Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

High-Frequency Oscillatory Ventilation for Acute Respiratory Distress Syndrome in Adults.

Derdak S, Mehta S, Stewart TE et al.

Am J Resp Crit Care Med 2002; 166:801-808. [abstract]

Reviewed by Kathleen Ventre MD , Children's Hospital, Boston, MA

Review posted January 7, 2003

---

I. What is being studied?:

The study objective:

To demonstrate the safety and effectiveness of high-frequency oscillatory ventilation (HFOV) and to determine whether it was comparable to conventional ventilation (CV) for the treatment of ARDS in an adult population.

The study design:

Randomized, controlled trial.

The patients included:

Adult patients with early-phase ARDS defined a: patients 16 years of age or more who were mechanically ventilated, with paO2/FiO2 ≤ 200 mm Hg while on PEEP ≥ 10 cm H2O, bilateral radiographic pulmonary infiltrates, and no clinical evidence of left atrial hypertension (PAOP ≤ 18 mm Hg if PA catheter data available) (3).

The patients excluded:

Those weighing less than 35 kg, those with severe obstructive lung disease (COPD/asthma), those with severe airleak (defined as having more than one chest tube per hemithorax with a persistent airleak of more than 120 hours), FiO2 more than 0.8 for more than 48 hours, intractable shock, a nonpulmonary terminal diagnosis with an estimated 6-month mortality of more than 50%, and those who had participated in other investigational trials for ARDS or septic shock within 30 days of the present study (3).

The interventions compared:

Ventilator strategies for patients in each treatment arm. Treatment strategies for both HFOV and CV were directed at achieving similar physiologic targets: SaO2 at least 88% on FiO2 0.6 or less while maintaining mean airway pressure (HFOV group) or PEEP (CV group) until FiO2 could be reduced to 0.6 or less, and PaCO2 40-70 or higher as long as serum pH was 7.15 or greater. Bicarbonate could be administered for severe respiratory acidosis (pH < 7.15) at the discretion of the clinicians (3).

Patients randomized to the HFOV arm of the study were placed on the 3100B high-frequency oscillatory ventilator (SensorMedics, Yorba Linda, CA) which in comparison to the 3100A model commonly used for infants and children has a higher maximal bias flow (which allows for the delivery of higher mean airway pressures), higher maximum pressure amplitude of oscillation ("delta P"), and a more powerful electromagnet that allows faster acceleration to delta P (10). In this study, HFOV was initiated at FiO2 0.8-1.00, oscillation frequency 5 Hz, I/E ratio 1:2, and bias flow 40 L/ minute. Mean airway pressure (MAP) was initially set at 5 cm H2O greater than the MAP on CV just before conversion to HFOV - a maneuver intended to produce an ideal volume/"open lung" strategy of ventilation which has been shown in human studies to maximize the oxygenation benefits of HFOV (11, 7, 1, 5), and data from animal studies has shown that this method results in decreased histologic evidence of barotrauma (1, 6, 9, 2).

If FiO2 > 0.6 was needed to produce an SaO2 of at least 88%, the MAP was increased in increments of 2 to 3 cm H2O every 20-30 minutes to a maximum of 40-45 cm H2O. Delta P was initially set to achieve chest wall vibration to the level of mid-thigh. Delta P and Hz were sequentially adjusted to maintain the target PaCO2 and serum pH. If these target values could not be achieved with maximal delta P and minimal Hz, an endotracheal tube cuff leak was allowed in order to enhance CO2 egress. While on HFOV, patients were given sedation and neuromuscular blocking agents.

Patients were switched from HFOV to CV when FiO2 was 0.5 or less, and MAP was 24 cm H2O or less with an SaO2 of 88% or more. On transition to CV, the conventional ventilator was in pressure control mode, with settings intended to achieve a MAP comparable to the MAP on HFOV just before conversion: Peak inspiratory pressure (PIP) was adjusted to achieve a delivered tidal volume of 6-10 cc/kg actual body weight, PEEP 10 cm H2O, and I/E ratio initially 1:1.

Patients in the CV arm of the study were managed in the pressure control mode, targeting a delivered tidal volume of 6-10 cc/kg actual body weight, RR adjusted to meet pH guidelines (max 35), PEEP ≥ 10 cm H2O, and initial I/E ratio 1:2 (Table 1). The weaning phase employed the pressure support mode as well as spontaneous breathing trials (3).

The outcomes evaluated:

The primary outcome measure was survival without the need for mechanical ventilation at 30 days following study entry. Secondary outcomes included mortality at 6 months, mechanical ventilation at 6 months, intractable hypotension, oxygenation failure, ventilation failure, developing or worsening airleak, and mucus plugging requiring endotracheal tube change (3).

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes. 148 patients meeting enrollment criteria were randomized by computer at local sites. The authors do not provide specific information describing the randomization process. 75 patients were assigned to the HFOV arm and 73 patients were assigned to receive conventional ventilation. The sample size was calculated in order to detect a 20% difference in "key adverse outcomes" and to establish with 95% confidence that HFOV was not more than 10% worse than CV when applied to this patient population for the management of ARDS (3).

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Nearly. Analyses were based on intention to treat, and the authors report a study attrition of 2 patients in the HFOV group and 5 patients in the CV group (3).

Were patients analyzed in the groups to which they were randomized?

Yes. Analyses were based on intention to treat.

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

Most of the patients may well have been blinded to the treatment (mode of ventilation) because of the sedation needed to manage them with complex strategies on sophisticated ICU ventilators - especially during the early days of the study. Clinicians would not be blinded to the ventilator mode on study participants because the use of a high-frequency ventilator instead of a conventional ventilator would be obvious. The authors do mention that overall results were reviewed annually by an independent safety monitoring committee (3). Overall, the results were blinded to investigators and the study sponsors.

4. Were the groups similar at the start of the trial?

Nearly. There were no significant differences in any of the baseline physiologic (respiratory and hemodynamic) parameters between the two study groups (3). The authors do report trends toward a significant difference between the study groups in several areas. Of particular interest is the trend toward increased pre-enrollment days on CV in patients assigned to the CV arm of the study. There was also a trend toward decreased pH and increased OI in the patients randomized to receive CV.

5. Aside from the experimental intervention, were the groups treated equally?

Not exactly. Among critical care clinicians and ICU's there is wide variation in the clinical management of ARDS. Accordingly, the authors mention that patients randomized to either arm of this study could receive adjunctive therapies at the discretion of the attending physicians. Seven of 75 patients in the HFOV arm received "rescue therapies:" 4 received inhaled nitric oxide (iNO), 2 were proned, 1 received high-dose corticosteroids, and 4 were treated with conventional ventilation. Twelve of 73 patients in the CV arm received rescue therapies: 8 received iNO, 9 received HFOV, three were proned, and 4 received high-dose corticosteroids.. There was no statistically significant difference in mortality between the 2 treatment groups among those patients who received these adjunctive therapies (3).

The authors do not provide information on management of other organ systems in the study patients. Specifically, it is reasonable to suspect significant differences in intravascular volume support of patients in the HFOV arm (especially early on) versus those managed with CV. To optimize oxygenation it is often necessary to volume load patients when transitioning to HFOV in order to offset regional conditions where alveolar pressure exceeds pulmonary arterial and pulmonary venous pressures on sustained application of high mean airway pressures (13). The authors point out that among invasively monitored patients in the HFOV group, CVP was increased at 2 hours vs baseline values (p = 0.003). PAOP was significantly increased at 2 hours vs baseline in the HFOV group, as well (p = 0.001). Other investigators have noted early changes in these parameters when patients are transitioned to HFOV (4, 10).

The fact that patients in the present study as well as in previous studies demonstrate significant differences in these parameters only early on, while maintaining stable indices of cardiac output, supports the notion that these changes are attributable to changes in intravascular volume, and not to the sustained application of high mean airway pressures. One would not expect poorly compliant lungs to transmit significant amounts of pressure to neighboring intrathoracic structures.

III. What were the results?

1. How large was the treatment effect?

This study was powered only to evaluate comparability of the primary and secondary outcome measures. Accordingly, 36% of those randomized to receive HFOV were alive at 30 days without need for mechanical ventilation (the primary outcome measure), and 31% of CV patients met this standard (p= 0.686; CI -12 to 22%). There is a trend toward increasing difference in the two treatment groups in those alive on mechanical ventilation at 30 days (26% in the HFOV group and 16% in the CV group [p = 0.19 and CI -4 to 24%]) and those who are dead at 30 days (37% in the HFOV group and 52% in the CV group (p = 0.102 and CI -32 to 2%).

Causes of death are similar between the groups, with multiple organ failure being the most common cause cited. There were no significant differences in the secondary outcome measures between the two study groups, although there was a nonsignificant trend toward decreased 6-month mortality in the HFOV group (47% vs 59%). The mean duration of mechanical ventilation was 22 ± 21 and 20 ± 31 days, in the HFOV group and CV group, respectively. The mean duration of HFOV was 6 ± 6 days (3).

It is interesting to examine some of the oxygenation and ventilation data for each study group over the first 3 study days. The differences between MAP and PaCO2 for the HFOV group and the CV group over the 3 days were significant (p = 0.001 and 0.01, respectively). An early rise in PaO2/FiO2 on transition to HFOV has been demonstrated in other adult HFOV studies (4, 10). Oxygenation index (OI = FiO2 x MAP x 100/PaO2), first reported in an adult ARDS study by Fort, et al. to be predictive of mortality in this population (4), decreased to a similar degree over the study period in both the HFOV and CV groups. Post-study multivariate analysis revealed that the OI trend was the most significant post-treatment predictor of survival regardless of treatment group: survivors showed an improvement over the first 72 hours of study time and nonsurvivors did not (p = 0.014). Other investigators have demonstrated this phenomenon in adult ARDS (10). Although the OI is not commonly used as a physiologic parameter in adults, following its trajectory in the adult ARDS population may prove to be an important and informative clinical tool.

2. How precise was the estimate of the treatment effect?

Again, this study was powered only to evaluate comparability of the primary and secondary outcome measures. Accordingly, the p values associated with the differences in primary outcomes do not fall within the realm of statistical significance. In addition, each of the confidence intervals associated with the differences in each of the primary outcome variables span zero, indicating that the authors' point estimates do not represent a stable pattern of outcomes with 95% confidence when HFOV is compared to CV for the management of ARDS in adults: one cannot exclude either an important HFOV benefit or worse outcomes associated with its use in a similar patient population. A 2 x 2 table constructed to represent the primary outcome measure would appear as the following (14):

	Alive at 30 days    without MV	Alive on MV or not alive	total n
HFOV	27	48	75
CV	23	50	73
Total	50	98	148
	Value	95% C.I.
Relative Risk	1.143	(0.700 to 1.877)
RRI*	16%	(-83% to 26%)
ARI*	0.05	(-0.10 to 0.20)
NNH*	20	(5 to infin)

(* one is usually looking for disease reduction from a therapy; in this case, we are looking for a risk "increase" - the "risk" of a better outcome of being alive at 30 days without MV.)

One could also construct a 2 x 2 table to represent another outcome measure, mortality (14):

	Dead at 6 months	Alive at 6 months	n
HFOV	35	40	75
CV	43	30	73
Total	78	70	148
	Value	95% C.I.
Relative Risk	0.792	0.57-1.103
RRR (1-RR)	20%	(-8% to 41%)
ARR	0.12	0.04 to 0.28
NNT (1/ARR)	8	4->infin

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

Pediatric critical care physicians should be interested in this study as they consider using high-frequency oscillatory ventilation to support pediatric and adolescent patients > 35 kg who develop ARDS and show evidence of failing conventional mechanical ventilation. Although the mean age of the patients in this study is 48 ± 17 years in the HFOV group and 51 ± 18 in the CV group, the authors do include patients as young as 16 years of age admitted to university-affiliated medical centers who meet physiologic criteria commonly encountered in pediatric respiratory failure and ARDS (3). Nonetheless, the pediatric clinician might wonder if the older age range of the patients in this study might imply additional comorbidities that could alter treatment outcomes.

Possibly the most compelling reason to interpret the results of this study with caution is the ventilator strategy used in the patients who were randomized to receive conventional ventilation. The patients in the "control" arm were exposed to a pressure-limited mode of ventilation targeted to achieve relatively large tidal volumes (up to 10 cc/kg actual body weight) without attention to plateau pressures, % inspiratory times of up to 66%, and weaning to extubation (pressure supported ventilation) that was not protocolized (3). The authors acknowledge that the recently published ARDS Network Trial (12) has meanwhile produced a 10% decrement in mortality in patients assigned to receive 6cc/kg tidal volumes based on ideal body weight (IBW) rather than 12cc/kg IBW, using volume assist-control ventilation in both groups, where plateau pressures in the low tidal volume group are limited to < 30 cm H2O. The high-tidal volume ("control") group in the ARDSnet study had a mortality rate of approximately 40%, which is considerably less than that quoted in the present study, but their study population had a PaO2/FiO2 ratio of < 300 at study entry, which is greater than that seen in the present study (12). The ARDS Network Trial has had considerable clinical influence on mechanical ventilation strategies in practice, and it is reasonable to argue that Derdak, et al. employ a conventional ventilation strategy that is no longer in widespread use and could potentially contribute to additional lung injury in a way that might influence outcomes. In addition, a transition strategy that exposes patients to conventional ventilator settings designed to produce a MAP that is comparable to the one just prior to leaving HFOV may be unnecessarily conservative, and may potentially offset some of the physiologic benefits of high-frequency ventilation.

2. Were all clinically important outcomes considered?

The authors consider several outcomes in both the primary and secondary measures that may be appropriate to a study that intends to demonstrate comparability of two different modes of mechanical ventilation in adult ARDS. As future trials may be considered to compare a protocolized application of HFOV against more recent consensus guidelines for conventional ventilation in ARDS, it may be interesting to consider ventilator-free days and days free of nonpulmonary organ system dysfunction as outcome measures. Finally, as the authors acknowledge at the end of the study, it may be interesting to incorporate serologic indices of inflammation which have been associated with ventilator-induced lung injury (8) into outcome measures (2). Again, wide variations among clinicians in the management of complex disease processes such as ARDS can always be expected to produce confounding influences on outcomes in trials intended to evaluate strategies of mechanical ventilation.

3. Are the likely treatment benefits worth the potential harms and costs?

Although this study was not powered to demonstrate a difference in mortality between the group randomized to HFOV and the group randomized to CV, it is tempting to anticipate that a larger trial may be able to associate some of the early favorable physiologic trends observed here and in other studies (4, 10), in the HFOV group (increased PaO2/FiO2 ratio, and possibly an OI decrement) with improved clinical outcomes in adult patients with ARDS who are supported with HFOV. Until there is an appropriately powered study that compares the early and standardized application of HFOV with a protocolized, "lung protective" strategy of conventional ventilation in a carefully defined population, it is difficult for clinicians to conclude on the basis of this trial alone that supporting adult ARDS patients with HFOV may be a superior strategy. Despite the numerous technical challenges that may preclude the enactment of a definitive HFOV trial in ARDS patients, the data in animals and humans supportive of its numerous potential advantages suggests a role for HFOV in the clinical arena, so the costs associated with acquisition of high-frequency oscillatory ventilators and training staff in the use of these machines may still be justified.

References

1. Arnold JH, Hanson JH, Toro-Figuero LO, Gutierrez J, Berens RJ, Anglin DL. Prospective, randomized comparison of high-frequency oscillatory ventilation and conventional mechanical ventilation in pediatric respiratory failure. Crit Care Med 1994;22:1530-1539. [abstract]; PedsCCM EBJC Review
2. Bond DM, Froese AB. Volume recruitment maneuvers are less deleterious than persistent low lung volumes in the atelectasis-prone rabbit during high-frequency oscillation. Crit Care Med 1993;21:402-412. [abstract]
3. Derdak S, Mehta S, Stewart TE, Smith T, Rogers M, Buchman BC, Lowson S, Granton J, et al. High- frequency oscillatory ventilation for acute respiratory distress syndrome in adults: a randomized, controlled trial. Am J Respir Crit Care Med 2002;166:801-808. [abstract]
4. Fort P, Farmer C, Westerman J, Johannigman J, Beninati W, Dolan S, Derdak S. High-frequency oscillatory ventilation for adult respiratory distress syndrome: a pilot study. Crit Care Med 1997;25:937-947 [abstract]
5. Gerstmann DR, Minton SD, Stoddard RA, et al. The Provo multicenter high-frequency oscillatory ventilation trial improved pulmonary and clinical outcome in respiratory distress syndrome. Pediatrics 1996; 98:1044-1057. [abstract]
6. Hamilton PP, Onayemi A, Smyth JA, et al. Comparison of conventional and high-frequency ventilation: oxygenation and lung pathology. J Appl Physiol 1983;55:131-138. [citation]
7. HiFO Study Group. Randomized study of high-frequency oscillatory ventilation in infants with severe respiratory distress syndrome. J Pediatr 1993;122:609-619. [abstract]
8. Imai Y, Kawano T, Miyasaka K, Takata M, Imai T, Okuyama K. Inflammatory chemical mediators During conventional ventilation and during high frequency oscillatory ventilation. Am J Respir Crit Care Med 1994;150:1550-1554. [abstract]
9. McCulloch PR, Forkert PG, Froese AB. Lung volume maintenance prevents lung injury during high-frequency oscillatory ventilation in surfactant-deficient rabbits. Am Rev Respir Dis 1988; 137:1185-1192. [abstract]
10. Mehta S, Lapinsky SE, Hallet DC, Merker D, Groll RJ, Cooper AB, MacDonald RJ, Stewart TE. A prospective trial of high frequency oscillatory ventilation in adults with acute respiratory distress syndrome. Crit Care Med 2001;29:1360-1369. [abstract]
11. Ogawa Y, Miyasaka K, Kawano T et al. A multicenter randomized trial of high-frequency oscillatory ventilation as compared with conventional mechanical ventilation in preterm infants with respiratory failure. Early Hum Dev 1993;32:1-10. [abstract]
12. The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301-1308. [abstract]; PedsCCM EBJC Review
13. West JB. Respiratory Physiology: The Essentials. 4th ed. Baltimore: Williams and Wilkins, 1990, p. 41.
14. http://www.healthcare.ubc.ca/calc/clinsig.html for online calculations of statistical variables

Comments

Back to the EB Journal Club Index