—To investigate the importance of splanchnic ischemia in patients with acute circulatory failure by comparing gastric intramucosal pH as measured by tonometry with conventional methods of assessing adequacy of tissue oxygenation.
—Prospective cohort of patients with acute circulatory failure in first 24 hours after admission to the intensive care unit.
—Two general intensive care units in London, England.
—Consecutive sample of 83 patients of varying diagnostic categories that required pulmonary artery catheterization.
Main Outcome Measures.
—Gastric intramucosal pH and hemodynamic, oxygen transport, and metabolic variables were measured on admission and at 12 hours and 24 hours after admission. Prediction of outcome (death or survival) by each measurement was assessed by sensitivity, specificity, and logistic regression analysis.
—Mean 24-hour Acute Physiology and Chronic Health Evaluation (APACHE II) score was 20.3. There were significant differences in mean gastric intramucosal pH between survivors and nonsurvivors on admission and at 24 hours, (7.40 vs 7.28, 7.40 vs 7.24, respectively; P<.001). Admission heart rate was higher (116 vs 101 beats per minute; P<.003) and mean arterial pressure lower (82 vs 97 mm Hg; P<.01) in nonsurvivors. There were no consistent differences in cardiac index, oxygen delivery, and oxygen uptake between survivors and nonsurvivors. Admission arterial pH was significantly lower (7.3 vs 7.36; P<.003), base excess more negative (-5.3 vs -1.9; P<.001), and lactate concentration higher (3.14 vs 1.91 mmol/L; P<.03) in nonsurvivors. Gastric intramucosal pH had a sensitivity of 88% for predicting death and a likelihood ratio of 2.32, higher than for any other variable. Only gastric intramucosal pH at 24 hours independently predicted outcome.
—Gastric intramucosal pH was the most reliable indicator of adequacy of tissue oxygenation in this group of patients. Inadequate regional blood flow as detected by a reduction in gastric intramucosal pH, but not by systemic measures, is an important contributor to morbidity and mortality in intensive care units.(JAMA. 1993;270:1203-1210)