Introduction

The importance of delivering warm, humidified gas to infants ventilated through an endotracheal tube is widely accepted [1, 2, 3, 4, 5, 6]. Mechanical ventilation with endotracheal intubation bypasses the upper airway and the normal heat and moisture exchanging process of inspired gases. Medical gases are dried to avoid condensation damage to valves and regulators in the hospital distribution network. If patients are ventilated with cold and dry gases, a continuous loss of moisture and heat occurs and predisposes to serious airway damage: thick secretions, alteration of mucociliary transport, reduced airway defense, reduced airway patency and lung compliance, increased work of breathing and hypothermia [1, 2, 3, 4, 5, 6, 7]. Neonates are very prone to develop hypothermia, due to their large “surface-area-to-body-weight” ratio. This phenomenon is accelerated by poorly developed epidermis associated with high evaporative loss and the inability to shiver in response to cold. When dry gases are used, heat loss is also greater in infants because the respiratory minute volume relationship to body surface is twice that of adults.

Complications after ventilation with dry and cold gases may be prevented by the addition of exogenous heat and humidity using heated hot-water systems. Heated humidifiers (HH) have some disadvantages, namely, condensation of water that may be a source of infection, malfunctions, high maintenance cost, and increased workload for the nursing staff [8, 9]. Furthermore, in neonates nursed in incubators the inspired gas humidity is altered by two different environmental temperatures: the temperature of the air in the room and the incubator temperature [10, 11]. When the incubator is set at lower temperatures (mature babies or fever) a relationship has been demonstrated between inspired gas humidity and the incubator temperature when the heated humidifier is set at a given temperature [10, 11]. This observation is explained by the fact that 50 cm or more of the inspiratory tubing is inside the incubator and directly exposed to its temperature variations [10, 11]. Passive airway humidification with selected hygroscopic heat and moisture exchangers (HMEs) has proven efficient and safe in adults [9, 12, 13, 14]. In infants and neonates, HMEs could be a reliable and simple alternative to active devices for short-term mechanical ventilation [15]. Heat and humidity are captured by the chemical elements of the HME during expiration and restored to the inspired gases during inspiration. Thus, endotracheal tube leak, a potential complication during neonatal ventilation, should be kept as low as possible during passive humidification. Schiffman et al. [15] studied a heterogeneous group of pediatric intensive care unit patients ventilated with a HME in the ventilatory circuit. They did not present specific data on neonates. The aim of the present study was therefore to evaluate the efficacy and possible side effects of passive inspiratory gas conditioning by a selected HME in mechanically ventilated neonates and to compare with the gold standard, HH.

Materials and methods

Study design

The institutional review board and the committee on human research of our institution approved this study, and parental informed consent was obtained. Fourteen consecutive neonates who required mechanical ventilation were studied. Inclusion criteria were an expected period of mechanical ventilation > 24 h, a stable hemodynamic status and no life-threatening hypoxemia.

Indications for mechanical ventilation were recorded along with the Score for Neonatal Acute Physiology (SNAP) [16] (Table 1). All patients were stable and without hemodynamic support.

Table 1 Characteristics of the study patients

All patients were nursed in incubators (model 8000 IC/SC; Dräger, Lübeck, Germany). Incubator temperature was automatically regulated to maintain neonatal abdominal temperature at 37 °C.

Patients were ventilated with continuous flow, time and pressure-controlled ventilators (model IV-100B; Sechrist, Gambro Engstrom SA, Sweden). Respiratory rates were set between 30 and 40 breaths/min, the gas flow was 8 l/min, and the inspiratory pressure was 15 cmH2O. Transcutaneous PCO2 was maintained between 50 and 55 mmHg.

For passive gas humidification, a HME Humidvent® mini (Gibeck, Upplands Vasby, Sweden) was used (dead space 2.4 ml, weight 4 g, flow resistance 0.6 cm H2O at 7 l/min). The HME was placed between the Y connector of the ventilator tubing and the endotracheal tube (Fig. 1).

Fig. 1
figure 1

Ventilator circuit with a heated humidifier (HH; (1)) or a heat and moisture exchanger (HME; (2))

For active humidification, an HH MR 730® AGM (Fisher & Paykel, New Zealand) was used. The heated humidifiers were set to deliver a temperature of 37 °C and an absolute humidity of 40 mgH2O /l at the incubator A point (Fig. 1).

For the two devices, the length of the inspiratory tubing in the incubator was 50 cm. In each patient, starting measurements with HME or HH were assigned randomly. Measurements were performed at 5, 10, and 15 min. After the third measurement the other mode of humidification was set. Respirator tubes were changed with each mode of humidification. At the end, all patients had measurements with the two devices.

Measuring temperatures and absolute humidity of inspired gas

The measuring system consisted of a hollow device (dead space 3 ml) inserted between the Y piece and the endotracheal tube or between the HME and the endotracheal tube (Fig. 1).

Inserted in the device were two thermal probes and a tube connected to a differential pressure transducer (Fig. 2). The two thermal probes, one wet and one dry, had the following characteristics: thermocouple T, class 1, diameter 0.5 mm, time response 5 ms (TC SA, Dardilly, France).

Fig. 2
figure 2

Measuring device

The wet probe was kept wet with a low continuous water supply (0.1 ml/h, 11 drops/h of 9 μl) by an automatic pump ( B-D Pilote C, Franklin Lakes, USA) (Fig. 2).

The psychometric method used in this study is based on comparing the temperatures obtained with the two thermal probes. The dry probe measures the gas temperature. On the wet probe, evaporation is directly proportional to the dryness of the inspired gas. The temperature gradient measured between the two probes increases when inspired gas humidity decreases. When the inspired gas is fully saturated with water (100% relative humidity) no thermal gradient is measured.

With the use of a psoriometric diagram, relative humidity (RH) was obtained. Then, absolute humidity at saturation point (AHs) was obtained from the following formula: AHs = 16.41563–0.731 T + 0.03987 T2, where AHs = absolute humidity at saturation point (100% of RH) in mgH2O/l, and T (°C) = dry probe temperature. AHs was used to calculate AH from the following formula: \( \text{AH}_{(\text{mgH}_{2}\text{O/l})}=(\text{AHs} \times \text{RH})/100 \)  [17].

Measuring transpulmonary pressures

Airway opening pressure (Pao) is measured directly through the third entrance in the special device. According to the Milic-Emili et al. [18] technique, esophageal catheter pressure (Pes) was measured with a latex balloon (2 cm long, endocavity probe latex cover, GE Medical System Accessories and Supplies, USA) connected to a ± 20 cmH2O differential pressure transducer (Validyne MP-45-871, Northridge, CA). The balloon was located in the middle one-third of the esophagus and filled with 0.2–0.4 ml of air. The correct position was identified by seeking the point with the least artifacts in the range with the tip at less than 10 cm from the dental arch. Transpulmonary pressure (Ptp) was taken to be the pressure difference between the Pes and Pao: Ptp = Pao–Pes.

All data (room temperature, incubator temperature, wet probe temperature, dry probe temperature, Pes, Pao) were recorded with a MP100® (Biopac Systems, Goleta, CA) at 200 MHz and stored on a Macintosh® computer using Acknowledge® software.

Statistical analysis

The Statview® software package was used. Results are presented as mean ±  SD.

Statistical analysis was performed using the Wilcoxon test, the Friedman test, and linear regression where appropriate. A p value of less than 0.05 was considered significant.

Results

The clinical characteristics of the patients are presented in Table 1. Age ranged from 1 to 37 days, gestational age from 24 to 40 weeks and weight from 610 to 3350 g. Indications for mechanical ventilation were respiratory distress syndrome in nine neonates, meconium aspiration in three, encephalopathy in one, and gastroschisis in one.

Temperature

Measurement of inspired temperature were performed after a steady state was obtained (1–2 min). Individual values are presented in Table 2, and averaged values (mean ±  SD) in Fig. 3. No significant differences were observed for a given device throughout the study period. The values of inspired gas temperatures were significantly higher using HH at 5, 10, and 15 min ( p < 0.001). A significant positive correlation was observed between incubator temperatures and those of inspired gases delivered by the HH (T°(inspired gas) = 13.9 + 0.63 T°(incubator), R2 = 0.61, p < 0.001) (Fig. 4). A significant correlation was observed between incubator temperature and those of inspired gases delivered by the HME (T°(inspired gas ) = 6.3–0.83 T°(incubator), R2 = 0.5, p < 0.001) (Fig. 5).

Fig. 3
figure 3

Mean values ±  SD of temperature (°C) of inspired gas. Temperatures were significantly higher at the three measurement periods with heated humidifiers (HH) than heat and moisture exchangers (HME)

Fig. 4
figure 4

Correlation between HH mean inspiratory temperature and incubator temperature

Fig. 5
figure 5

Correlation between HME mean inspiratory temperature and incubator temperature

Table 2 Individual values of mean temperature (T° insp, in °C) and absolute humidity (AH, in mgH2O/l) of inspired gas in ventilated neonates with heat and moisture exchanger (HME) or heated humidifier (HH)

Absolute humidity

Individual values are presented in Table 2 and averaged values (mean  ± SD) in Fig. 5. No patient had an AH below 28 mgH2O/l at any measurement time (Table 2). For the HME and the HH there were no significant differences between the values obtained at 5, 10, and 15 min. The values of absolute humidity were significantly higher using heated humidifiers at the three measurements periods (p < 0.001) (Fig. 6). There was a positive correlation between incubator temperature and those of AH of inspired gases delivered by HH (Absolute humidity(inspired gas) = 7.1 + 0.9 T°(incubator), R2 = 0.5, p < 0.001).

Fig. 6
figure 6

Mean values ±  SD of absolute humidity of inspired gas. Absolute humidification was higher at the three measurement periods with heated humidifiers (HH) than with heat and moisture exchangers (HME)

Transpulmonary pressure

No significant changes were observed throughout the study period with any of the tested devices; no significant differences were observed between the two devices (Fig. 7).

Fig. 7
figure 7

Variation of the transpulmonary pressure with heat and moisture exchanger (HME) or heated humidifier (HH) at 5, 10 and 15 minutes. No significant differences were observed between the HH group and the HME group at any measurement time

No significant changes were observed in the hemodynamic or respiratory status of the neonates throughout the study period and the hours following the study.

Discussion

The main results of the present study are:

  1. 1.

    The use of a HME in neonates makes it possible to achieve a conditioning of inspired gases with an absolute humidity of 28 mgH2O/l or more and a temperature of 30 °C or more.

  2. 2.

    Higher values of absolute humidity and temperatures were obtained using a heated humidifier.

  3. 3.

    No significant changes were observed in transpulmonary pressures with any of the tested devices.

The optimal humidity of the inspired gas of intensive care unit patients submitted to mechanical ventilation has not been well evaluated, and the minimal acceptable level is still the matter of controversy. Values published in the literature range from 17 to 44 mgH2O/l. Some data suggest that 23–33 mgH2O/l is a desirable range [4, 19, 20, 21] with a tracheal temperature of 30 °C. However, others have suggested that higher temperatures (35–37 °C) are adequate, leading to absolute humidity of up to 44 mgH2O/l [22, 23].

In any case, the values obtained in the present study with a HME are higher than the minimal expected for adequate inspired gas conditioning in adult use.

For neonates, adequate values are not known and should be determined by means of further studies. The efficacy of HMEs in neonates for long-term mechanical ventilation is unknown. In intubated newborn infants without tracheal leakage, inspired gas humidity < 31 mgH2O/l is associated with endotracheal plugging [10, 24]. This is equivalent to 70% saturation at 37 °C. In the present study, the lowest level of humidification 28 mgH2O/l was obtained with a temperature of 32.5 °C at the Y piece. Further heating and humidification (4 mg/l H2O) is expected only from the endotracheal tube. An average amount of 6–8 mgH2O/l was added to the absolute humidity measured at the Y piece when the inspired gas was at 30 °C [22].

Major endotracheal tube leakage can occur in neonates. A loss of 15% or more of inspiratory humidity content because of endotracheal tube leak can result in a humidity of 30 mgH2O/l or less [15]. Endotracheal tube leak in prolonged mechanical ventilation can therefore result in insufficient humidification, the effects of which on lung function in neonates are not known.

The use of an active humidifier is not problem-free in neonates. As demonstrated by O'Hagan and coworkers, substantial fluctuations in inspired gas humidity are observed in infants nursed in incubators [10]. A correlation was observed between inspired gas humidity and incubator temperature when infants were managed in cooler incubators [10]. This is the case in premature babies kept in the incubator for more than 2 weeks, in mature babies, or in the case of fever. The cool incubator causes condensation in the inspiratory tubing. A long portion (50 cm or more) of the inspiratory limb is inside the incubator and directly exposed to its temperature variations. The temperature and the humidity of the inspired gas decreases between the entry into the incubator (where the temperature probe of the HH is located) and the Y piece. The present study shows that HH efficacy greatly depends on environmental factors such as incubator temperature. When the incubator is cold (mature babies, fever), the absolute humidity delivered by the HH is very low (T° incubator = 29 °C, AH = 32 mgH2O /l). This low absolute humidity was observed despite the fact that the HH was set to deliver an inspiratory gas at 37 °C and 40 mg/l H2O. Such a situation is an inherent disadvantage of current active humidification systems which monitor temperature, but not humidity. On the contrary, setting higher temperature on the HH may result in excessive humidification with warmer incubators [23]. The systematic use of incubator in the present study may explain some of the differences observed with Schiffmann et al. [19], in addition to the fact that they studied a heterogeneous group of pediatric patients and not only neonates.

The present study also investigated the potential side effects of HMEs and HHs by measuring end-expiratory transpulmonary pressure. Flow resistance generated by the HME may result in an increased end-expiratory transpulmonary pressure. This potential complication was not observed in the present and no change in this parameter was observed throughout the study period.

Dead space expansion should be avoided, especially in mechanical ventilation of neonates with small tidal volumes. PaCO2 was not measured in the present study. According to Schiffman et al., in mechanically ventilated infants there is no difference in PaCO2 whether active or passive humidifiers are used [19]. However, it would appear to be wise to use a transcutaneous measurement of PCO2 in all mechanically ventilated premature infants [25, 26]. One limitation of the present preliminary study is the short duration of the study period. With HMEs, inspiratory humidification may improve with time, and in Fig. 6 a very small increase in AH is seen in the HME group. However, this increase is clinically irrelevant and unlikely to offset the difference between the two groups. Further studies are needed to clarify this point. Other factors that may influence the humidification effects of HH have been reported in the literature. The major negative effect of ambient air on HH performance was studied by Lellouche et al. [27]. Absolute humidity of inspired gas was strongly and inversely correlated to the inlet chamber temperature, both in experimental conditions and in patients. Inlet chamber temperature is influenced by ambient air and ventilator output temperatures. The higher the ambient air temperature, the higher the inlet temperature was, and high inlet temperatures were associated with low values of AH [27]. In an ICU with no air conditioning and high ambient air temperature, heated-wire humidifiers should be avoided. In our study, all neonates were ventilated with the same continuous-flow pressure-controlled ventilator with the gas flow set at 8 l/min but ambient air temperature was not monitored. Similar findings were observed by Carter et al. [28]. Davies et al. emphasized the place where inspired gas temperature should be measured [29]. It would seem prudent to measure temperature as close as possible to the patient. Measurements, as in the present study, should be performed not at the circuit exit, but at the airway opening [29]. Similar conclusions were reached by Todd et al. [30, 31]. Finally, the type of mechanical ventilation may influence airway humidification, and today high-frequency oscillatory ventilation is being used increasingly in neonates. Schiffmann et al. [32] studied the efficiency of HMEs and HHs used in this type of ventilation in an artificial lung model. They concluded that HHs and HMEs are suitable for airway humidification. Also, they emphasized the fact that dead-space expansion due to the HME, which may limit their use in neonates, is probably not important with high-frequency oscillatory ventilation [32].

Conclusion

The use of HMEs in neonates makes it possible to achieve, in inspired gases, an absolute humidity of 28 mgH2O/l or more, and a temperature of 30 °C or more. Higher values were obtained using a HH. The optimal values in neonates are not known, but the values observed in the present study with HMEs are higher than the minimum required for adequate inspired gas conditioning for use in adults. The differences observed in inspiratory temperature and absolute humidity are probably not clinically significant, unless further large clinical studies should demonstrate an influence on, for instance, the incidence of pulmonary complications, such as atelectasis.

The present study was designed to evaluate HMEs during short-term ventilation. Other clinical trials should be undertaken to ascertain the safety of extending the duration of use of HMEs in neonates.