Although acute peritonitis is a common and severe complication associated with peritoneal dialysis, the culture-based test used as the diagnostic criterion for this disease is often too slow to allow appropriate point-of-care diagnosis of specific bacterial infection. To address this problem, Zhang et al. report the efficacy of a novel set of immune biomarkers derived from a machine-learning algorithm applied to patient data. This fingerprint could predict major pathogenic causes of peritonitis.
Copyright © 2017 International Society of Nephrology. Published by Elsevier Inc. All rights reserved.