Early prediction of outcome of respiratory failure. Comparison of high-frequency jet ventilation and volume-cycled ventilation.

Data from a prospective randomized investigation comparing volume-cycled ventilation and high-frequency jet ventilation were reexamined to determine whether improvement of respiratory and hemodynamic function, as well as ultimate outcome (death or survival), could be predicted early in the course of the disease. End points were selected for the ratio of the arterial oxygen pressure over the fractional concentration of oxygen in the inspired gas (PaO2/FIO2), the arterial oxygen saturation (SaO2), the arterial carbon dioxide tension (PaCO2), and the cardiac index. Patients were assigned to "success" or "failure" groups, according to the values recorded for those end points 24 hours after institution of mechanical ventilation. Values obtained from initiation of mechanical ventilation to 16 hours later were divided into four time groups. Differences between patients who "succeeded" and "failed" were compared at each time interval. Ultimate outcome was also compared. The PaCO2 and cardiac index were poor predictors of survival. Early values did not foretell the progression of these variables. The PaO2/FIO2 and SaO2 effectively discriminated, at all time intervals, between patients who succeeded and failed on volume-cycled ventilation. On high-frequency jet ventilation, significant differences were evident only after eight hours of support. With both types of ventilator, patients who reached the end point of oxygenation at 24 hours survived in far greater numbers than those who did not. On the basis of this investigation, it appears justified to attempt high-frequency jet ventilation in patients who do not rapidly improve on volume-cycled ventilation. Institution of high-frequency jet ventilation as the initial support method may not be advisable, since failure does not become apparent for many hours.