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 in adult acute respiratory distress syndrome.

David M, Weiler N, Heinrichs W, et al.

Intensive Care Med. 2003 Oct;29(10):1656-65 [abstract]

Reviewed by James O'Rourke MD, Nilesh Mehta MD, Medical/Surgical ICU, Children's Hospital Boston.

Review posted November 23, 2005

---

I. What is being studied?:

The study objective:

The study examined whether adult patients who met criteria for the Acute Respiratory Distress Syndrome (ARDS) could be safely transitioned from optimized pressure-controlled ventilation (PCV) to high frequency oscillatory ventilation (HFOV) without adverse incidents.

The second objective of the study was to assess the efficacy of HFOV and whether it was superior to conventional controlled ventilation in the setting of adult patients with a diagnosis of ARDS who are failing oxygenation parameters with conventional ventilation strategies.

The study design:

This was a prospective observational study carried out over a three year period in a single 14 bed intensive care unit.

The patients included:

Forty-two consecutive patients with ARDS (PaO2/FiO2 < 200 with no signs of heart failure) who failed oxygenation criteria, defined as a failure of the PaO2/FiO2 ratio to improve by 50 after 2 hours on an optimal pressure controlled mode of ventilation (including recruitment maneuvers, maximum PEEP of 15 cm H2O and maximum inspiratory pressure of 35 cm H2O).

The patients excluded:

• Absence of consent
• Pregnancy
• Anticipated death
• Withdrawal of life support at 24 hours because of poor prognosis

The interventions compared:

This was not a controlled study; all patients were placed on HFOV. The group of patients who responded to HFOV was compared to the group who failed the trial of HFOV after 24 hours, and survivors were compared to non-survivors.

The outcomes evaluated:

Variables such as oxygenation index and PaO2/FiO2 ratios were compared at three time points before the trial of HFOV: at -12 hours, -6 hours and immediately before transitioning to HFOV and at 6 time points after commencing HFOV: baseline, 1 hour, 6 hours, 12 hours, 24 hours, end.

The other primary outcome variables compared were the adverse effects encountered while on HFOV.

Patients were divided into two groups based on their oxygenation response to HFOV. Subset analysis was then performed to compare the 30 day mortality in responders versus non-responders not identified a priori.

Responders to HFOV (N=25)

• Improved PaO2/FiO2 ratio by greater than 50.

Non-responders to HFOV (N=17)

• Failure of oxygenation, defined as a rise in PaO2/FiO2 ratio by less than 50 after 24 hours of HFOV treatment despite a continuous distending pressure (CDP) or mean airway pressure of 40 cm / H2O, an FiO2 of 1.0, a Delta P of 90 and a frequency of 3 hertz.
• Failure of ventilation, defined as a PCO2 > 80 mmHg with a pH of < 7.2 and bicarbonate < 19 mmol/L
• Persistent hemodynamic instability defined as a fall in mean arterial pressure to < 60mmHg despite the optimal use of pressors and inotropes.

Morbidity related to high mean airway pressures was assessed in both groups and data collected on events such as air leaks, hypotension or endotracheal tube obstruction.

The authors attempted to ascertain whether the patient's chances of survival improved if they had a significant response to HFOV and whether their chances of survival related to the duration of mechanical ventilation before institution of HFOV.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

No, all 42 patients were assigned to the treatment with HFOV as they were deemed to have failed conventional ventilation strategies. This study was never intended as a randomized controlled trial, but was a prospective observational study.

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

Was followup complete?

Yes, all patients were followed and all outcomes reported until hospital discharge.

30 day mortality was 43% (18/42). These patients were further subdivided into those who died from respiratory causes (n=6) and those who died from non-respiratory causes (n=12). A further 4 patients died outside of the 30 day monitoring period during their hospital admission. Therefore the overall mortality for the hospital stay was 52% (22/42).

Eighteen patients were removed from HFOV within 24 hours. Seven patients improved rapidly and were returned to pressure controlled ventilation, five patients failed oxygenation criteria and were returned to pressure controlled ventilation. There were 5 deaths from bleeding or progressive septic shock and one patient was changed to conventional ventilation to for a surgical procedure.

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

Not applicable.

Secondary questions:

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

This study was not blinded, as each patient received high frequency oscillatory ventilation.

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

All subjects underwent treatment; there were no groups in this sense to compare. However, the post hoc subgroup analysis did reveal statistically significant differences in patients who responded to HFOV (N=25) versus those who did not (N=17).

Responders: those who tolerated HFOV and increased their PaO2/FiO2 ratio by greater than 50 at 24 hours.

• Were more likely to have received mechanical ventilation for a shorter period of time before the study
• Required lower Pmax inflation pressures at baseline while on conventional mechanical ventilation (32 vs. 38 cm H2O).
• Had lower SAPS II scores at baseline (44 vs. 62).
• Had higher PaO2/FIO2 ratios at baseline (105 vs. 70).
• Finally as might be expected from the above had lower oxygen indices at baseline (18 vs. 32).

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

All subjects underwent treatment; there were no groups in this sense to compare.

All patients were treated equally from a general intensive care management perspective; all patients were given nutrition, pressors and dialysis as required. Fluid management strategies were aimed at a zero or negative balance. Although sedation regimes were standardized, muscle relaxants were not given to patients in this trial. This is an unusual management option, especially in the four patients with head injury in whom coughing and ventilator dyssynchrony might be particularly detrimental to intracranial pressure.

III. What were the results?

1. How large was the treatment effect?

In the post hoc analysis two groups were identified as described above - responders and non-responders. The responders had lower APACHE scores and were less critically ill on the basis of many criteria before transitioning to HFOV. The success of HFOV may be a reflection of this fact rather than a treatment effect itself.

Survivors were also compared with non-survivors in the post hoc analysis. As one might expect, survivors were statistically more likely to have higher PaO2/FiO2 ratios and lower oxygenation indices, APACHE II and SAPS II scores at baseline. Survivors were also much more likely to have been ventilated for a shorter time before they were transitioned to HFOV. The mortality for patients who were ventilated for greater than three days was 64% versus 43% for the group as a whole. This is in keeping with the views of several other authors who contend that the damaging effects of ventilation tend to occur early in the disease process. A trial of HFOV as a rescue strategy after conventional ventilation is deemed to have failed is perhaps less useful than if it were employed earlier in the disease process (1).

The results of the study seem to indicate that response to HFOV is more a predictor of survival than a marker of HFOV's efficacy. 76% of responders survived (19/25), whereas only 29% of the patients who failed to improve their oxygenation while on HFOV survived (5/17). We can construct a 2 by 2 table to ascertain whether a patient's oxygenation response to HFOV does have an impact on survival. These results are displayed on table 1 below.

Similar to this trial, both Arnold and Derdak et al have demonstrated that survivors have lower oxygenation indices at 24 hours than non survivors (2,3). It is evident from the earlier discussion that patients who responded to HFOV were less critically ill than those who did not respond. The results are confounded by small sample sizes and subgroups that were not clarified at the outset and therefore it is difficult to identify the true efficacy of the treatment.

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

There was no treatment effect per se, since all patients received the treatment. In assessing the mortality risk of responders vs. non-responders, the confidence intervals are very wide, and therefore not very precise.

Table 1. Mortality risks in patients who responded to HFOV (compared to non-responders)

Clinical significance Measurement	Value	Lower 95% confidence interval	Upper 95% confidence interval
Absolute Risk Reduction (ARR)	0.46	0.19	0.73
Relative Risk Reduction (RRR)	66%	27%	86%
Odds Ratio	0.13	0.03	0.56

The results presented above represent an unconventional use of these metrics and have been included simply to highlight the different mortality risks in those who do and do not respond to HFOV.

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

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

Yes. This is an adult ARDS study and results should not be extrapolated to the pediatric population. APACHE scoring is not used in the pediatric ICU and we would encounter difficulties in extrapolating measures of patient severity of illness to our population. Objective parameters of bedside oxygenation apart from oxygen saturations and routine blood gases were utilized. The authors present information on both PaO2/FiO2 ratios and the oxygenation index; or as the authors state "the pressure cost of oxygenation". The oxygenation index is commonplace in the pediatric ICU setting and is one value and trend that an ECMO center would request if considering a referral.

The results of this study indicate that if an adult patient with ARDS fails to improve after 2 hours of the maximal safe ventilation parameters: recruitment maneuvers, PEEP 15, limited Pmax to 35 cm H2O and an FiO2 1.0 then it is safe to transition this patient to HFOV. Typical initial settings for the oscillator might be a CDP 5 cm greater than the last MAP observed while on conventional ventilation. Settings may be optimized by increments in CDP to a maximal setting of 40 cm H2O to improve oxygenation. A progressive reduction in ventilation frequency to 3 Hz and increasing delta P to a maximal value of 90 may be effective if CO2 clearance is inadequate.

Significant similarities and overlap exist between pediatric and adult patients with ARDS to implement a similar protocol for management of HFOV in patients who are deemed to be failing with a conventional ventilation strategy. The results of this study indicate that HFOV has a similar safety profile to conventional ventilation and on the basis of other similar studies may possibly confer a small treatment benefit in patients with ARDS.

2. Were all clinically important outcomes considered?

One of the primary outcome variables assessed in this study was the oxygenation response to HFOV. However, the utility of oxygenation as a predictive tool of mortality in ARDS has been questioned. Two examples of this are how both prone positioning and the use of nitric oxide can produce rapid improvements in oxygenation in some patients with ARDS, yet neither treatment had a demonstrable impact on clinical outcome. The second example is how low tidal volume ventilation may actually worsen oxygenation during the first few days of therapy when compared with a higher tidal volume but is associated with better clinical outcomes (4). Others have reiterated the need to investigate patient important outcomes such as longer term pulmonary dysfunction and mortality.

This study was primarily a safety study, it focused on short term outcomes and morbidity related to the use of the high frequency oscillator. No information is given in relation to longer term sequelae among responders such as oxygen requirements at 30 days. Little data is presented on how the four patients with head injury responded to the change in ventilation mode given how fluctuations in PCO2 may impact intracranial compliance. Although each patient who had a head injury had an intracranial pressure monitoring device, none were adversely affected by the higher intrathoracic pressures of HFOV. The data pertaining to pressure fluctuations and HFOV in these patients is presented in a more recent publication (5).

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

The most recent evidence in the management of ARDS focuses on attaining an open lung concept of ventilation; optimizing the FRC with the judicious application of recruitment maneuvers and PEEP and allowing hypercapnic acidosis (6). In doing this we attempt to limit barotrauma by keeping the plateau pressure ≤ 30 cm H20 and limit volutrauma to tidal volumes to 5 - 6 ml/kg. "Atelect-trauma" is avoided by optimizing end expiratory lung volume and preventing the cyclical opening and closure of airways. With this strategy we aim to ventilate patients in the steepest part of their pressure volume curve.

One problem in ventilating patients with ARDS is that it is not a homogenous condition, areas of over-distention co-exist with atelactatic segments and our goals are to attempt to recruit atelectatic segments while preventing over-distention of the well aerated lung segments. The smaller phasic volume and pressure changes that characterize HFOV make this a theoretically very safe mode of ventilation.

Most of the literature regarding the use of high frequency oscillation has been in the neonatal population with respiratory distress syndrome or in the adult ARDS population where it has been employed primarily as a rescue strategy. To date there have been two randomized prospective trials comparing CMV and HFOV in children and adults.

The adult trial by Derdak et al enrolled 148 patients, this trial was primarily a safety and efficacy study. Safety aspects such as ventilation related complications: hypotension, air leaks and endotracheal tube obstruction rates were similar in both groups.

The pediatric trial by Arnold et al enrolled 70 patients, 58 patients were subsequently randomized into two groups of 29 patients. The patients crossed over from one group to the other if they met criteria for treatment failure in either group such as intractable cardiogenic or respiratory failure. Arnold's results indicate that despite the use of higher mean airway pressures HFOV results in less barotrauma. The HFOV group had a lower requirement for oxygen at 30 days. In this trial survivors had lower oxygenation indices after 24 hours of HFOV than non survivors (26.2 versus 41.4).

Critics of both of these trials emphasized that the tidal volumes in the control group were higher than one would seek based on current thinking and results from the ARDS net trial. While neither of the trials by Arnold et al or Derdak et al was powered to detect differences in mortality, both did demonstrate a non-significant trend towards a reduction in mortality when HFOV was employed (7).

Therefore in answering the question "are the treatments worth the potential harms?" the answer at present is that we do not know. The theoretical advantages of HFOV in terms of safety are supported by the other literature where it has been used as a rescue strategy in cases where conventional ventilation has been deemed to have failed. In this study there was one pneumothorax while on the oscillator, the rate of pneumothorax in adult patients receiving conventional mechanical ventilation for ARDS is typically 7-14% (5,8). There were no cases of endotracheal tube obstruction an no patients needed to be removed from the oscillator because of hemodynamic compromise. This is a low rate of complications when considering the population involved.

References:

1. Chan KP, Stewart TE. Clinical use of high frequency oscillatory ventilation in adult patients with acute respiratory distress syndrome. Crit Care Med 2005; 33(suppl):S170-174 [abstract]
2. Arnold JH, Hanson JH, Toro Figuero LO et al. Prospective randomized comparison of high frequency oscillatory ventilation and conventional mechanical ventilation in pediatric respiratory failure. Crit Care Med. 1994; 22(10) 1530-9 [abstract]; [PedsCCM EBJC Review]
3. Derdak S, Mehta S, Stewart TE.. High frequency oscillatory ventilation for acute respiratory distress syndrome in adult patients Am J Resp Crit Care Med 2002;166: 801-808 [abstract]
4. Ware LB. Prognostic determinants of acute respiratory distress syndrome in adults: impact on trial design. Crit Care Med 2005; 33(suppl.): S217-S222 [abstract]
5. David M, Karmrodt J, Weiler N, et al High frequency oscillatory ventilation in adults with traumatic brain injury and acute respiratory distress syndrome. Acta Anaesthesiol Scand. 2005; 49(2) 209-214 [abstract]
6. Marraro GA. Protective lung strategies during artificial ventilation in children. Pediat Anaesth. 2005; 15: 630-7 [citation]
7. Wunsch H, Mapstone J. High frequency ventilation versus conventional ventilation for treatment of acute lung injury and acute respiratory distress syndrome. The Cochrane library 2004; 3: Art No CD 004085 [abstract]
8. Fort P, Farmer C, Westerman J. et al. High frequency oscillatory ventilation for adult respiratory distress syndrome Ð a pilot study. Crit Care Med. 1997;25: 937-947 [abstract]

Comments

Back to the EB Journal Club Index