Time for capnography – everywhere

Capnography has been in routine use in anaesthetic rooms and operating theatres since 1988 [1, 2], and has proved invaluable in preventing harm from accidental oesophageal intubation and other airway management problems. Besides monitoring lung ventilation, capnography can provide safety-critical information about the patient’s circulation and metabolism, and can aid the diagnosis of low cardiac output states and pulmonary embolism [3]. Despite these significant contributions to patient safety, and in contrast to its rapid adoption and ubiquity in UK operating theatres, its use is not universal in clinical areas outside the operating theatre [4]. The intensive care unit (ICU) would seem an obvious place for the routine use of capnography. There are approximately 2000 Level-3 beds [5, 6] in the UK, and about 64 000 patients a year have their tracheas intubated and lungs ventilated in ICUs [7]. However, survey results show that only 20% of ICUs had capnography available for every bed in the UK in 2002 [8], and only 25% of ICUs used it for the continuous monitoring of patients requiring lung ventilation in 2009 [9]. The routine use of capnography in ICUs is more common in some other European countries, e.g. the Netherlands, but even there no formal standards exist for its routine use. During the development of intensive care, many of the monitoring advances such as continuous ECG monitoring, pulse oximetry, central venous cannulation, pulmonary artery catheterisation and currently, transoesophageal echocardiography, have all transferred from anaesthesia, particularly cardiothoracic anaesthesia, and it is astonishing that capnography has yet to achieve a similar successful transfer. It is tempting to speculate why capnography seems to differ in this respect from other successful monitoring techniques, and the academic point of imperfect correlation between the Paco2 and the end-expiratory fractional concentration of carbon dioxide (FEco2) in diseased lungs appears to have had a disproportionately inhibitory effect on the use of capnography in the critical care setting, whereas more important and fundamental patient safety issues have been seriously overlooked. Approximately 500 patient safety incidents associated with the use of airway devices occur in ICUs every year, of which 10% are associated with more than temporary harm [10]. These data, from the National Patient Safety Agency (NPSA) database, almost certainly represent an underestimate [11]. The majority of the incidents occur during the ‘maintenance phase’ of intensive care (88% in adults or older children; 80% in infants) rather than at the start of treatment, so continuous capnography would seem to be particularly important. The accidental removal of a tracheal tube can be seen or identified by a disconnection alarm, but the partial dislodgement of a tube, e.g. during airway procedures or when moving the patient, is more than twice as likely to cause harm [10], and this can often be identified by continuous capnography as a change in the capnograph waveform. Only a minority of critical incidents lead to litigation [12], but it is interesting to reflect that about 55 incidents per year, in which more than temporary harm results from problems with airway devices in critical care, are reported to the NPSA [10], although Cook et al.’s analysis of 12 years of the NHS Litigation Authority database showed that only about five litigation cases per year relate to airway complications in anaesthesia [13]. In this period, more than 10 times the number of patients per year would have undergone tracheal intubation in anaesthesia compared with critical care [7], and capnography would have been routine for all of them. Problems with tracheostomy tubes are responsible for 22% of airway incidents recorded in the NPSA database [10]. Capnography can be used for continuous monitoring and also as a safety check during the percutaneous insertion of tracheostomy tubes or when exchanging them. Patients in ICU are often the sickest patients in the hospital, likely to be the least tolerant of hypoxia, and are therefore exactly the group for whom one would expect the greatest precautions to be taken to minimise its occurrence. Similarly, incorrect placement of nasogastric tubes can cause harm to patients, and capnography can be used to exclude tracheal or bronchial placement, or as part of a technique to confirm their correct placement [14–17]. An evaluation of capnography monitoring during the apnoea test in brainstem dead patients has recommended that the FEco2 be systematically monitored and recorded, not least for medicolegal considerations, during the test [18]. It is important not to forget that although capnography primarily evaluates lung ventilation, it also reveals changes in cardiac output, and has been used to guide the efficacy of chest compressions in cardiac arrest [19] (see below). Data from the US National Registry of Cardiopulmonary Resuscitation (CPR) show that out of 86 748 in-hospital cardiac arrests, 40 050 (46%) occurred in ICUs [20]. The ICU was not only the most common location for cardiac arrests: ICU patients also had some of the worst outcomes; i.e. only 15.5% survival to discharge [20]. Having continuous capnography already in place and fully functioning for ICU patients receiving controlled ventilation may well help with the diagnosis, treatment, management and outcome of cardiac arrest, when it occurs. For all these reasons, the availability and routine use of FEco2 monitoring for every ICU bed space would seem essential [4]. There are now many intensive care monitors that include capnography, and a number of ICU ventilators used in Europe have had built-in FEco2 monitoring for some time. Another scenario in which patients’ tracheas are intubated is during resuscitation, and the same logic supporting the use of capnography elsewhere would seem to apply [21]. The value of having capnography at out-of-hospital cardiac arrests to confirm correct tracheal tube placement has been demonstrated and, in addition, the FEco2 trace can aid assessment of the effectiveness of chest compression, and when the person performing it is beginning to tire. This is confirmed in the Resuscitation Council (UK)’s 2010 Guidelines, which state that “there is increased emphasis on the use of capnography to confirm and continually monitor tracheal tube placement, quality of CPR and to provide an early indication of return of spontaneous circulation” [22]. It has been suggested that there is little point in having capnography at a cardiac arrest because one may not obtain any FEco2 from a pulseless patient to confirm tracheal tube placement, but this shows a lack of rational understanding. When a pulseless patient presents, the 2010 Resuscitation Guidelines say that CPR should be started first and then a laryngeal mask airway (LMA) may be considered as first choice to manage the airway, thereby avoiding any delay in providing effective cardiac massage [23]. Once the patient’s circulation has been restored, even if it is just by CPR, carbon dioxide will be detected from the LMA, albeit at low levels. When it is thought necessary to intubate the patient’s trachea as a further step in their management, the levels of carbon dioxide by now being provided by CPR or by the patient from their own return of spontaneous circulation can be used to confirm tracheal placement. It has been suggested that FEco2 values may be an indicator of overall successful outcome, although current evidence does not support the use of any particular FEco2 value as an indicator that resuscitation has been unsuccessful and can cease [22]. Most manufacturers are now producing defibrillators that include built-in FEco2 monitoring. An increasing number of patients with tracheal tubes in place are transferred both within and between hospitals, and Association of Anaesthetists of Great Britain & Ireland (AAGBI) recommendations have always included the use of continuous capnography for these transfers [24–26]. This approach has additional benefits when transporting trauma patients, especially those with brain injuries who are undergoing deliberate hyperventilation as part of their treatment [25]. Most patients whose tracheas have been intubated have the tubes removed in the operating theatre before transfer to recovery units, but the increasing use of the LMA has resulted in many anaesthetists’ leaving these airway devices in place until the patient is fully awake in the recovery room. The AAGBI guidelines say the recommended monitoring devices should continue to be used until the patient has recovered from the effects of anaesthesia [2]. It therefore seems illogical not to continue to monitor FEco2 in the recovery room until the patient has recovered from the immediate effects of the anaesthetic and the airway device has been removed [27]. Capnography has been shown to be useful for monitoring patients who are using patient-controlled analgesia (PCA) devices, as it provides an early warning of respiratory depression. The use of capnography in this situation has shown an incidence of PCA-related respiratory depression that is higher than that previously reported [28]. Using capnography in this way will help identify respiratory depression and other complications earlier, and should prevent serious adverse events and the need for potentially costly interventions. Similarly, capnography has been used to monitor patients who have been given postoperative spinal or epidural opioids [29, 30]. Respiratory rate is a core clinical observation [31] but the way it is measured varies. It is normally recommended that it is counted over one whole minute or two 30-s intervals, which can represent a significant investment in nursing time in the ward setting, and accurate rates may only be recorded as little as 37% of the time [32]. Capnography currently offers the most accurate and reliable means of measuring respiratory rate, with the availability of instant readings and, perhaps more importantly, trends. For patients already being given additional oxygen, capnography monitoring is easy and not an additional burden. In addition to the value of monitoring FEco2 when the airway is instrumented, capnography [33–35] also has a place for monitoring patients’ breathing when the airway is potentially compromised during sedation for diagnostic or therapeutic procedures. Anaesthetists have adapted existing equipment for this purpose, including the use of bacterial filters to protect the capnography monitors and subsequent patients from infection. Several commercial devices are now available that are specifically designed for the purpose [36]. The absolute values of FEco2 in this context will not directly correlate with Paco2 but the breathing pattern and respiratory rate are important safety assessments, particularly when the anaesthetist or operator cannot easily see the patient’s chest and abdomen move in a darkened X-ray room or cardiac catheter laboratory. The detection of exhaled carbon dioxide is a sensitive and accurate technique for the confirmation of correct tracheal tube placement in very low birth weight infants during neonatal resuscitation. In a study of attempted tracheal intubations in 40 neonates, FEco2 monitoring correctly identified all 40 tracheal and all 11 oesophageal intubations with 100% accuracy, and did so faster than concurrent clinical evaluation, which incorrectly identified the placement of the tube on three occasions [37]. It has been used successfully during resuscitation to confirm tracheal tube placement in extremely low birth weight premature babies (< 1000 g) [37], and Clifford and Hunt have suggested that routine use of capnography may facilitate neonatal clinical assessment during active resuscitation [38]. There is increasing interest in the use of capnography to monitor patients presenting with acute exacerbations of respiratory disease in the emergency department, without tracheal intubation. Changes in the slopes of the expiratory capnographic waveform can reflect changes in airway calibre in acute asthma [39]. Using capnography in this way has several advantages when compared with the use of peak flowmeters, as it is effort-independent and provides continuous monitoring of normal tidal breathing. This use of capnography has been proposed for the routine monitoring of asthmatics in the emergency department [40]. Although the measurement of FEco2 in the emergency department does not, as expected, seem to predict Paco2 accurately in patients presenting with acute dyspnoea [41], it will, as mentioned above, provide an efficient and reliable measure and trend of the respiratory rate. The British Thoracic Society’s guideline for use of emergency oxygen recommends that respiratory rate should be included in the assessment of all acutely breathless patients [42] and it is said to be the single best predictor of severe illness [43]. Professor Bill Runciman, pioneer of the Australian Incident Monitoring Study (AIMS), said in 2007 at a National Patient Safety Agency (NPSA) workshop on the National Reporting and Learning System (NRLS) that doctors as a group seem to stand alone in requiring a robust evidence base for everything they do and, specifically in the area of safety, a time must come when judgement and common sense have to be used and action taken in the absence of overwhelming evidence. If the brakes on a particular model of car failed, the automotive engineers would do something about it; they would not wait for the results of a large, randomised, controlled trial. Professor Sir Michael Rawlins, Chairman of the National Institute for Health and Clinical Excellence (NICE) in his Harveian Oration in 2008 [44], exposed the hierarchies of evidence, saying again that where safety was concerned, randomised, controlled trials were unlikely to yield the right answer quickly, and judgment was often required. Although using historical controls in Britain, ganciclovir had been shown to prevent blindness in cases of cytomegalovirus retinitis. Some national authorities insisted on a large, randomised, controlled trial before its routine introduction and, after two years in which many patients in the placebo arm had gone blind, they eventually agreed to its use [45]. Harm from thalidomide was first identified in a general practitioner’s letter to the Lancet [46], and bovine spongiform encephalopathy (BSE) at a veterinary surgeons’ lunchtime meeting in Sussex [47]. The speed at which drugs and devices are widely adopted by the medical profession is another whole topic in itself. However, it is clear that capnography is a reliable, safe, well-established and fully understood technology with an excellent 23-year track record in anaesthesia, whose cost is low and for which the case for its much wider application for patient safety has been made. This is yet another safety issue on which anaesthetists can show a lead. In 2011, it’s time for capnography everywhere. DKW is a Past President of the AAGBI and was a co-author of the AAGBI Safety Statement [4]. He is currently Chairman of the European Board of Anaesthesiology Safety Committee, and a member of the European Board of Anaesthesiology/European Society of Anaesthesiology Patient Safety Task Force. No external funding or competing interests declared. Since this Editorial was written, the 4th National Audit Project (NAP4) of the Royal College of Anaesthetists and the Difficult Airway Society has been published [48], which looked at major complications of airway management in three clinical areas: operating theatres; ICUs; and emergency departments. The report now provides the best estimates currently available of the incidence and impact of airway-related morbidity and mortality [49] and it has already reached the attention of the media [50]. The data show that in 2008–09 there were 16 airway deaths from the 3 million patients undergoing general anaesthesia monitored with continuous capnography, giving a death rate of 1:180 000. Similarly, there were 18 deaths from a much smaller number of ICU patients receiving ventilation: 48 000 that year [7], which gives an ICU airway death rate of 1:2700. This appears an extremely high death rate, and the NAP4 investigators commented that it is likely that 74% of these ICU deaths could have been prevented had continuous capnography been used [48]. As only 25% of ICUs monitor patients routinely with capnography [9], this provides further evidence that capnography should become universal practice in the ICU as a matter of urgency. Indeed, NAP4 states as one of its recommendations that: “continuous capnography should be used in all ITU patients with tracheal tubes (including tracheostomy) who are intubated and ventilator dependent” [48]. The authors similarly recommend that it should be used routinely in emergency departments. Many of these incidents also involve human factors, training and other issues but for the team involved, never having the vital information from capnography available to them handicaps their response right from the start. Looking at the two groups of patients, one in theatre with 100% capnography and the other in ICU with only 25% capnography [9], NAP4 has virtually reported on the controlled trial of capnography that was never done in the 1980s and which would never get ethical approval today. In this way, the report now provides robust evidence – for anyone who was still in any doubt – that continuous capnography should be universal in ICU.