The tracheal tube: gateway to ventilator-associated pneumonia

Ventilator-associated pneumonia (VAP) is a major healthcare-associated complication with considerable attributable morbidity, mortality and cost. Inherent design flaws in the standard high-volume low-pressure cuffed tracheal tubes form a major part of the pathogenic mechanism causing VAP. The formation of folds in the inflated cuff leads to microaspiration of pooled oropharyngeal secretions into the trachea, and biofilm formation on the inner surface of the tracheal tube helps to maintain bacterial colonization of the lower airways. Improved design of tracheal tubes with new cuff material and shape have reduced the size and number of these folds, which together with the addition of suction ports above the cuff to drain pooled subglottic secretions leads to reduced aspiration of oropharyngeal secretions. Furthermore, coating tracheal tubes with antibacterial agents reduces biofilm formation and the incidence of VAP. In this Viewpoint article we explore the published data supporting the new tracheal tubes and their potential contribution to VAP prevention strategies. We also propose that it may now be against good medical practice to continue to use a 'standard cuffed tube' given what is already known, and the weight of evidence supporting the use of newer tube designs.


Introduction
Ventilator-associated pneumonia (VAP) is defi ned as pneumonia occurring in a mechanically ventilated patient after 48 hours of endotracheal intubation [1]. Despite signifi cant advances in managing intubated patients, VAP remains a common and occasionally fatal complication in the ICU [2]. A systematic review of published data since 1990 showed the incidence of VAP to be 10 to 20%, with a possible two-fold increase in mortality attributable to VAP [3]. Th e ICU length of stay was also signifi cantly increased by a mean of 6.1 days with an attributable cost of $10,019 per case [3]. A recent Canadian study estimated an additional 4.3 ICU days attributable to VAP, occupying 2% of all ICU days and an estimated national cost of CAN$43 million per year [4]. Similar fi ndings were reported by a North American study with increased unadjusted ICU length of stay and mortality in patients with VAP (50% mortality in VAP patients versus 34% in non-VAP) with an estimated $11,897 attributable cost [5]. Furthermore, the burden of VAP takes up a signifi cant portion of antibiotic dispensing in the ICU [6] and may well be a contributor to the development of multiresistant bacteria [7].
Importantly, many units are recently reporting a reduction in VAP incidence following implementation of various prevention measures, as well as programs that increase compliance with such care bundles [8][9][10].

Pathogenesis
Th e pathogenesis of VAP mainly stems from the introduction of microbial pathogens by microaspiration past the tracheal tube cuff and into the lower respiratory tract ( Figure 1). Subsequent colonization and overwhelming of the host mechanical, humoral and cellular defence mecha nisms lead to the development of VAP [11]. Th e tracheal tube forms the essential fi rst part of this mechanistic pathway by breeching the anatomic barriers formed by the glottis and larynx. Suppression of the cough refl ex as a result of sedation further hampers natural refl exes [2]. Th e oropharynx, nasal sinuses and the stomach have been proposed as potential reservoirs of infective material [11]. Furthermore, bacterial biofi lm formation on the inner aspect of the tracheal tube is another potential portal of bacteria [12]. But perhaps more importantly, this biofi lm formation, which takes place within days of intubation, functions to maintain bacterial colonization of the trachea [13]. Th e biofi lm is inaccessible to antibiotic therapy unless aerosolised, which likely reduces bacterial shedding, but does not lead to full eradication and may even promote antibiotic Abstract Ventilator-associated pneumonia (VAP) is a major healthcare-associated complication with considerable attributable morbidity, mortality and cost. Inherent design fl aws in the standard high-volume lowpressure cuff ed tracheal tubes form a major part of the pathogenic mechanism causing VAP. The formation of folds in the infl ated cuff leads to microaspiration of pooled oropharyngeal secretions into the trachea, and biofi lm formation on the inner surface of the tracheal tube helps to maintain bacterial colonization of the lower airways. Improved design of tracheal tubes with new cuff material and shape have reduced the size and number of these folds, which together with the addition of suction ports above the cuff to drain pooled subglottic secretions leads to reduced aspiration of oropharyngeal secretions. Furthermore, coating tracheal tubes with antibacterial agents reduces biofi lm formation and the incidence of VAP. In this Viewpoint article we explore the published data supporting the new tracheal tubes and their potential contribution to VAP prevention strategies. We also propose that it may now be against good medical practice to continue to use a 'standard cuff ed tube' given what is already known, and the weight of evidence supporting the use of newer tube designs.
resistance [14]. Th e fact that the NASCENT trial, using a silver-coated tracheal tube, showed a signifi cant reduction in microbiologically diag nosed VAP suggests that intraluminal biofi lm and the intratracheal route of infection contribute signifi cantly to the aetiology of VAP [15].
Th is article explores the contribution of the tracheal tube to the development of VAP and looks at how the emergence of new designs may help prevent this complication. A question is posed: is it good medical practice to continue to use a 'standard cuff ed tube' given what is already known?

The contribution of the tracheal tube cuff
Th e main function of the tracheal tube cuff is to produce a seal between the tube and the tracheal mucosa in order to allow the institution of positive pressure ventilation. It has become apparent, however, that pooling and leakage around the endotracheal tube cuff leads to aspiration of contaminated oropharyngeal secretions, stomach contents and bacteria into the trachea and the lower respiratory tract [16]. Th is is supported by the fi nding that persistent intra-cuff pressures of <20 cmH 2 O in intubated patients are independently associated with pneumonia [17]. Furthermore, frequent microaspiration of stomach contents as evidenced by the presence of pepsin in sequential tracheal aspirates is an independent predictor of pneu monia in intubated patients [18]. Th ese fi ndings highlight the importance of adequate sealing of the lower respira tory tract from soiling by oropharyngeal secretions.
Some small studies have looked at the use of automated devices to control cuff pressures but have produced varying results. In an 'underpowered' randomised controlled trial where cuff pressures were maintained either by an automatic device at 25 to 30 cmH 2 O, or 8-hourly nurse-adjusted cuff pressures, no diff erence in the incidence of VAP was observed [19]. Conversely, a more recent proof of concept study showed reduced pepsin and micro-aspiration in the tracheal samples and lower microbiologically confi rmed VAP in patients where an automated device was used to maintain cuff pressures [20].
Interestingly, despite correctly pressurised cuff s in commonly used standard high-volume low-pressure   [21][22][23]. Th is stems from the basically fl awed design of these cuff s. Th e diameter of the fully infl ated cuff is up to twice as large as the diameter of the tracheal lumen, so that once infl ated to the correct pressure in the trachea, these cuff s remain only partially infl ated and result in the formation of folds along the cuff that then channel oropharyngeal secretions into the trachea [21]. Bench studies have clearly demonstrated these fi ndings ( Figure 2) and identifi ed major design features that infl uence the size of these leaky folds and channels.

Excess cuff material and shape of the cuff
Newer tracheal tube cuff designs such as the Lo-Trach™ (Intravent) low-volume low-pressure cuff and the Microcuff ™ tube (Kimberly-Clark) have shown promising results with signifi cant reduction in leakage of fl uid placed above the cuff s [24][25][26]. Th e Microcuff ™ tube has an elongated cylindrical-shaped cuff that results in a fully infl ated cuff in situ with minimal excess cuff material at acceptable intracuff pressures of 20 to 30 cmH 2 O. Th e fully infl ated cuff leads to limited or even absence of channel formation and the shape of the cuff results in a larger surface area in contact with the trachea. Microaspiration of blue dye placed above the cuff (as demonstrated by bronchoscopy) was also shown to be reduced and delayed in patients intubated with the Mallinckrodt/ Covidien SealGuard™ tube with an inverted pear-shaped (conical) cuff [27]. Th is pear-shaped design was fi rst described in 1999 by Young and Blunt [28], and subsequently shown to provide better sealing properties across a wider range of tracheal diameters when compared to the cylindrical-shaped cuff [29].

Tracheal tube cuff material
A major advance in the tracheal tube cuff design has been the introduction of thinner (7 μm thick) polyurethane (PU) material [29]. Th ese cuff s have consistently been shown to form narrower folds/channels with reduced leak age than tubes with thicker (50 μm) polyvinyl chloride (PVC) cuff s [21,25,30]. Clinical studies have also confi rmed reduced leakage of subglottic material [27], reduced pepsin levels in tracheal secretions [31], and lower rates of nosocomial post-operative pneumonia in patients following cardiac surgery randomised to endotracheal tubes with cuff s made from PU [32]. A retrospective analysis of intubated patients following the intro duction of tracheal tubes with PU cuff s showed a reduction in VAP from 5.3 to 2.8 per 1,000 ventilator days [33]. In an elegant in vitro study of six commercially available tracheal tubes, Zanella and colleagues [30] showed that all cuff s made from PVC (be they conical or cylindrical in shape) demonstrated signifi cant leakage over a 24-hour period, except when a positive

Subglottic secretion drainage and ventilatorassociated pneumonia
Removal of oropharyngeal secretions that have pooled above the tracheal tube cuff by subglottic secretion drainage (SSD) further reduces microaspiration [34][35][36][37] ( Figure 3). Specially designed tracheal tubes are widely available, with a separate lumen or lumens that open above the cuff and allow intermittent or continuous drainage of the pooled secretions. In 2005, a metaanalysis examining fi ve prospective studies using SSD showed a 50% reduction in the incidence of pneumonia in patients randomised to SSD [38]. Th is eff ect was more pronounced in those who were intubated for more than 72 hours and for early onset VAP. A reduction in the length of mechanical ventilation and ICU stay (2 and 3 days, respectively) makes this a favourable intervention, and has now been included in many national VAP prevention bundles [39]. Further convincing evidence has also been published since this 2005 meta-analysis [34,36,37]. However, the uptake of SSD into clinical practice has been slow [40], and this is likely because of (i) confl icting clinical trial evidence, (ii) safety concerns surrounding laryngeal/tracheal damage caused by the stiff er nature of these tubes, (iii) suction damage to the tracheal mucosa, (iv) the higher cost of the tubes, and (v) the fact that the studies often only show that early VAP is reduced. Of nine prospective, randomised controlled studies of SSD, six did not look at adverse eff ects of SSD, two reported no adverse events and one reported a signifi cant increase in the risk of laryngeal oedema requiring re-intubation in patients intubated with an SSD tube [13,41]. Furthermore, Dragoumanis and colleagues [42] described herniation of tracheal mucosa into the dorsal SSD suction port of the fi rst generation Hi-Lo® Evac endotracheal tube (Hi-Lo Evac, Mallinckrodt, Athlone, Ireland), resulting in failure of subglottic suction and the risk of mucosal injury due to tissue ischaemia. Of concern also is the fact that in sheep, signifi cant tracheal mucosal injury has been observed with continuous aspiration of subglottic secretions [43].
Th ese observations have encouraged the development of second and third generation tube designs, which have overcome some of these concerns. In some tubes the position of the SSD suction port has been placed adjacent to the tracheal tube cuff , lifting its opening away from the tracheal mucosa. Also, improved cuff designs that produce a better tracheal seal (tapered/conical PU or lowvolume controlled-pressure cuff s) have allowed hourly intermittent suctioning, instead of continuous suction [37]. Avoiding continuous suctioning strongly suggests that ischaemic lesions of the posterior tracheal wall can now be avoided. Th e LoTrach™ tube has further addressed these concerns by incorporating three subglottic suction ports adjacent to the tracheal tube cuff , making subglottic suctioning more effi cient and less traumatic. Furthermore, its fl exible design conforms to the shape of the upper airway, potentially reducing laryngeal injury [24].
In order to accommodate the SSD channel into the wall of the tracheal tube, the outer diameter of the tube in some of the designs has had to be enlarged by approximately 1 mm on average and some designs are slightly stiff er. Th is needs to be taken into consideration when sizing these tracheal tubes in order to avoid laryngeal injuries, and is a potential area for future refi nement in design.

Tracheal tube biofi lm management
Microbial biofi lms are present on the luminal surface of endotracheal tubes of all patients ventilated in the ICU and form within hours of tracheal intubation, becoming abundant at 96 hours [12,44,45]. Whilst the exact sequence of tube colonisation and infection is unclear, it is thought that the microbial biofi lm may act as a reservoir of pathogens causing recurrent infections [45]. Adair and colleagues [12] showed that 70% of patients with VAP had the identical pathogen isolated from their tracheal tube and lower respiratory tract. Furthermore, biofi lms are associated with developing microbial bacterial resistance [46].
Endotracheal tubes coated with anti-microbial silver hygrogel showed delayed and reduced bacterial colonisation in a mechanically ventilated animal model [47]. In one of the largest multi-centre prospective studies of VAP prevention (NASCENT study), Kollef and colleagues [15] compared rates of VAP in 2,003 patients randomised to intubation with either a silver-coated tube (Agento IC, CR Bard Inc., Covington, GA, USA) or the Hi-Lo Endo tracheal Tube (Mallinckrodt, St Louis, MO, USA). Th is study showed a statistically signifi cant relative risk reduction of 36% in the occurrence of VAP in patients intubated with silver-coated tubes, but failed to show a reduction in length of mechanical ventilation, ICU stay or mortality. Interestingly, the VAP rates in this study were lower than previously quoted rates (5 to 10% versus the 10 to 20% in other studies frequently quoted).

Tracheotomies and ventilator-associated pneumonia
Another frequently debated point is the infl uence of the timing of tracheotomy on the development of VAP. While Rumbak and colleagues [48] showed a signifi cant reduction in the development of VAP with early tracheotomy (within 48 hours of mechanical ventilation) versus delayed tracheotomy (after 14 to 16 days), a later but larger multicentre study showed no diff erence in the rate of pneumonia in patients that underwent early tracheotomy versus patients who had more prolonged endotracheal intubation [49]. Th e largest multicentre study on this question showed no diff erence in the incidence of VAP with early tracheotomy (after 6 to 8 days of tracheal intubation) versus later tracheotomy (after 13 to 15 days) [50]. One interpretation of these observations is that while a patient has any artifi cial cuff ed airway in place (whether endotracheal or tracheostomy tube), this predisposes to some degree of microaspiration and biofi lm formation on the inner lumen, which increases the risk for VAP. However, many of the novel design features already mentioned are also now being incorporated into new tracheostomy tubes.

Other strategies
It is clear that reducing the incidence of VAP requires new strategies incorporating guidelines, resources, educa tion and leadership [51,52]. Patient positioning, sedation holds, adequate and frequent oral hygiene, SSD, maintenance of cuff pressures and guidelines on stress ulcer prophylaxis are listed in a high impact care bundle to prevent VAP that has recently been published [39]. Th e integration of new and proven technology into this strategy will further improve its success. Current evidence highlighting the role of tracheal tubes in the pathogenic processes leading to the development of VAP may be ethically diffi cult to ignore. Th e newer tracheal tubes with tapered or cylindrical cuff s made from thin PU material and incorporating subglottic ports for intermittent suction of subglottic secretions should be an addition to this VAP bundle. Furthermore, automating cuff pressures using devices such as the disposable Portex PressureEasy® cuff controller or the Venner™ PneuX PY™, comprising the Venner tracheal seal monitor device in conjunction with the LoTrach™ ET Tube and LoTrach™ T Tube, would limit exposure to low (<20 cmH 2 O) and high (>30 cmH 2 O) cuff pressures (that is, by continuous monitoring and maintaining of cuff pressure). Interestingly, in a study discussed earlier, using a constant cuff pressure controller device failed to show statistically signi fi cant reduction in VAP when compared to an 8-hourly nurse monitored method (22% VAP in intervention group versus 29% in control group) [19]. However, the investigators used the leaky PVC high-volume low-pressure cuff ed tubes for both groups, and highlight the point that using single interventions to reduce VAP is unlikely to be as successful as multiple interventions [53].

Attributable mortality and ventilator-associated pneumonia
Some clinicians remain concerned about the lack of eff ect on mortality in the published studies of VAP reduction interventions. Mortality is clearly the hardest end-point for clinical trials in the critically ill, but softer outcomes such as VAP should not be ignored [54]. It is important to distinguish the consequences of VAP from the progression of an underlying illness, and a key variable is the attributable mortality of developing VAP. Th e literature is not clear in answering this question, since it may well depend on the severity of underlying disease and acute respiratory failure as a result of pneu monia [55,56], the case-mix of the population [57], the adequacy of initial empiric treatment [58], and the infect ing agent [59]. Methodological diff erences, such as varia bles and covariates used to match control patients in the various studies, add to this uncertainty [3].
Th e attributable mortality of VAP is widely quoted as 0 to 36% [60][61][62][63]. By way of illustration, let us consider a hypothetical VAP trial involving 800 patients, and a proposed new intervention that reduces VAP by 50%. In this population the control arm event rate (VAP) is about 20%, and the 28-day mortality is also about 20%. If the trial showed that VAP was reduced to 10% in the intervention arm (a 50% reduction), and if it was assumed that the attributable mortality due to VAP is about 10%, then at very best mortality could only be reduced by a statistically non-signifi cant (but clinically important) four patients. Hence the reason why so many of the current studies are not powered for mortality.

Should the newer tubes be used for all intubated patients admitted to the ICU?
Th e incidence of VAP increases with length of mechanical ventilation [64,65], and the evidence presented above points towards more benefi t being gained by patients intubated for prolonged periods [15,38]. Various tools have been proposed to predict length of mechanical ventilation, but these are only applicable at 24 to 48 hours following institution of mechanical ventilation [66,67]. It could therefore be argued as reasonable to intubate all patients admitted to the ICU who are expected to be intubated for longer than 24/48 hours with a newer generation tracheal tube.

Can we justify the higher cost of the newer tubes?
When compared to the old generation (leaky) tubes without subglottic suction, which cost about $2 each, and considering that each new case of VAP leads to an increased estimated cost of approximately $5,000 to $26,000 [5], then it would be fi nancially benefi cial to pay a lot more for the 'interface' (that is, the tube) between the ventilator and the patient -an 'interface' that can potentially contribute to or prevent VAP. A cost-analysis performed by Shorr and colleagues [68] found a $12,840 reduction in hospital costs per single case of VAP prevented as a result of introducing $90 silver-coated tubes compared to using standard $2 non-coated tubes, making this a very fi nancially viable inter vention. In fact, the break-even cost of the silver-coated tubes was calculated to be $388. Even 'back of the envelope' conservative calculations that assume the cost of a VAP to be just $5,000 would support an investment of $49 per patient in a new intervention if it merely reduces the rate of VAP by just 1% absolute!

Conclusion
In a recent editorial, Valles, Blanch and Respiratorias [69] applied Sutton's law to interventions that prevent VAP. Willy Sutton was a prolifi c bank robber, having stolen $2 million over his career. When asked by a reporter why he continued to rob banks, he replied, 'Because that is where the money is' . Application of Sutton's law -'Go where the money is' -to the paradigm of VAP prevention strongly favours multi-faceted strategies aimed at reducing aspiration of oropharyngeal secretions. With the increasing weight of evidence pointing at the role of the tracheal tube design and maintenance of adequate cuff pressures, is it really good medical practice to continue to use 'standard cuff ed tubes'? Abbreviations PEEP, positive end-expiratory pressure; PU, polyurethane; PVC, polyvinyl chloride; SSD, subglottic secretion drainage; VAP, ventilator-associated pneumonia.
Competing interests PSZ has no competing interests. In the past 3 years, DW has acted as a paid consultant or given lectures on VAP for Covidien, Kimberly-Clark and Bard.