NephroPOC - Risk assessment and prediction of acute kidney injury in emergency patients with suspected organ dysfunction: Secondary analysis from the prospective observational LifePOC study.

Abstract

We aimed to assess markers and risk factors for imminent acute kidney injury (AKI) in emergency patients, as risk stratification in the emergency department is currently not widely used. Using data from a sub-cohort (440 patients) of the prospective multicentre LifePOC study (1434 patients), proenkephalin A 119-159 (penKid) was assessed for early identification of subclinical kidney damage compared to serum creatinine in emergency patients with a qSOFA score ≥1. Logistic regression was applied to assess the usefulness of penKid, four further biomarkers (midregional pro-adrenomedullin, bioactive adrenomedullin, dipeptidyl-peptidase-3, procalcitonin) and clinical risk factors to predict AKI within 24 h, 48 h and 72 h after admission, need for organ support and 28-day mortality. PenKid and bio-adrenomedullin performed moderately to predict AKI within 48 h (AUC 0.645, 95% CI: 0.582-0.703 and AUC 0.647, 95% CI: 0.583-0.707, respectively). Pre-existing chronic kidney disease (OR 2.36, 95% CI: 1.06-5.27), confirmed sepsis (OR 2.41, 95% CI: 1.28-4.56), mechanical ventilation (OR 3.03, 95% CI: 1.48-6.19), and elevated levels of penKid (OR 2.21, 95% CI: 1.60-3.07) at admission were associated with an increased risk of AKI whereas a restrictive fluid management (OR 0.43, 95% CI: 0.26-0.71) was associated with a lower risk of AKI. Patients at high AKI risk may be identified based on specific risk factors, bio-ADM and penKid. The trial was registered in the German Registry for Clinical Trials (DRKS00011188) on 20 October 2016.