Group A (Mechanical Ventilation <72 h): 2-H Wean
IMV: The IMV rate was progressively decreased over 2 h. This consisted of 30 minutes of an IMV rate of 6, 30 minutes of an IMV rate of 4, and 1 h of an IMV rate of 0. Blood gas levels were measured 30 minutes after each ventilator change, and the patient was changed to the next step in the protocol if clinically stable while blood gas results were pending.
T-Piece: A 30-minute T-piece trial was followed by 30 minutes on the original ventilator settings, and then with a 1-h T-piece trial (total, 2 h). At the end of each T-piece trial, a blood gas analysis was obtained.
Group B (Mechanical Ventilation > 72 h): 7-H Wean
IMV: The IMV rate was progressively decreased over 7 h, the rate being decreased every 2 h from a rate of 8 to 6 to 4; this was followed by 1 h at an IMV of 0. Blood gas values were measured 30 minutes after IMV rate changes were instituted.
T-Piece: A T-piece trial of 30 minutes q 2 h x 3 with blood gas values obtained after the first and third T-piece trial was followed by a 1-h T-piece trial with a blood gas analyses obtained at the end of the trial (total, 7 h). other
Group C (Never Satisfied Weaning Criteria After Seven Days or Failed Three Attempts at Weaning [2- and 7-h Wean]): 3-Day Wean
IMV: The IMV rate was decreased each morning; on day one the rate was set at 8, day 2 the rate was decreased to 6, and on day 3 the rate was set at 4. A blood gas analysis was obtained 30 minutes after each ventilator change was instituted. At the beginning of day 4, the ventilator rate was decreased to 0 for 1 h and blood gas values obtained.
T-Piece: (3-Day progressive T-piece trial): On day 1 there was a 15-minute T-piece trial q 4 h x 2 followed by a 30 minute T-piece trial q 4 h X 2; for the remaining 8 h the patient was maintained on the original ventilator settings. On day 2 there was a 15-minute T-piece trial q 2 hx4 for the first 8 h, then T-piece trials of 30 minutes. This was again followed by an 8-hour rest period. On day 3, there was a 1-h T-piece trial q 2 h x 4 followed by a 2 h T-piece trial q 4 h x 2, with blood gas values obtained after the first and third T-piece trial in each of the first two 8-h periods. At the beginning of the fourth day, the patient was given a T-piece trial for 1 h and a blood gas analysis obtained. If blood gases and vital signs were acceptable, the patient progressed to potential extubation by the primary physician. If the patient failed all three or four attempts, then one trial in the opposite mode (IMV or T-piece) was attempted.
All data between groups were analyzed by both parametric and nonparametric methods. Parametric analysis consisted of a t test matrix for analysis of variance. Non-parametric analysis employed the Kruskal-Wallis analysis of variance using x2 distribution with one degree of freedom (1 df). A significant p value was <0.05.