Twelve infants, median gestational age 27 weeks and postnatal age 1 day, were examined to determine whether oxygenation improves on transfer to high frequency oscillation (HFO). Lung volume was assessed before transfer to HFO by measuring functional residual capacity (FRC) using a helium gas dilution technique and specially designed infant circuit. On transfer to HFO, the inspired oxygen was initially kept constant, but the mean airway pressure (MAP) increased until maximum oxygenation was achieved (optimal MAP). The median FRC of the 12 infants before HFO was 8-1 ml/kg (range 4*7 to 28.7) and their median alveolararterial oxygen gradient (A-aDO2) 484 mm Hg. On transfer to HFO, oxygenation did not improve in two infants, but, overall, the A-aDO2 fell to a median of 289 mm Hg (p<005). The median optimal MAP was 18-5 cm H20 (range 106 to 24.4) and this had an inverse correlation with the Transfer to high frequency oscillation (HFO) can improve oxygenation in infants whose respiratory failure has responded poorly to conventional ventilation.' When a 'high volume' strategy is pursued the improvements in oxygenation seen when the mean airway pressure (MAP) is increased may be attributable to the opening up of atelectatic lungs and the maintenance of the lung above its closing volume.2 If that supposition is correct then infants with the lowest lung volume would require the greatest increase in MAP to optimise oxygenation. To test that hypothesis we assessed lung volume by measuring functional residual capacity (FRC) before transfer to HFO and related this to the absolute and change in MAP level necessary to maximise oxygenation.
MethodsInfants were transferred to HFO if their respiratory failure was deemed by the clinician in charge to be responding poorly to conventional ventilation. This was defined as a requirement for an inspired oxygen concentration (FIO2) of 0-5 or greater and/or a MAP of at least 10 cm H20 to maintain the pH between 7-25 and 7 40, arterial oxygen tension (Pao2) between 6-67 and 12-0 kPa (50 and 90 mm Hg), and arterial carbon dioxide tension (Paco2) between 4-67 and 6-67 kPa (35 and 50 mm Hg). Their respiratory status was confirmed by measuring a blood sample taken from an indwelling arterial line sited for clinical purposes.Immediately before transfer to HFO, FRC was estimated twice in each infant, with an interval of 10 minutes between measurements. Infants were measured in a supine position. A helium gas dilution technique and specially designed infant circuit with a circuit volume of 95 ml was used.3 The measurement technique has been described in detail before.4 The FRC system contains a rebreathing bag -the system reservoir -enclosed in an airtight cylinder. During measurement, the infant's endotracheal tube is connected to the rebreathing bag via a three-way valve. The ventilator is also connected to the three-way valve and the valve is connected to a side port on the airway cylinder. The infant can therefore be ventilated directly or, once the position of the three-way valve ...