SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Modern ventilators provide capnography monitoring in patients with tracheal tubes, in compliance with national and international recommendations. This technology is often not used when patients’ lungs are non-invasively ventilated; however, it should be accessed immediately following tracheal intubation to confirm tube placement. This study assessed the effect of ventilation interface design on the speed with which capnography can be activated by comparing the Dräger Evita 4 and Dräger V500 before and after a specific training episode. We configured the V500 to have a capnography activation button on the front screen in contrast to the Evita 4 which requires a sequence of actions to access capnography monitoring. We used a randomised crossover design, measuring time to monitoring activation, and repeated the study after 3 months. Survival analysis showed significantly quicker activation associated with ventilator choice (V500, p < 0.0001) and training (p = 0.0058). The training improved activation speed with both machines, though this was only significant for the Evita 4 (p = 0.0097).

Capnography provides a measurement of the partial pressure of carbon dioxide in inspired and exhaled respiratory gases and has been used in clinical practice since the 1960s [1]. Waveform capnography has been the accepted standard to ensure correct tracheal tube positioning during induction and maintenance of anaesthesia for many years [2, 3]. The Royal College of Anaesthetists (RCoA) states that where capnography is not used, the clinical reasons must be clearly documented [4]. Recommendations for anaesthetic practice were made following studies reporting anaesthetic deaths and neurological injuries as a result of hypoxia associated with unintentional oesophageal intubations, unrecognised extubation and disconnection from mechanical ventilation [5–9]. These incidents are now less common [10, 11] following improvements in monitoring during anaesthesia, including the adoption of capnography as a standard [3].

Surveys have shown that the routine use of waveform capnography is less consistently applied in intensive care practice [12, 13]. An Intensive Care Society survey indicated that 44% of intubations performed in UK intensive care units (ICUs) were carried out without capnography to confirm tube placement [14]. Thomas and McGrath reviewed airway incidents in critical care related to 44 675 patient safety incidents submitted to the UK National Patient Safety Agency between 2005 and 2007 [15]. Of 1085 airway related incidents, 61% were associated with harm and 25 (2%) contributed to death. Capnography was not identified as being used in any of the incidents of tube displacement or obstruction and was only referred to three times in the entire series of 1085 airway incidents [15]. Furthermore, the fourth National Audit Project (NAP4) of the RCoA and Difficult Airway Society revealed that failure to use capnography contributed to 74% of airway related deaths or neurological injury during anaesthesia, within ICUs and in emergency departments [16, 17].

Until recently, capnography during airway placement in critical care had not been formally covered by guidelines [18]. However, the Association of Anaesthetists of Great Britain and Ireland did recommend the use of capnography in critical care during intubation and throughout mechanical ventilation [19] and the Australasian Joint Faculty of Intensive Care Medicine advised that capnography should be used for intubation and then be available at each bed area in intensive care [18].

Newly revised Intensive Care Society recommendations [18] now state that ‘capnography should be used for all critically ill patients during the procedures of tracheostomy or endotracheal intubation when performed in the intensive care unit’. Capnography devices available for use in critical care units may be either integral to the ventilator or part of a multiparameter monitor.

Anaesthesia induction is normally performed in a controlled environment with a dedicated capnograph that is checked and working before the start of an operating list. Non-invasive ventilation (NIV) is increasingly used as first line therapy in acute respiratory failure, but normally requires deactivation of the capnograph as expiratory gases are vented to the atmosphere without passing into the expiratory limb of the ventilator tubing. Many critically ill patients now undergo a trial of NIV and only require tracheal intubation if this fails. Patients failing on NIV are a particularly challenging group for tracheal intubation. It is often an urgent procedure and needs to be performed in a timely manner with immediate confirmation of successful intratracheal placement. Bag and mask ventilation may not be effective and waking the patient and restarting is not an option. We noted, in our institution, a number of critical incidents relating to the delayed activation of capnography during intubation in the ICU. These involved experienced nurses and doctors fully familiar with and trained in the use of the ventilator (all ventilators on the ICU were the Evita 4 (Dräger, Lubeck, Germany)). Events occurred following urgent intubations when ventilators set in the NIV mode had the capnograph disabled; the route to activation of this monitoring on the ventilator interface confused the attending nursing and medical staff. It seems likely that a more intuitive interface would make activation of the capnography simpler and therefore quicker.

The Dräger Evita 4 and V500 are both mechanical ventilators used for invasive and NIV in ICUs. However, the two ventilators differ in terms of their interface. The Evita 4 has a number of fixed/hard-wired buttons in combination with a touch screen and the V500 has a fully configurable touch screen alone. The route to capnography activation for the Evita 4 requires the ‘alarm limits’ button to be pressed (one of 20 fixed buttons on the face of the ventilator), followed by the touch screen ‘monitoring’ button and then the touch screen ‘CO2 on’ button. The V500 has a locally configurable interface so that the touch sensitive button can be placed on the main screen for immediate activation of capnography in clinical emergencies. In consideration of our critical incidents relating to delayed capnography activation, as a risk management strategy, we chose to evaluate whether changing the interface of the configurable V500 would improve the ability of staff to activate capnography in a timely manner. In addition, we decided to evaluate whether or not specific staff training would effectively improve this.

We studied the speed and ability with which nursing and medical staff, who were familiar with the Evita 4 but unfamiliar with the newer V500, successfully activated capnography. We also compared differences between the Evita 4 and the V500 before and after a specific episode of individualised training and assessment.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The study was approved by our institution’s Research and Development Committee, which felt there were no material ethical issues, and that Research Ethics Committee authorisation was not required under the terms of the governance arrangements for UK research ethics committees [20]. The study was logged with the institution’s audit department. All subjects provided written informed consent to study participation, and understood that this was voluntary.

A convenience sample of critical care staff who were familiar with, and declared themselves competent with, the Evita 4 ventilator participated in the study, which used a randomised crossover design. The Evita 4 had been used for eight years in our ICU. Participants had no or minimal experience with the V500.

The following scenario was read to each participant: “Your patient has just been tracheally intubated following an episode of NIV. The anaesthetist urgently requires capnography to verify correct placement of the tracheal tube. The capnograph has been deactivated in the NIV mode and you must turn it back on in the ventilator menus as quickly as possible.” The ventilators, Evita 4 and V500, were presented in a random order and in controlled ventilation mode with the default front screen visible and with capnography disabled. Each staff member was required to activate capnography as quickly as possible. The time taken to activate capnography successfully was recorded for each ventilator. The procedure was terminated at 120 s if capnography was not activated. Following the first study period (for both ventilators) each participant was formally taught how to activate each of the two ventilators by means of a hands-on demonstration, and the background and importance were explained. Specific training was given to each participant whether they had been successful or not in activating capnography, although they were not told of the intention to re-test them in the future. The experience of participants in using the Evita 4 (in years) was recorded. Three months following the test and teaching episode, the participants were re-tested on their ability to turn on the capnography in both ventilators.

Survival analysis was performed on the time to event data from the two ventilators (V500 and Evita 4) and two timeframes (pre- and post-training) using the Cox proportional hazards regression model. Data were right-censored at 120 s if activation of capnography had not been completed by this stage. Further analyses were then performed on the ventilator subsets of the data to try to identify if training affected the two subsets. Analysis was performed using the R project statistical software [21].

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

For the ‘pre-training study’ (Fig. 1), 31 members of staff (17 nursing staff, 10 medical staff, four operating department practitioners) were compared using the Evita 4 and the V500.

image

Figure 1.  Survival analysis for the time taken to activate capnography with the Evita 4 (black line) and the V500 (grey line) before training.

Download figure to PowerPoint

The staff had a median (IQR [range]) of 6.0 (1.8–10.0 [0.1–12.0]) years of experience of using the Evita 4 ventilators. None had experience with the V500.

For the ‘post-training study’ (Fig. 2), 19 of the same participants (13 nursing staff, five medical staff, one operating department practitioner) were available for the follow-up study. The staff had a median (IQR [range]) of 5.5 (1.6-10.0 [0.1–12.0]) years of experience in using the Evita 4 ventilators. Other than the training episode three months previously, none had experience with the V500.

image

Figure 2.  Survival analysis for the time taken to activate capnography with the Evita 4 (black line) and the V500 (grey line) after training.

Download figure to PowerPoint

Both the model of ventilator used and training process were independent significant covariants of hazard. When analysed separately (Figs. 3 and 4), the faster activation associated with training was significant for the Evita 4, but not for the V500 (Table 1).

image

Figure 3.  Survival analysis for the time taken to activate capnography using the V500, before training (black line) and after training (grey line).

Download figure to PowerPoint

image

Figure 4.  Survival analysis for the time taken to activate capnography using the Evita 4, before training (black line) and after training (grey line).

Download figure to PowerPoint

Table 1.   Data on time of activation of capnography for the Evita 4 and V500 ventilators before and after training. Cox proportional hazards model coefficients for the initial model (all data), and the two secondary models (separated by ventilator used to investigate the effect of training on each). Values are number (95% CI).
DataCovariantCoefficientp value
AllVentilator0.346 (0.216–0.553)<0.0001
Training1.876 (1.200–2.934)0.0058
V500Training1.545 (0.855–2.794)0.150
Evita 4Training2.501 (1.249–5.009)0.0097

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Our results showed that despite undergoing the standard training as approved by Dräger, and with a median of 6 years of experience using the Evita 4 ventilation unit, 45% of the individuals were unable to activate capnography on the pre-training trial within 120 s; 16% still failed to turn on capnography after specific training related to this problem. Furthermore, when capnography was activated on the Evita 4, activation was slower than when staff activated capnography on the V500 for both pre- and post-training trials.

Thus, the results showed two effects. First, activation of capnography is quicker on the V500 (with an appropriately configured interface so that the capnography button is on the main screen). Secondly, activation of capnography is quicker following specific training, although the success and speed of activation for the Evita 4 (even after training) never reached that of the V500 (even when participants had no experience of the V500, and had not yet received the training). The difference between these ventilators in terms of capnography is in the design of the ventilator interface. We have shown that the V500, when configured appropriately, is much more intuitive than the Evita 4. Configuring the screen of the V500 so that the capnography button is on the main screen increased both the speed and the ability to activate capnography. Our results suggest that the design of the ventilator interface is an important factor to be considered by manufacturers due to its important role in safety when a patient’s trachea is being intubated.

Cook et al. observed more serious incidents when staff were less experienced and inadequately trained with equipment [17]. However in our study, if the interface is poor then specific training is only partially effective; 16% of experienced Evita 4 users were still unable to activate capnography within 120 s, even after training. Therefore, it appears that using equipment with good intuitive and ergonomic design (such as the V500), coupled with specific training, is the best method for ensuring the safety of patients in ICU.

In conclusion, our research indicates that the design of the ventilator interface affects capnography activation. A poorly designed interface is only modestly improved by staff training and this should not be relied upon. It is essential that this is considered by manufacturers during their design of mechanical ventilators for the ICU. When purchasing ventilators this should be taken into consideration. Alternatively, stand-alone capnographs, present at every bed space and ready for immediate use, could be considered.

Acknowledgements

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Dräger supplied the V500 ventilators that were used in this study. No external funding and no competing interests declared.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References