High frequency ventilation for acute respiratory distress syndrome


High Freqency Oscillation (HFO)

Continue Reading

High Freqency Ventilation (HFV)

1. Description of the problem

1. General description of the procedure

Most patients with acute respiratory distress syndrome (ARDS) can be managed with conventional modes of mechanical ventilation. In a minority of cases (<10%), however, routine ventilator strategies fail to achieve adequate oxygenation using lung protective settings and alternative strategies are required. These can include neuromuscular paralysis, prone positioning, inhaled vasodilators, extracorporeal membrane oxygenation (ECMO) and alternative modes of mechanical ventilation, including airway pressure release ventilation (APRV) and high frequency ventilation (HFV).

There are currently two ongoing randomized clinical trials testing the effectiveness of HFV in patients with ARDS. Until results from these trials are available, HFV is generally regarded as a salvage mode of ventilation for severe cases with refractory hypoxemia from severe ARDS, and should be considered when escalating conventional support requires airway pressures and FiO2 exposures associated with ventilator induced lung injury (VILI).

With this ventilator modality, much higher-than-normal breathing frequencies are used (>100 breaths/min in the adult) along with much smaller tidal volumes (eg <1 ml/kg in the alveolar regions) that are often less than anatomic dead space. The rationale for such an approach is that the smaller tidal pressure swings, coupled with appropriate mean airway pressure applications, create a conceptually ideal lung-protective strategy in ARDS that limits over-distention and collapse-reopening VILI (Figure 1).

Figure 1.

Conceptual rationale for high-frequency ventilation (HFV). Left panel is the pressure-volume relationship plot in patient with acute lung injury. Points a and b represent alveolar de-recruitment pressure and overdistention pressure, respectively. Right panel illustrates how a high-frequency ventilator with intrinsic positive end-expiratory pressure (PEEP) and smaller tidal pressure swings might be expected to ventilate between inflection points more readily than conventional ventilation (CV) with applied PEEP.

Although there are different devices that can be used to generate such high frequencies, including jets and oscillators, only oscillators are approved for adult use (specifically, only the SensorMedics 3100b is FDA approved) and this discussion will thus be limited to use of this device.

2. Mechanism of how HFV works

HFV using the oscillator (HFO) operates by using a piston mechanism to generate a “to and fro” application of pressure to the airway opening. Fresh gas is supplied into the ventilator circuit as a “bias flow,” and mean airway pressure is adjusted by the relationship between fresh gas inflow and any positive or negative pressure placed on the gas outflow from the bias flow circuit (Figure 2).

Figure 2.

Schematic diagram of a high-frequency oscillator (HFO) for delivering high-frequency ventilation. Airway pressure oscillations are produced by a piston or membrane or microprocessor flow controller at a selected rate and displacement. Fresh gas inflow occurs through a bias flow. Gas outflow and circuit pressure are regulated by adjusting bias flow and/or pressure near the exhalation port. (Reprinted with permission from Tobin M: Principles and Practice of Mechanical Ventilation. New York: McGraw-Hill, 1994.)

With oscillators, clinicians usually have the capability to set oscillator frequency, oscillator displacement (volume), inspiratory to expiratory time and bias flow. The actual delivered fresh gas volume to the lung depends on oscillator displacement volume as it interacts with respiratory system mechanics, and both the magnitude and the locations of the bias flow. In general, the bias flow location should be as deep as possible in the airways to minimize dead space.

With HFO, peak and baseline pressures in the proximal airway are dampened considerably by the time the oscillations reach alveolar units. However, mean airway and alveolar pressures are probably comparable during HFO in the adult (Figure 3). Because delivered volumes are very difficult to monitor with these systems, ventilation parameters are set using pressure measurements, visual inspection of chest motion and arterial blood gases, as described below in discussion on mechanisms of gas transport in HFO.

Figure 3.

Airway pressure profiles over time at different sites in the tracheobronchial tree. The left panel reflects HFO; the right panel reflects conventional ventilation (CV). Note that with HFO, very large pressure swings in the ventilator circuitry are considerably dampened near the alveolar structures. In contrast, with CV, circuit pressures equilibrate with alveolar pressures near end inspiration.

3. Implementing HFO

A. Indications

HFO use is generally considered in adult respiratory failure when adequate gas exchange cannot be provided using conventional ventilator settings that are “lung protective.” In practical terms, this means the arterial PO2 is below 50-55 mmHg despite end-inspiratory pressures (plateau airway pressures corrected for any chest wall effects) approaching 35 cm H2O and FiO2requirements above 0.60 to 0.70. As stated previously, the timing of when to implement HFO relative to other salvage therapies must be individualized.

The other indication for use of HFO in adults is for broncho-pleural fistula, to decrease the air leak owing to a reduced amplitude of pressure swings.

B. Relative contraindications

Increased intracranial pressure, due to the increased mean airway pressures used. In addition, patients with any type of obstructive lung disease should not be placed on this mode as HFO effects on gas transport and alveolar pressures are less well understood or predictable under these circumstances.

C. Set up: initial ventilator settings

The important settings are frequency, power (the piston pressure swings generating the tidal volume), FiO2, and the mean airway pressure (mPaw). Most authorities recommend some form of recruitment maneuver (e.g. 30-40 cm H2O mPaw for 30-40 seconds) and a subsequent mPaw setting 5 cm H2O higher than what had been delivered with the conventional ventilator.

  • Initial frequency at 5 Hz.

  • FiO2 is 1.0.

  • Power (piston displacement force) at 5-6 and adjust subsequently to see an obvious “chest wiggle.”

  • Inspiratory:expiratory ratio at 1:2.

  • Circuit bias flow at 30-40 L/min to assure CO2 clearance and maintenance of mPaw.

D. Subsequent ventilator settings

To adjust ventilation

  • To decrease PaCO2

  • Increase the power setting in single increments.

  • If power setting is maximal, decrease frequency by increments of 0.5 (this increases tidal volume).

  • If PaCO2 is still too high, consider deflating the endotracheal tube cuff to facilitate CO2 clearance. Note that this may decrease mPaw and decrease oxygenation.

  • To increase PaCO2 decrease the power setting in single increments or increase the frequency in 0.5 Hz increments.

To improve oxygenation

As in other modes of mechanical ventilation, to improve oxygenation one can either increase the mPaw and/or the FiO2. This can be accomplished in a protocolized fashion by referring to mean airway pressure/FiO2 tables similar to the PEEP/FiO2 tables employed in large clinical trials of conventional ventilation. An example of one such titration is given in Table I, with the strategy being to move down the table if below oxygenation goals and to move up the table if oxygenation goals are exceeded:

Table I.
Goal: 55 ≤ PaO2 ≤ 80 Goal: 88 ≤ SpO2 ≤ 95

E. Other considerations when using HFO

Additional adjustment to vent settings to minimize VILI

Once the patient has stabilized on HFO, some authorities suggest increasing the frequency as much as possible to further minimize alveolar cyclical pressures. This is done by adjusting the power settings in 0.5Hz increments to maximal or until CO2 clearance becomes unacceptable (due to decreased tidal volumes).

Use of sedation and paralysis

Deep sedation and neuromuscular paralysis are employed commonly in the first few hours of HFO to minimize or eliminate ventilator asynchrony while optimizing ventilator settings to improve gas exchange. Thereafter, paralysis should be discontinued if at all possible. Surprisingly, most patients tolerate HFO quite well without excessive sedative/analgesic requirements. It is important to note that spontaneous breaths can occur during HFO although inspiratory gas flow is limited by the set bias flow.


Significant complications can occur using this mode and an extensive “learning curve” is required before operators become skilled at delivering vent support appropriately and safely. The most common complications include the following:

Insufficient airway humidification:Humidification is difficult to achieve when using high gas flows. Therefore, effective systems that provide adequate heat and humidity are required along with frequent assessments of airway function and sputum consistency.

Hypotension: Since high mean airway pressures are used with HFO this may reduce venous return. As with other ventilation strategies, the first intervention should be to give intravenous fluids.


As the patient improves, the FiO2 and mPaw requirements will decrease. When the mPaw is 20-22 cm H2O and the FiO2is 0.40-0.50, consideration can be given to returning the patient to conventional ventilation.

Mechanism of gas transport

The tidal breaths in HFO are usually small and often are less than anatomic dead space. For effective CO2 and O2 transfer to take place between alveoli and the environment under these circumstances, mechanisms other than conventional bulk flow transport (i.e., nonconvective gas transport) must be invoked. This is because the traditional relationship between effective alveolar ventilation (VA) and the frequency (f), tidal volume (VT) and dead-space volume (VD) (i.e., VA = f x [VT – VD]) becomes meaningless when VT is less than VD.

A number of different mechanisms have been proposed to explain gas transport under these seemingly “unphysiologic” conditions.

These include:

  • If the gas-flow profile during one phase of the ventilatory cycle is parabolic and square during the other phase, a net flow of gas can occur in one direction through the center of the airway and in the other direction via the periphery (coaxial flow). Measurements in models of the human tracheobronchial tree have demonstrated such asymmetric flow profiles, but they are quite complex and depend heavily on airway geometry (especially bifurcations) and gas velocity during different phases of the ventilatory cycle.

  • Taylor dispersion is a complex physical concept that describes gas dispersion along the front of a high-velocity gas flow. The dispersion characteristics are different depending on whether flow is turbulent or laminar. Dispersion also is affected by bifurcations in the airway and development of flow eddies. Net gas transport occurs as a result of this dispersion of gas molecules beyond the bulk flow front.

  • Molecular diffusion is responsible for gas mixing within alveolar units during conventional ventilation. Molecular diffusion is also likely to serve this role during conventional ventilation. Molecular diffusion is also likely to serve this role during HFV as well. It is unclear whether augmented molecular diffusion serves any additional role during HFV.

  • Pendelluft is the phenomenon of intra-unit gas mixing due to impedance differences. This intra-unit mixing also can involve airway gas and thus produce effective alveolar ventilation. Pendelluft may be particularly pronounced when HFV is used in a lung with heterogeneous impedances.

The relative importance of each of these mechanisms is not clear. In fact, because these mechanics are not mutually exclusive, all may be operative simultaneously and to varying degrees depending on HFV parameters as well as the effects of lung disease on regional mechanics.

Predicting gas exchange as a function of ventilator parameters when these nonconvective flow mechanisms are operative during HFV can be difficult. In general, as nonconvective flow mechanisms become more important, alveolar ventilation becomes increasingly a function of frequency times the square of tidal volume (f x VT2). Thus VT has more effect than f in determining VA. Indeed, increases in frequency usually results in VT reduction, which can actually reduce effective VA.

The proportionality constant between VA and fx VT2 is quite small and thus, during nonconvective flow HFV, the f x VT product needs to be quite high for effective alveolar ventilation to occur. This is why typical HFV “output” is generally severalfold higher than conventional mechanical ventilation. Importantly, however, these pressure and volume changes in the major airways are considerably dampened by the time they reach the alveoli, and thus alveolar pressure and volume changes are small. Because of this alveolar pressure profile of small oscillations around a substantial mean during HFO, some have termed HFO as being simply “CPAP with a wiggle.”

Alveolar capillary gas transport during HFV depends on matching effective ventilation with perfusion (V/Q), just as it does with conventional ventilatory strategies. Thus, the alveolar-arterial oxygen difference during HFV remains largely dependent on mean alveolar pressure (and functional residual capacity), just as it does with conventional strategies.

Interestingly, observational trials with HFV (especially HFO) described below have often safely utilized mean pressures higher than conventional ventilation and sometimes higher than what is considered a “safe” maximal pressure during conventional ventilation (i.e. >35 cm H2O). This may be possible because the tidal pressure swings are so small with HFO (low lung “strain”) and the application of the higher mean pressure is often a gradual process. Because of this, alveolar epithelial cells may be able to “adapt” to this higher mean stretch.]

Special considerations for nursing and allied health professionals.


What's the evidence?

Clinical trials of HFV

The strongest clinical data supporting the use of HFV come from studies in neonatal and pediatric populations. In these populations, jet breathing frequencies in the range of 250-600 breaths per minute or oscillatory frequencies of 500-1000 breaths per minute produce adequate gas exchange and often a lower incidence of chronic lung disease in survivors. Several of these studies emphasize the need for HFV to have adequate volume recruitment for successful application.

However, not all studies favor HFV, although most of the negative trials only showed HFV to be no worse than conventional ventilation (i.e. they are equivalent). Importantly, a recent review of the neonatal/pediatric experience with HFV emphasized that while HFV does provide benefit, the magnitude of that benefit has shrunk over the years as new modalities (e.g. surfactant) and a better appreciation for lung protective conventional ventilation has emerged.

Adult experience with HFV has been limited because devices with adequate ventilation capabilities for adults are few. The largest adult trial to date utilized an oscillator and demonstrated strong trends in improved mortality in favor of HFO. A concern with this trial was the fact that the control group had a tidal volume generally considered now to be excessive.

In 2010 the McMaster University Evidenced Based Medicine Group updated a meta-analysis of HFO in ARDS. They analyzed eight clinical trials of HFO in patients with ARDS. This population included some pediatric patients as well who met the criteria for ARDS. In this analysis, six of the eight studies applied HFO within 48 hours of intubation and in five of the eight studies the ARDS Network low tidal volume strategy was used as the control group. 419 patients were included in these studies.

None of the trials on their own showed a significant reduction in mortality but the meta-analysis of the group did. The summary results showed that HFO produced a significant reduction in mortality with a risk ratio of .77 and a 95% confidence interval range from 0.61 to 0.98.

This certainly suggests that there may be a role for HFO in severe respiratory failure from ARDS. However, the numbers are still small, pediatric patients were included, and the combined results just barely reached statistical significance. There are two large on-going trials of HFO, one in Canada and one in Europe, that should supply much more solid data in the future.


Plotz, FB, Slutsky, AS, van Vught, AJ. “Ventilator induced lung injury and multiple system organ failure: a critical review of facts and hypotheses”. Intensive Care Med. vol. 30. 2004. pp. 1865-72. (A great review of both animal and human data from the last two decades that have formed the basis of our understanding of ventilator induced lung injury [VILI].)

Imai, Y, Slutsky, AS. “High-frequency oscillatory ventilation and ventilator-induced lung injury”. Crit Care Med.. vol. 33(3 Suppl). 2005. pp. S129-34. (A review of the rationale for why very small pressure oscillations around a substantial mean airway pressure should minimize VILI.)

Fort, P, Farmer, C, Westerman, J. “HFOV ventilator for ARDS – a pilot study”. Crit Care Med. vol. 25. 1997. pp. 937-47. (The first description of an HFO device in adult humans.)

Simon, BA, Weinmann, GC, Mitzner, W. “Mean airway pressure and alveolar prssure during high-frequency ventilation”. J Appl Physiol. vol. 57. 1984. pp. 1069-78. (A careful analysis of the relationships of mean and delta pressures from the circuit to the alveoli during HFO.)

Chang, HK. “Mechanisms of gas transport during ventilation by high frequency oscillation”. J Appl Physiol Respir Environ Exerc Physiol. vol. 56. 1984. pp. 553-63. (A comprehensive review of all the postulated non-convective gas transport mechanisms at play during HFO and how they likely co-exist in injured lungs.)

Protti, A, Cressoni, M, Santini, A, Langer, T, Mietto, C. “Lung stress and strain during mechanical ventilation: any safe threshold?”. Am J Respir Crit Care Med. vol. 183. 2011. pp. 1354-62. (An indepth analysis evaluating the likely combined role of "maximal" stretch and repetitive "tidal" stretch in producing VILI.)

Hubmayr, RD. “Cellular stress failure in ventilator-injured lungs”. Am J Respir Crit Care Med. vol. 171. 2005. pp. 1328-42. (An insightful analysis of alveolar epithelial stretch injury at the cellular level.)

Bollen, E. “Cumulative meta-analysis of high frequency vs conventional ventilation in premature neonates”. Am J Resp Crit Care Med. vol. 168. 2003. pp. 1150(A meta-analysis of infant HFV over time demonstrating that the improved outcomes from HFV have become less apparent in the setting of more advanced support strategies emphasizing lung protection.)

Derdak, S, Metha, S, Steward, T. “High-frequency oscillatory ventilation for acute respiratory distress syndrome in adults: A randomized, controlled trial”. Am J Respir Care Med. vol. 166. 2002. pp. 801-8. (The largest randomized trial to date of HFO in ARDS showing a non-significant survival trend from HFO – study likely underpowered.)

Sud, S, Sud, M, Friedrich, J, Meade, M, Ferguson, N. “High frequency oscillation in patients with acute lung injury and acute respiratory distress syndrome (ARDS): systematic review and meta-analysis”. BMJ. vol. 340. 2010. pp. c2327(The most recent meta-analysis of HFO for ARDS suggesting improved survival using the oscillator.)