High-frequency oscillatory ventilation and pediatric cardiac surgery: Yes, we can!

In the previous issue of Critical Care, Bojan and colleagues reported their experiences with high-frequency oscillatory ventilation (HFOV) after pediatric cardiac surgery. A total of 120 patients were treated with HFOV on the day of surgery, thus excluding rescue HFOV use. The main finding of the authors was that the duration of mechanical ventilation was significantly shorter in patients in whom HFOV was initiated on the day of surgery. Especially interesting about this work is that the authors used HFOV when there was evidence of pulmonary hypertension or right ventricular (RV) failure in their patients. This is an interesting approach as it is often assumed that high intra-thoracic pressures increase RV afterload and thus may enhance RV dysfunction. The findings of Bojan and colleagues may be explained by the fact that they were able to decrease the pulmonary vascular resistance by finding a proper balance between atelectasis and overdistension of the lung. It can be argued that it is possible to do so by applying positive end-expiratory pressure. But, at the same time, this may coincide with the delivery of high inspiratory pressures (>30 cm H2O). As HFOV is, in fact, a continuous positive airway pressure system, its advantage is that it is possible to maintain sufficient lung volume without large injurious pressure swings. Although the observations by Bojan and colleagues need to be confirmed in a prospective randomized trial, they have provided arguments not to rule out the early use of HFOV in pediatric cardiac surgery patients.

High-frequency oscillatory ventilation (HFOV), as used by Bojan and colleagues, is, at least in theory, an ideal tool for lung-protective ventilation as it allows eff ective pulmonary gas exchange with the delivery of a very small tidal volume (V T ) below dead space and diminished risk of atelectrauma [1,2]. Numerous animal and clinical studies have shown clearly that mechanical ventilation (MV) itself can initiate or exacerbate lung injury, termed ventilator-induced lung injury. Th e application of a V T and the repetitive opening and closure of alveoli have been identifi ed as important pathophysiological mechanisms [3].
Until now, only two randomized controlled trials investi gating the eff ect of HFOV on patient outcome have been performed [4,5] and the larger of these included only 58 children [4]. Th e main fi nding of this study was that HFOV did not signifi cantly improve survival (66% with HFOV versus 59% with conventional MV) or total ventilator days (20 ± 22 days with HFOV versus 22 ± 17 days with conventional MV) [4]. As a conse quence, HFOV is not universally employed in pediatric critical care. Also, it seems very unrealistic that any new pediatric HFOV trial will be initiated within the next few years. Th is means that well-designed observational studies are needed to shed light on some of the many other aspects of pediatric HFOV which remain to be explored, including the identifi cation of patients who are most likely to benefi t from HFOV, the timing of HFOV (early versus rescue), optimal oscillator settings, and monitoring during HFOV.
In the previous issue of Critical Care, Bojan and colleagues [1] reported their experiences with HFOV after pediatric cardiac surgery. A total of 120 patients were treated with HFOV on the day of surgery, thus excluding rescue HFOV use. Patients were transitioned to HFOV when hypoxemia and acidosis occurred despite increasing alveolar ventilation on conventional MV (if V T exceeded 10 mL/kg) or when there was evidence of pulmo nary hypertension and right ventricular (RV) failure. Th e main fi nding was that the duration of MV was signifi cantly shorter in patients in whom HFOV was

Abstract
In the previous issue of Critical Care, Bojan and colleagues reported their experiences with highfrequency oscillatory ventilation (HFOV) after pediatric cardiac surgery. A total of 120 patients were treated with HFOV on the day of surgery, thus excluding rescue HFOV use. The main fi nding of the authors was that the duration of mechanical ventilation was signifi cantly shorter in patients in whom HFOV was initiated on the day of surgery. Especially interesting about this work is that the authors used HFOV when there was evidence of pulmonary hypertension or right ventricular (RV) failure in their patients. This is an interesting approach as it is often assumed that high intra-thoracic pressures increase RV afterload and thus may enhance RV dysfunction. The fi ndings of Bojan and colleagues may be explained by the fact that they were able to decrease the pulmonary vascular resistance by fi nding a proper balance between atelectasis and overdistension of the lung. It can be argued that it is possible to do so by applying positive end-expiratory pressure. But, at the same time, this may coincide with the delivery of high inspiratory pressures (>30 cm H 2 O). As HFOV is, in fact, a continuous positive airway pressure system, its advantage is that it is possible to maintain suffi cient lung volume without large injurious pressure swings. Although the observations by Bojan and colleagues need to be confi rmed in a prospective randomized trial, they have provided arguments not to rule out the early use of HFOV in pediatric cardiac surgery patients.
initiated on the day of surgery. Th is is a very positive fi nding in a relatively large sample size.
However, there are some concerns related to the methodology of the study and the practical use of HFOV in the authors' institution. Some of the limitations inherent to their work have been properly addressed. But their study was designed as a retrospective one introducing confounding by indication. Hence, the authors have calculated a propensity score to minimize this, although the variables chosen for calculating it raise some concern. Included were all demographic and postoperative variables that yielded a P value of less than 0.10 in the univariate analysis between patients with HFOV and those without HFOV. However, it seems more rational to defi ne, beforehand, which variables infl uence the decision of the attending physician to transition to HFOV and therefore which variables should be used for calculating the propensity score. As a consequence of this approach, hemodynamic and ventilator parameters have not been used. If the authors had taken these into account when calculating their propensity score, the results of the study might have been even more fi rm. Also, it is not customary in their institution to routinely use an open lung strategy. Arterial partial pressure of oxygen (PaO 2 ) increases linearly with lung volume during HFOV [6]. It is necessary, therefore, to recruit the lungs after switching to HFOV. Animal work has shown improved lung compliance and less hyaline membrane formation when such a strategy was applied [7,8]. Th e eff ect of increased lung volume on PaO 2 is even more profound when oscillating on the defl ation limb of the pressure-volume curve [9][10][11]. Furthermore, there is more simultaneous alveolar recruitment and overstretching on the inspiratory than on the expiratory limb of the pressure-volume curve [12]. Th e hysteresis of the lungs provides an additional benefi t: when oscillating on the expiratory limb, less pressure is required to maintain lung volume in comparison with when oscillating on the inspiratory limb.
Nevertheless, what is especially interesting about the work of Bojan and colleagues is that they used HFOV when there was evidence of pulmonary hypertension or RV failure in their patients. Th is is an interesting approach as it is often assumed that high intra-thoracic pressures increase RV afterload and thus may enhance RV dysfunction. However, in a small study of 5 children after Fontan cardiac surgery [13], the use of HFOV did not result in an increase in pulmonary vascular resistance (PVR). So how can the positive fi ndings of the study by Bojan and colleagues be explained? It is appreciated that there is a relationship between the PVR and lung volume. Th is typically U-shaped curve shows that the PVR is high at both residual volume and total lung capacity [14]. Th e curve is at its optimum (that is, lowest PVR) at 'normal' functional residual capacity. It is not uncommon for ventilated patients after cardiac surgery to suff er from pulmonary dysfunction caused by atelectasis or even acute lung injury [15]. One of the main approaches in such disease conditions is to maintain suffi cient endexpiratory lung volume by the application of positive end-expiratory pressure (PEEP). However, the required level of PEEP is diffi cult to establish and may coincide with the delivery of high inspiratory pressures (>30 cm H 2 O). Th is is where HFOV comes into play. HFOV is, in fact, nothing more than a continuous positive airway pressure system with superimposed small oscillations delivering the small V T . As a consequence, it is possible to maintain suffi cient lung volume without large injurious pressure oscillations.
It is important to stress that the fi ndings of Bojan and colleagues warrant prospective confi rmation in a welldesigned randomized controlled trial. In the mean time, the authors are to be congratulated for their work as their study has provided arguments not to rule out the early use of HFOV in pediatric cardiac surgery patients.