当前位置: 首页 > 期刊 > 《美国呼吸和危急护理医学》 > 2005年第11期 > 正文
编号:11417358
Impact of Humidification Systems on Ventilator-associated Pneumonia
http://www.100md.com 《美国呼吸和危急护理医学》
     Medical, Surgical, and Neurosurgical Intensive Care Units, and Department of Anesthesiology, Henri Mondor Hospital; AP-HP, Universite Paris 12, and INSERM U651, Creteil; and Infectious Diseases Intensive Care Unit, Bichat Claude Bernard Hospital, Paris, France

    ABSTRACT

    Rationale and objectives: The respective influence on the incidence of ventilator-associated pneumonia of currently available systems used for warming and humidifying the gases delivered to mechanically ventilated patients, that is, heated humidifiers and heat and moisture exchanger filters, remains controversial.

    Methods: We addressed this question in a multicenter randomized study comparing heated humidifiers (with heated circuits) and filters in an unselected population of 369 intensive care patients receiving mechanical ventilation for more than 48 h.

    Main Measurements and Results: The diagnosis of pneumonia was confirmed according to strict microbiologic criteria. There was no difference in pneumonia rate between the two groups (53 of 184 [28.8%] versus 47 of 185 [25.4%] for humidifiers versus filters; p = 0.48), or in the incidence density of pneumonia (27.4/1,000 ventilatory days versus 25.3/1,000 ventilatory days for humidifiers versus filters; p = 0.76). The mean duration of mechanical ventilation did not differ between the two groups (14.9 ± 15.1 versus 13.5 ± 16.3 days for humidifiers versus filters, p = 0.36). Endotracheal tube occlusion occurred, respectively, in five patients and one patient in the humidifier and filter groups (p = 0.12). Intensive care mortality was identical in the two groups (about 33%).

    Conclusion: These results suggest that both heated humidifiers and heat and moisture exchanger filters can be used with no significant impact on the incidence of ventilator-associated pneumonia and that other criteria may justify their choice.

    Key Words: mechanical ventilation endotracheal tube nosocomial infection

    Two systems are currently available for warming and humidifying the gases delivered to mechanically ventilated patients. Heated humidifiers (HHs) comprise an external source producing heat and water vapor from sterile water. Heat and moisture exchangers combined with microbiological filter (HMEFs), also called artificial noses, work by passively retaining warmth and humidity leaving the trachea during expiration and by recycling them during the next inspiration.

    A controversy exists concerning the possible influence of these systems on the incidence of ventilator-associated pneumonia (VAP) (1–4). A significant reduction in pneumonia rate (6 versus 16%) when using HMEFs was found in a single-center study conducted with trauma patients (5). However, seven other studies found no significant difference between the two systems with regard to VAP rates, although a trend to a reduction of VAP was observed when using HMEFs (6–12). A higher colonization rate of ventilatory circuits has also been demonstrated with HHs. Of note, the only study using strict microbiological criteria for the diagnosis of pneumonia found no difference in pneumonia rates (8). Most of the studies, however, were not powered enough to adequately answer the question. The CDC guidelines reported that there was not enough evidence to conclude on the superiority of one system over the other (4), and two more recent systematic reviews (3, 13) found that the evidence was inconclusive, although one suggested HMEFs might be preferred on the basis of practicality and cost considerations (13). A metaanalysis on this aspect suggested that a decrease in VAP rate could exist with the use of filters (14).

    Because both systems may be useful in specific clinical indications, we designed a large multicenter randomized controlled study, to assess the impact of these two humidification and gas-warming systems on the incidence of VAP in an unselected population of intensive care unit (ICU) patients, using strict microbiological criteria based on invasive respiratory secretion sampling to confirm the diagnosis of VAP. We used the most recent generations of devices, both for HMEFs and for HHs with heated circuits, which, by reducing condensation, may also reduce circuit colonization. Some of the results of this study have been previously reported in the form of an abstract (15).

    METHODS

    See the online supplement for more details on methods.

    Study Design

    The study was conducted in five ICUs located in two French university–affiliated teaching hospitals.

    During the study period, patients expected to require mechanical ventilation for more than 48 h were eligible to be randomly assigned to receive either an HMEF or an HH. Patients already ventilated, patients having contraindications to the use of an HMEF or of an HH, patients admitted after cardiac arrest, patients already enrolled in a clinical trial, and patients with early decision of treatment withdrawal were not included.

    Patients were randomized according to a computer-generated randomization list, stratified by participating ICU.

    The HMEF was the DAR Hygrobac filter device (Tyco Healthcare/Nellcor, Pleasanton, CA). The HH was the MR730 device (Fisher & Paykel Healthcare Ltd, Auckland, New Zealand), which includes wire-heated circuits.

    HMEF devices were changed at 48-h intervals as per the manufacturer's recommendation and the hospital's infection control guidelines. The ventilator circuits were changed for every new patient in either randomized group with no other scheduled circuit changes during the time of ventilation.

    The protocol was approved by the ethics committee of the French Society of Intensive Care Medicine (Paris, France) and consent was waived.

    Diagnosis of Ventilator-associated Pneumonia

    Patients enrolled were screened daily for the occurrence of VAP. Pneumonia diagnosed within 48 h of ventilation was not considered ventilator associated.

    A clinical suspicion of VAP was based on the presence of a recent and persistent infiltrate on chest X-ray and two of the following criteria: fever or hypothermia, leukocytosis or leukopenia, and purulent tracheal secretions. The diagnosis of VAP was confirmed according to strict microbiological criteria based on invasive respiratory secretion samplings cultured quantitatively, using a protected telescoping catheter or bronchoalveolar lavage, obtained before any change in antibiotics administered (16–18). Microorganisms associated with VAP were recorded.

    Data Collection and Definitions

    The following parameters were recorded prospectively for each patient: age, sex, organ failure at admission, diagnostic category, Simplified Acute Physiology Score II (19), major comorbidities, PaO2:FIO2 ratio, and Logistic Organ Dysfunction score (20) at initiation of mechanical ventilation. During ventilation, the following data were recorded: maximal inspired oxygen fraction, positive end-expiratory pressure, duration of mechanical ventilation before pneumonia, total duration of mechanical ventilation (until death or weaning), need for tracheostomy, duration of ICU stay, and outcome at ICU discharge. Other potential risk factors for VAP were also recorded daily during the follow-up.

    Outcome Measures

    The primary end point was comparison of the VAP rate between the two randomized groups. Secondary outcome variables included duration of mechanical ventilation before the occurrence of VAP, ICU mortality rate, duration of mechanical ventilation, duration of ICU stay, tracheostomy rate, and occurrence of endotracheal tube occlusion.

    Statistical Analysis

    Comparisons between groups were done by Student t test for continuous variables and the chi-square statistic for categorical variables. A multivariate logistic regression analysis was performed including VAP as the dependent variable, and randomized group and other pertinent variables that differed between the groups (with p value below 0.20 in the univariate analysis) as potentially independent variables. The time to occurrence of VAP was also analyzed by the Kaplan-Meier method and tested by log-rank test. (See the online data supplement for additional details.)

    RESULTS

    Patient Characteristics

    The trial was performed from February 2000 to June 2002.

    During the study period, 1,147 patients received mechanical ventilation in the five participating ICUs. Of the 589 patients expected to receive mechanical ventilation for more than 48 h, 370 patients could be included, 186 in the HMEF arm and 184 in the HH arm (see Figure E1 in the online supplement). One patient in the HMEF group was secondarily excluded because of an early decision to limit active therapy, including a specific diagnostic procedure for pneumonia. The mean number of included patients per center participating to the trial was 74 ± 47.

    Of the 370 patients, only 12 (3%) received mechanical ventilation for less than 48 hours (9 in the HMEF group and 3 in the HH group; p = 0.14).

    Demographic data and markers of acute illness among the randomized patients are shown in Table 1. The characteristics of the two groups were similar for most variables. Patients assigned to the HMEF group had, slightly more often, respiratory failure as the primary diagnosis than did patients in the HH group (p = 0.06), and patients assigned to the HH group had a slightly higher Simplified Acute Physiology Score II score than those assigned to the HMEF group (p = 0.06); however, the Logistic Organ Dysfunction score did not differ between the two groups. There were also more patients with HIV infection in the HH group than in the HMEF group (p = 0.001).

    Twenty patients developed pneumonia within the first 48 h of mechanical ventilation (12 in the HMEF group and 8 in the HH group, p = 0.49). These early pneumonia episodes were not included in the VAP rate (see METHODS), but these patients were retained in the analysis.

    Risk Factors and Incidence of VAP

    The distribution of risk factors for VAP is shown in Table 2 and was comparable between the two groups. These data were not available for 7.5% of patients included in the HMEF arm and for 4.4% of patients in the HH arm. The rates at which patients received antacids or H2 blockers and enteral feeding appeared similar for the two groups.

    The rate of reintubation was not significantly different between the two groups despite a trend toward a higher rate with use of HH (12.9% in the HMEF group and 20.3% in the HH group, p = 0.08). During the follow-up, 60% of patients required transport out of the ICU. The overall use of antibiotic therapy was similar in the two groups. About 20% of the patients had received antibiotic treatment before the introduction of mechanical ventilation, and 50% began antibiotic therapy on the day of introduction of ventilation. In addition, about 40% received a new antibiotic during the follow-up (Table 2).

    A clinical suspicion of pneumonia occurred in 74 patients (40%) in the HMEF group and in 77 patients (42%) in the HH group, and a similar number of samples (106 and 112, respectively) were obtained from each group. There was no difference in the type of sample (bronchoscopic or nonbronchoscopic) obtained from each group (see Table E1). A total of 100 episodes of microbiologically confirmed VAP occurred during the trial, 47 in the HMEF group and 53 in the HH group (p = 0.48), yielding an incidence rate of 25.3 and 27.4 episodes of VAP per 1,000 ventilator days (p = 0.76), respectively (Table 3). The duration of mechanical ventilation before the occurrence of VAP was similar in the two groups (8.0 ± 4.5 d in the HMEF group versus 8.9 ± 5.7 in the HH group). The rate of early-onset pneumonia (10.8% in the HMEF arm versus 13% in the HH arm), and of late-onset pneumonia (14.6 and 15.8%, respectively) were similar in both groups.

    A logistic regression analysis was performed to adjust for possible differences between groups and confounding variables. The adjusted odds ratio assessing the relationship between ventilator-associated pneumonia and treatment group assignment (receiving HMEF compared with HH) was 0.95 (95% confidence interval, 0.58–1.57; p = 0.85) (Table 4); none of the variables introduced in this model was significantly associated with a higher risk of VAP.

    Figure 1 shows the Kaplan-Meier curve analysis comparing the proportion of patients remaining without VAP in each group; there was no significant difference between groups (p = 0.90, log-rank test).

    Secondary Outcomes

    (See Table 5.) The tracheostomy rate (13.4% in the HMEF group and 19.6% in the HH group), the duration of mechanical ventilation (13.5 ± 16.3 versus 14.9 ± 15.1 d, respectively), and the duration of ICU stay (21.4 ± 20.8 versus 25.3 ± 30.1 d, respectively) were similar between the two groups. The death rate in the ICU was also similar: 32.8% for the HMEF group and 34.2% for the HH group.

    Occlusion of the endotracheal tube requiring an emergency reintubation occurred six times during the study, five times in the HH group and once in the HMEF group (p = 0.12). These reintubations were performed without apparent clinical consequences.

    Microorganisms Causing Ventilator-associated Pneumonia

    Microbiological analysis showed that 90 episodes of VAP were monomicrobial, whereas the others were polymicrobial episodes. The 10 polymicrobial VAP cases occurred exclusively in the HH group (p < 0.01). Eight of these 10 polymicrobial VAP cases were early-onset pneumonia with no multidrug-resistant bacterial strains.

    One hundred and eight pathogens were isolated (Table 6), 47 in the HMEF arm and 61 in the HH arm (p = 0.08). Gram-negative bacilli accounted for 58.3% and gram-positive cocci for 38%; other organisms included fungi (1.9%) and polymorphic flora (1.9%). The distribution of microorganisms was similar between the two groups. There was no difference between the proportion of Pseudomonas aeruginosa and methicillin-resistant Staphylococcus aureus between groups.

    DISCUSSION

    In this randomized multicenter trial, we found no significant difference in VAP rate in ICU patients ventilated more than 48 h when using heat and moisture exchanger filters compared with heated humidifiers. The overall rates of ventilator-associated pneumonia appeared to be high in both study groups, 25.4% in the HMEF group and 28.8% in the HH group. The proportions of early- and late-onset pneumonia were also similar in both groups, and the incidence density of VAP did not differ between the two groups, at, respectively, 25.3 and 27.4 episodes of VAP per 1,000 ventilator days.

    The strength of our study included its multicenter design, its relatively large sample size (the largest conducted on this controversial issue), the unselected patient population included (medical and surgical patients), the concealment of the allocation process, the complete follow-up of included patients, and the intention-to-treat analysis. We also recorded risk factors for VAP during the follow-up, which allowed controlling for potential confounding factors and strengthened the crude results of the trial. Among previously published studies, only those conducted by Kollef and coworkers (11) and Dreyfuss and coworkers (8) reported such data.

    In our trial, the diagnosis of VAP was based on the conjunction of a clinical suspicion and a significant culture of at least one microorganism from a distal airway sample (distal protected sample or bronchoalveolar lavage), many of them being performed under bronchoscopy. These samples were also obtained before any introduction or change in antibiotics administered. Among all published studies on this topic, such a strategy was used only in the (negative) study by Dreyfuss and coworkers (8). Among the seven remaining trials, one used clinical criteria only (7), whereas the six others (5, 6, 9–12) used clinical suspicion of VAP and endotracheal aspirate cultures to confirm the presence of pneumonia, a sampling method largely open to bias due to its low specificity (18).

    The lack of reduction of VAP rate found in our study, when using HMEF as compared with HH, is, however, consistent with most of the previously published randomized studies (6–12). The most recent randomized study addressing this topic, published during the course of our study, found a comparable result (11.4% VAP rate in the HMEF group and 15.8% in the HH group; p = 0.3) in a heterogeneous ICU population (12). Our results diverge from those of Kirton and coworkers (5), who found a statistically significant reduction in pneumonia rate with the use of HMEFs (6.4 versus 15.7% with HHs) in a prospective randomized study of 280 trauma patients. A striking feature of that study was the low incidence of VAP recorded in the HMEF group. Indeed, in a review addressing the epidemiology of VAP in a large U.S. database, a 17.5% rate of VAP was recorded among 1,262 trauma patients (21). In addition to the difference in population studied and in method used to diagnose pneumonia, differences in brand, type, and management of HMEFs existed between their study and ours and could have affected the different results. HMEFs were changed every 24 h in the study by Kirton and coworkers, and three times per week in ours. Several clinical studies have shown, however, that extended use of HMEFs, from 24 to 48 h and up to 4 or 7 d, was not associated with increased risk of VAP (1, 22–24). On the other hand, the Pall BB-100 filter has clearly lower humidification properties than the filter used in our study (25, 26). Whether dried secretions with this system could explain a lower rate of clinical suspicion is also a possible hypothesis.

    A higher colonization rate of the ventilator circuits has been consistently described in previous studies with the use of HHs compared with HMEFs (8). The lack of reduction in VAP rate with the use of HMEFs found in our study suggests that circuit colonization plays little or no role in the occurrence of VAP, as already suggested (8). Furthermore, reduction in the rate of change of ventilator circuit does not adversely affect occurrence of VAP, as shown by several randomized trials (27–29). A practical problem with HHs is the risk of spilling over contaminated condensate accumulated in the water trap directly into the patient's tracheobronchial tree during simple procedures such as turning the patients or raising the bed rail. Such an inoculation of large amounts of contaminated secretions is a possible way to overwhelm pulmonary defense and cause pneumonia. New generations of HH such as those used in our trial include heated ventilator circuits, which markedly reduce the formation of condensate; and with no water trap on the ventilator circuit, the risk of massive inoculation is reduced (18). Therefore, our results may not be applicable to centers not using the same type of device including heated—inspiratory and expiratory—circuits. Of note, only one of the eight studies published on this topic used heated tubing (11).

    In our trial, endotracheal tube occlusion with need for reintubation occurred in six patients, one in the HMEF group and five in the HH group. This low rate of occlusion (below 1%) with the use of HMEFs confirms the safety of HMEFs exhibiting hygroscopic properties (5, 11, 12, 24, 30). A higher rate of endotracheal occlusion has been found in earlier studies comparing HMEFs and HHs, in which purely hydrophobic HMEFs were used (6, 7). We observed a relatively high rate of endotracheal tube occlusion with HH systems (3%, compared with < 1% in the other studies) (5–12). This observation raises another limit of the performance of HHs. The quality of gas humidification is directly dependent on the gas temperature at the entry of the HH. Thus, as shown by Lellouche and coworkers under situations of high gas temperature at the entry of the heated humidifier, the short heating period required to achieve the target temperature (37°C with the HH used in our study) with a warm gas could result in poor humidification of the tracheobronchial tree (31). On the other hand, in a study assessing the impact of HMEFs and HHs on the endotracheal tube patency of patients requiring mechanical ventilation for more than 48 h, tube resistance increased more with HMEFs, suggesting better long-term humidification (32).

    Several limitations of our study should be noted. Despite the concealed randomization and the size of our trial, there were imbalances in baseline characteristics between the groups (e.g., a difference concerning the numbers of HIV patients, most of whom were in the same center). There was no statistically significant difference between HIV-infected patients and non–HIV-infected patients concerning the rate of VAP (respectively, 26.1 and 27.1%; p > 0.999), the duration of mechanical ventilation (p = 0.65), or mortality rate (p = 0.45). The adjusted analysis, including all significantly different or clinically relevant confounding variables, reached the same conclusion, strengthening the validity of the study results.

    Another limitation is the fact that clinicians and study investigators were not blinded to treatment, the microbiologists being blinded only to the randomization assignment. The importance of blinding, however, can be considered as moderate in a trial in which the main outcome, that is, occurrence of pneumonia, is based on strictly objective criteria (33).

    We failed to show a protective effect produced by the use of HMEFs, compared with HHs, on the occurrence of VAP. Among all previous studies, only one of these found a statistically significant reduction in the risk of VAP when using HMEFs in a population of trauma patients (5). A new, large randomized controlled study may be needed to confirm this result in this specific population. The studies, including ours, using strict microbiological criteria found no significant influence of the device used. Therefore, we suggest that both systems can be used with no significant impact on the incidence of VAP in unselected ICU populations. To what extent our results are specific to HHs using heated circuits is not known, however. Other aspects may be worth considering, including circuit dead space and costs. In France, the mean daily acquisition cost per patient was 2.3 versus 0.92 ($2.75 versus $1.1) when using HHs or HMEFs, respectively.

    Several systematic reviews have addressed the question of the relative risk of pneumonia associated with HMEFs or HHs. Although a trend toward a reduced incidence of pneumonia with the use of HMEFs was noted in two of them (13, 14), the other noted that this effect was driven by one study only (3), and that additional trials were warranted. It was also concluded that cost consideration may favor the use of HMEFs. That HMEFs can be associated with lower costs will depend in large part on the frequency of change of the devices, and associated costs.

    Acknowledgments

    Fisher & Paykel Healthcare provided equipment necessary for the study.

    FOOTNOTES

    Supported by Fisher & Paykel Healthcare.

    This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org

    Originally Published in Press as DOI: 10.1164/rccm.200408-1028OC on August 26, 2005

    Conflict of Interest Statement: J.-C.L., M.A, C.C., A.V.d.L., L.S., Y.R., S.R., J.-D.R., F.L., and C.B-B. do not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Fisher & Paykel Healthcare provided equipment necessary for the study; and L.B., as Director of the Research Laboratory of the Creteil Department of Medical Intensive Care Medicine, received a grant from Fisher & Paykel Healthcare of 8,700 in 2001 and 10,000 in 2002 to the research laboratory.

    REFERENCES

    Cook D, Ricard JD, Reeve B, Randall J, Wigg M, Brochard L, Dreyfuss D. Ventilator circuit and secretion management strategies: a Franco-Canadian survey. Crit Care Med 2000;28:3547–3554.

    Kollef MH. The prevention of ventilator-associated pneumonia. N Engl J Med 1999;340:627–634.

    Collard HR, Saint S, Matthay MA. Prevention of ventilator-associated pneumonia: an evidence-based systematic review. Ann Intern Med 2003;138:494–501.

    Tablan OC, Anderson LJ, Besser R, Bridges C, Hajjeh R. Guidelines for preventing health-care–associated pneumonia, 2003: recommendations of CDC and the Healthcare Infection Control Practices Advisory Committee. MMWR Recomm Rep 2004;53(RR-3):1–36.

    Kirton OC, DeHaven B, Morgan J, Morejon O, Civetta J. A prospective, randomized comparison of an in-line heat and moisture exchange filter and heated wire humidifiers: rates of ventilator-associated early-onset (community-acquired) or late-onset (hospital-acquired) pneumonia and incidence of endotracheal tube occlusion. Chest 1997;112:1055–1059.

    Martin C, Perrin G, Gevaudan MJ, Saux P, Gouin F. Heat and moisture exchangers and vaporizing humidifiers in the intensive care unit. Chest 1990;97:144–149.

    Roustan JP, Kienlen J, Aubas P, Aubas S, du Cailar J. Comparison of hydrophobic heat and moisture exchangers with heated humidifier during prolonged mechanical ventilation. Intensive Care Med 1992;18: 97–100.

    Dreyfuss D, Djedaini K, Gros I, Mier L, Le Bourdelles G, Cohen Y, Estagnasie P, Coste F, Boussougant Y. Mechanical ventilation with heated humidifiers or heat and moisture exchangers: effects on patient colonization and incidence of nosocomial pneumonia. Am J Respir Crit Care Med 1995;151:986–992.

    Boots RJ, Howe S, George N, Harris FM, Faoagali J. Clinical utility of hygroscopic heat and moisture exchangers in intensive care patients. Crit Care Med 1997;25:1707–1712.

    Hurni JM, Feihl F, Lazor R, Leuenberger P, Perret C. Safety of combined heat and moisture exchanger filters in long-term mechanical ventilation. Chest 1997;111:686–691.

    Kollef MH, Shapiro SD, Boyd V, Silver P, Von Harz B, Trovillion E, Prentice D. A randomized clinical trial comparing an extended-use hygroscopic condenser humidifier with heated-water humidification in mechanically ventilated patients. Chest 1998;113:759–767.

    Memish ZA, Oni GA, Djazmati W, Cunningham G, Mah MW. A randomized clinical trial to compare the effects of a heat and moisture exchanger with a heated humidifying system on the occurrence rate of ventilator-associated pneumonia. Am J Infect Control 2001;29:301–305.

    Dodek P, Keenan S, Cook D, Heyland D, Jacka M, Hand L, Muscedere J, Foster D, Mehta N, Hall R, et al. Evidence-based clinical practice guideline for the prevention of ventilator-associated pneumonia. Ann Intern Med 2004;141:305–313.

    Kola A, Eckmanns T, Gastmeier P. Efficacy of heat and moisture exchangers in preventing ventilator-associated pneumonia: meta-analysis of randomized controlled trials. Intensive Care Med 2005;31:5–11.

    Lacherade J, Auburtin M, Soufir L, Van de Low A, Cerf C, Rebufat Y, Reizzega S, Ricard J, Lellouche F, Lemaire F, et al. Ventilator-associated pneumonia: effects of heat and moisture exchangers and heated water humidificators . Intensive Care Med 2002;28:S132.

    Pham LH, Brun-Buisson C, Legrand P, Rauss A, Verra F, Brochard L, Lemaire F. Diagnosis of nosocomial pneumonia in mechanically ventilated patients: comparison of a plugged telescoping catheter with the protected specimen brush. Am Rev Respir Dis 1991;143:1055–1061.

    Fartoukh M, Maitre B, Honore S, Cerf C, Zahar JR, Brun-Buisson C. Diagnosing pneumonia during mechanical ventilation: the clinical pulmonary infection score revisited. Am J Respir Crit Care Med 2003; 168:173–179.

    Chastre J, Fagon JY. Ventilator-associated pneumonia. Am J Respir Crit Care Med 2002;165:867–903.

    Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993;270:2957–2963.

    Le Gall JR, Klar J, Lemeshow S, Saulnier F, Alberti C, Artigas A, Teres D; ICU Scoring Group. The Logistic Organ Dysfunction system: a new way to assess organ dysfunction in the intensive care unit. JAMA 1996;276:802–810.

    Rello J, Ollendorf DA, Oster G, Vera-Llonch M, Bellm L, Redman R, Kollef MH. Epidemiology and outcomes of ventilator-associated pneumonia in a large US database. Chest 2002;122:2115–2121.

    Djedaini K, Billiard M, Mier L, Le Bourdelles G, Brun P, Markowicz P, Estagnasie P, Coste F, Boussougant Y, Dreyfuss D. Changing heat and moisture exchangers every 48 hours rather than 24 hours does not affect their efficacy and the incidence of nosocomial pneumonia. Am J Respir Crit Care Med 1995;152:1562–1569.

    Boisson C, Viviand X, Arnaud S, Thomachot L, Miliani Y, Martin C. Changing a hydrophobic heat and moisture exchanger after 48 hours rather than 24 hours: a clinical and microbiological evaluation. Intensive Care Med 1999;25:1237–1243.

    Boyer A, Thiery G, Lasry S, Pigne E, Salah A, de Lassence A, Dreyfuss D, Ricard JD. Long-term mechanical ventilation with hygroscopic heat and moisture exchangers used for 48 hours: a prospective clinical, hygrometric, and bacteriologic study. Crit Care Med 2003;31:823–829.

    Beydon L, Tong D, Jackson N, Dreyfuss D; Groupe de Travail sur les Respirateurs. Correlation between simple clinical parameters and the in vitro humidification characteristics of filter heat and moisture exchangers. Chest 1997;112:739–744.

    Taille S, Lefrancois F, Lellouche F, Ricard JD, Jouvet P, Brochard L. Comparison of humidification performances of heat and moisture exchangers on a bench test. Intensive Care Med 2001;27:S211.

    Dreyfuss D, Djedaini K, Weber P, Brun P, Lanore JJ, Rahmani J, Boussougant Y, Coste F. 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;143:738–743.

    Kollef MH, Shapiro SD, Fraser VJ, Silver P, Murphy DM, Trovillion E, Hearns ML, Richards RD, Cracchilo L, Hossin L. Mechanical ventilation with or without 7-day circuit changes: a randomized controlled trial. Ann Intern Med 1995;123:168–174.

    Long MN, Wickstrom G, Grimes A, Benton CF, Belcher B, Stamm AM. Prospective, randomized study of ventilator-associated pneumonia in patients with one versus three ventilator circuit changes per week. Infect Control Hosp Epidemiol 1996;17:14–19.

    Thiery G, Boyer A, Pigne E, Salah A, De Lassence A, Dreyfuss D, Ricard JD. Heat and moisture exchangers in mechanically ventilated intensive care unit patients: a plea for an independent assessment of their performance. Crit Care Med 2003;31:699–704.

    Lellouche F, Taille S, Maggiore SM, Qader S, L'Her E, Deye N, Brochard L. Influence of ambient and ventilator output temperatures on performance of heated-wire humidifiers. Am J Respir Crit Care Med 2004; 170:1073–1079.

    Jaber S, Pigeot J, Fodil R, Maggiore S, Harf A, Isabey D, Brochard L, Louis B. Long-term effects of different humidification systems on endotracheal tube patency: evaluation by the acoustic reflection method. Anesthesiology 2004;100:782–788.

    Cook DJ, Hebert PC, Heyland DK, Guyatt GH, Brun-Buisson C, Marshall JC, Russell J, Vincent JL, Sprung CL, Rutledge F. How to use an article on therapy or prevention: pneumonia prevention using subglottic secretion drainage. Crit Care Med 1997;25:1502–1513.(Jean-Claude Lacherade, Marc Auburtin, Ch)