Published ahead of print on July 21, 2004, doi:10.1164/rccm.200309-1245OC
© 2004 American Thoracic Society doi: 10.1164/rccm.200309-1245OC
Influence of Ambient and Ventilator Output Temperatures on Performance of Heated-Wire HumidifiersService de Réanimation Médicale, Hôpital Henri Mondor, INSERM U492, Université Paris XII Créteil, France; and Istituto di Anestesiologia e Rianimazione, Università Cattolica Policlinico A. Gemelli, Rome, Italy Correspondence and requests for reprints should be addressed to Laurent Brochard, M.D., Service de Réanimation Médicale, Hôpital Henri Mondor, 51 Avenue du Maréchal de Lattre de Tassigy, 94010 Créteil, France. E-mail: laurent.brochard{at}hmn.ap-hop-paris.fr
Although heated humidifiers are considered the most efficient humidification devices for mechanical ventilation, endotracheal tube occlusion caused by dry secretions has been reported with heated-wire humidifiers. We tested the hypothesis that inlet chamber temperature, influenced by ambient air and ventilator output temperatures, may affect humidifier performance, as assessed by hygrometry. Hygrometry was measured with three different humidifiers under several conditions, varying ambient air temperatures (high, 28–30°C; and normal, 22–24°C), ventilators with different gas temperatures, and two  E levels. Clinical measurements were performed to confirm bench measurements. Humidifier performance was strongly correlated with inlet chamber temperature in both the bench (p < 0.0001, r2 = 0.93) and the clinical study. With unfavorable conditions, absolute humidity of inspired gas was much lower than recommended (approximately 20 mg H2O/L). Performance was improved by specific settings or new compensatory algorithms. Hygrometry could be evaluated from condensation on the wall chamber only when ambient air temperature was normal but not with high air temperature. An increase in inlet chamber temperature induced by high ambient temperature markedly reduces the performance of heated-wire humidifiers, leading to a risk of endotracheal tube occlusion. Such systems should be avoided in these conditions unless automatic compensation algorithms are used.
Key Words: endotracheal tube occlusion • hygrometric performance • humidification device • mechanical ventilation During mechanical ventilation, humidification of delivered gases is frequently achieved with heated humidifiers (HHs) (1). Inadequate humidification can lead to bronchial inflammation, cell damage, mucociliary clearance impairment, and endotracheal tube occlusion (2). Endotracheal tube occlusion is a potentially life-threatening complication described mainly with hydrophobic heat and moisture exchangers (HMEs) characterized by poor humidification performance (3–7). In a recent large prospective study, Kapadia noted 13 episodes of endotracheal tube occlusion over a 3-year period, of which eight required cardiopulmonary resuscitation (7). HMEs used in this study were poorly performing ones. With newer hygroscopic HMEs, endotracheal tube occlusion seems unusual (8–15), even with prolonged use (16–18). HHs have been shown to provide better gas humidification performances than HMEs (19–23), although no data have demonstrated a better clinical outcome. However, we have also shown that partial undetected obstruction can markedly increase the patient's work of breathing and potentially impede weaning (24, 25). Anecdotal cases of endotracheal tube occlusion have been described with HHs and may be more common when settings are suboptimal (3, 5, 8). Indeed, in these studies, temperature at the Y-piece was 31–32°C, whereas recommended temperature settings for the most recent HHs are 37°C for the outlet chamber and 40°C at the Y-piece. In theory, these recommended settings ensure excellent performance with delivery to the patient of saturated gas at 37°C (44 mg H2O/L). Miyao and colleagues described the first cases of endotracheal tube occlusion with HHs based on heated wire circuits (26). In several recent studies involving the use of heated-wire humidifiers in large numbers of patients (11, 12), no cases of endotracheal tube occlusion were noted, although partial occlusions occurred in another study (27). Unexpectedly, a recent randomized controlled study comparing ventilator-associated pneumonia rates with an HME (Hygrobac DAR) and an HH found a substantial rate of endotracheal tube occlusion with the HH (5 of 169 patients compared with 1 of 172 patients in the HME group, p = 0.12) (28). Using the recommended settings, we also observed several episodes of endotracheal tube occlusion with the new heated-wire humidifiers recently introduced in our intensive care unit. These episodes seemed to be associated with higher ambient air temperatures and with specific ventilator types.
The general working principles of the HHs are to heat water contained in the humidification chamber and to humidify the gas passing through. A heated wire is present all along the inspiratory and expiratory circuits to avoid water condensation in the tubings. Consequently, new heated wire humidifiers need to regulate both the Y-piece and the outlet chamber temperatures. For the latter, the only regulated parameter is the temperature and not the humidity by itself. Therefore, the principles underlying the regulation of these HHs (Figure 1) led us to hypothesize that their performances may be impaired by high inlet chamber temperature. Indeed, in such a case, the heater plate may stop to heat, leading to insufficient energy to humidify gases coming from the ventilator. Because of the potentially severe consequences of unrecognized underhumidification, the goal of this study was to determine whether factors such as ambient air temperature, type of ventilator, and
Bench Study Protocol. The humidification performance of the HH MR 730 (Fisher & Paykel, Auckland, New Zealand) was assessed using the following temperature settings for the outlet chamber/Y-piece in 37°C/40°C and 40°C/40°C (additional details can be found in the online supplement) (Figure 1 and Table 1).
The MR 850 model (Fisher & Paykel), which has a new automatic compensation system designed to increase automatically the set outlet chamber temperature in case of underheating, was also tested. Inspired gas hygrometry was measured with the ventilator and airway circuit connected to a test lung under the following conditions:
Measurements.
TEMPERATURE MEASUREMENTS.
Clinical Study We used the psychometric method to measure delivered gas humidity in 20 consecutive patients ventilated in our intensive care unit with the HH MR 850. The aim of this clinical study was to assess in patients the impact of the inlet chamber temperature on the HH performance and to evaluate the efficacy of a new algorithm, developed to avoid low humidifications performances. After 2 hours at the steady state, hygrometry was determined during three successive periods: with the new automatic compensation system on, after this system was switched off (with the outlet chamber temperature at 37°C and the Y-piece at 40°C), and with the compensation system switched back on.
Statistical Analysis
Hygrometric Results (Bench Study) Fully saturated air at 37°C has an absolute humidity of 44 mg H2O/L. This would be the value observed with HHs capable of fully humidifying the inspired gas. In the first part of the study, with the settings recommended by the manufacturers (37°C at the outlet chamber and 40°C at the Y-piece), absolute humidity was 35.8 ± 1.9 mg H2O/L in the best condition (Table E2) and 19.0 ± 0.8 mg H2O/L in the worst condition (Table E2). Hygrometric data expressed as absolute (mg H2O/L) and relative (%) humidities are shown in Table 2 and Figure 2. To simplify the presentation, absolute humidity data are displayed in the tables as means ± SD obtained by pooling the values obtained at 10 and 20 L/minute. The results for each individual condition are reported in the online supplement.
Main hygrometric results. Humidification performance of HH was markedly decreased by several factors.
These differences were found for humidification performance expressed as both absolute humidity (mg H2O/L) (Tables 2 and E2) and relative humidity (%) (Figure 2).
Inlet chamber temperature.
Inlet chamber temperature varied from 27.4 ± 0.8°C (with normal ambient air temperature, EVITA 4, and low E) to 37.9 ± 1.3°C (with high ambient air temperature, T-Bird, and high E). Ambient air temperature had the largest effect on the inlet chamber temperature.
Output ventilator temperature ranged from 29.8 ± 1.0°C (with normal ambient air temperature, EVITA 4, and high
The combination of high ambient air temperature and high outlet ventilator temperature (with T-Bird and high E) led to the highest inlet chamber temperatures (Table 3) and was associated with the worst HH performances. In these conditions, HH performance was extremely poor, with an absolute humidity of less than 20 mg H2O/L (Tables 2 and E2).
Optimized settings.
Hygrometric Results (Clinical Study)
This study is the first to highlight the major negative effect of ambient air and ventilator output temperatures on HH performance. The most striking finding is that the performance of a recent-generation HH was greatly influenced by inlet chamber temperature, becoming extremely poor when this temperature was high (Tables 2 and E2 and Figures 2 and 3). Absolute humidity of inspired gas was strongly and inversely correlated to inlet chamber temperature in the bench (Figure 3) and in the clinical study. Inlet chamber temperature was influenced by ambient air temperature and ventilator output temperature (Table 3). The weakness point of these new-generation heated-wire humidifiers lies in the mechanism that regulates outlet chamber temperature (Figure 1). When inlet chamber temperature is high, the heater plate stops heating, supposedly maintaining the set outlet chamber temperature. The water contained in the chamber remains too cold for evaporation to occur, leading to extremely low levels of relative and absolute humidity.
Inlet chamber temperature is influenced by both ambient air temperature and ventilator output temperature (Table 3). High ambient air temperature prevents the gas from cooling in the circuitry between the ventilator output and the humidification chamber. Ventilator output gas temperature is also influenced by Under unfavorable conditions (high ambient air temperature, high ventilator output gas temperature), performance was extremely poor, with absolute humidities of less than 20 mg H2O/L. This is lower than measured with the HMEs reported to be responsible for endotracheal tube occlusion (3–6, 19). These results are consistent with the specific mechanism that regulates HH function (Figure 1) but are nevertheless troublesome given their potential for inducing adverse clinical effects. Very few data in the literature can be directly compared with this study. One recent publication concluded that inlet gas temperature had little influence on HH performance (31). This study was performed using pediatric ventilators with very low flows and provided few hygrometric data. In contrast, we found a marked influence of external conditions with a major effect of ambient air temperature. The HH never provided the theoretic ideal value of 44 mg H2O/L when the outlet chamber temperature was set at 37°C. With this setting, normal ambient air temperature, and low ventilator output temperature, the highest measured humidity value was 35.8 mg H2O/L (Table E2). In a recent publication (32), with the same settings and using a dew point mirror hygrometer, a similar result of 36.2 mg H2O/L was obtained. These results emphasize the need for independent assessment of humidification devices (33). Gas humidity values measured using psychrometry have been obtained over the last 10 years in several clinical and bench studies, with a good reproducibility across research groups (18, 19, 21, 23, 26, 34–39). The hygrometry values of inspired gas in this study (from less than 20 mg H2O/L of absolute humidity under the worst conditions to more than 40 mg H2O/L under the best conditions with best settings: 40/40) are comparable to previously reported psychrometric values obtained using various humidification devices. Ricard and colleagues measured the absolute humidity produced by the BB2215 (19), an HME used in several studies and associated with endotracheal tube occlusion (3–6), and found a mean value of 21.8 ± 1.5 mg H2O/L. Under the worst conditions of this study, the HH provided equivalent or even lower humidification values than the BB2215. This may explain the clinical occurrence of endotracheal tube occlusion, because this event is closely related to insufficient performance of humidification devices (3–6). The American Association for Respiratory Care recommends a minimum of 30 mg H2O/L of absolute humidity provided by humidification devices for prolonged mechanical ventilation (40). Absolute inspired gas humidity values obtained with the most efficient hygroscopic and hydrophobic HMEs were 29.1 ± 1.8 mg H2O/L (21) and 30.8 ± 1.5 mg H2O/L (18), respectively, as measured using the psychrometric method. We are aware of only three reported cases of endotracheal tube occlusion with these HMEs, including one in a patient who experienced massive tracheal bleeding before the occlusion (9, 13). This suggests that an absolute humidity of 30 mg H2O/L (with the psychrometric method) or even a bit less may be sufficient to prevent most cases of endotracheal tube occlusion. In this study, absolute inspired gas humidity was well below this threshold under many conditions (Tables 2 and E2). Endotracheal tube occlusion is a very late and insensitive index of inadequate humidification during mechanical ventilation. Most studies evaluating the effects of inadequate humidification on the bronchial mucosa were performed during anesthesia, and few data on patients receiving prolonged mechanical ventilation have been reported. Chalon and colleagues showed that after less than 4 hours of mechanical ventilation, anesthetized patients who received inadequately humidified gas (23°C and 60% of relative humidity, i.e., 12.5 mg H2O/L of absolute humidity) had significantly more postoperative complications and cytologic tracheobronchial tree damage than did patients who received saturated gas at 32°C (33.9 mg H2O/L of absolute humidity) (41). Thus, the low levels of humidification measured in our study (approximately 20 mg H2O/L) may lead to mucosal damage, especially in the event of prolonged mechanical ventilation. Williams and colleagues suggested that airway mucosal dysfunction may be related to the time spent by the mucosa at a given level of humidity (2). In their model, mucosal dysfunction arose after approximately 1 hour of exposure to inspired gas humidity values lower than 25 mg H2O/L. When exposure was approximately 24 hours, mucosal damage occurred with humidity levels below 30 mg H2O/L. After exposure for 5 days, Hurni found moderate airway epithelial damage in patients ventilated with a HH (Fisher & Paykel) set at 31°C at Y-piece or a high-performance HME (Hydroster-DAR), both devices delivering approximately 30 mg H2O/L (8). Using a heated-wire humidifier, we found a wide range of inspired gas humidity values (from 19.0 ± 0.8 to 41.5 ± 1.2 mg H2O/L). Performances of earlier generation HHs, which do not share the same regulation mechanism, have been reported to vary between 29.2 ± 1.4 and 34.3 ± 1.3 mg H2O/L absolute humidity measured using the psychrometric method (19, 21, 23). The main drawback of these earlier HHs is that the water traps must be emptied repeatedly, which increases the nurse workload and carries a risk of cross-contamination (42, 43). Heated-wire humidifiers heat both the inspiratory and the expiratory line, thereby avoiding condensation in the circuitry and obviating or minimizing the need for water traps. Condensation remains a problem, however, related to some compensating systems (30, 44). We did not compare new generation to earlier generation HHs. We used two very similar recent HHs: Aerodyne Kendall and MR 730 Fisher & Paykel. We found that humidification performance was slightly but significantly better with the Aerodyne Kendall under several conditions (Table E1), but both the humidification chamber and the circuits used were different, which may have influenced the inlet chamber temperature. The most important finding was that a high inlet chamber temperature was associated with poor performance for both HHs. This suggests that the mechanism for regulating outlet chamber temperature, which is the same for both HHs, was the relevant characteristic. Thus, the influence of external conditions demonstrated in our study does not seem to be specific to one manufacturer but may apply to HH with heated wire circuits having the same mechanism for regulating outlet chamber temperature. Therefore, the use of these HHs should be avoided in case of high intensive care unit room air temperature, unless equipped with automatic compensation.
Air conditioning of the room is obviously important to avoid high inlet temperatures, but even with normal temperatures (22–24°C in this study), unfavorable conditions (e.g., turbine ventilator at high We tested the automatic compensating system of the MR 850, which is based on a theoretic calculation of the energy needed to warm and humidify a given volume of gas. This system requires that the HH measures the flow delivered by the ventilator. When the heater plate is not sufficiently warm (i.e., when the energy delivered by the HH is insufficient), the outlet chamber temperature is automatically increased until the heater delivers the level of energy determined by the theoretic calculation. The system can automatically increase the set temperature of the outlet chamber by 3°C (from 37 to 40°C, for intubated patients). Under unfavorable conditions, it avoided very low humidification levels in our study. Values obtained with this system were generally greater than 30 mg H2O/L with high ambient air temperatures, except under the worst conditions (Tables 2 and E4). A simple way to avoid low humidification was to set the temperatures at 40/40 with the MR 730 (Table 2 and Figure 2), but this setting cannot be recommended under all environmental conditions: indeed, when the air temperature decreases (e.g., at night), it may lead to considerable condensation in the circuitry. We also found that with turbine ventilators and an FIO2 set at 21%, the absolute humidity of inspired gas was significantly higher than when dry gases were used (100% FIO2) (Table E4). Turbine ventilators predominantly use ambient air, whose hygrometry level is greater than that of wall medical gases.
Another important finding from this study is that condensation is not a reliable clinical marker for adequate humidification when the ambient air temperature is high. When the ambient air temperature was normal, between 22°C and 24°C, we found a good correlation ( In conclusion, the performance of the new heated-wire humidifiers is greatly influenced by inlet gas temperature. Under unfavorable conditions, especially in case of high ambient air temperature (which is common in intensive care units without air conditioning), inspired gas humidification is extremely poor, suggesting a high risk of endotracheal tube occlusion, and the heated-wire humidifiers should then be avoided in these situations. This risk may be lowered using specific settings or automatic compensation systems that partially correct the dysfunction. For the present, a simple alternative in these special conditions is to use a HME. Other types of HHs, despite presenting other drawbacks, or "active" HMEs, both of which are not influenced to the same amount by external conditions, could also be an alternative.
Supported by Fisher and Paykel and Tyco Healthcare, which supplied free of charge the humidifiers for this study. Presented in part at the 2002 European Society of Intensive Care Medecine Congress, September 29–October 2, Barcelona, Spain and at the 2003 International Conference of the American Thoracic Society, May 16–21 Seattle, WA. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Conflict of Interest Statement: F.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; S.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; S.M.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; S.Q. has received grants and salary from Hudson Company for a5,617 in 2002 and from Fisher Paykel for a13,237 in 2002 and 2003; E.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; N.D. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; L.B. as Director of the research laboratory of the Créteil Department of Medical Intensive Care Medicine, received a grant from Hudson Company of a10,000 in 2002 and, as director, a grant from Fisher & Paykel Company of a8,700 in 2001 and a10,000 in 2002 to the same laboratory. Received in original form September 8, 2003; accepted in final form July 15, 2004
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
E on ventilator output and inlet chamber temperature (expressed in °c)
= 0.85) between inspired gas hygrometry and condensation on the HH chamber wall. In contrast, the correlation did not exist at high ambient air temperature (Figures E1 and E2). In this situation, the gas cannot cool and therefore cannot condense on the wall of the HH chamber. Unfortunately, this corresponds to impaired HH function. There is no other means of checking the humidification level. A correlation between inspired gas hygrometry and condensation has been found (
