During mechanical ventilation with an endotracheal tube (ETT) in place, the air conditioning functions of the nose and upper airways are bypassed. To ensure adequate water content and temperature of the gas reaching the trachea, these functions must be taken over by an external device.[1-3] For that purpose, it has been common for decades to use heated water-bath humidifiers (HWBH), which are external active sources of heat and water. A more recent solution is the combined heat and moisture exchanger filter (HMEF), which passively retains the heat and humidity, leaving the trachea during expiration, and recycles it during the next inspiration. HMEFs are designed to combine this air conditioning function with bacterial filtration. They have several possible advantages over HWBHs: ease of ventilatory circuit management, reduced cost,[4,5] and avoidance of potential complications associated with HWBHs, such as airways burns[6] and overhydration of the respiratory tract.[2,7]

However, while a properly set HWBH guarantees physiologic gas temperature and moisture at the tracheal level, this is not always true of HMEFs, the performances of which depend on a host of factors, including device type, external temperature, tidalvolume, and possible leaks around the ETT cuff.[8,9] Therefore, when using HMEFs, inappropriate conditioning of the inspired gas may be a concern, especially in the case of long-term mechanical ventilation. Potential risks include hypothermia and shivering induced by respiratory heat loss,[10] impaired mucociliary function,[11] structural damage to the respiratory epithelium,[12] and obstruction of airways or ETT by thickened secretions.[13-15]

Relatively few clinical data are available on the actual safety of HMEFS when employed for long-term mechanical ventilation. Some studies appear to substantiate the concerns expressed above,[13-15] but were conducted using HMEFS made of purely hydrophobic material. Although highly effective as antibacterial filters, these are presently known to have less than ideal performances as heat and moisture exchangers.[8,16-18] By contrast, the more recently marketed HMEFS, in which the air conditioning function is taken over by a hygroscopic membrane, are able to maintain the temperature and water content of inspired gas at levels closer to those achieved with an HWBH.[8,16-18] To our knowledge, safety data relevant to the long-term use of these newer devices in mechanically ventilated patients are limited to the very recent and carefully designed work by Dreyfuss and associates:[5] these authors recorded various clinical end points and found no adverse effect associated with the HMEF.

The present study had the following objectives: (1) to confirm the clinical safety of HMEFS for prolonged mechanical ventilation (ie, [is greater than] 48 h), and (2) to provide additional information regarding the integrity of tracheobronchial epithelium in such conditions.

MATERIALS AND METHOD

Study Design

This study took place during an 18-month period in the medical ICU of a large teaching hospital. Advantage was taken of a recent institutional decision to introduce HMEFs into the daily practice of respiratory care in this ICU, so that approval by the local ethical committee was not sought. During the study period, each patient with an indication for mechanical ventilation was considered for inclusion at the time of intubation or at ICU admission if already intubated. Subjects who were hypothermic (central or rectal temperature [is less than] 36 [degrees] C), or who had been intubated for 12 h before ICU admission were excluded.

Eligible patients were randomized to inhale air conditioned with either an HWBH (HWBH group) or an HMEF (HMEF group). The same conditioning method was strictly maintained throughout the time while receiving mechanical ventilatory assistance, unless crossover from HMEF to HWBH was deemed necessary by the clinical staff Crossover in the opposite direction was prohibited. Only patients who continued to receive ventilatory assistance for at least 48 h were retained for analysis.

Respiratory Equipment and Care

Intubation was performed via the orotracheal route, with tubes of 7.5 or 8 mm outer diameter. Three brands of ventilators were used: Servo 900c (Siemens-Elema; Solna, Sweden), MA-2 (Puritan-Bennett; Carlsbad, Calif), and Veolar (Hamilton Medical; Rhazuns, Switzerland).

In the HWBH group, the inspired air was water saturated to achieve a temperature of 32 [degrees] C at the Y-connector of ventilator tubing, using a standard device (Fisher Paykel; Auckland, New Zealand, or Puritan-Bennett). In the HMEF group, unconditioned gas was passed through a recently marketed HMEF (Hygroster; DAR; Mirandola, Italy) inserted between the ETT and the Y-connector. The HMEF consisted of a hygroscopic membrane for heat and moisture exchange (cellulose, patient's side of the device), and of a hydrophobic pleated membrane for bacterial filtration (fiberglass, ventilator side). In critically ill ventilated patients, the average temperature and humidity outputs of this device were previously documented at 30 mg [H.sub.2]O per liter of inspired air and 32 [degrees] C, respectively.[8] The advertised dead space added to the circuit by the HMEF we used (Hygroster) is 95 mL.