0.591) and a significant Bartlett’s test (p < 0.001), showed that the first component grouped renal variables, creatinine levels, and BUN as the primary determinants. This finding is consistent with previous evidence that positions kidney dysfunction and the retention of nitrogenous wastes as independent predictors of hospital mortality, reflecting reduced glomerular filtration that contributes to systemic toxicity. (10,11) Furthermore, components associated with respiratory function and acid-base status were identified. These components influence tissue oxygenation and the risk of multiple organ dysfunction. (12,13)

Based on this factor structure, five key variables associated with hemodynamic instability were identified: respiratory rate, oxygen saturation, pulse, SBP and DBP, which were used to construct the HII-AKI. The reduction in SBP and DBP, along with the compensatory increase in heart rate and respiratory rate, reflects a physiological response to hypoperfusion and circulatory compromise characteristic of AKI. In this context, the HII-AKI demonstrated a strong discriminatory ability to predict in-hospital mortality (AUCROC = 0.742), surpassing general scales such as the SOFA score or the Elixhauser Comorbidity Index, as well as isolated parameters such as the anion gap level or white blood cell count.

Compared to the SOFA score, the HII-AKI offers significant advantages, as it is based on basic, noninvasive, and easily reproducible hemodynamic monitoring parameters, making it particularly useful in resource-limited settings. While the SOFA score assesses multiple physiological systems, the IHII-AKI focuses on cardiovascular instability, a prognostic axis that is often underestimated for AKI progression. (14,15) This specificity explains its greater discriminatory power, suggesting that it could become a practical tool for risk stratification and therapeutic prioritization within general and critical care units.

Patients with elevated IHII-AKI (≥ 72.2) presented greater metabolic and inflammatory instability, elevated anion gap level, lower bicarbonate, higher BUN, and leukocytosis, along with clear signs of cardiovascular impairment: hypotension, tachycardia, tachypnea, and hypoxemia. They also had a longer length of hospital stay and a lower mean age, which could reflect more aggressive forms of hemodynamic dysfunction in young patients. These results are consistent with studies that have associated early circulatory instability with adverse outcomes and mortality in AKI. (16-18) By integrating these objective variables, the HII-AKI quantitatively synthesizes the impact of hemodynamic imbalance on prognosis, providing a clinical tool that can be immediately applied and validated in different healthcare settings.

Likewise, Kaplan-Meier survival analysis showed that patients with elevated HII-AKI had a lower probability of in-hospital survival, a difference that was statistically significant. In accordance with these findings, the Cox proportional hazards model revealed that a high HII-AKI was associated with an elevated risk of in-hospital mortality, even after adjusting for clinical variables and prognostic scales such as the SOFA score, the Elixhauser Comorbidity Index, and GFR, confirming the independent value of the tool as a prognostic predictor. This finding reinforces the role of the HII-AKI as a prognostic marker and suggests that acute cardiovascular impairment plays a central role in the progression and mortality associated with AKI, consistent with the mechanisms of systemic hypoperfusion and dysregulated inflammation described in recent literature. (19,20)

The HII-AKI could have relevant implications for the clinical management of patients with acute kidney injury. By integrating simple hemodynamic parameters, this index could contribute to estimating the risk of in-hospital mortality and the early identification of patients with a higher probability of adverse outcomes. It could also facilitate more accurate risk stratification and guide monitoring or interventions according to individual clinical profiles. The use of this tool in intensive care units could help prioritize the allocation of resources to cases requiring close monitoring or adjustments in hemodynamic support. In addition, its prospective use could facilitate the evaluation of the clinical course and therapeutic response in realtime settings. Finally, the HII-AKI has the potential to serve as a valuable research instrument in examining the association between hemodynamic instability and clinical outcomes in acute kidney injury. However, external validation is necessary before its implementation in routine clinical practice.

Among the limitations of this study, despite the amount of data available, the retrospective analysis of a hospital cohort could restrict the generalizability of the findings to other settings. Additionally, while PCA enabled the identification of key pathophysiological axes, the selection of variables and the methodology may influence the interpretation of results. Further research is necessary to externally validate the HIIAKI and explore its usefulness for the early management of AKI.

In conclusion, a high HII-AKI is associated with greater hemodynamic instability, inflammation, metabolic disturbances, and prolonged length of hospital stay, supporting its usefulness as a prognostic marker. This index represents an innovative tool for assessing hemodynamic instability in AKI, with a discriminatory ability superior to other scores and biomarkers. Its implementation in clinical practice could improve risk stratification and optimize the therapeutic management of this vulnerable population. It is imperative to conduct further research in diverse cohorts to evaluate the generalizability of these findings.

Conflicts of interest

None declared.

(See authors' conflict of interests forms on the web).