BACKGROUND: The hypothesis that gastric intramural pH (pHi) is predictive of outcome in haemodynamically stable, mechanically ventilated patients was tested in 25 patients on assisted mechanical ventilation for respiratory failure. METHODS: Simultaneous samples of arterial blood and gastric juice were obtained from patients on assist control, synchronised intermittent and pressure control ventilation during the first 48 hours of mechanical ventilation. Gastric pHi was calculated from the equation: pHi= 6.1 + log HCO3/(gastric PCO2 X 0.03). The outcome was survival or death due to respiratory or circulatory failure within 45 days of admission. RESULTS: Gastric pHi proved to be a better predictor of outcome than all presently utilised parameters. Although all patients included in this study were haemodynamically stable and were similar for all laboratory indices, the only variable capable of accurately predicting outcome was gastric pHi. Patients with a normal arterial pH but a gastric intramural pH of less than 7.25 had an observed mortality of 66%. Standard severity of illness scores grossly underestimated mortality rates. The sensitivity and specificity of a gastric pHi value of less than 7.25 in predicting death were 86% and 83%, respectively. A receiver operator curve for all variables exaggerates the superiority of gastric pHi as a predictor of outcome. CONCLUSION: Low gastric pHi, a marker of gastrointestinal ischaemia, may occur in the presence of normal haemodynamics and may be used to predict severity of illness and mortality accurately.
BACKGROUND: The hypothesis that gastric intramural pH (pHi) is predictive of outcome in haemodynamically stable, mechanically ventilated patients was tested in 25 patients on assisted mechanical ventilation for respiratory failure. METHODS: Simultaneous samples of arterial blood and gastric juice were obtained from patients on assist control, synchronised intermittent and pressure control ventilation during the first 48 hours of mechanical ventilation. Gastric pHi was calculated from the equation: pHi= 6.1 + log HCO3/(gastric PCO2 X 0.03). The outcome was survival or death due to respiratory or circulatory failure within 45 days of admission. RESULTS: Gastric pHi proved to be a better predictor of outcome than all presently utilised parameters. Although all patients included in this study were haemodynamically stable and were similar for all laboratory indices, the only variable capable of accurately predicting outcome was gastric pHi. Patients with a normal arterial pH but a gastric intramural pH of less than 7.25 had an observed mortality of 66%. Standard severity of illness scores grossly underestimated mortality rates. The sensitivity and specificity of a gastric pHi value of less than 7.25 in predicting death were 86% and 83%, respectively. A receiver operator curve for all variables exaggerates the superiority of gastric pHi as a predictor of outcome. CONCLUSION: Low gastric pHi, a marker of gastrointestinal ischaemia, may occur in the presence of normal haemodynamics and may be used to predict severity of illness and mortality accurately.