Metabolic Acidosis Assessment in High-Risk Surgeries: Prognostic Importance.

ABSTRACT

BACKGROUND: Metabolic acidosis frequently is present in surgical patients; however, different types of metabolic acidosis (hyperlactatemia, hyperchloremia, and others) may have different relationships to perioperative outcomes. We hypothesized that in postoperative surgical patients, distinctive types of metabolic acidosis would correlate differently with the outcomes of high-risk surgeries. METHODS: A prospective, multicenter observational study was performed in 3 different tertiary care hospitals. Patients who required postoperative admission to the intensive care unit (ICU) were included in this study. Patients with a short life expectancy (those with untreated cancer and limited treatment), hepatic failure, renal failure, or a diagnosis of diabetes were excluded. Patients were classified at ICU admission according to the presence and type of metabolic acidosis into 4 groups those without acidosis, those with a base excess <-4 mmol/L and albumin-corrected anion gap ≤12 mmol/L (hyperchloremic), those with a base excess <-4 mmol/L and increased albumin-corrected anion gap >12 mmol/L, and those with a base excess <-4 mmol/L and hyperlactatemia >2 mmol/L. Furthermore, patients were reclassified 12 hours after admission to the ICU to verify the metabolic acidosis behavior and outcome differences among the groups. RESULTS: The study included 618 patients. The incidence of acidosis at ICU admission was 59.1%; 23.9% presented with hyperchloremia, 21.3% with hyperlactatemia, 13.9% with increased anion gap, and 40.9% of the patients presented without metabolic acidosis. Patients whose metabolic acidosis persisted for 12 hours had an incidence of ICU complications rates in hyperlactatemia group of 68.8%, increased anion gap of 68.6%, hyperchloremic of 65.8%, and those without acidosis over 12 hours of 59.3%. A Cox regression model for postoperative 30-day mortality showed in hyperlactatemic acidosis, hazard ratio (HR) = 1.74, 95% confidence interval (CI) = 1.02-2.96; increased anion gap acidosis, HR = 1.68, 95% CI = 0.85-3.81; hyperchloremic acidosis, HR = 1.47, 95% CI = 0.75-2.89, and 10.3% of 30-day mortality rate in patients without acidosis. An adjusted survival curve by Cox regression found a worse 30-day survival in the hyperlactatemic group compared with the other groups (P = .03). Furthermore, in multiple comparisons among groups, patients with hyperlactatemic acidosis were more likely to develop renal dysfunction (P < .001) up to the seventh day postoperatively. CONCLUSIONS: We found that among patients with different types of acidosis, patients who developed hyperlactatemic metabolic acidosis postoperatively showed greater rates of renal dysfunction within 7 days and hyperlactatemic acidosis represented an independent factor on 30-day mortality in high-risk surgical patients.