Objective: To study the value of gastric intramucosal pH and gastric intraluminal PCO2 measurements to predict weaning outcome from mechanical ventilation.
Design: Prospective clinical study.
Setting: Intensive care medicine department of a university hospital.
Patients: Nineteen adult critically ill patients who were mechanically ventilated because of acute respiratory failure and were considered ready to be weaned.
Interventions: The patients were weaned with: synchronized intermittent mandatory ventilation plus positive end-expiratory pressure (SIMV+PEEP) or continuous positive airway pressure with pressure support ventilation (CPAP+PSV). A gastric tonometer was placed in all the patients. Tonometric, respiratory, and hemodynamic variables were measured during the weaning process.
Measurements: Hemodynamic variables, respiratory mechanics, pulmonary gas exchange, respiratory muscle force, spontaneous pattern of breathing, and the central control of breathing were recorded. Simultaneously, the intramucosal pH and gastric intraluminal PCO2 were measured.
Main results: Eleven patients were successfully extubated and eight failed. The patients who failed showed higher values of mouth occlusion pressure, respiratory rate, and effective inspiratory impedance (mouth occlusion pressure/mean inspiratory flow). The intramucosal pH was initially 7.19 +/- 0.22 and decreased to 7.10 +/- 0.16 during the weaning process in patients who failed (p < .05). At the same time, the intramucosal pH showed a nonsignificant change from 7.36 +/- 0.07 to 7.32 +/- 0.07 in the patients who were successfully extubated. The intramucosal pH was statistically different when both groups were compared during the initial and the final evaluations (p < .05). For the initial evaluation, the sensitivity and specificity to predict weaning failure when the intramucosal pH was < or =7.30 were 0.88 (95% confidence interval [CI], 0.66-1) and 0.82 (95% CI, 0.59-1), respectively. The gastric intraluminal PCO2 was higher in patients who failed (p < .05). When gastric intraluminal PCO2 was . or =40 torr during the initial evaluation, weaning failure occurred with a sensitivity of 1 (95% CI, 0.31-1) and a specificity of 0.55 (95% CI, 0.26-0.84).
Conclusions: Weaning failure was associated with gastric intramucosal acidosis. The intramucosal pH and gastric intraluminal PCO2 may be helpful to predict weaning outcome. Further controlled clinical trials in a larger group of patients are needed.