PERFORMANCE COMPARISON OF 15 TRANSPORT VENTILATORS
gases and hemodynamics were evaluated to determine if the target blood gas values (PaO2 60–100 mm Hg, PaCO 30–50 mm Hg, pH 7.30–7.50) had been met, then the ventilator was adjusted as necessary to attempt to meet the targets. Once we determined whether the targets could be met, the next ventilator was attached to the animal’s air- way for evaluation. 2
We used 12 female Dorset sheep (21–31 kg), each fasted for 24 hours. Orotracheal intubation, with an 8-mm inner- diameter ETT, was performed during deep halothane an- esthesia via mask. The external jugular vein was then can- nulated, and an 8 French sheath introducer was inserted. After line placement, the anesthesia delivery was changed to intravenous only, with a loading dose of 10 mg/kg pentobarbital, 4 mg/kg ketamine, and 0.1 mg/kg pancuro- nium, followed by continuous infusion of pentobarbital (4 mg/kg/h), ketamine (8 mg/kg/h), and pancuronium (0.1 mg/kg/h) to provide surgical anesthesia with paraly- sis. After intubation, the basic ventilatory settings were volume control ventilation at a VT of 10 mL/kg, inspira- tory-expiratory ratio of 1:2, FIO2 of 1.0, and PEEP of 5 cm H2O, delivered by an intensive care ventilator (840, Puritan Bennett, Carlsbad, California). RR was adjusted to
achieve eucapnia (PaCO
35–45 cm H2O).
An 18-gauge catheter was then placed into the carotid artery for continuous measurement of arterial blood pres- sure and sampling of arterial blood gas values. Arterial and mixed venous blood samples were drawn for blood gas analysis. PO2, PCO2, pH, oxyhemoglobin saturation, and hemoglobin content were assessed with a blood gas ana- lyzer (282, Ciba Corning Diagnostics, Norwood, Massa- chusetts). Flow at the ETT was measured by a heated pneumotachometer (Hans Rudolph, Kansas City, Mis- souri) connected to a differential pressure transducer (MP-45 2 cm H2O, Validyne, Northridge, Califor- nia). Volume was determined via digital integration of the flow signal. A differential pressure transducer (MP-46 100 cm H2O, Validyne, Northridge, Califor- nia) was used to measure airway opening pressure. Car- diac output and pulmonary arterial pressure were mea- sured via a 7.5 French pulmonary artery catheter (831 HF 7.5, Edwards Life Sciences, Irvine, California) inserted into the left external jugular vein. Proper po- sition of the catheter was confirmed via pressure wave- form analysis before and after balloon occlusion. Fol- lowing instrumentation and a 30-min stabilization period, 5 transport ventilators were randomly applied.
All signals (flow at the ETT, airway opening pressure, arterial blood pressure, and pulmonary arterial pressure) were amplified (8805C, Hewlett Packard, Waltham, Mas- sachusetts), converted to digital signals with an analog-to-
digital converter (DI-220, Dataq Instruments, Akron, Ohio), and recorded at a sampling rate of 100 Hz, with data- acquisition software (Windaq/200, version 1.36, Dataq Instruments, Akron, Ohio). Ventilatory measurements made during the animal tests were all made at body-tem- perature-and-pressure-saturated conditions. All infusions, including the anesthetic, were given via volumetric infu- sion pump. A heating blanket was used to maintain a core temperature of 38–39°C. An orogastric tube was placed to empty the stomach.
Severe lung injury was produced with bilateral lung lavage via instillations of 1 L of isotonic saline, warmed to 39°C, repeated every 30 min, until PaO2 decreased to
100 mm Hg at an FIO2 of 1.0 and a PEEP of 5 cm H2O.
A stable lung injury was defined as a PaO2 change of 10% after 60 min. It took 2–4 lavages and 2–3 hours to estab- lish a stable lung injury. During development and stabili- zation of lung injury, the animals were ventilated with the Puritan Bennett 840 ventilator. After a stable lung injury was established, 5 transport ventilators were randomly ap- plied.
On completion of the protocol, the animals were sacri- ficed under deep anesthesia (10 mg/kg pentobarbital) with rapid infusion of 50 mL saturated potassium chloride so- lution. Electrocardiogram and arterial blood pressure read- ings confirmed cardiac standstill.
Formal statistical analysis was not performed. Lung model data were compared to the ventilator settings. A difference 10% was considered excessive, because most of the ventilator manufacturers indicate that the normal range of operation is within 10% of the set parameters. The mean SD VT was calculated from 5 breaths.
PIP, PEEP, and RR did not change with any ventilator during the bench evaluation. During the normal and in- jured-lung animal evaluations, the ability of each ventila- tor to achieve the target blood gas values was evaluated. The oxygen cylinder duration and battery life were re- corded in minutes.
The 15 ventilators evaluated can be classified as either “simple” or “sophisticated” transport ventilators, and as those that require compressed gas (pneumatic), those that can operate without compressed gas but require electrical power, and those that require both or either power source. The data in the tables and figures are organized with that
RESPIRATORY CARE JUNE 2007 VOL 52 NO 6