The effect of inner tube placement on resistance and work of breathing through tracheostomy tubes: a bench test


Correspondence to: S. Fletcher



To reduce the risk of tracheostomy tube blockage, a removable inner tube can be used. However, this will reduce the size of the lumen and will increase airflow resistance and work of breathing. The magnitude of this increase in workload is unknown. We undertook a bench test to measure the effect. A lung model was developed to ‘breathe’ through the tracheostomy tube. We created pressure–volume curves from which we calculated work of breathing with and without an inner tube using 6–10 mm tracheostomy tubes over a range of respiratory rates and tidal volumes. The inner tube increased the resistive work of breathing by an average factor of 2.2. The extra work of breathing imposed easily exceeded the normal total work of breathing. Our results will aid a risk–benefit analysis when deciding whether to use inner tubes. Selecting a larger tracheostomy tube is likely to aid weaning from mechanical ventilation.

Failure to wean critically ill patients from mechanical ventilation is largely due to an imbalance between the required work of breathing and weakness of the patient's respiratory pump (due to muscle wasting and critical illness neuromyopathy). Work of breathing is increased due to a combination of poor respiratory system compliance and the imposed resistive work of breathing due to the artificial airway and breathing system. In these patients, a tracheostomy may be formed with the aim of facilitating weaning mainly by reducing the need for sedation but also by reducing airway resistance when compared with trans-laryngeal tracheal tubes [1, 2]. Blockage of the tracheostomy tube with accumulated secretions may be life-threatening and the use of a replaceable inner tube is suggested as a way of reducing this risk [3]. However, inserting the inner tube decreases the internal diameter of the tracheostomy lumen and therefore will increase the resistance and resistive component of the work of breathing. According to the Hagen–Poiseuille law, assuming laminar flow, resistance is inversely proportional to the fourth power of the radius. Thus, a small change in radius may have a significant effect on work of breathing and may adversely affect weaning. If flow is turbulent then the effect of a change in radius is even more marked. We theorised that there was likely to be a mixture of laminar and turbulent flow in the native and artificial airways; furthermore, the proportion of turbulent flow would increase as the peak flow rates increased. For these reasons, the effect of an inner tube on resistance and work of breathing is not amenable to calculation. Therefore, we undertook a bench study to assess that effect of the introduction of the inner tube using one of the more commonly used tracheostomy tubes in the UK (Table 1).

Table 1. Internal diameters of Portex Blueline Ultra tracheostomy tubes
Tracheostomy tube sizeExternal diameter of tracheostomy tube; mmInternal diameter of tracheostomy tube without inner tube; mmInternal diameter of inner tube; mm


In the first test (steady-state), we looked at resistance to flow through the tracheostomy tube under conditions of constant flow. In the dynamic tests, we measured work of breathing using a lung model. Portex® Blueline Ultra® tracheostomy tubes of sizes 6–10 were studied (Smiths Medical International Ltd, Hythe, UK). Measurements were made using a pneumotachograph (VT Plus HF; Fluke Biomedical, Everett, WA, USA). As expiration is normally passive, we only considered inspiratory work of breathing. Five tracheostomy tubes of each size were tested and data were expressed as the mean driving pressure.

For the steady-state test, we attached a gas source capable of delivering variable flows to the upstream side of the pneumotachograph via 22-mm tubing. The tracheostomy tube to be tested was inserted via its 15-mm connector to the downstream side of the pneumotachograph. The driving pressure needed to create flows from 20 to 100 l.min−1 was then recorded at 20-l.min−1 steps for each size of tracheostomy tube, with and without the inner tube in place.

For the dynamic test, we used a model of spontaneous respiration as described previously (Fig. 1) [4, 5]. It comprises two mechanically linked bellows: the ‘master bellows’ are driven by a Siemens Servo 300A ventilator (Siemens Healthcare, Surrey, UK) using a volume control mode, and with an inspiratory to expiratory (I:E) ratio of 1:2. The ‘slave bellows’, which represent the ‘lung’, were connected via the pneumotachograph to the lower end of an anatomically accurate model trachea (taken from an intubation manikin), into which we inserted the test tracheostomy tube [2, 6]. The proximal end of the tracheostomy tube was open to atmosphere. The tracheostomy tube was inserted into the model trachea through an incision at a level approximating the space between the second and third tracheal rings. The cuff was inflated and the correct orientation of the tip of the tube parallel to the tracheal wall was confirmed visually.

Figure 1.

Diagram of the lung model used for the bench tests.

Synchronous pressure and flow data were sampled at 50 Hz throughout the respiratory cycle at different respiratory rates (10, 20, 30, 40 min−1) and different tidal volumes (300, 500, 700, 900 ml). The data were then uploaded to Microsoft Excel allowing us to calculate the area of the inspiratory portion of the pressure volume loops to obtain the resistive work of breathing, measured in Joules [7]. To enable comparison, these values were then indexed by tidal volume to Joules per litre (resistive work of breathing (J.l−1) = work for one breath (J)/tidal volume (l)).

In a bench test not involving biological material, the only sources of measurement variability and error reside in the test subject (in this case, tracheostomy tubes) and in the measuring equipment. The manufacturing tolerances of the Portex Blueline Ultra tracheostomy tubes are ± 0.05 mm on the diameter and ± 0.1 mm on the length (information supplied by Portex). The flow measurement error of the pneumotachograph is ± 2%. Therefore, the results derived from one exemplar tracheostomy tube should be valid. However, to confirm this, we ran the steady-state tests with five tracheostomy tubes of each size and calculated the mean and standard deviations for the driving pressure. The variability in the measurements was very small as predicted and the dynamic tests were therefore performed using one representative tube of each size.


Steady-state test

In the steady-state tests, the driving pressures required to produce a given flow increases in a curvilinear fashion confirming turbulent flow in the tracheostomy tubes (Fig. 2). Table 2 shows the mean (SD) driving pressures at each flow rate.

Table 2. Results of the steady-state tests. Figures are mean (SD) driving pressures in cmH2O
Tube size ± inner cannula20 l.min−140 l.min−160 l.min−180 l.min−1100 l.min−1
6 inner4.28 (0.08)14.28 (0.22)30.78 (0.4)51.94 (0.62)77.78 (0.95)
6 no inner1.96 (0.06)7.06 (0.09)15.04 (0.23)25.1 (0.36)37.38 (0.55)
7 inner3.22 (0.13)10.32 (0.24)21.2 (0.5)36.7 (1.19)54.8 (1.82)
7 no inner1.12 (0.04)3.66 (0.06)7.6 (0.16)12.82 (0.22)19.1 (0.54)
8 inner1.72 (0.05)5.56 (0.09)11.68 (0.22)19.84 (0.25)29.84 (0.45)
8 no inner0.62 (0.04)2.18 (0.11)4.64 (0.15)7.94 (0.29)11.92 (0.46)
9 inner1.02 (0.04)3.28 (0.04)6.9 (0.07)11.42 (0.16)16.94 (0.29)
9 no inner0.42 (0.04)1.5 (0)3.16 (0.05)5.18 (0.04)7.62 (0.08)
10 inner0.6 (0)2.02 (0.04)4.28 (0.04)7.1 (0.07)10.56 (0.06)
10 no inner0.3 (0)1.06 (0.05)2.24 (0.05)3.78 (0.04)5.52 (0.04)
Figure 2.

Pressure-flow characteristics of differing tracheostomy tubes with and without inner tubes: (image) 6 inner, (image) 6 no inner, (image) 7 inner, (image) 7 no inner, (image) 8 inner, (image) 8 no inner, (image) 9 inner, (image) 9 no inner, (image) 10 inner, (image) 10 no inner.

Dynamic test

Figure 3a–d show the measured imposed work of breathing in the tracheostomy tubes with and without their inner cannula at respiratory rates 10–40 min−1. The normal total work of breathing in healthy subjects (thought to be the optimal work of breathing) of 0.5 J.l−1 is also shown [7]. As the tidal volume and the respiratory rate rises, there is an increase in resistive work of breathing. This increase is more marked in the smaller tracheostomy tubes.

Figure 3.

(a–d) Imposed resistive work of breathing for differing tracheostomy tubes (with and without inner cannula) at increasing respiratory rates (a, 10 min−1; b, 20 min−1; c, 30 min−1; d, 40 min−1). Key for tube sizes ± inner cannula: (image) 6 inner, (image) 6 no inner, (image) 7 inner, (image) 7 no inner, (image) 8 inner, (image) 8 no inner, (image) 9 inner, (image) 9 no inner, (image) 10 inner, (image) 10 no inner. (– –) Normal total work of breathing.

Figure 4a–d show the additional imposed work of breathing due to insertion of the inner cannula. This was calculated by subtracting the resistive work of breathing of each tracheostomy tube without its inner cannula from the value with its inner cannula for each respiratory rate/tidal volume combination. It can be seen from these graphs that at low respiratory rates the imposed resistive work of breathing added by the inner cannula is modest. However, as the respiratory rates and tidal volumes increase, the added work by placing the inner cannula rapidly exceeds the optimal total work of breathing in the size 6 and 7 tubes.

Figure 4.

(a–d) Additional work of breathing due to inner tube for differing tracheostomy tubes at increasing respiratory rates (a, 10 min−1; b, 20 min−1; c, 30 min−1; d, 40 min−1). (– –) Normal total work of breathing. Key for tube sizes: (image) Size 6, (image) Size 7, (image) Size 9, (image) Size 10.

In an attempt to summarise these data, we calculated the average ratio of the work of breathing with and without inner tube for each size of tracheostomy, for each of the test conditions. For the size 6, the work of breathing on average increased by a factor of 2.2; for size 7, a factor of 2.8; for size 8, a factor of 2.3; for size 9, a factor of 1.9; and for size 10, a factor of 1.6.


The results from the steady-state test show that flow in the tracheostomy tubes is turbulent, and therefore resistance is inversely proportional to the fifth power of the radius. The driving pressures from the static tests are equivalent to the negative inspiratory pressure a spontaneously breathing patient would have to generate in the trachea to produce an equivalent gas flow. The pressure required to produce greater gas flows increases rapidly in a non-linear fashion, and the effect is amplified with smaller internal diameters. Inspiratory pressures of greater than 10 cmH2O are required to produce flows of 60 l.min−1 or more in tubes with an internal diameter of 6 mm or less. Flow rates of more than 100 l.min−1 are often seen in respiratory failure [5]. These data suggest inter alia that it would be almost impossible to sustain these flow rates when a size 6 or 7 tracheostomy of this type is used with its inner cannula.

The results from the dynamic tests show that placement of the inner cannula in Portex Blueline Ultra tracheostomy tubes significantly increases imposed inspiratory resistive work of breathing. The effect was greatest with the size 7.0 tube, with an increase in resistive work of breathing by a factor of 2.8 at clinically relevant respiratory rates and tidal volumes. This is because the inner cannula in the 6.0 tube only reduces the internal diameter by 1 mm compared with a 1.5 mm decrease in internal diameter in all other tubes tested (see Table 1). The 7.0 tube therefore has the greatest proportional change in internal diameter when the inner tube is inserted.

Total work of breathing in normal subjects is in the range of 0.3–0.6 J.l−1 [7]. The optimal work of breathing whilst weaning from a ventilator is difficult to assess; if too low, respiratory muscle wasting can develop; if too high, muscle fatigue and respiratory failure may be induced. The total work of breathing in the normal healthy range would seem to correspond to a reasonable muscle and energy load in both weaning and respiratory failure [7]. Our experiment only considers inspiratory imposed resistive work of breathing. Even so, the results show that in the smaller tubes with the inner cannula in place, the resistive work of breathing starts to exceed normal total work of breathing at fairly low tidal volumes and respiratory rates. Increased imposed work of breathing has been shown to cause failure to wean from mechanical ventilation [8]. In vivo, the imposed work of breathing will be cumulative with the resistive work of breathing of the rest of the bronchial tree and the elastic and viscous work of breathing used to expand the lungs. If the total work is already high due to poor lung compliance or airway narrowing, then even small increases in imposed work are highly likely to cause respiratory decompensation and failure to wean.

The increased imposed resistive work of breathing caused by use of the inner cannula does, however, need to be weighed against the important benefits that the inner tube provides. It can be removed readily and cleaned or replaced to remove secretions. Encrusted secretions can be difficult or impossible to remove with suctioning alone. It has been shown that organised secretions in tracheal tubes can reduce the effective internal tube diameter causing a significantly increased resistance and imposed resistive work of breathing [9].

Perhaps the major benefit of the inner cannula is the ability to remove it should the inner lumen become completely blocked. This makes treatment of a potentially life-threatening emergency situation much quicker and easier without the need for an emergency removal and/or replacement of a new tracheostomy tube.

In 2008, the UK Intensive Care Society published Standards for the care of adult patients with a temporary tracheostomy in which they state “Tracheostomy tubes with an inner cannula are inherently safer and are normally preferred” [3]. The document also states: “As a general rule, most adult females can accommodate a tube with an outside diameter of 10 mm, whilst a tube with an outside diameter of 11 mm is suitable for most adult males”. Using Portex Blueline Ultra tracheostomy tubes, this gives a maximum size of 7 in males and size 6 in females. With the inner cannula in situ, the internal diameter is 5.5 and 5 mm, respectively. Our results show that these tube sizes produce a marked restriction to flow and increased work of breathing and thus the guidelines may not be appropriate. Whilst a larger diameter lumen would appear beneficial, we are limited by the resulting outside diameter that the trachea is able to accommodate. Reducing the difference between the internal and external diameter would seem to be advantageous, but would require a thinner tube wall, whilst retaining wall strength. Portex Blueline Ultra tubes are made from thermosensitive PVC, which is designed to be rigid enough for percutaneous insertion and then soften once in situ to accommodate individual anatomy. Whilst thinner tubes are available, they tend to be rigid, for example, the Shiley and Trachoe. A more rigid tube will transmit more of the weight of the breathing system onto the tracheal wall, arguably increasing the risk of mucosal ischaemia and resultant fibrosis.

Imposed work of breathing during spontaneous respiration can be offset to a certain degree by using inspiratory assist. Our results suggest, however, that conventional pressure support ventilation may not fully overcome the resistive work imposed by smaller diameter tracheostomy tubes. Automated tube compensation employs a mathematical model to calculate the tube resistance for specific tube sizes at differing flow rates. The delivered pressure support can then be adjusted to offset the resistance. Automated tube compensation has been shown to be more effective at reducing imposed work of breathing compared with simple pressure support ventilation alone, whilst improving patient comfort, and maintaining cardiovascular stability [10-12]. The use of such ventilatory modes may improve weaning success, particularly when smaller tracheostomy tubes are employed. However, it is important to remember to set the actual internal diameter of the tube when an inner cannula is used, rather than the size written on the box or cuff e.g. a size 8.0 Portex Blueline Ultra with an inner cannula in situ has an inner diameter of 6.5 mm not 8 mm.

It is likely that the values for work of breathing for any given lumen size that we measured will be similar for the same sizes but different makes of tracheostomy tube. Differences may occur due to differing design (for example different radii, different lengths).

Our study has some limitations. At higher inspiratory pressures (> 30 cmH2O), the slave bellows tended to collapse; therefore, we could not generate data for the smaller tubes at the extremes of the test protocol. However, it is equally likely that a human would be unable to generate such negative pressures. Thus, loss of these data does not alter our conclusions. Another potential criticism is that we did not look at the expiratory work of breathing. Whilst expiration is normally passive, at expiratory pressures greater than 10 cmH2O, expiration becomes active [13]. Expiratory pressure will correlate with inspiratory pressure and thus we are likely to have underestimated total resistive work of breathing with smaller tubes and higher respiratory settings. However, this makes our findings more, not less, significant.

Our results will not guide decisions whether to use fenestrated tubes or deflate the tracheostomy cuff as part of the weaning process. However, if the tracheostomy tube is ‘capped off’, forcing the patient to breathe either through the fenestration and/or around the tube, then thought needs to be given to the size of the fenestration and the cross-sectional area of trachea between the tube and the tracheal wall. As we have demonstrated, small changes in airway dimensions can impose significant increases in respiratory workload.

In conclusion, although the decision of whether to use an inner cannula is complex in patients with weaning difficulty, our data provide an estimate of the increase in imposed work of breathing when they are used and thus should inform this decision making.

We would suggest the following:

  1. Select the largest possible tracheostomy tube.
  2. When a patient is failing to wean, consideration should be given to the possibility that the artificial airway is imposing excessive respiratory work.
  3. Do not routinely use inner tubes where there is adequate nursing and suitable medical supervision, but check regularly for accumulation of secretions.
  4. When a patient is to be discharged to a non-critical care environment with a tracheostomy in situ, then an inner tube should be inserted. A 24-h trial to demonstrate that the patient can cope with the extra work of breathing due to the inner tube should be performed before discharge.
  5. During weaning, automated tube compensation may help to offset the imposed work of breathing due to presence of an inner tube.

Competing interests

No external funding and no competing interests declared.