Critically ill patients with metabolic acidosis are twice as likely to die as patients who do not have metabolic acidosis. A study published today in the journal Critical Care shows that the mortality rate among patients with metabolic acidosis is highest for patients with lactic acidosis -- abnormally acidic blood pH due to an excess of lactate - but it is also considerable for patients with strong ion gap (SIG) acidosis -- acidosis due to an excess of unidentified ions. Both lactate and SIG should be carefully monitored in intensive care units (ICUs).
In the largest study of its kind, Kyle Gunnerson from Virginia Commonwealth University, Virginia, USA and colleagues from the University of Pittsburgh, Pennsylvania, USA, selected patients admitted to the University of Pittsburgh Medical Center ICUs who were suspected of having metabolic acidosis. Out of 851 patients screened, 548 patients had metabolic acidosis and 303 patients did not. The patients who did not were used as the control, non-acidosis group.
Among patients with metabolic acidosis, 44% of patients had lactic acidosis, 37% of patients had SIG acidosis and 19% had hyperchloremic acidosis (acidosis not associated with either lactate or SIG). The mortality rate was 45% for patients with metabolic acidosis compared with 26% for patients without. Patients with lactic acidosis had a mortality rate of 56% and for patients with SIG acidosis, the mortality rate was 39%. Patients with hyperchloremic acidosis had a mortality rate similar to the rate in the non-acidosis group: 29%.
A statistical analysis confirmed that both lactate and SIG were strong independent predictors of hospital mortality. The authors recommend that both should be monitored in critically ill patients.
Critical Care Manuscript: C50806P
Lactate vs. Non-Lactate Metabolic Acidosis: A Retrospective Outcome Evaluation of Critically Ill Patients
Kyle J Gunnerson, Melissa Saul, Shui He and John A Kellum
Critical Care 2006, 10:R22 (10 February 2006) doi:10.1186/cc3987
After the embargo, article available at: http://ccforum.com/content/10/1/R22
Cite This Page: