MO313: Haematological Ratios as Risk Factor for Acute Kidney Injury in Patients Suspected of an Infection at the Emergency Department

Abstract

Abstract BACKGROUND AND AIMS Evidence exists that inflammation plays an important role in the pathogenesis of acute kidney injury (AKI) in sepsis patients. Haematological ratios can be used to measure systemic and silent inflammation. Therefore, we investigated what haematological ratios could be identified as risk factors to predict AKI < 30 days in patients suspected of infection at the emergency department (ED). METHOD Data from the SPACE-cohort (SePsis in the Acutely ill patients in the ED) were used. This cohort consists of all patients ≥ 18 years, that present at the ED of the UMC Utrecht for the internal medicine with suspicion of infection. AKI was defined based on the KDIGO criteria. Haematological markers were determined with the Abbott Cell-DYN Sapphire haematology analyser. In total, 7 haematological ratios were included. A Cox regression was performed after the ratios were divided into tertiles. Consequently, Kaplan–Meier curves for 30-day mortality were constructed for the significant predictors from the Cox analysis. RESULTS In univariate Cox regression, neutrophil-to-lymphocyte ratio (NLR), segmented neutrophil-to-monocyte ratio (SeMo) and neutrophil-to-lymphocyte-and-platelet ratio (NLPR) were all significant risk factors for the occurrence of AKI < 30 days after ED presentation (Fig. 1). After correction for multiple confounders, NLR, SeMo and NLPR remained significant. There were significant differences in survival for all these three ratios (Fig. 2). CONCLUSION In this study, we show that several haematological ratios are risk factors for AKI and mortality in patients suspected of infection at the ED. Using these risk factors in prediction models might enable ED physicians to treat AKI in an early phase and prevent any short- and long-term complications. A limitation of this study is the retrospective character. Because of this, not all patients in the cohort had follow-up creatinine levels, and the occurrence of AKI could not be determined in these patients.