Development of the Hemodynamic Instability Index in Acute Kidney Failure and its association with in-hospital mortality

pp. 340-347

Abstract

Background: Hemodynamic instability increases the risk of in-hospital mortality in patients with acute kidney injury (AKI), but there is no specific tool to quantify this risk.

Objective: The aim of the present study was to develop the Hemodynamic Instability Index in Acute Kidney Injury (IIH-AKI), analyze its association with in-hospital mortality in patients hospitalized for AKI and compare its discriminatory ability with other established prognostic scores.

Methods: We conducted an analytical study based on a secondary database derived from a clinical record of 5060 patients hospitalized with AKI. The outcome analyzed was in-hospital mortality. Principal component analysis (PCA) was used to develop the HII-AKI model based on five key parameters: pulse, systolic blood pressure, diastolic blood pressure, respiratory rate, and oxygen saturation. The HII-AKI performance was evaluated using the area under the ROC curve (AUC-ROC), Kaplan-Meier curves, and Cox regression analysis.

Results: The HII-AKI prsented an AUC-ROC of 0.742 (95% CI 0.722-0.762; p<0.001) for predicting in-hospital mortality, surpassing the SOFA score (AUC-ROC=0.723) and the Elixhauser comorbidity index (AUC-ROC=0.465). Patients with high HII-AKI were younger and had a longer hospital stay. They also had more acidosis, lower bicarbonate levels, higher urea nitrogen levels, and lower creatinine levels. In Cox regression analysis, a high HII-AKI was associated with higher in-hospital mortality (HR=2.394; 95% CI: 2.008-2.855; p<0.001).

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 of mortality in AKI. Its implementation in clinical practice could improve risk stratification and optimize the therapeutic decisions. Further studies are necessary for external validation.