Clin Perinatol 34 (2007) 1933

Respiratory Gas Conditioning

and Humidication
Andreas Schulze, MDa,b,*
a

Division of Neonatology, Dr. von Hauner Childrens Hospital, Munich, Germany

Department of Obstetrics and Gynecology, Klinikum Grosshadern, Ludwig Maximilian
University, Marchioninistr. 15, D-81377, Munich, Germany

Although there have been great advances in our understanding of pulmonary pathophysiology in the management of neonatal lung disorders, the
proper humidication and conditioning of delivered gas remains a mystery
to most clinicians managing these devices. This article attempts to unravel
the mystery and describe the physiologic bases for humidication and gas
conditioning.

Basic physics of heat and humidity

Medical grade gases have virtually no water content at room temperature. There are three options to deliver water into inspired gas or directly
into the airways:
Vaporization of water using heated humidiers or heat and moisture exchangers (articial noses).
Nebulization of water using jet or ultrasonic devices.
Periodic instillation of bolus water or normal saline solution into the endotracheal or tracheostomy tube as is commonly done prior to suctioning procedures.
Fundamental dierences exist between nebulized and vaporized water.
Nebulization creates a dispersion of small droplets of water in air. These particles may vary in size from about 0.5 to 5 mm. They are visible as mist because
they scatter light and may carry infectious agents or other particulate matter.
* Division of Neonatology and Department of Obstetrics and Gynecology, Klinikum
Grosshadern, Ludwig Maximilian University, Marchioninistr. 15, D-81377 Munich,
Germany.
E-mail address: andreas.schulze@med.uni-muenchen.de
0095-5108/07/$ - see front matter 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.clp.2006.12.009
perinatology.theclinics.com

20

SCHULZE

In contrast, vaporization generates a molecular (ie, gaseous) distribution of

water in air. Water vapor is thus invisible and unable to carry infectious
agents. It exerts a gaseous pressure, which amounts to a partial pressure of water of 47 mm Hg when air is fully saturated with water vapor at 37 C. This corresponds to a water vapor mass of 44 mg of water per liter of gas. The partial
pressure of water vapor at saturation depends solely on the temperature. The
fraction of water vapor pressure at saturation, 37 C, and 760 mm Hg ambient
pressure is therefore FH2 O 47 mm Hg=760 mm Hg 0:062 6:2%. The
term absolute humidity (AH) is dened as the amount of water vapor (in milligrams) per gas volume (in liters) at a given temperature. Relative humidity
(RH) is the actual water vapor content of the gas volume (in milligrams) relative to the water vapor content (in milligrams) of this same gas volume at saturation at the same temperature. There is a xed relationship between AH,
RH, and temperature (Fig. 1).
Air takes up energy when nebulized water or liquid water is converted
into water vapor. Conversely, heat is generated in the process of rainout
of water vapor (condensation). There are two components to the total energy content of air: a sensible and a latent heat content. The air temperature
solely reects the former, whereas the water vapor mass reects the latter. It
is important to understand that changing the air temperature alone without
changes in water vapor mass constitutes a small change in total energy content compared with changing the humidity of the gas. The dierence may
approach an order of magnitude. Warming of frigid inspiratory air does little cooling to the upper airway lining compared with the amount of heat loss
that occurs when this air is particularly dry. Conversely, if air is inhaled at
39 C without being fully saturated with water vapor, it quickly cools to core
body temperature without signicant heating of the airway lining. The

40

50
44 mg/L

37

40

35
30 mg/L

30

30
25

20
10

Temperature (centigrade)

Absolute Humidity (mg/L)

60

0
65

75

85

Relative Humidity (

95

Fig. 1. The relative humidity of a gas depends on its absolute water content and gas temperature. At 37 C and 100% relative humidity, the respiratory gas has 44 mg/L absolute water content. If the gas is saturated (100% relative humidity) at 30 C, its water content is only 30 mg/L.
When the gas is then warmed to 37 C, its relative humidity falls to less than 70%.

RESPIRATORY GAS CONDITIONING AND HUMIDIFICATION

21

following example illustrates this issue quantitatively. If air is saturated with

water vapor at 37 C in a humidier chamber and subsequently dry heated to
39 C in a heated breathing circuit, this dry heating adds almost no energy to
the gas. It contains 143 J/g at 37 C, 100% relative humidity versus 145 J/g at
39 C, 90% relative humidity. If, however, air enters the airway at much
higher than core body temperature and full saturation, there may be a risk
for thermal injury.
Structure and function of the airway lining
The lining of the respiratory tract consists of three layers: a basal cellular layer,
an aqueous (sol) layer, and a viscoelastic gel (mucus) layer at the luminal surface.
Most of the upper airways and the entire tracheobronchial tree down to the respiratory bronchioles are covered by ciliated epithelial cells. Each cell carries
about 200 cilia at its apex. The cilia are 5 to 6 mm in length and carry a crown
of short claws at their tip. When extended, the cilia may reach through the aqueous layer and get their claws entangled with the macromolecular network of the
mucus, which oats at the surface of the aqueous layer. A movement cycle of
a single cilium consists of two parts. During the eective stroke, the fully extended cilium moves in a plane approximately perpendicular to the cell surface
and pulls the mucus layer forward. It then bends down toward the cell surface,
thereby disengaging from the gel layer, and swings back near the surface of
the cell to reach its starting position (recovery stroke through a less viscous medium). Under normal circumstances, the orientation of the eective strokes is always such that the mucus is propelled in a cephalad direction. Neighboring cilia
beat in a coordinated fashion so that waves of aligned cilia move through the airway lining uid. There are various other specic cell types, such as serous cells
and Clara cells, in the cellular layer of the airway lining, each of which may be
concerned with a specic function [1]. Mucus is secreted by goblet cells and submucosal mucous glands as globules of 1 to 2 mm in diameter. They absorb water
from the aqueous layer and subsequently swell rapidly. Mucus consists of glycoproteins, proteoglycans, lipids, and 95% water [2]. It forms small akes [3], larger
plaques, or blanket-like covers [4,5] that oat on the luminal surface and trap inhaled particles and other compounds, such as bacteria, macrophages, and cell debris [6]. Experimental evidence exists that mucus secretion is stimulated when
particles come into contact with the ciliated surface. Such particles are subsequently encapsulated by mucus and carried away by the ciliary escalator [7].
The structure and function of the various components of this intricate system
may easily become disturbed when the inspired gas is inadequately conditioned
in infants who have an articial airway.
Inspired gas conditioning in the respiratory tract
The alveolar air is fully saturated with water vapor at core body temperature, whereas ambient air is cooler and contains less water. This gradient in

22

SCHULZE

heat and water vapor pressure is maintained under physiologic conditions

along the nose and upper airways [8]. They function as a counter-current
heat and moisture exchanger. The inspired air gains heat and water vapor
from the upper airway lining, which is partly recovered when the expired
gas loses heat and water condenses back to the airway surface. This recovery
occurs because the upper airway temperature remains below core body temperature during expiration. Breathing is associated with a net heat and water
loss because the expired air temperature is higher than ambient temperature
under normal circumstances [9]. The greater this dierence in temperature between the inspired and expired gas, the higher are the losses. They must be replenished by the airway epithelium, which in turn is supplied by the bronchial
circulation. It is entirely unclear under which circumstances the capacity of the
airway lining to humidify cold and dry gas becomes overcharged. This capacity is likely dierent in healthy and diseased states. Water transport through
the mucosa itself into the aqueous layer of the airway lining is possibly rate limiting. Another potential limitation of water supply to the airway lining may result from decreased blood perfusion to the mucosa. It was suggested that this
occurs when cold air is inhaled and mucosal blood vessels constrict in a response similar to the reduction in skin perfusion in a cold environment [10].
It could be argued teleologically that keeping the upper airway cool during expiration by restricting blood ow facilitates heat and water recovery. It has
been shown in hyperventilated anesthetized dogs, however, that tracheobronchial blood ow increases with breathing cold or warm dry air [11]. This
response is not mediated by the autonomic nervous system [12].
The level at which the inspired air reaches core body temperature and full
saturation with water vapor is called the isothermic saturation boundary
[13]. It is located at the level of the main stem bronchi in the adult during normal quiet breathing of room air [14]. Its position varies with the heat and moisture content of the inspired air and depends on the pattern of breathing. For
example, it moves cephalad with slow and shallow breathing [15], and oral
breathing moves it downward compared with nasal breathing [16]. Incompletely conditioned air may penetrate deep into the distal parts of the tracheobronchial tree during high minute ventilation of frigid air because of the large
volume of air to be conditioned and the short residence time at any given point
[14,17]. Bypassing the upper airway (for example, by using a tracheostomy
tube) also is associated with a much greater burden for humidication on
the distal airways. Overall, however, under normal physiologic circumstances
only a small segment of the airway surface is exposed to a temperature below
core body temperature and less than full saturation with water vapor.
Risks of using poorly conditioned medical gases in infants who have an
articial airway
The optimal temperature and humidity of the inspired gas for infants undergoing mechanical ventilation has been a matter of controversy. Also, the

RESPIRATORY GAS CONDITIONING AND HUMIDIFICATION

23

minimal acceptable level of temperature and humidity has not been clearly
established by clinical studies. International and United States standards
recommend an absolute humidity level of inspired gas greater than 33 mg/L
in patients whose supraglottic airway is bypassed [1820]. There is no
doubt that inadequate humidication may lead to progressive airway dysfunction and systemic eects, depending on the degree of under-humidication and coolness, the exposure time, and the underlying disease.
The mucociliary transport system is probably the most sensitive respiratory function to changes in inspired gas humidity and temperature [21].
Transport velocity depends on mucus rheology, the depth of the aqueous
airway lining layer, and cilia beat frequency. Dry inspired gas changes the
viscosity gradient of the airway lining. Dehydrated mucus slows the transport rate and cilia beat frequency. Thinning of the aqueous layer impairs
the recovery stroke of cilia [22,23]. The mucociliary machinery may recover
with rehumidication after short periods of desiccation. Cessation of cilia
activity for 3 hours is irreversible, however, and is followed by inammation
and sloughing of the mucosa [24]. Such damage deprives the upper airway of
its function as a heat and moisture exchanger. Subsequently, the isothermic
saturation boundary retreats into smaller airways and the area of damaged
mucosa progressively extends into the periphery of the lung.
Because the depth of the airway lining uid is only about 7 to 12 mm [25],
water loss by evaporation may increase the osmolarity. This increase may
induce bronchial smooth muscle contraction in patients who have exercise-induced asthma independently of the inhaled air temperature [2629].
The mechanism by which a change in the osmotic environment can lead
to bronchoconstriction is unknown. Airway irritant receptors have been implemented as possible mediators. It is also unknown whether a critical
threshold of respiratory water loss exists to induce a stimulus for bronchoconstriction, or whether this mechanism occurs in preterm infants who have
chronic lung disease.
In a retrospective study, mechanically ventilated infants weighing less
than 1500 g at birth had signicantly more air leaks and more severe chronic
lung disease if exposed during their rst 4 days of life to inspired gas less
than 36.6 C and with less than 37 mg H2O/L [30]. This association could
not be corroborated for more mature infants, suggesting that the resistance
to poor inspiratory gas conditioning might be a function of gestational age.
The cause of necrotizing tracheobronchitis in mechanically ventilated infants and adults has not been fully elucidated, but it is rational to assume
that inadequate humidication is a signicant contributor [3133].
The mucociliary elevator moves secretions up to the tip of the endotracheal tube. This material accumulates and resides there for a variable period
of time awaiting removal by suctioning. It is subjected to desiccation by any
inspiratory gas leaving the endotracheal tube with a capacity to accommodate more water vapor. Inspissation is faster the lower the humidity and the
higher the airow at the endotracheal tube outlet. High minute ventilation

24

SCHULZE

and an endotracheal tube leak hasten this process. It has been suggested that
humidity levels below a critical threshold of 31 mg H2O/L are associated
with a high risk for endotracheal tube plugging in infants [34]. At least theoretically, however, this does not imply that any humidity level above this
threshold and below full saturation at core body temperature is safe in
regard to upper airway obstruction by inspissated secretions.
Mechanical ventilation with dry gas results in a mean decrease in rectal
temperature by 1.4 C within one hour in neonates. Even moderate warming
to 31.5 C and humidifying the inspiratory gas reduces the insensible water
loss in intubated preterm infants signicantly and consistently to levels
below those of extubated infants in room air at 27 C [35].
Severe underhumidication nally leads to impaired surfactant activity,
decreased functional residual capacity, atelectasis, and compromised pulmonary mechanics [36].

Devices for inspiratory gas conditioning and humidication procedures

Heated humidiers
For various physiologic reasons, it is rational to deliver the inspiratory
gas to endotracheally intubated infants at or close to core body temperature
and full saturation. Heated humidiers (Fig. 2) can achieve this goal. The
respiratory gas is warmed inside the humidication chamber to a set target
temperature and water vapor is added from the heated water reservoir. Inspiratory circuit tubing containing a heated wire is then used to maintain or
slightly raise the gas temperature to prevent water rainout before the gas
reaches the infant. At a set humidier chamber temperature of 37 C, the
gas absorbs 44 mg/L of water, which corresponds to full saturation
(100% RH) unless the vaporizing capacity is low relative to the size of the
gas ow. The vaporizing capacity of the humidication chamber depends
on its water surface area and temperature. Recording the water consumption of the chamber over time is a simple test to check for sucient vaporization. Because most infant ventilators use continuous circuit ow of
known rate, the absolute and relative humidity delivered at the chamber
outlet can be calculated from the humidiers water consumption rate.
Any decrease in gas temperature along the way from the humidication
chamber to the wye adaptor induces condensation if the gas was saturated
at the chamber outlet. This implies that if the chamber temperature was set
at or below 37 C, any rainout in the tubing indicates moisture loss from the
respiratory gas. The gas reaching the infant is underhumidied. It must be
emphasized that rainout in the inspiratory limb of the ventilator circuit
does not indicate proper humidication in such situations. To the contrary,
condensation is necessarily associated with underhumidication. The inspiratory gas can rapidly cool in unheated segments of the circuit. This cooling
is promoted by the large outer surface area of small diameter tubing

RESPIRATORY GAS CONDITIONING AND HUMIDIFICATION

5 L/min

Respirator

25

Incubator

5 L/min

5 L/min
= temperature probes

Humidifier
Chamber
Maximum water consumption
5 L/min x 44 mg H2O/L
= 220 mg/min = 13.2 mg/h
= 13.2 mL/h

Fig. 2. Position of three temperature probes of a heated-wire humidication system for infants.
The user sets the target temperature to be reached at the endotracheal tube adaptor. This
temperature is commonly set at or slightly greater than 37 C. The temperature inside the
humidier chamber must be high enough to vaporize an amount of water near the absolute
water content of gas saturated at 37 C (44 mg/L). The water consumption rate of a humidier
chamber required to reach a target respiratory gas humidity can be calculated from the circuit
ow rate. Observation of this water consumption rate can be used as a simple test of the
eciency of a humidier.

(particularly when corrugated), by drafts around the tubing (air conditioned

rooms), and by low room temperatures. The decrease in temperature is
larger with smaller circuit gas ow rates because of the longer contact
time. Insulating unheated segments of the inspiratory circuit may partly
obviate these problems. Rainout should also be avoided for other reasons.
Condensate is easily contaminated, may be ushed down the endotracheal
tube (with risks of airway obstruction and nosocomial pneumonia), and
may disturb the function of the ventilator. In a regular application for an
intubated subject, therefore, a heated humidier should be set up with
a chamber temperature of 37 C to saturate the gas with 44 mg/L of water.
To avoid loss of moisture in the inspiratory limb of the heated circuit, the
target gas temperature at the wye adaptor can be set at 39 C so that the
gas arrives with slightly less than full saturation. The gas quickly cools to
core body temperature inside the adapter and the endotracheal tube.
The technology required for preterm infants is slightly more complex because the ventilator circuit passes through two dierent environments: the
room and the incubator or radiant warmer. The temperature probe close
to the patient connection serves to monitor the respiratory gas temperature.
It is commonly part of a servo-control, which aims to maintain the set gas
temperature at the wye adaptor by controlling the circuits heated wire
power output. If the temperature probe is in the presence of a heated eld,
it may register a temperature higher than the actual respiratory gas temperature as a result of radiation or convection from the warmer environment.
This may signal to the servo to decrease the heating output of the ventilator

26

SCHULZE

circuit and lead to loss of gas temperature and rainout. Such a problem may
arise when the incubator temperature is higher than the targeted gas temperature or from a radiation heat source. Insulating the temperature probe by
a light reective patch or other material can improve the performance of the
system. Another way to alleviate this problem is to place the temperature
probe just outside the heated eld and use an unheated extension adaptor
tubing to carry the gas through the heated eld to the infant. The extension
tube does not need to incorporate heated wires because its temperature is
maintained by the heated eld. If cooler incubator temperatures are used
(as usually used for older preterm infants), rainout occurs in the unheated
segment, particularly at low circuit gas ow rates. A circuit should then
be used that is equipped with a heated wire along the entire length of its inspiratory limb. Another suitable type of circuit is that with two temperature
probes, one outside the heated eld and the other close to the wye adaptor.
These circuits can perform well over a range of incubator temperatures both
greater than and less than the target respiratory gas temperature. This is because the heated wire servo-control can be programmed to select the lower
of the two recorded temperatures to drive the power output. The maximum
heat output of any heated wire circuit may not be sucient to meet target
gas temperatures under extremes of room and incubator temperatures.
Also, generic circuits have been on the market that may not be fully compatible with the humidier and its power source. There has been a warning that
covering heated wire circuits with drapes or other material for insulation
may involve a risk for melting or charring of circuit components [37].
Articial noses
Heat and moisture exchangers (HMEs) are designed to recover part of
the heat and moisture contained in the expired air. A sponge material of
low thermal conductivity inside the clear plastic housing of these devices
absorbs heat and condenses water vapor during expiration for subsequent
release during inspiration. HMEs are an attractive alternative to heated humidiers for several reasons, such as simplication of the ventilator circuit,
passive operation without requirement of external energy and water sources,
no ventilator circuit condensate, low risk for circuit contamination [38], and
low expense. Additionally, some HMEs are coated with bacteriostatic substances and equipped with bacterial or viral lters (HMEF). Devices called
hygroscopic condenser humidiers (HCHs) use hygroscopic compounds,
such as CaCl2, MgCl2, LiCl, or others, to increase the water retention capacity. Small HME/HCHs for neonatal applications are commercially available, but data on their use are sparse. Theoretically, an application in
small subjects may be particularly eective because mean delivered inspiratory humidity increases inversely with tidal volume [39]. Also, the HME
membranes may become saturated before a large volume expiration has
been completed, which increases respiratory water loss. A pediatric HME

RESPIRATORY GAS CONDITIONING AND HUMIDIFICATION

27

maintained an average inspiratory humidity greater than 30 mg/L in a clinical study involving neonates for a test period of 6 hours [40]. Other clinical
studies on pediatric and neonatal application conrmed the ability of HME/
HCHs to conserve heat and to provide humidity levels that are appropriate
for short-term conventional mechanical ventilation [4143]. Bench studies
on high-frequency oscillatory ventilation in a neonatal lung model showed
that a neonatal HME was able to provide more than 35 mg/L of mean humidity at the proximal end of the endotracheal tube adapter. The HME
dampened the oscillatory pressure amplitude less than a neonatal endotracheal tube of 3.5 mm inner diameter [44]. The safety and eectiveness of
HME/HCH for long-term mechanical ventilation is controversial in adults
[4549] and has not been established in infants. Depending on their actual
water load and duration of use, HMEs add a variable resistance and dead
space to the circuit [5052]. A risk for airway occlusion from clogging
with secretions or from a dislodgement of HME internal components [53]
has been reported for infants even during short-term application. Also, an
expiratory air leak impairs the barrier eect against moisture loss [54,55].
HME must not be used in conjunction with heated humidiers, nebulizers,
or metered dose inhalers, which may cause a hazardous increase in device
resistance [56] or wash o the hygroscopic coating [57]. Dierent brands
of HME/HCHs may vary widely in their performance characteristics. Their
eectiveness is not reliably reected by indirect clinical measures, such as the
occurrence of nosocomial pneumonia, number of endotracheal tube occlusions, or frequency of tracheal suctioning and instillation [58]. Visual evaluation of the amount of moisture in the adapter segment between the
endotracheal tube and the HME/HCH was found to closely correlate with
objective measurements of the delivered humidity, however [59]. Device performance has improved much during recent years, and further advances can
be expected to facilitate neonatal applications. Microprocessor-controlled
active release of additional external heat and water into the airway between
an HME and the endotracheal tube adapter increased the inspired gas humidity to 100% at 37 C without obvious untoward eects in 24-hour studies
in adults. Such hybrid systems, however, require further study and renement [60].
Aerosol application
Aerosol water particles that range in size from about 1 to 10 mm may deposit on the airway by impaction (larger particles) or sedimentation (smaller
particles). Sedimentation occurs as a gravitational eect when airow velocity declines in the smaller airways. An aerosol cannot contribute to respiratory gas conditioning downstream to the isothermic saturation boundary
because the gas is already fully saturated. For this same reason, aerosol water particles cannot be eliminated in this airway region through evaporation
and exhalation. They therefore become a water burden on the mucosa that

28

SCHULZE

needs to be absorbed by the airway epithelium to maintain an appropriate

periciliary uid depth. An increase in depth of the airway linings aqueous
layer may make it impossible for the cilia to reach the mucous layer and
thus impair mucus transport. Furthermore, if the aerosol deposition rate
exceeds absorption capacity, this may lead to increased airway resistance
[61,62] and possibly narrowing or occlusion of small airways. Severe systemic overhydration subsequent to ultrasonic aerosol therapy has been described in a term newborn infant [63], and similar occurrences have been
reported in adults [64]. If an aerosol stream meets the airway proximal to
the isothermic saturation boundary, the particulate water can theoretically
contribute to the gas conditioning process by evaporation before and after
deposition. The droplets, however, contain only sensible heat and the mucosa needs to supply most of the latent heat for vaporization. This process
cools the airway. If the inspiratory gas is supplied through the endotracheal
tube at body core temperature and close to full saturation, the isothermic
saturation boundary is located near the tip of the endotracheal tube. In
such a situation, any aerosol may contribute little, if anything, to appropriate gas conditioning. Water or normal saline nebulization, therefore, seems
to oer no signicant benet for inspiratory gas conditioning and may entail
a risk for overhumidication [65].
Irrigation of the airway
It is common clinical practice to instill small amounts of water or normal
saline solution into the endotracheal tube before suctioning procedures in
the belief that this provides moisture and loosens tenacious secretions. A randomized controlled trial showed a lower endotracheal tube resistance and
fewer hemorrhagic secretions when periodic bolus instillation of normal saline
solution was used before suctioning in adults on mechanical ventilation with
HMEs [66]. It has been shown, however, that more than 80% of the instillate
may remain inside the airway after suctioning and probably later is absorbed
or removed by the mucociliary system [67]. Suggested amounts of uid to be
used in infants vary widely from 0.1 mL/kg to 0.5 mL [37]. The safety and effectiveness of this practice, under conditions of appropriate warming and humidication of respiratory gases, remain dubious [68].

Inspiratory gas conditioning and the nosocomial infection risk

Water vapor as such cannot transmit an infection. It has been shown,
however, that cold water and hot water humidier chambers, reservoirs of
nebulizers, ventilator circuits, and circuit condensate may all become colonized with infectious agents [6971]. Theoretically, the airway may then become contaminated from colonized condensate ushed inadvertently into
the airway, from colonized aerosol particles, and directly from a colonized
circuit. A bubble-through humidier chamber may not only produce water

RESPIRATORY GAS CONDITIONING AND HUMIDIFICATION

29

vapor but also some aerosol capable of dispersing infectious particles. The
colonization risk may be reduced by the use of sterile closed delivery systems
or by maintaining a water reservoir temperature greater than 60 C (pasteurization) [36]. It has been shown in adults, however, that such colonization
usually originates from the patients own respiratory ora and may occur
within a few hours of connecting a sterile circuit [72,73]. It was subsequently
observed that the incidence of nosocomial pneumonia in adults was not increased when ventilator circuits were changed less frequently than every
24 hours or even between patients [50,74,75]. Nevertheless, although these
studies indicate that ventilator circuit changes may be extended to more
than 48 hours, the optimal rate of circuit changes for infants is unknown.
Changing a ventilator circuit is not a benign procedure because it may disrupt ventilation in a potentially dangerous way, and medical personnel may
become a vector for cross-contamination between patients [76]. Arguments
can therefore be made for weekly circuit changes or for no circuit changes at
all, except between patients.
In 2005, a meta-analysis of randomized controlled trials found a protective eect against ventilator-associated pneumonia in adults with the use of
HMEs compared with heated humidiers, particularly in patients ventilated
for 7 days or longer [77]. The size of this eect, however, is likely small; it
could not be corroborated in two more recent studies on a large number
of patients [50,78].

Summary
There is a strong physiologic rationale for delivering the inspiratory gas
at or close to core body temperature and saturated with water vapor to infants who have an articial airway undergoing longer-term mechanical ventilatory assistance. Cascade humidiers with a heated wire ventilatory circuit
may achieve this goal safely. Whenever saturated air leaves the humidier
chamber at 37 C and condensate accumulates in the circuit, the gas loses humidity and acquires the potential to dry airway secretions near the tip of the
endotracheal tube. Heat and moisture exchangers and hygroscopic condenser humidiers with or without bacterial lters have become available
for neonates. They can provide sucient moisture output for short-term
ventilation without excessive additional dead space or ow-resistive loads
for term infants. Their safety and ecacy for very low birthweight infants
and for long-term mechanical ventilation have not been established conclusively. A broader application of these inexpensive and simple devices is
likely to occur with further design improvements. When heated humidiers
are appropriately applied, water or normal saline aerosol application oers
no additional signicant advantage in inspiratory gas conditioning and may
impose a water overload on the airway or even systemically. Although airway irrigation by periodic bolus instillation of normal saline solution before

30

SCHULZE

suctioning procedures is widely practiced in neonatology, virtually no data

exist on its safety and ecacy when used with appropriately humidied inspired gas. There is no evidence that conditioning of inspired gas to core
body temperature and full water vapor saturation may promote nosocomial
respiratory infections.
References
[1] Jeery PK, Reid LM. The respiratory mucus membrane. In: Brain JD, Proctor DF, Reid
LM, editors. Respiratory defense mechanisms. Part I. New York: Marcel Dekker; 1977.
p. 193245.
[2] Yeager H. Tracheobronchial secretions. Am J Med 1971;50:493501.
[3] Iravani J, van As A. Mucus transport in the tracheobronchial tree of normal and bronchitic
rats. J Pathol 1972;106:8193.
[4] Aiello E, Sleigh MA. Ciliary function of the frog oropharyngeal epithelium. Cell Tissue Res
1977;178:2678.
[5] Puchelle E, Petit A, Adnet JJ. Fine structure of the frog palate mucociliary epithelium. J Submicrosc Cytol 1984;16:27382.
[6] Brain JD, Godlaski JJ, Sorokin SP. Quantication, origin and fate of pulmonary macrophages. In: Brain JD, Proctor DF, Reid LM, editors. Respiratory defense mechanisms.
Part II. New York: Marcel Dekker; 1977. p. 84992.
[7] Spungin B, Silberberg A. Stimulation of mucus secretion, ciliary activity and transport in
frog palate epithelium. Am J Physiol 1984;247:C299308.
[8] Rouadi P, Baroody FM, Abbott D, et al. A technique to measure the ability of the human
nose to warm and humidify air. J Appl Physiol 1999;87:4006.
[9] Walker JEC, Wells RE, Merrill EW. Heat and water exchange in the respiratory tract. Am J
Med 1961;30:25967.
[10] McFadden ER. Respiratory heat and water exchange: physiological and clinical implications. J Appl Physiol 1983;54:3316.
[11] Baile EM, Dahlby RW, Wiggs DR, et al. Role of tracheal and bronchial circulation in respiratory heat exchange. J Appl Physiol 1985;58:21722.
[12] Baile EM, Osborne S, Pare PD. Eect of autonomic blockade on tracheobronchial blood
ow. J Appl Physiol 1987;62:5205.
[13] Dery R, Pelletier J, Jacques A, et al. Humidity in anaesthesiology. 3. Heat and moisture patterns in the respiratory tract during anaesthesia with the semi-closed system. Can Anaesth
Soc J 1967;14:28798.
[14] McFadden ER, Pichurko BM, Bowman HF, et al. Thermal mapping of the airways in
humans. J Appl Physiol 1985;58:56470.
[15] Dery R. The evolution of heat and moisture in the respiratory tract during anaesthesia with
a non-rebreathing system. Can Anaesth Soc J 1973;20:296309.
[16] Primiano FP, Saidel GM, Montague FW, et al. Water vapor and temperature dynamics in
the upper airways of normal and CF subjects. Eur Respir J 1988;1:40714.
[17] McFadden ER, Dension DM, Waller JR, et al. Direct recordings of the temperatures in the
tracheobronchial tree in normal man. J Clin Invest 1982;69:7005.
[18] American Society for Testing and Materials. Standard specication for humidiers for
medical use. ASTM F1690-96 (2004).
[19] Miyoshi E, Fujino Y, Uchiyama A, et al. Eects of gas leak on triggering function, humidication, and inspiratory oxygen fraction during noninvasive positive airway pressure
ventilation. Chest 2005;128:36918.
[20] International Organization for Standardization. Humidiers for medical use. ISO 8185
(1997).

RESPIRATORY GAS CONDITIONING AND HUMIDIFICATION

31

[21] Williams R, Rankin N, Smith T, et al. Relationship between the humidity and temperature of
inspired gas and the function of the airway mucosa. Crit Care Med 1996;24:19209.
[22] Puchelle E, Zahm JM, Quemada D. Rheologic properties controlling mucociliary frequency
and respiratory mucus transport. Biorheology 1987;24:55763.
[23] Sleigh MA, Blake JR, Liron N. The propulsion of mucus by cilia. Am Rev Respir Dis 1988;
137:72641.
[24] Hirsch JA, Tokayer JL, Robinson MJ, et al. Eects of dry air and subsequent humidication
on tracheal mucus velocity in dogs. J Appl Physiol 1975;39:2426.
[25] Kilburn KH. A hypothesis for pulmonary clearance and its implications. Am Rev Respir Dis
1968;98:44963.
[26] Anderson SD, Schoeel RE, Black JL, et al. Airway cooling as the stimulus to exerciseinduced asthmada re-evaluation. Eur J Respir Dis 1985;67:2030.
[27] Anderson SD, Schoeel RE, Follet R, et al. Sensitivity to heat and water loss at rest and during exercise in asthmatic patients. Eur J Respir Dis 1982;63:45971.
[28] Freed AN, Anderson SD, Daviskas E. Thermally induced asthma and airway drying. Am J
Respir Crit Care Med 2000;161:21123.
[29] Hahn A, Anderson SD, Morton AR, et al. A reinterpretation of the eect of temperature and
water content of the inspired air in exercise-induced asthma. Am Rev Respir Dis 1984;130:
5759.
[30] Tarnow-Mordi WO, Reid E, Griths P, et al. Low inspired gas temperature and respiratory
complications in very low birth weight infants. J Pediatr 1989;114:43842.
[31] Boros SJ, Mammel MC, Lewallen PK, et al. Necrotizing tracheobronchitis: a complication
of high frequency ventilation. J Pediatr 1986;109:95100.
[32] Circeo LE, Heard SO, Griths E, et al. Overwhelming necrotizing tracheobronchitis due
to inadequate humidication during high-frequency jet ventilation. Chest 1991;100:
2689.
[33] Hanson JB, Waldstein G, Hernandez JA, et al. Necrotizing tracheobronchitis. Am J Dis
Child 1988;142:10948.
[34] Lomholt N, Cook R, Lunding M. A method of humidication in ventilator treatment of
neonates. Br J Anaesth 1968;40:33540.
[35] Sosulski R, Polin RA, Baumgart S. Respiratory water loss and heat balance in intubated
infants receiving humidied air. J Pediatr 1983;103:30710.
[36] Shelly MP, Lloyd GM, Park GR. A review of the mechanisms and methods of humidication of inspired gases. Intensive Care Med 1988;14:19.
[37] Quinn W, Sandifer L, Goldsmith JP. Pulmonary care. In: Goldsmith JP, Karotkin EH,
editors. Assisted ventilation of the neonate. 3rd edition. Philadelphia: W.B. Saunders;
1996. p. 10123.
[38] Dreyfuss D, Djedaini K, Gros I, et al. Mechanical ventilation with heated humidiers or heat
and moisture exchangers: eects on patient colonization and incidence of nosocomial pneumonia. Am J Respir Crit Care Med 1995;151:98692.
[39] Mebius C. A comparative evaluation of disposable humidiers. Acta Anaesthesiol Scand
1983;27:4039.
[40] Schimann H, Rathgeber J, Singer D, et al. Airway humidication in mechanically ventilated neonates and infants: a comparative study of a heat and moisture exchanger vs. a heated
humidier using a new fast-response capacitive humidity sensor. Crit Care Med 1997;25:
175560.
[41] Bissonnette B, Sessler DI. Passive or active inspired gas humidication increases thermal
steady-state temperatures in anesthetized infants. Anesth Analg 1989;69:7837.
[42] Bissonnette B, Sessler DI, LaFlamme P. Intraoperative temperature monitoring sites in infants and children and the eect of inspired gas warming on esophageal temperature. Anesth
Analg 1989;69:1926.
[43] Bissonnette B, Sessler DI, LaFlamme P. Passive and active inspired gas humidication in
infants and children. Anesthesiology 1989;71:3504.

32

SCHULZE

[44] Schimann H, Singer S, Singer D, et al. Determination of airway humidication in high-frequency oscillatory ventilation using an articial neonatal lung model. Comparison of a heated
humidier and a heat and moisture exchanger. Intensive Care Med 1999;25:9971002.
[45] Branson RD, Davis K Jr, Campbell RS, et al. Humidication in the intensive care unit. Prospective study of a new protocol utilizing heated humidication and a hygroscopic condenser
humidier. Chest 1993;104:18005.
[46] Kollef MH, Shapiro SD, Boyd V, et al. A randomized clinical trial comparing an extendeduse hygroscopic condenser humidier with heated-water humidication in mechanically
ventilated patients. Chest 1998;113:75967.
[47] Misset B, Escudier B, Rivara D, et al. Heat and moisture exchanger vs. heated humidier
during long-term mechanical ventilation. A prospective randomized study. Chest 1991;
100:1603.
[48] Nakagawa NK, Macchione M, Petrolino HM, et al. Eects of a heat and moisture exchanger
and a heated humidier on respiratory mucus in patients undergoing mechanical ventilation.
Crit Care Med 2000;28:3127.
[49] Ricard JD, Le Miere E, Markowicz P, et al. Eciency and safety of mechanical ventilation
with a heat and moisture exchanger changed only once a week. Am J Respir Crit Care Med
2000;161:1049.
[50] Boots RJ, George N, Faoagali JL, et al. Double-heater-wire circuits and heat-and-moisture
exchangers and the risk of ventilator-associated pneumonia. Crit Care Med 2006;34:68793.
[51] Briassoulis G, Paraschou D, Hatzis T. Hypercapnia due to a heat and moisture exchanger.
Intensive Care Med 2000;26:147.
[52] Iotti GA, Olivei MC, Palo A, et al. Unfavorable mechanical eects of heat and moisture exchangers in ventilated patients. Intensive Care Med 1997;23:399405.
[53] Casta A, Houck CS. Acute intraoperative endotracheal tube obstruction associated with
a heat and moisture exchanger in an infant. Anesth Analg 1997;84:93940.
[54] Gedeon A, Mebius C, Palmer K. Neonatal hygroscopic condenser humidier. Crit Care Med
1987;15:514.
[55] Tilling SE, Hayes B. Heat and moisture exchangers in articial ventilation. An experimental
study of the eect of gas leakage. Br J Anaesth 1987;59:11818.
[56] Barnes SD, Normoyle DA. Failure of ventilation in an infant due to increased resistance of
a disposable heat and moisture exchanger. Anesth Analg 1996;83:193.
[57] Rathgeber J, Zuchner K, Burchardi H. Conditioning of air in mechanically ventilated
patients. In: Vincent JL, editor. Yearbook of intensive care and emergency medicine. Berlin:
Springer; 1996. p. 50119.
[58] Djedaini K, Markowicz P, Mier L, et al. Comparisons between three types of heat and moisture exchangers changed every 48 hours during mechanical ventilation. Am J Respir Crit
Care Med 1997;155:A769.
[59] Ricard JD, Markowicz P, Djedaini K, et al. Bedside evaluation of ecient airway humidication during mechanical ventilation of the critically ill. Chest 1999;115:164652.
[60] Larsson A, Gustafsson A, Svanborg L. A new device for 100 per cent humidication of
inspired air. Crit Care 2000;4:5460.
[61] Cheney FW, Butler J. The eects of ultrasonically produced aerosols on airway resistance in
man. Anestesiology 1968;29:1099106.
[62] Melville GN, Josenhans WT, Ulmer WT. Changes in specic airway resistance during
prolonged breathing of moist air. Can J Physiol Pharmacol 1970;48:5927.
[63] Tamer MA, Modell JH, Rieel CN. Hyponatremia secondary to ultrasonic aerosol therapy
in the newborn infant. J Pediatr 1970;77:10514.
[64] Sladen A, Laver MB, Pontoppidan H. Pulmonary complications and water retention in
prolonged mechanical ventilation. N Engl J Med 1968;279:44853.
[65] Williams RB. The eects of excessive humidity. Respir Care Clin N Am 1998;4:21528.
[66] Tenaillon A, Boiteau R, Humbert M, et al. Eect of periodical instillation of isotonic salt
solute on tracheal tubes permeability. Intensive Care Med 1990;16:S112.

RESPIRATORY GAS CONDITIONING AND HUMIDIFICATION

33

[67] Hanley MV, Rudd T, Butler J. What happens to intratracheal saline instillations. Am Rev
Respir Dis 1978;117:124S.
[68] Ackerman MH. The use of bolus normal saline instillations in articial airways: is it useful or
necessary. Heart Lung 1985;14:5056.
[69] Christopher KL, Saravolatz LD, Bush TL, et al. The potential role of respiratory therapy
equipment in cross infection. A study using a canine model for pneumonia. Am Rev Respir
Dis 1983;128:2715.
[70] Craven DE, Goularte TA, Make BJ. Contaminated condensate in mechanical ventilator
circuits: a risk factor for nosocomial pneumonia. Am Rev Respir Dis 1984;129:6258.
[71] Redding PJ, McWalter PW. Pseudomonas uorescence cross infection due to contaminated
humidier water. Br Med J 1980;281:275.
[72] Craven DE, Connolly MG, Lichtenberg DA, et al. Contamination of mechanical ventilators
with tubing changes every 24 or 48 hours. N Engl J Med 1982;306:15059.
[73] Lareau SC, Ryan KJ, Diener CF. The relationship between frequency of ventilator circuit
changes and infectious hazard. Am Rev Respir Dis 1978;118:4936.
[74] Dreyfuss D, Djedaini K, Weber P, et al. Prospective study of nosocomial pneumonia and of
patient and circuit colonization during mechanical ventilation with circuit changes every 48
hours versus no change. Am Rev Respir Dis 1991;143:73843.
[75] Hess D, Burns E, Romagnoli D, et al. Weekly ventilator circuit changes. A strategy to reduce
costs without aecting pneumonia rates. Anesthesiology 1995;82:90311.
[76] Cadwallader HL, Bradley CR, Aylie GA. Bacterial contamination and frequency of changing ventilator circuitry. J Hosp Infect 1990;15:6572.
[77] Kola A, Eckmanns T, Gastmeier P. Ecacy of heat and moisture exchangers in preventing
ventilator-associated pneumonia: meta-analysis of randomized controlled trials. Intensive
Care Med 2005;31:511.
[78] Lacherade JC, Auburtin M, Cerf C, et al. Impact of humidication systems on ventilator-associated pneumonia: a randomized multicenter trial. Am J Respir Crit Care Med 2005;172:
127682.