Closed Suctioning System Reduces Cross-Contamination Between Bronchial System and Gastric Juices

Werner Rabitsch, MD* , Wolfgang J. Ko stler, MD* , Wolfgang Fiebiger, MD* , Christoph Dielacher, RN* , Heidrun Losert, MD* , Camillo Sherif, MD* , Thomas Staudinger, MD* , Edith Seper , Walter Koller, MD , Florian Daxbo ck, MD , Ernst Schuster, PhD , Paul Kno bl, MD* , Heinz Burgmann, MD* , and Michael Frass, MD*
*Intensive Care Unit, Department of Internal Medicine I, Department of Hospital Hygiene, and Department of Medical Computer Sciences, University of Vienna, Vienna, Austria

In this prospective, randomized study, we evaluated whether a closed suctioning (CS) system (TrachCare) influences crossover contamination between bronchial system and gastric juices when compared with an open suctioning system (OS). The secondary aims were an analysis of the frequency of ventilatorassociated pneumonia (VAP) and an analysis ofalteration in gas exchange. Antibiograms were performed from tracheal secretions and gastric juice aspirates on Days 1 and 3 of intubation in 24 patients in a medical intensive care unit. Five cross-contaminations were observed in the OS group onDay 3 versus Day1; the 5 strains shared common genotypes as determined by random amplification of polymorphic DNA. No

cross-contaminations were seen in the CS group ( P 0.037). VAP occurred in 5 patients of the OS group but in none of the CS group patients ( P 0.037). Spao 2 decreased significantly in the OS group compared with presuctioning valuesthe opposite of the CS group. Whereas presuctioning values were comparable between groups, postsuctioning Spao was signifi2 cantly higher in the CS group. CS significantly reduced cross-contamination between bronchial system and gastric juices and reduced the incidence of VAP when compared with OS. Hypoxic phases can be reduced by the help of CS. (Anesth Analg 2004;99:88692)

icrobiological control is one of the foremost tasks in the intensive care unit (ICU). This is especially important with respect to the avoidance of ventilator-associated pneumonia (VAP) in critically ill patients, which may be associated with increased mortality or a prolonged ICU stay (1). In one study, oral and sputum specimens were collected from 20 subjects who were orally intubated for at least 24 h and required mechanical ventilation (2). After 24 h, all subjects had potential pathogens in the mouth, and 67% had sputum cultures positive for pathogens. Suctioning devices were colonized with many of the same pathogens that were present in the mouth within 24 h. We conclude that the presence of

The visualized endotracheal tubes were provided by Pulmonx (Palo Alto, CA). No funding was granted. Accepted for publication March 12, 2004. Address correspondence and reprint requests to Michael Frass, MD, Intensive Care Unit, Department of Internal Medicine I, University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria. Address e-mail to michael.frass@akh-wien.ac.at. DOI: 10.1213/01.ANE.0000143353.85428.39

pathogens in oral and sputum specimens in most patients supports the notion that microaspiration of secretions occurs (2). One of the problems is the suctioning procedure, through which secretions of the patient may cause a high risk of contamination of the trachea by oral or gastric bacteria. Conversely, contamination may occur during the procedure by allowing direct communication between room air and the patient. Classic open suctioning (OS) is conventionally performed by disconnecting the patient from the ventilator and introducing a regular suctioning catheter through the endotracheal tube (ETT) into the upper airways (3). Alternatively, suctioning can be accomplished with a closed suctioning system (CS) attached between the ETT and ventilatory tubing, allowing introduction of the suctioning catheter into the airways without disconnecting the patient from the ventilator (4,5). CS may have some advantages compared with the conventional OS technique. It is suggested that CS could be helpful in limiting environmental, personnel, and patient contamination and in preventing the loss of lung volume and the alveolar derecruitment associated with standard suctioning in severely hypoxemic
2004 by the International Anesthesia Research Society 0003-2999/04

886

Anesth Analg 2004;99:88692

ANESTH ANALG 2004;99:88692

CRITICAL CARE AND TRAUMA RABITSCH ET AL. CONTAMINATION WITH CLOSED SUCTIONING

887

patients (6 9). A high risk of intrinsic positive endexpiratory pressure and extreme negative pressures during suctioning have been described with the volumecontrolled mode (7). However, the effect of CS on VAP and its cost-effectiveness and the influence of such devices on ventilatory support remain to be assessed (3). Patients admitted to ICUs often require prolonged ventilatory support. Most patients airways are secured by cuffed ETTs, and two problems are typically encountered: 1) tracheal tube placement and maintenance of correct ETT positioning and 2) timing and effectiveness of tracheobronchial suctioning. A new visualized ETT (VETT) has been designed to provide fiberoptic control of ETT positioning and visual estimation of the amount of tracheobronchial secretions (10). There are no studies comparing OS and CS in adults with respect to microbiological cross-contamination between the bronchial system and stomach. Therefore, we initiated a study assessing the value of both systems with respect to microbiological cross-contamination between bronchial system and gastric juices in both groups (OS and CS) and to the incidence of nosocomial infections in critically ill patients in a medical ICU. To quantify secretions, the VETT was used in all patients (10). The primary aim of the study was to evaluate whether the CS system would decrease the risk of cross-contamination in ventilated, critically ill patients. The secondary aims were an analysis of the frequency of VAP and an analysis of alterations in gas exchange.

Methods
The study was approved by the Ethical Committee of the University of Vienna. Consent was obtained from patients or their next of kin. Patients not able to give informed consent were instructed about the study after recovery from sedation. Twenty-four consecutive patients undergoing controlled mechanical ventilation were enrolled into the study. Inclusion criteria were an estimated length of ventilation of 3 days and an age of 18 yr. The Acute Physiology and Chronic Health Evaluation II score was evaluated on Day 1. Exclusion criteria included bleeding diathesis, participation in another study, and severe respiratory distress not allowing replacement of an ETT by VETT. Patients were randomized by using sealed envelopes with consecutive numbers into an OS group ( n 12) and a CS group ( n 12). Randomization was performed by using a list provided by the Department of Medical Computer Sciences, University of Vienna, Austria. All patients were initially intubated with an ETT before evaluation for inclusion in the study. No longer than 12 h after the initial intubation, the ETT was replaced within 30 s by an experienced intensivist with a VETT (Pulmonx, Palo Alto, CA) by using a

tracheal tube exchanger (Cook Critical Care, Bloomington, IN) in patients selected for inclusion. There have been no adverse effects observed in more than 100 patients. As a prerequisite, siliconizing of the tracheal tube exchanger is mandatory. Afterward, with the help of a laryngoscope, it is very easy and in no way dangerous to replace an ETT under direct visualization. Information pertaining to the study was provided to the patients next of kin before the patients inclusion into the study and to the patient immediately after he or she regained consciousness. No patient declined to participate in the study. Information was collected on the antibiotic regimen and proton pump inhibitor administration. OS is performed by disconnecting the patient from the ventilator and introducing a suctioning catheter into the ETT. Nursing personnel using sterile gloves are necessary for this procedure. One person opens the connection between the ETT and the ventilatory tubing system; the second person introduces a suctioning catheter and performs flushes. Usually, two to three suctioning catheters are necessary to complete suctioning. Different catheters are used for suctioning the trachea and the oropharynx. A CS system (TrachCare; Tyco Healthcare, Neustadt/Donau, Germany) is attached between the ETT and the ventilatory tubing system. This system, included in the ventilatory circuit, allows introduction of suctioning catheters into the patients airway without the necessity of disconnecting the patient from the ventilator (4,5). The CS system is changed routinely every 24 h. The CS was introduced into our medical ICU 6 mo before the start of the study, and all participating nurses were trained in its use (35). In both groups, regular suctioning took place every 4 h, and additional suctioning was performed whenever nurses decided that it was needed clinically. The VETT system (Pulmonx) consists of a disposable polyvinyl chloride tube embedded with light- and image-transmitting fiberoptic bundles; a high-volume, low-pressure cuff; and a compact video monitoring system (10). Illumination fibers transmit light from a standard light source to the tip of the tube, and the image-transmitting bundle equipped with a lens carries the image to a video monitoring system. An airflow lumen is integrated within the tubes wall, making it possible to rinse the lens with normal saline in case of soiling with tracheal secretions. The system enables visualization of the airway during the intubation process and continuous observation of the lower airways. VETTs with an inner diameter of 8.0 mm and an outer diameter of 11.7 mm were used. Evaluations included recording the number of suctionings per day and the number of flushes during the suctioning maneuvers. Visual assessment of the quantity of secretions inside the trachea was performed

888

CRITICAL CARE AND TRAUMA RABITSCH ET AL. CONTAMINATION WITH CLOSED SUCTIONING

ANESTH ANALG 2004;99:886 92

with the help of fiberoptic monitoring via the VETT system. The quantity of mucus inside the trachea was divided into 3 levels (little 1, moderate 2, and profuse 3). Intraobserver and interobserver variability were evaluated by judging pictures made from the monitoring system. Furthermore, oxygen saturation before and immediately after suctioning was recorded. Patients breathed oxygen for 2 min before suctioning with 100% oxygen. All patients were mechanically ventilated with either continuous positive pressure or bilevel intermittent positive airway pressure. The protocol for oropharyngeal care in the unit includes suctioning of the oropharynx before suctioning of the trachea. A different suctioning catheter is used solely for the oropharynx. No antiseptic is used in our unit. In addition, microbiological samples were obtained by aspiration via a tracheal catheter or the feeding tube. Samples were collected in vials attached between the outer end of the catheter or feeding tube and the suctioning system. Antibiograms were performed from the vials containing tracheal secretions and from gastric juice aspirates on Days 1 and 3. In case of appearance of a bacterium in the other specimen (bronchial system or stomach) on Day 3, examinations of genetic identity were performed. A rapid and simple method was used for template DNA preparation (11). Strains were grown overnight at 35C on Columbia agar plates (Biomerieux, Marcy LEtoile, France). Two colonies were suspended in 150 L of sterile distilled water and boiled for 15 min. After centrifugation (13,000 rpm; 10 min; 4C), 3 Lof the supernatant was used as a template for random amplification of polymorphic DNA (RAPD) polymerase chain reaction (PCR). A RAPD setup initially described for molecular typing of Gram-negative rods was applied to Enterococcus faecalisand Candida spp. (12). RAPD was performed by using the primer P3 (5'-AGA CGT CCA C-3') E for . faecalis and the primer P15 (5'-AAT GGC GCA G3') for Candida albicans and C. tropicalis. The PCR contained Ready to Go RAPD analysis beads (Amermixture sham Biosciences, Little Chalfont, UK), 2 L of primer (25 pmol/ L), 3 L of target DNA, and 20 L of sterile distilled water, for a final volume of 27 L. PCR was performed with a TGradient thermocycler (Biometra, Gottingen, Germany) and included an initial denaturation step (94C for 5 min) and 40 cycles, each consisting of denaturation at 94C for 30 s, annealing at 36C for 30 s, and extension at 72C for 90 s. The amplified products were analyzed by electrophoresis in a 1.5% NuSieve 3:1 agarose gel (BioWhittaker Molecular Applications, Rockland, ME) at 50 V for3h30min and were visualized by ethidium bromide staining. RAPD typing was used to prove microbiological cross-contamination.

The diagnostic criteria of VAP were adapted from criteria established by the American College of Chest Physicians. We did not include bronchial alveolar lavage fluid cultures in the criteria because routine sampling of lower-airway secretions by bronchoscope is not performed in our unit (1315). VAP was diagnosed when a new or progressive radiographic infiltrate developed in conjunction with one of the following signs: radiographic evidence of cavitation; histological evidence of pneumonia; a positive blood culture finding without another source evidence of infection; a purulent tracheal aspirate; or a positive pleural fluid culture finding with 2 of the following symptoms or signs: fever (increase in rectal temperature to 38.0C), leukopenia, or leukocytosis. Leukopenia and leukocytosis were defined as leukocyte 6 /L and 10 10 6 /L, respectively. counts 3 10 Tracheal secretions were considered purulent when the aspirate showed 25 leukocytes per field. A radiologist blinded to the group assignment interpreted all chest radiographs. Statistical analysis was performed with the SAS software package (SAS Institute, Cary, NC). Comparison of oxygen saturation values (observed before and after suctioning) between groups was performed with the Mann-Whitney U -test. Comparison of changes in oxygen saturation values during the suctioning procedure within patient groups was performed with Wilcoxons signed rank test. Comparison of crosscontaminations was performed with Fishers exact test. Using 12 patients in each group, we would have detected (on a P 0.05 level of significance) a 6 times more frequent incidence of cross-contamination in either group with a probability of 0.75 (power of test). In clinical studies, it is customary to keep the sample size small. Because there is little evidence available in the literature on the actual incidence of cross-contamination, we based our assumptions on our local experience. A sixfold more frequent incidence in the OS group is quite plausible considering that opening the ventilatory system during suctioning increases the chance of microbiological transfer. A value of P 0.05 was considered significant.

Results
The median age and sex distributions were not substantially different between groups (Table 1). The reason for intubation and ventilation was respiratory failure on the basis of pulmonary infiltrations in 18 cases. In 3 cases, the underlying disease was chronic obstructive lung disease; in 2 patients, cardiac arrest; and in 1 patient, an inhalational burn trauma. Acute Physiology and Chronic Health Evaluation (APACHE) II scores did not differ between groups. Antibiotic regimen and proton pump inhibitor administration did

ANESTH ANALG 2004;99:88692

CRITICAL CARE AND TRAUMA RABITSCH ET AL. CONTAMINATION WITH CLOSED SUCTIONING

889

Table 1.

Age, Sex Distribution, and Oxygen Saturation as Evaluated by Pulse Oximetry (Sao Variable OS group CS group

) on Days 1 and 3 P value

Age, y, median (range) 63 (5079) 64 (5175) Sex distribution (M:F) 8:4 7:5 Sao at beginning of suctioning, Day 1 97.2 1.9 96.3 1.4 0.196 2 Sao at end of suctioning, Day 1 89.6 2.5* 96.8 1.0 0.0001 2 Sao at beginning of suctioning, Day 3 96.8 1.0 97.0 1.4 0.656 2 Sao at end of suctioning, Day 3 89.6 1.9* 96.4 0.8 0.0001
2

Patients were randomized into an open suctioning group (OS; * P 0.0001; Sao 2 at beginning of suctioning versus Sao

n 12) and a closed suctioning group (CS; at end of suctioning in the same group.

12).

not differ between groups. No antifungal drugs were used. There were no differences in the number of suctionings, the number of flushes, or the quantity of secretions (Table 2). With respect to assessing the quantity of mucus, intraobserver variability was 2%, and interobserver variability was 4%. The results of antibiograms of bronchial secretions and gastric juice aspirates on Day 1 and Day 3 are shown in Table 3. Cross-contaminations are marked by bold letters. Five cross-contaminations could be found in the OS group, whereas no cross-contamination was seen in the CS group ( P 0.037). VAP occurred in the 5 patients of the OS group who had cross-contamination but in none of the CS group (P 0.037). VAP was diagnosed on the basis of progressively developed radiographic infiltrates, fever, leukocytosis, and purulent tracheal aspirates in all 4 patients and on the basis of a positive blood culture finding in 2 patients. Oxygen saturation showed no differences between groups before suctioning, either on Day 1 ( P 0.196) or Day 3 ( P 0.656). However, immediately after suctioning, oxygen saturation decreased significantly in the OS group ( P 0.0001), whereas oxygen saturation values did not change after suctioning in the CS group ( P 0.436 on Day 1 and P 0.188 on Day 3). Thus, patients in the OS group had significantly lower postsuctioning oxygen saturation values compared with patients in the CS group on Day 1 and Day 3 ( P 0.0001). Overall, the 11 investigated isolates from the 5 patients with cross-contamination were as follows: 3 E faecalis strains from Patient 7 (aspirates from right and . left main bronchus, Day 1; gastric juice aspirate, Day 3), 2 C. tropicalis strains from Patient 5 (gastric juice aspirate, Day 1; aspirate from the right main bronchus, Day 3), 2 C. albicans strains from Patient 6 (aspirates from the right main bronchus, Day 1; gastric juice aspirate, Day 3), 2 C. albicans strains from Patient 9 (gastric juice aspirate, Day 1; bronchial aspirate, Day 3), and 2 C. albicans strains from patient 12 (gastric juice aspirate, Day 1; bronchial aspirate, Day 3). The respective strains isolated from gastric juice aspirates and bronchial aspirates of the same patient always shared common genotypes, as determined by RAPD.

Discussion
This is the first study to evaluate the use of CS versus OS with respect to potential bacterial crosscontamination between bronchial system and gastric juices in adult, critically ill patients. In contrast to OS, CS appeared to prevent bacterial cross-contamination. In 5 cases, cross-contamination could be observed in either direction: 2 times from the bronchial system to gastric juices and 3 times in the opposite direction. Since contamination by air during disconnection would bring exogenous bacteria into the lung but not cross-contamination between lung and stomach, we suggest that bacteria from the trachea travel along the tube to the oropharynx and are aspirated from there into the stomach. Similarly, bacteria from the stomach might go into the oropharynx and then along the tube, thereby taking the opportunity to invade the trachea via the open tube during disconnection. In addition, the handling of the disconnected tube enhances the chance of inadvertent cross-contamination, because the gloves of the nurses are touching the tube as well as the feeding tube and oral mucosa. A direct invasion of the trachea by microorganisms via the gastrointestinal tract is not known. Rello and Diaz (16) stated in a recent review that there are only four routes through which bacteria can reach the lower respiratory tract to cause VAP: contiguous spread, hematogenous spread, inhalation, and aspiration. Hematogenous or contiguous routes of invasion are very rare. Therefore, it is assumed that aspiration is the main route. The mechanism of aspiration is assigned to the supine (versus semiupright) position, pooling of secretions above the inflated ETT cuff (with inadequate drainage of subglottic secretions), lack of adequate swallowing and cough mechanism, and so on. Contamination of the ventilator circuits is universal and has no clinical implications. Therefore, the ventilator circuit change interval does not affect the incidence of VAP. However, improper manipulation of the circuits may allow condensate from the warm humidified air that ventilates patients to precipitate, and aerosolization of bacteria is possible. This means that great care must be taken during manipulation of circuits (16). Similar to Rello and Diaz

890

CRITICAL CARE AND TRAUMA RABITSCH ET AL. CONTAMINATION WITH CLOSED SUCTIONING

ANESTH ANALG 2004;99:886 92

Table 2. Quantity of Tracheal Secretions (Little 1, Moderate 2, and Profuse 3), Number of Flushes, and Number of Suctionings Variable Open suctioning group Closed suctioning group P value (OS versus CS)

Day 1: quantity of secretions Little 2 2 Moderate 7 6 0.896 Profuse 3 4 Day 3: quantity of secretions Little 2 2 Moderate 6 6 1.000 Profuse 4 4 P value, Day 1 versus Day 3 0.896 1.000 Day 1: No. flushes, median (range) 2 (13) 2 (23) Day 3: No. flushes, median (range) 2 (13) 2 (23) Day 1: No. suctionings, median (range) 8 (69) 7 (69) Day 3: No. suctionings, median (range) 8 (610) 8 (69)
OS open suctioning; CS closed suctioning.

Table 3.

Antibiograms of Bronchial Secretions and Gastric Juices on Days 1 and 3

Variable group Bronchial A, Day 1 Gastric juice A, Day 1 Bronchial A, Day 3 Gastric juice A, Day 3 Open suctioning 1 Ca Ca, Ct, Efm Ca Ca, Cg, Ct, Efm 2 3 Ca, Sa, Skg, Sv Ca, Sa, Skg Ca, Pa, Sa, Skg Sa 4 Ca, Skg Ca, Skg Ca, Skg, Sv Ca, Skg 5 Ef, Sh Ct, Sh, Skg 6 Ca Ca 7 Ca, Ef, Sm, Skg Ca, Ec Skg, Sm Ec, 8 Ca, Sa Ca, Sa Pa, Sa 9 Ca Ca 10 11 Ca, Skg Ca, Skg Ca, Sv Ca, Sv 12 Ef Ca Ca

Ct , Ef, Skg Skg Ca Ef Skg ,Ef Ca

Bronchial A, Day 1 Bronchial A, Day 3 Gastric juice A, Day 1 Gastric juice A, Day 3 Closed suctioning 1 Sa, Skg Ca, Sa Ca 2 Ca Ca Ca, Ef Ca 3 Ca, Pa Ca, Pa Ca, Pa Ca, Pa 4 Ec, Skg Ec 5 Ca, Ef, Skg Ca, Ef Ca, Skg Ca, Ec 6 Skg Skg Skg Skg 7 8 Ca, Ko, Sv, Sa Ca, Ko, Sa Ko Ko 9 Ca, Sa, Skg, Sv Ca Ca, Sa Ca 10 Ec, Ko Ca 11 Ca, Skg Ca, Ko, Skg Ca, Skg Ca 12 Ef Ef Ef, Ko Ef, Ko
Random amplification of polymorphic DNA (RAPD) typing was used to prove microbiological cross-contamination. Cross-contaminations are marked in bold.

(16), we believe that care during manipulation deserves the utmost attention. The similarity in the quantity of tracheal secretions and the number of interventions needed despite the major differences in VAP is surprising. However, we have observed that the amount of secretions in patients suffering from pneumonia may vary substantially.

In a previous article (17), 84 intubated and mechanically ventilated patients were prospectively evaluated for the incidence of colonization and nosocomial pneumonias depending on whether they received endotracheal suctioning by an OS method versus a CS (TrachCare system) method. Results showed that CS was associated with a significant (67% versus 39%; P 0.02) increase in colonization compared with OS.

ANESTH ANALG 2004;99:88692

CRITICAL CARE AND TRAUMA RABITSCH ET AL. CONTAMINATION WITH CLOSED SUCTIONING

891

This study showed that suctioning performed with the Trach Care CS system increased the incidence of colonization but not the incidence of nosocomial pneumonia and may actually decrease mortality when compared with OS systems. In another article (18), the VAP incidence rates were compared in mechanically-ventilated patients according to the type of endotracheal suctioning (CS versus OS). The nonadjusted incidence rate of VAP was less for the Stericath than for the OS group (7.32 versus 15.89 per 1000 patient-days; P 0.07), without any adverse effects. Data suggest that CS may be helpful to prevent bacterial cross-contamination in adult critically ill patients. Further studies are needed to demonstrate that CS may also be useful in patients with antibiotic-resistant bacterial strains to prevent cross-contamination of other patients in the same room. To determine whether ventilated, low-birth-weight infants treated with CS versus OS in a neonatal ICU (NICU) differ as to airway bacterial colonization, nosocomial pneumonia, bloodstream infection, incidence and severity of bronchopulmonary dysplasia, neonatal mortality, frequency of suctioning, reintubation, and nurse preference, 175 low-birth-weight infants, intubated and ventilated in the delivery room, were randomized on admission to a CS (TrachCare) or OS group (19). Nosocomial pneumonia was diagnosed in 5 patients from each group. A total of 28% of CS patients and 27% of OS patients died. A total of 40 of 44 NICU nurses considered CS to be easier to use, less time-consuming, and better tolerated by the patient (19). Furthermore, pulmonary function decreases during OS as compared with CS. This finding is in accordance with previous studies. Baun et al. (5) found statistically significant differences in 12 anesthetized and oleic acidinjured animals regarding arterial carbon dioxide tension, arterial oxygen saturation, airway pressure, right atrial pressure, intrathoracic pressure, arterial pressure, and right ventricular afterload. Other potential advantages of the CS have been reported previously. The CS system maintains the connection with the mechanical ventilator during tracheal suctioning and is claimed to limit loss in lung volume and oxygenation. Cereda et al. (4) compared changes in lung volume, oxygenation, airway pressure, and hemodynamics during endotracheal suctioning performed with CS and OS systems in a prospective, randomized study in 10 patients in the ICU. They performed 4 consecutive tracheal suctioning maneuvers2 with CS and 2 with OSat 20-minute intervals. Loss in lung volume during OS was significantly more frequent than during CS. During OS, they observed a marked decrease in Sao , whereas 2 during CS the change was only minor. During CS,

ventilation was not interrupted. The authors concluded that avoiding suctioning-related lung volume loss can be helpful in patients with an increased tendency for alveolar collapse. In another study, the physiologic consequences and costs associated with the two methods of endotracheal suctioning (OS and CS) were examined (20). In this prospective, randomized, controlled study, 35 trauma/general surgery patients (16 OS and 19 CS) were included. Physiologic data collected after hyperoxygenation, immediately after suctioning, and 30 seconds after suctioning were compared with baseline values. OS resulted in significant increases in mean arterial blood pressure throughout the suctioning procedure. Both methods resulted in increased mean heart rates. However, 30 seconds after the procedure, the OS method was associated with a significantly more rapid mean heart rate than the CS method. CS was associated with significantly fewer dysrhythmias. Arterial oxygen saturation and systemic venous oxygen saturation decreased with OS. In contrast, arterial oxygen saturation and systemic venous oxygen saturation increased with the CS. There was no difference between the methods in the occurrence of nosocomial pneumonia. OS was calculated to cost $1.88 more per patient per day and required more nursing time and resources. The authors stated that the CS method resulted in significantly fewer physiologic disturbances and appeared to be an effective and cost-efficient method of endotracheal suctioning that was associated with fewer suctioning-induced complications. Although CS systems are more expensive, the open system requires more human resources and disposable equipment (two nurses, sterile gloves, and a series of suctioning catheters). One limitation of our study is the relatively small number of patients. Furthermore, differences in the quality of nursing could not be assessed in this study. Other possible sources of contamination, such as hands and gloves of the nursing personnel or the outer surface of the ETTs, were not investigated. Another feature to be discussed is the correct technique for OS. A comparison with other institutions where a single trained nurse performs tracheal suctioning would be interesting. Therefore, further studies are warranted to investigate these questions. In conclusion, in contrast to the OS group, no crosscontaminations or VAP were seen in the CS group. Spao 2 decreased significantly in the OS group compared with presuctioning values, unlike in the CS group. Whereas presuctioning values were comparable between groups, postsuctioning Spao was signif2 icantly higher in the CS group.
The authors thank Nadja El-Madani for laboratory assistance and careful evaluation of antibiograms.

892

CRITICAL CARE AND TRAUMA RABITSCH ET AL. CONTAMINATION WITH CLOSED SUCTIONING

ANESTH ANALG 2004;99:886 92

References
1. Leone M, Bourgoin A, Giuly E, et al. Influence on outcome of ventilator-associated pneumonia in multiple trauma patients with head trauma treated with selected digestive decontamination. Crit Care Med 2002;30:17416. 2. Sole ML, Poalillo FE, Byers JF, Ludy JE. Bacterial growth in secretions and on suctioning equipment of orally intubated patients: a pilot study. Am J Crit Care 2002;11:1419. 3. Maggiore SM, Iacobone E, Zito G, et al. Closed versus open suctioning techniques. Minerva Anestesiol 2002;68:3604. 4. CeredaM, VillaF, Colombo E, et al.Closed system endotracheal suctioning maintains lung volume during volume-controlled mechanical ventilation. Intensive Care Med 2001;27:64854. 5. Baun MM, Stone KS, Rogge JA. Endotracheal suctioning: open versus closed with and without positive end-expiratory pressure. Crit Care Nurs Q 2002;25:1326. 6. Taggart JA, Dorinsky NL, Sheahan JS. Airway pressures during closed system suctioning. Heart Lung 1988;17:53642. 7. Stenqvist O, Lindgren S, Karason S, et al. Warning! Suctioning: a lung model evaluation of closed suctioning systems. Acta Anaesthesiol Scand 2001;45:16772. 8. Maggiore SM, Lellouche F, Pigeot J, et al. Prevention of endotracheal suctioning-induced alveolar derecruitment in acute lung injury. Am J Respir Crit Care Med 2003;167:121524. 9. Choong K, Chatrkaw P, Frndova H, Cox PN. Comparison of loss in lung volume with open versus in-line catheter endotracheal suctioning. Pediatr Crit Care Med 2003;4:6973. 10. Frass M, Kofler J, Thalhammer F, et al. Clinical evaluation of a new visualized endotracheal tube (VETT). Anesthesiology 1997; 87:12623.

11. Grundmann HJ, Towner KJ, Dijkshoorn L, et al. Multicenter study using standardized protocols and reagents for evaluation of reproducibility of PCR-based fingerprinting of Acinetobacter spp. J Clin Microbiol 1997;35:30717. 12. Maroye P, Doermann HP, Rogues AM, et al. Investigation of an outbreak of Ralstonia pickettii in a pediatric hospital by RAPD. J Hosp Infect 2000;44:26772. 13. Pingletton SK, Fagon JY, Leeper KV Jr. Patient selection for clinical investigation of ventilator-associated pneumonia: criteria for evaluating diagnostic techniques. Chest 1992;102(Suppl 1):553S556S. 14. Jolliet P, Chevrolet J. Bronchoscopy in the intensive care unit. Intensive Care Med 1992;18:1609. 15. Niederman MS, Torres A, Summer W. Invasive diagnostic testing is not needed routinely to manage suspected ventilator associated pneumonia. Am J Respir Crit Care Med 1994;150:5659. 16. Rello J, Diaz E. Pneumonia in the intensive care unit. Crit Care Med 2003;31:254451. 17. Deppe SA, Kelly JW, Thoi LL, et al. Incidence of colonization, nosocomial pneumonia, and mortality in critically ill patients using a Trach Care closed-suction system versus an opensuction system: prospective, randomized study. Crit Care Med 1990;18:138993. 18. Combes P, Fauvage B, Oleyer C. Nosocomial pneumonia in mechanically ventilated patients: a prospective randomised evaluation of the Stericath closed suctioning system. Intensive Care Med 2000;26:87882. 19. Cordero L, Sananes M, Ayers LW. Comparison of a closed (TrachCare) with an open endotracheal suction system in small premature infants. J Perinatol 2000;20:1516. 20. Johnson KL, Kearney PA, Johnson SB, et al. Closed versus open endotracheal suctioning: costs and physiologic consequences. Crit Care Med 1994;22:65866.