From January 1998 to February 2002, 156 patients with ARDS and severe hypoxemia underwent 171 trials of HFOV (Table 1). The mean (± SD) duration of ARDS prior to HFOV was 3.5 ± 4.3 days (median duration, 2 days; 25 to 75% confidence interval [CI], 0.2 to 5 days). The etiology of ARDS included sepsis (88 patients), pneumonia (26 patients), and aspiration of gastric contents (8 patients). All patients had severe ARDS, as evidenced by a mean lung injury score (LIS) of 3.5 ± 0.4, a mean Pa02/Fl02 ratio of 91 ± 48 mm Hg, and an OI of 31.2 ± 13.7. Airway pressures during CV immediately prior to HFOV are presented in Table 1.
MPaw and Gas Exchange
The applied mPaw was significantly higher during HFOV than was mPaw at baseline during CV (Fig 1, top left, A). As part of a general HFOV strategy, Fio2 was increased immediately after HFOV initiation (Fig 1, top right, B), and consequently this first value was significantly higher than the Fio2 during CV. Within 3 h of receiving HFOV, and for up to 72 h, there was a significant reduction in Fio2 compared with CV (Fig 1, top right, B). Compared with the baseline value during CV, the mean Pao2/Fio2 ratio continued to improve for the first 12 h and remained stable beyond that time (Fig 1, bottom left, C). At 12 h, the mean improvement of the group in Pao2/ Fio2 ratio compared with baseline was 63 mm Hg (ie, a 70% improvement). OI was significantly lower than baseline at 12 h, and remained lower for the remainder of the study (Fig 1, bottom right, D). Paco2 was well-controlled during HFOV, and by 48 h it was significantly lower than the baseline value (Paco2 during CV, 53 mm Hg; Paco2 at 48 h, 50 mm Hg; p = 0.026).
Figure 2 illustrates the hemodynamic variables immediately prior to and during the first 72 h of therapy with HFOV. As a group, heart rate decreased, systolic pressure increased, and diastolic pressure was unchanged throughout the study period.
Figure 1. mPaw (top left, A), Fio2 (top right, B), Pao2/Fio2 ratio (bottom left, C), and OI (bottom left, D) plotted over the study duration. CV represents values observed during CV immediately prior to initiating HFOV therapy. All subsequent measurements were made during HFOV therapy. Time 0 represents values observed within 30 min of HFOV initiation. Values are given as the mean ± SD. The numbers adjacent to each data point represent the number of trials. The p values for trends in physiologic parameters are as follows: mPaw, p < 0.001; Fio2 p < 0.001; Pao2/Flo2 ratio, p = 0.0002; OI, p < 0.0001 (all by repeated-measures analysis of variance).
Figure 2. Systolic BP, diastolic BP, and heart rate (top, A), and CVP, PAOP, and CO (bottom, B) plotted over the duration of the study. CV represents values observed during CV immediately prior to initiating HFOV therapy. All subsequent measurements were made during HFOV therapy. Time 0 represents values observed within 30 min of HFOV initiation. Values are given as the mean ± SD. The numbers adjacent to each data point represent the number of trials. The p values for trends in physiologic parameters are as follows: systolic BP, p = 0.03; heart rate, p < 0.0001; CVP, p = 0.01; CO, p = 0.004.
Table 1—Patient Characteristics
|HFOV trials, No.||171|
|Age, yr||47.8 ± 18.3|
|Weight, kg||71.2 ± 19.4|
|1st 24 h in ICU||23.8 ± 7.5|
|24 h prior to HFOV||23.9 ± 7.9|
|MODS||9.4 ± 3.6|
|LIS||3.5 ± 0.4|
|ARDS prior to HFOV, d||3.5 ± 4.3|
|Ventilation prior to HFOV, d||5.6 ± 7.6|
|Gas exchange during conventional ventilation|
|Paco2, mm Hg||53.0 ± 18.9|
|Fio2||0.86 ± 0.17|
|Pao2/Fio2 ratio, mm Hg||91.2 ± 47.6|
|OIt||31.2 ± 13.7|
|Airway pressures during conventional ventilation, cm H2O|
|Plateau||36.0 ± 6.5|
|PEEP||14.0 ± 3.2|
|mPaw||24.0 ± 5.1|