Development of a respiratory dysfunction Index to predict early mortality in acute kidney injury patients

Introduction: Acute kidney injury can lead to respiratory dysfunction, yet no specific index currently exists to assess it comprehensively. Objective: To develop and evaluate the performance of the Respiratory Dysfunction Index in Acute Kidney Injury for predicting short-term mortality. Methods: A retrospective cohort study using secondary data from 5,060 hospitalized Acute Kidney Injury (AKI) patients. Clinical and biochemical variables were analyzed, including respiratory rate, oxygen saturation, creatinine, anion gap, bicarbonate, hemoglobin, blood pressure, and hemodynamic parameters. Principal component analysis (PCA) was applied, with model adequacy assessed using the Kaiser-Meyer-Olkin index and Bartlett’s test of sphericity. Based on PCA, the Respiratory Dysfunction Index in AKI was constructed and evaluated using the area under the curve (AUC), Cox regression for 14-day mortality, and Kaplan-Meier survival analysis. Results: PCA showed moderate adequacy (Kaiser-Meyer-Olkin = 0.522; p< 0.001) and identified seven components. The Respiratory Dysfunction Index in AKI was developed based on respiratory rate and oxygen saturation (AUC = 0.735; 95% CI: 0.710–0.760) to predict 14-day mortality. In Kaplan-Meier analysis, patients with a high index had lower survival (161.5 vs. 275.9 days; p< 0.001). In Cox regression, a high index was associated with increased mortality risk (HR = 2.323; 95% CI: 1.895–2.848; p< 0.001). Conclusions: The Respiratory Dysfunction Index in Acute Kidney Injury proved to be a useful tool for predicting mortality in patients with AKI.