• Katsunori Mochizuki, Tomoko Fujii, Eldho Paul, Matthew Anstey, Shigehiko Uchino, David V Pilcher, and Rinaldo Bellomo.
• Australian and New Zealand Intensive Care Research Centre, Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia; Department of Emergency and Critical Care Medicine, Shinshu University School of Medicine, Matsumoto, Nagano, Japan. Electronic address: katsunori.mochizuki@monash.edu.
• J Crit Care. 2021 Aug 1; 64: 10-17.

PurposeTo study the prevalence, characteristic, outcome, and acid-base biomarker predictors of outcome for different acidemia subtypes.MethodsWe used national intensive care databases from three countries and classified acidemia subtypes as metabolic (standard base excess [SBE] < -2 mEq/L only), respiratory (PaCO2 > 42 mmHg only), and combined (both SBE < -2 mEq/L and PaCO2 > 42 mmHg) based on blood gas analysis in the first 24 h after ICU admission. To investigate acid-base predictors for hospital mortality, we applied the area under the receiver operating characteristic curve approach.ResultsWe screened 643,689 ICU patients (2014-2018) and detected acidemia in 57.8%. The most common subtype was metabolic (42.9%), followed by combined (30.3%) and respiratory (25.9%). Combined acidemia had a mortality of 12.7%, compared with 11% for metabolic and 5.5% for respiratory. For combined acidemia, the best predictor of hospital mortality was pH. However, for metabolic or respiratory acidemia, it was SBE or PaCO2, respectively.ConclusionsIn ICU patients with acidemia, mortality differs according to subtype and is highest in the combined subtype. Best acid-base predictors of mortality also differ according to subtype with best performance for pH in combined, SBE in metabolic, and PaCO2 in respiratory acidemia.Copyright © 2021 Elsevier Inc. All rights reserved.

13 keywords...

hide…