Back
Timely identification of acute kidney injury (AKI), whether occurring in the hospital or in the community, is key to improving its management and long-term outcomes. The current KDIGO definition of AKI relies on the rise in serum creatinine (sCr) and fails to identify patients with community-acquired AKI for whom hospitalisation coincides with recovery of AKI. The extended KDIGO (eKDIGO) definition which incorporates the fall in sCr may be better suited to capturing the full temporal spectrum of AKI.
We conducted a retrospective, observational study using all admissions in the MIMIC-IV database from 2008 to 2019. Incidence, staging and timing of AKI were evaluated using both traditional and eKDIGO definitions. Patients diagnosed with AKI by a decrease in sCr were labelled as deKDIGO. Clinical characteristics and outcomes were compared for each group of patients.
Overall, 254,303 admissions were analysed. Incidence of AKI was 22.5% using KDIGO and 36.2% using eKDIGO (Figure 1). Those in the eKDIGO group had a greater proportion of stage 1 AKI (66% vs 60% in KDIGO patients) with the deKDIGO peak AKI cases occurring predominantly on days 2 and 3 of admission (Figure 2). Compared to those without AKI, patients in the eKDIGO group had worse kidney function on admission, higher rates of ICU admission (40% vs 17%), invasive mechanical ventilation (18% vs 14%) and in-hospital death (6% vs 1%).
Conclusions
The eKDIGO definition of AKI resulted in a significantly higher detection rate in this population. These additional cases appear to be community-acquired and, while milder in severity than KDIGO cases, have worse outcomes than patients without AKI.